\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: UNITED STATES of America, Plaintiff - Appellee, v. James Ryan TAYLOR, Defendant - Appellant. United States of America, Plaintiff - Appellee, v. Steven Vincent Smith, Defendant - Appellant.\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-28\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: Yes\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 17-14915 Date Filed: 08/28/2019 Page: 1 of 56\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 17-14915\\n ________________________\\n\\n D.C. Docket No. 2:16-cr-00203-KOB-JEO-1\\n\\n\\n\\nUNITED STATES OF AMERICA,\\n\\n Plaintiff - Appellee,\\n\\n versus\\n\\nJAMES RYAN TAYLOR,\\n\\n Defendant - Appellant.\\n\\n\\n\\n\\n ________________________\\n\\n No. 18-11852\\n ________________________\\n\\n D.C. Docket No. 4:16-cr-00312-VEH-JHE-1\\n\\n\\n\\nUNITED STATES OF AMERICA,\\n\\n Plaintiff - Appellee,\\n\\n versus\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 2 of 56\\n\\n\\n\\n\\nSTEVEN VINCENT SMITH,\\n\\n Defendant - Appellant.\\n\\n ________________________\\n\\n Appeals from the United States District Court\\n for the Northern District of Alabama\\n ________________________\\n\\n (August 28, 2019)\\n\\nBefore TJOFLAT and NEWSOM, Circuit Judges, and ANTOON, * District Judge.\\n\\nNEWSOM, Circuit Judge:\\n\\n James Taylor and Steven Smith are the latest in a long line of child-\\n\\npornography consumers to argue that the evidence of their crimes should be\\n\\nsuppressed because the warrant that led to its discovery—issued by a magistrate\\n\\njudge in the Eastern District of Virginia but purporting to authorize a nationwide,\\n\\nremote-access computer search—violated the Fourth Amendment. By our count,\\n\\nwe become today the eleventh (!) court of appeals to assess the constitutionality of\\n\\nthe so-called “NIT warrant.” Although the ten others haven’t all employed the\\n\\nsame analysis, they’ve all reached the same conclusion—namely, that evidence\\n\\ndiscovered under the NIT warrant need not be suppressed. We find no good\\n\\n\\n\\n\\n* Honorable John Antoon II, United States District Judge for the Middle District of Florida,\\nsitting by designation.\\n\\n 2\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 3 of 56\\n\\n\\nreason to diverge from that consensus here, but the case nonetheless calls for\\n\\ncareful consideration, as it implicates several important issues.\\n\\n As an initial matter, did the NIT warrant violate Federal Rule of Criminal\\n\\nProcedure 41(b), which specifies where and in what circumstances a magistrate\\n\\njudge may issue a warrant—and relatedly, if the warrant did violate Rule 41(b),\\n\\nwas that violation of constitutional magnitude? We hold that because the\\n\\nmagistrate judge’s actions exceeded not only Rule 41(b) but also her statutorily\\n\\nprescribed authority under the Federal Magistrates Act, 28 U.S.C. § 636(a)—which\\n\\ncircumscribes the scope of a magistrate judge’s jurisdiction—the warrant was void\\n\\nab initio, rendering any search purporting to rely on it warrantless and thus\\n\\npresumptively unlawful under the Fourth Amendment.\\n\\n That leads us to the question of remedy, which we take in two parts: First, is\\n\\nexclusion required—without regard to the reasonableness of the officers’\\n\\nreliance—where, as here, the warrant was void from the outset, as Taylor and\\n\\nSmith urge? Or, as the government contends, should a void warrant be treated no\\n\\ndifferently from other defective warrants, such that the good-faith exception to the\\n\\nexclusionary rule can still apply? We hold that, because the exclusionary rule is\\n\\nconcerned solely with deterring culpable police misconduct—and not at all with\\n\\nregulating magistrate judges’ actions—void and voidable warrants should be\\n\\n\\n\\n\\n 3\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 4 of 56\\n\\n\\ntreated no differently; accordingly, an officer’s reasonable reliance on the former,\\n\\nlike the latter, can provide the basis for applying the good-faith exception.\\n\\n Second, even if the good-faith exception can apply when an officer relies on\\n\\na void warrant, should the exception apply in the particular circumstances of this\\n\\ncase? We hold that the officers’ warrant application here adequately disclosed the\\n\\nnature of the technology at issue and the scope of the intended search, that the\\n\\nofficers reasonably relied on the magistrate judge’s determination that the search\\n\\nwas permissible, and, accordingly, that the good-faith exception applies in this\\n\\ncase.\\n\\n I\\n\\n A\\n\\n We begin with a bit of context. In the normal world of web browsing, an\\n\\ninternet service provider—Comcast or AT&T, for example—assigns an IP address\\n\\nto every computer that it provides with internet access. An IP address is a unique\\n\\nnumerical identifier, tantamount to a computer’s name. (OK, in the laptop era it’s\\n\\nslightly more complicated than that, because the “name” changes as the computer\\n\\nmoves around and connects to different service providers’ networks—but you get\\n\\nthe picture.) Websites can log IP addresses to keep track of the computers that\\n\\nvisit, in essence creating a digital guest book. Internet browsing, therefore, isn’t\\n\\nquite as private as most people think—it’s actually pretty easy, for instance, for\\n\\n\\n 4\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 5 of 56\\n\\n\\nlaw enforcement to find out who visited what sites, when, and for how long simply\\n\\nby subpoenaing IP-address logs from service providers.\\n\\n Not so when it comes to the “dark web,” the part of the internet “only\\n\\naccessible by means of special software, allowing users and website operators to\\n\\nremain anonymous or untraceable.” Blog.OxfordDictionaries.com. 1 “The Onion\\n\\nRouter”—usually abbreviated “Tor”—is one such software program. Tor, which\\n\\nwas the brainchild of the U.S. Navy but has since been released to the public,\\n\\nworks by routing a user’s webpage requests through a series of computer servers\\n\\noperated by volunteers around the globe, rendering the user’s IP address essentially\\n\\nunidentifiable and untraceable. In the words of the folks who currently administer\\n\\nthe “Tor Project,” a Massachusetts-based § 501(c)(3) organization responsible for\\n\\nmaintaining Tor, you might think of what Tor does as “using a twisty, hard-to-\\n\\nfollow route in order to throw off someone who is tailing you—and then\\n\\nperiodically erasing your footprints.” 2\\n\\n\\n\\n1 See also Ahmed Ghappour, Searching Places Unknown: Law Enforcement Jurisdiction on the\\nDark Web, 69 Stan. L. Rev. 1075, 1087 (2017) (“The dark web is a private global computer\\nnetwork that enables users to conduct anonymous transactions without revealing any trace of\\ntheir location.”).\\n2 See Lee Matthews, What Tor Is, and Why You Should Use It to Protect Your Privacy, Forbes\\n(Jan. 27, 2017 2:30 p.m.), https://www.forbes.com/sites/leemathews/2017/01/27/what-is-tor-and-\\nwhy-do-people-use-it/#3186d5387d75 (last visited Aug. 27, 2019); see also Tor Project,\\nhttps://2019.www.torproject.org/projects/torbrowser.html.en (“[Tor] prevents somebody\\nwatching your Internet connection from learning what sites you visit, it prevents the sites you\\nvisit from learning your physical location, and it lets you access sites which are blocked.”) (last\\nvisited Aug. 27, 2019).\\n\\n 5\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 6 of 56\\n\\n\\n As you can imagine, Tor has plenty of legitimate uses—think military and\\n\\nlaw-enforcement officers carrying out investigations, journalists seeking to\\n\\nmaintain anonymity, and ordinary citizens researching embarrassing topics. As\\n\\nyou can also imagine, Tor has spawned—and effectively enables—a cache of\\n\\nunsavory sites for black-market trading, child-pornography file-sharing, and other\\n\\ncriminal enterprises. This is so because, in addition to allowing users to access\\n\\npublic websites without leaving a trail, Tor also hosts a number of so-called\\n\\n“hidden services,” i.e., sites accessible only through Tor. You can’t just Google a\\n\\nhidden service; rather, a user can access one of these Tor-specific sites only by\\n\\nknowing its exact URL address. Most Tor-site addresses comprise a random\\n\\njumble of letters and numbers followed by the address “.onion”—in place, say, of\\n\\n“.com” or “.org”—and are shared via message-board postings on the regular\\n\\ninternet or by word of mouth.\\n\\n The hidden-service page at issue here, “Playpen,” was a child-pornography-\\n\\ndistribution site accessible only through Tor. At the time the FBI began\\n\\nmonitoring Playpen, the site contained more than 95,000 posts, had 160,000\\n\\nmembers, and hosted up to 1,500 visitors per day. The FBI monitored the site for\\n\\nseveral months until, based on a foreign-government tip, it found and arrested the\\n\\nadministrator. Rather than shuttering Playpen immediately, the FBI covertly took\\n\\n\\n\\n\\n 6\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 7 of 56\\n\\n\\ncontrol of the site and began operating it out of a government server in Newington,\\n\\nVirginia, hoping to snare more users.\\n\\n As a means of ferreting out Playpen visitors whose identities were masked\\n\\nby Tor, the FBI sought to deploy government-created malware—specifically, a\\n\\ncomputer code called the Network Investigative Technique (“NIT”)—that would\\n\\ntransmit user information back to the FBI. Here’s how the NIT worked: When a\\n\\nPlaypen user downloaded images from a Tor-based site, the NIT would essentially\\n\\n“hitchhike” along, invade the host computer, and force it to send to the FBI\\n\\n(among other information) the computer’s IP address, the computer’s host name,\\n\\nand the username associated with the computer. Based on that information, the\\n\\nFBI could identify the user’s internet service provider and the computer affiliated\\n\\nwith the account that accessed Playpen, thereby unmasking the user and providing\\n\\nprobable cause for the FBI to seek a warrant to seize computers and hard drives.\\n\\n B\\n\\n To effectuate this plan, FBI Agent Douglas Macfarlane submitted a search-\\n\\nwarrant application to a magistrate judge in the Eastern District of Virginia,\\n\\nrequesting authorization to deploy the NIT. The application wasn’t a model of\\n\\nclarity or precision, particularly regarding the issue that most concerns us here—\\n\\nnamely, the geographic scope of the requested search authority. In the case\\n\\ncaption, the application described the “property to be searched”—seemingly\\n\\n\\n 7\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 8 of 56\\n\\n\\nwithout territorial restriction—as “COMPUTERS THAT ACCESS\\n\\nupf45jv3bziuctml.onion,” which we now know to be associated with Playpen. Just\\n\\nbelow, however, in the body, the application asserted a reasonable belief that\\n\\nevidence of child-pornography-related crimes was contained on property “located\\n\\nin the Eastern District of Virginia.” As part of the same statement—regarding the\\n\\n“property to be searched”—the application referred to an “Attachment A.”\\n\\nAttachment A in turn stated that the NIT was “to be deployed on the computer\\n\\nserver . . . operating the [Playpen] website” and specified that the server was\\n\\n“located at a government facility in the Eastern District of Virginia.” Attachment\\n\\nA then went on to state, though, that the goal of deploying the NIT was to obtain\\n\\ninformation from “[t]he activating computers . . . of any user or administrator who\\n\\nlogs into [Playpen] by entering a username and password.”\\n\\n As is often the case, the NIT application also referenced an attached\\n\\naffidavit. Agent Macfarlane’s affidavit summarized the applicable law, explained\\n\\nnumerous technical terms of art, and described Tor and the “Target Website”—i.e.,\\n\\nPlaypen. On page 29 of 31, under the bolded heading “SEARCH\\n\\nAUTHORIZATION REQUESTS,” the affidavit stated, for the first time expressly,\\n\\nthat “the NIT may cause an activating computer—wherever located—to send to a\\n\\n\\n\\n\\n 8\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 9 of 56\\n\\n\\ncomputer controlled by or known to the government” certain information,\\n\\nincluding the IP address and host name. 3\\n\\n A magistrate judge in the Eastern District of Virginia signed the warrant and\\n\\nthe FBI deployed the NIT.\\n\\n C\\n\\n Not long thereafter, NIT-transmitted data revealed to the FBI that a certain\\n\\nPlaypen user was linked to a computer with the host name “RyansComputer.”\\n\\nAfter the user accessed several images of child pornography, the FBI sent an\\n\\nadministrative subpoena to the user’s internet service provider and discovered that\\n\\nthe IP address associated with the computer was assigned to James Taylor in\\n\\nBirmingham, Alabama. A magistrate judge in the Northern District of Alabama\\n\\nthen authorized a search warrant for Taylor’s residence, where the FBI seized\\n\\nTaylor’s laptop, hard drive, and USB drive. After analyzing the hardware twice,\\n\\nthe FBI found what it was looking for.\\n\\n Steven Smith’s Playpen activities were discovered in a nearly identical way.\\n\\nAs in Taylor’s case, the NIT revealed that someone had used Smith’s computer\\n\\nand IP address to log into Playpen. Based on the NIT data, the FBI subpoenaed\\n\\n\\n\\n3 The warrant also explained that the NIT would send the following information: the unique\\nidentifier that distinguishes the data on the host computer from that of other computers, the type\\nof operating system the host computer is running, whether the NIT has already been downloaded\\nto the host computer, an active operating system username, and a Media Access Control address.\\n\\n 9\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 10 of 56\\n\\n\\nrecords from an internet service provider and used that information to secure a\\n\\nwarrant from a magistrate judge in the Northern District of Alabama, allowing\\n\\nofficers to search Smith’s residence in Albertville, Alabama. The search revealed\\n\\nchild-pornography images on a thumb drive. After arresting Smith, the officers\\n\\nobtained a search warrant for his office and seized his work computer, which also\\n\\ncontained child pornography.\\n\\n Taylor and Smith were charged with receiving child pornography under 18\\n\\nU.S.C. § 2252A(a)(2) and with possessing and accessing child pornography with\\n\\nthe intent to view it under 18 U.S.C. § 2252A(a)(5)(B) & (b)(2). They both moved\\n\\nto suppress the evidence against them, asserting, as relevant here, that the NIT\\n\\nwarrant violated the Fourth Amendment, Federal Rule of Criminal Procedure\\n\\n41(b), and the Federal Magistrates Act, 28 U.S.C. § 636(a), and, accordingly, that\\n\\nthe seized images should be suppressed as fruit of the poisonous tree. The district\\n\\ncourt in each case denied the motion to suppress. Both courts agreed that the NIT\\n\\nwarrant violated the Fourth Amendment—and was thus void—but declined to\\n\\nsuppress the evidence on the ground that the searches, and the resulting seizures,\\n\\nfell within the good-faith exception to the exclusionary rule. Both defendants\\n\\nappealed, and their cases were consolidated for review and decision.\\n\\n\\n\\n\\n 10\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 11 of 56\\n\\n\\n II\\n\\n All here agree that the NIT’s extraction and transmission of Taylor’s and\\n\\nSmith’s information was a “search” within the meaning of the Fourth Amendment.\\n\\nU.S. Const. amend. IV. 4 All likewise agree that no exigency or other exception\\n\\nexempted the FBI from the usual requirement to obtain a search warrant. See\\n\\nUnited States v. Cooks, 920 F.3d 735, 741 (11th Cir. 2019) (“[W]arrantless\\n\\nsearches are presumptively unreasonable, ‘subject only to a few specifically\\n\\nestablished and well-delineated exceptions.’” (quoting Katz v. United States, 389\\n\\nU.S. 347, 357 (1967))). There, the agreement ends. The parties vigorously dispute\\n\\nwhether the NIT warrant was valid and, if not, whether (and to what extent) that\\n\\nfact should bear on the admissibility of the evidence found. Accordingly, we are\\n\\nfaced with the following issues, each with its own twists and turns: (1) Did the NIT\\n\\nwarrant violate Federal Rule of Criminal Procedure 41(b) and, if so, did it likewise\\n\\n\\n\\n\\n4 That Taylor and Smith used Tor to download child pornography is important because it takes\\nthis case out of third-party-doctrine land. See Smith v. Maryland, 442 U.S. 735 (1979). Instead\\nof traveling along the equivalent of “public highways” (by browsing the open internet) or leaving\\nthe equivalent of a calling card at each website visited (as with a normal internet search), Tor\\nusers purposefully shroud their browsing, such that they have a reasonable expectation of privacy\\nin their online “movements.” See United States v. Davis, 785 F.3d 498, 507 (11th Cir. 2015)\\n(explaining that the Fourth Amendment’s protections apply where an individual has exhibited “a\\nsubjective expectation of privacy” that society recognizes as reasonable (citation omitted)).\\n\\n\\n 11\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 12 of 56\\n\\n\\nviolate the Fourth Amendment? And (2) if the NIT warrant did run afoul of the\\n\\nFourth Amendment, does the exclusionary rule apply? 5\\n\\n A\\n\\n 1\\n\\n Federal Rule of Criminal Procedure 41(b), titled “Venue for a Warrant\\n\\nApplication,” both outlines the situations in which a magistrate judge may issue a\\n\\nwarrant for a search within her district and specifies the more limited\\n\\ncircumstances in which she may issue a warrant for a search outside her district.\\n\\nWith respect to the former, Rule 41(b)(1) states that “a magistrate judge with\\n\\nauthority in the district . . . has authority to issue a warrant to search for and seize a\\n\\nperson or property located within the district.” Fed. R. Crim. P. 41(b)(1). It is\\n\\nundisputed, though, that the NIT warrant sought authority to search for information\\n\\noutside the territorial confines of the Eastern District of Virginia. And the parties\\n\\nagree that, for present purposes, Rule 41(b)(4)—which authorizes “tracking\\n\\ndevice” warrants—is the only provision that could have empowered the magistrate\\n\\njudge to authorize the specific out-of-district search in this case. That rule permits\\n\\na magistrate “to issue a warrant to install within the district a tracking device” to\\n\\n\\n5 In reviewing a district court’s denial of a motion to suppress, we review factual findings for\\nclear error and the application of law to those facts de novo. United States v. Ramirez, 476 F.3d\\n1231, 1235 (11th Cir. 2007). Where, as here, the facts are undisputed, we simply review the\\nlegality of a search de novo. United States v. Phillips, 834 F.3d 1176, 1179 (11th Cir. 2016).\\n\\n\\n 12\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 13 of 56\\n\\n\\n“track the movement of a person or property located within the district, outside the\\n\\ndistrict, or both.” Fed. R. Crim. P. 41(b)(4) (emphasis added). 6 Accordingly, the\\n\\nNIT warrant complies with Rule 41(b) only if we conclude that it was issued in\\n\\naccordance with subsection (b)(4). 7\\n\\n We find two mismatches—one formal (but telling) and the other substantive.\\n\\nInitially, as a matter of form, although the government now defends the NIT\\n\\nwarrant on a tracking-device basis, it conspicuously didn’t seek the warrant under\\n\\nRule 41(b)(4). Tracking-device warrants issued under subsection (b)(4) are\\n\\ngenerally requested pursuant to a specialized “Application for a Tracking\\n\\nWarrant.” 8 Here, though, the FBI seems to have sought the NIT warrant under\\n\\nRule 41(b)(1)’s general provision for warrants authorizing in-district searches. The\\n\\nwarrant application’s cover sheet represented that the FBI wished to search\\n\\nproperty “located in the Eastern District of Virginia,” and neither the application\\n\\nnor the accompanying affidavit mentioned the term “tracking device” or otherwise\\n\\nindicated that the application sought authorization under subsection (b)(4). The\\n\\n\\n\\n6 As it turns out, Rule 41(b) has since been amended to add a provision—subsection (b)(6)—for\\nremote electronic searches of the sort at issue in this case. See infra Section II.B.2.\\n7No court of appeals has found that the NIT warrant fits within the tracking-device exception,\\nalthough this argument has persuaded a few district courts. See United States v. Taylor, 250 F.\\nSupp. 3d 1215, 1222–23 (N.D. Ala. 2017) (compiling district and appellate court holdings on\\nNIT-warrant searches).\\n8 See, e.g., Administrative Office of U.S. Courts, Criminal Forms AO 102 (2009) & AO 104\\n(2016), http://www.uscourts.gov/forms/criminal-forms (last visited Apr. 26, 2019).\\n\\n 13\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 14 of 56\\n\\n\\ngovernment’s revisionism on appeal—invoking Rule 41(b)(4) to defend what was,\\n\\nby all accounts, a Rule 41(b)(1) application—undermines its position that the\\n\\nRule’s tracking-device provision sanctions the NIT warrant.\\n\\n Moreover, and in any event, we reject the government’s tracking-device\\n\\nargument on the merits. For Rule 41 purposes, a “tracking device” is “an\\n\\nelectronic or mechanical device which permits the tracking of the movement of a\\n\\nperson or object.” 18 U.S.C. § 3117(b); see also Fed. R. Crim. P. 41(a)(2)(E)\\n\\n(explaining that “‘[t]racking device’ has the meaning set out in 18 U.S.C. §\\n\\n3117(b)”). The government contends that the NIT constitutes a tracking device\\n\\nbecause “just as a GPS tracker attached to a car will send a receiver coordinates or\\n\\nother signals with locational information, the NIT augmented the content of\\n\\nPlaypen and sent locational information back to a government-controlled\\n\\ncomputer.” Br. of Appellee at 15.\\n\\n We disagree. The NIT didn’t reveal “locational information” at all—it\\n\\ndidn’t even send a locational snapshot, let alone the type of ongoing, GPS-\\n\\ncoordinate transmissions that would “permit[] the tracking of the movement of a\\n\\nperson or object” within the meaning of Rule 41(b)(4). Rather, the NIT performed\\n\\na one-time extraction of non-locational information—including a computer’s IP\\n\\naddress, username, and other identifying material—which it transmitted to the FBI.\\n\\nOf course, the identifying information that the NIT extracted and sent was then\\n\\n\\n 14\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 15 of 56\\n\\n\\ntraced to a physical address using an internet service provider’s records, but only in\\n\\nthe same way that a person’s name might be traced to a physical address using a\\n\\nphone book. In other words, that the FBI eventually used the NIT-transmitted\\n\\ninformation to discover additional facts that, in turn, enabled it to then determine a\\n\\nPlaypen user’s location in no way transformed the initial information transmittal\\n\\ninto “tracking.” 9\\n\\n To be clear, it’s not just that the NIT isn’t exactly a tracking device—it’s\\n\\nthat it’s exactly not a tracking device. A GPS tracker stuck to the bottom of a car\\n\\ncan’t tell you the car’s make and model, its owner, or its place of registration—but\\n\\nit can tell you whether the car is parked at Starbucks or cruising down I-20. By\\n\\ncontrast, the NIT malware can and did transmit the equivalent of a computer’s\\n\\nname, to whom it was registered, and other identifying information—but it didn’t\\n\\n(and couldn’t) reveal whether the computer was at the owner’s home or office, at\\n\\nStarbucks, or in the car on the move. In short, while a tracking device transmits\\n\\n\\n\\n\\n9 The government also points out that the NIT was deployed from a computer in the Eastern\\nDistrict of Virginia—which, it says, is the equivalent of a tracking device being “installed within\\nthe district.” But a GPS tracker that is physically attached to an item within the territorial\\nconfines of a particular district is clearly “install[ed] within” that district. By contrast, the NIT\\nsoftware, although deployed and activated from a government computer in the Eastern District of\\nVirginia, was not “installed within” that district—it was installed on suspects’ computers outside\\nof the district.\\n\\n\\n 15\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 16 of 56\\n\\n\\nlocation but not identifying information, the NIT sent identifying information but\\n\\nnot location. 10\\n\\n To sum up, we hold that the NIT is not a “tracking device” within the\\n\\nmeaning of Federal Rule of Criminal Procedure 41(b), and we reject the\\n\\ngovernment’s post hoc attempts to classify it as such. Because the NIT warrant\\n\\nwas not authorized by any of Rule 41(b)’s applicable subsections, the warrant\\n\\nviolated the Rule.\\n\\n 2\\n\\n So, what effect? While constitutional violations may merit suppression—\\n\\nmore on that later—mere “technical noncompliance” with a procedural rule results\\n\\nin the exclusion of evidence only when (1) “there was ‘prejudice’ in the sense that\\n\\nthe search might not have occurred or would not have been so abrasive if the rule\\n\\nhad been followed,” or (2) “there is evidence of intentional and deliberate\\n\\ndisregard of a provision in the Rule.” United States v. Williams, 871 F.3d 1197,\\n\\n1203 (11th Cir. 2017) (citation omitted).\\n\\n Which do we have here—a constitutional violation or just a technical one?\\n\\nThe government says that the violation in this case was merely technical because\\n\\n\\n\\n10If the term “tracking device” included every gadget capable of acquiring and transmitting\\ninformation that could somehow, in some way, aid in identifying a person’s location, the term\\nwould be unimaginably broad, including any phone or camera capable of sending a photo, as\\nimages of buildings, street signs, or other landmarks can surely be used to identify a location.\\n\\n 16\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 17 of 56\\n\\n\\nRule 41(b) is just a venue provision—it has nothing to do with a magistrate’s\\n\\npower or jurisdiction. The government points out, for instance, that as of 2016,\\n\\nRule 41(b) is no longer titled “Authority to Issue a Warrant,” but rather “Venue for\\n\\na Warrant Application.” See Fed. R. Crim. P. 41(b). And, the argument goes, if\\n\\nRule 41(b) is an ordinary venue provision, a breach of its provisions would not rise\\n\\nto the level of a constitutional violation.\\n\\n Fair enough. As we’ve recently been at pains to emphasize—following the\\n\\nSupreme Court’s lead—not every mandatory proclamation or prohibition creates a\\n\\njurisdictional bar, and we are loath to “jurisdictionalize” issues unnecessarily. See,\\n\\ne.g., Orion Marine Constr., Inc. v. Carroll, 918 F.3d 1323, 1328–29 (11th Cir.\\n\\n2019); Sec’y, U.S. Dep’t of Labor v. Preston, 873 F.3d 877, 881–82 (11th Cir.\\n\\n2017). Here, though, jurisdiction is squarely in play: While Rule 41(b) itself may\\n\\naddress only venue, the statute behind the rule—the Federal Magistrates Act, 28\\n\\nU.S.C. § 636—imposes clear jurisdictional limits on a magistrate judge’s power.\\n\\nSection 636(a) states that magistrate judges “shall have within [their] district[s]”\\n\\nthe “powers . . . conferred . . . by law or by the Rules of Criminal Procedure.” 28\\n\\nU.S.C. § 636(a)(1) (emphasis added). Because no one contends that any law or\\n\\nRule other than Rule 41(b) gave the magistrate judge the authority to issue the NIT\\n\\nwarrant in this case, when the magistrate issued the warrant outside of Rule 41(b)’s\\n\\nambit, she necessarily transgressed the limits of her jurisdiction.\\n\\n\\n 17\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 18 of 56\\n\\n\\n We aren’t breaking any new ground here. As now-Justice Gorsuch\\n\\nexplained during his tenure on the Tenth Circuit, § 636(a) “expressly—and\\n\\nexclusively—refers to the territorial scope of a magistrate judge’s power to\\n\\nadjudicate” and, further, is “found in Title 28 of the U.S. Code—the same title as\\n\\nthe statutes that define a district court’s jurisdiction.” United States v. Krueger,\\n\\n809 F.3d 1109, 1122 (10th Cir. 2015) (Gorsuch, J., concurring). Or, as the Ninth\\n\\nCircuit put it, “federal magistrates are creatures of [§ 636(a)], and so is their\\n\\njurisdiction.” N.L.R.B. v. A-Plus Roofing, Inc., 39 F.3d 1410, 1415 (9th Cir. 1994);\\n\\nsee also United States v. Hazlewood, 526 F.3d 862, 864 (5th Cir. 2008) (“In the\\n\\nFederal Magistrates Act, 28 U.S.C. § 636, Congress conferred jurisdiction to\\n\\nfederal magistrate[]judge[s].”). Thus, as § 636(a) is the sole source of a magistrate\\n\\njudge’s warrant authority, a warrant issued in defiance of its jurisdictional\\n\\nlimitations is void—“no warrant at all.” Krueger, 809 F.3d at 1118 (Gorsuch, J.,\\n\\nconcurring).\\n\\n To be fair, Krueger was an easier case—there, a magistrate judge in one\\n\\ndistrict purported to authorize a search in an adjacent district, in which she clearly\\n\\nhad no jurisdiction. The magistrate judge here, by contrast, issued a warrant\\n\\npurporting to allow a search of computers “wherever located”—which, of\\n\\nnecessity, included her own district. But the fact that the warrant in its overbreadth\\n\\nhappened to sweep in the Eastern District of Virginia along with the rest of the\\n\\n\\n 18\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 19 of 56\\n\\n\\nnation doesn’t cure the fact that it was issued outside of the magistrate judge’s\\n\\nstatutorily prescribed (and proscribed) authority in the first place. Indeed, the idea\\n\\nthat a warrant may be issued partially from a place of statutorily-granted authority\\n\\nand partially from the great beyond (with one foot inside and one foot outside the\\n\\nlines, so to speak) strikes us as nonsensical. Rather, it seems to us that a magistrate\\n\\njudge must act either pursuant to the authority granted her by statute or not, and\\n\\nthus have the authority either to issue a warrant (in toto) or not. 11\\n\\n Because the NIT warrant was void at issuance, the ensuing search was\\n\\neffectively warrantless and therefore—because no party contends that an exception\\n\\nto the presumptive warrant requirement applies here—violative of the Fourth\\n\\nAmendment. Accord United States v. Werdene, 883 F.3d 204, 214 (3d Cir.), cert.\\n\\ndenied, 139 S. Ct. 260 (2018); United States v. Horton, 863 F.3d 1041, 1050 (8th\\n\\n\\n\\n\\n11 Nor do we see a persuasive case for “severing” the NIT warrant, so to speak, along\\njurisdictional lines—such that it might be deemed valid in the Eastern District of Virginia, even\\nif invalid everywhere else, and thus not void ab initio and in toto (to really pour on the Latin).\\nWe are aware, of course, that several courts have held that a warrant can be severed along what\\nmight loosely be called subject-matter lines—i.e., with respect to probable cause or particularity.\\nSee, e.g., United States v. George, 975 F.2d 72, 79 (2d Cir. 1992) (“When a warrant is severed\\n(or redacted) the constitutionally infirm portion—usually for lack of particularity or probable\\ncause—is separated from the remainder and evidence seized pursuant to that portion is\\nsuppressed; evidence seized under the valid portion may be admitted.”). But the flaws in the two\\nsituations, it seems to us, are fundamentally different. Subject-matter severance addresses an\\nerror made by a properly empowered official; the error that plagues the NIT warrant is more\\nfundamental—it implicates the magistrate judge’s power to act in the first instance.\\n\\n\\n 19\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 20 of 56\\n\\n\\nCir. 2017), cert. denied, 138 S. Ct. 1440 (2018); United States v. Henderson, 906\\n\\nF.3d 1109, 1116 (9th Cir. 2018), cert. denied, 139 S. Ct. 2033 (2019). 12\\n\\n B\\n\\n So the search carried out under the NIT warrant violated not just Rule 41 but\\n\\nalso the Fourth Amendment. But again: What effect? At last we come to the\\n\\nquestion at the heart of the remedy that Taylor and Smith seek. Can the good-faith\\n\\nexception to the exclusionary rule apply in a situation like this, where officers rely\\n\\non a warrant that is later determined to have been void ab initio? And more\\n\\nspecifically, does the good-faith exception apply in the particular circumstances of\\n\\nthis case?\\n\\n 1\\n\\n The “exclusionary rule”—which operates to bar the admission of evidence\\n\\nobtained in violation of the Fourth Amendment—appears nowhere in the\\n\\nConstitution’s text. It is, the Supreme Court has said, not “a personal\\n\\n\\n12The government also contends—in nearly identical terms in both cases—that “[b]ecause the\\nsearch of Taylor’s [and Smith’s] computer[s] would have been valid if a magistrate judge in the\\nNorthern District of Alabama had signed the NIT Warrant, any Rule 41(b) violation did not\\ncause [them] prejudice” and suppression is not necessary. Br. of Appellee at 34 (emphasis\\nadded) (Taylor); see also Br. of Appellee at 29 (Smith). “Taylor [and Smith] suffered no more\\nof an intrusion of [their] privacy,” the government contends, “than [they] would have if the FBI\\nhad searched [their] computer[s] under a valid warrant.” Br. of Appellee at 31 (Taylor); see also\\nBr. of Appellee at 28 (Smith). No. Had the magistrate judge in the Eastern District of Virginia\\nacted within her jurisdiction, the warrant could not have extended to Alabama and the FBI would\\nnot have identified Taylor or Smith, nor would it have had probable cause to apply for a second\\nwarrant to search their homes.\\n\\n\\n 20\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 21 of 56\\n\\n\\nconstitutional right,” but rather a “judicially created” remedy, whose purpose is to\\n\\n“deter future Fourth Amendment violations” and “compel respect for the\\n\\nconstitutional guaranty.” Davis v. United States, 564 U.S. 229, 236–37, 238\\n\\n(2011) (citation omitted). This remedy, however, doesn’t follow automatically;\\n\\nsociety must swallow the “bitter pill” of suppression when necessary, id. at 238,\\n\\nbut only when the “benefit” of exclusion outweighs its “substantial social costs,”\\n\\nIllinois v. Krull, 480 U.S. 340, 352–53 (1987). The dual pillars of the exclusion\\n\\ndecision, the Supreme Court recently emphasized, are deterrence and culpability:\\n\\n“Police practices trigger the harsh sanction of exclusion only when they are\\n\\ndeliberate enough to yield ‘meaningfu[l]’ deterrence, and culpable enough to be\\n\\n‘worth the price paid by the justice system.’” Davis, 564 U.S. at 240 (alteration in\\n\\noriginal) (quoting Herring v. United States, 555 U.S. 135, 144 (2009)); see also id.\\n\\n(suppression not warranted because officer did not act “deliberately, recklessly, or\\n\\nwith gross negligence”).\\n\\n The good-faith exception is a “judicially created exception to this judicially\\n\\ncreated rule.” Id. at 248. 13 In United States v. Leon, the Supreme Court explained\\n\\n\\n13Although “good faith” is most often framed as an “exception” to the exclusionary rule, it is\\nprobably more accurately described as a reason for declining to invoke the exclusionary rule in\\nthe first place. Compare, e.g., Davis, 564 U.S. at 238 (“The Court has over time applied this\\n‘good-faith’ exception across a range of cases.” (emphasis added)), with, e.g., id. at 239 (“The\\nquestion in this case is whether to apply the exclusionary rule when the police conduct a search\\nin objectively reasonable reliance on binding judicial precedent.” (emphasis added)), and\\nHerring v. United States, 555 U.S. 135, 139 (2009) (characterizing the question presented as\\n\\n\\n 21\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 22 of 56\\n\\n\\nthat exclusion is not warranted when police act “in objectively reasonable reliance”\\n\\non a subsequently invalidated search warrant—in other words, when they act in\\n\\n“good faith.” 468 U.S. 897, 922 (1984). “‘[O]ur good-faith inquiry is confined to\\n\\nthe objectively ascertainable question whether a reasonably well trained officer\\n\\nwould have known that the search was illegal’ in light of ‘all of the\\n\\ncircumstances.’” Herring, 555 U.S. at 145 (quoting Leon, 468 U.S. at 922 n.23).\\n\\n To date, the Supreme Court has applied the good-faith exception when,\\n\\namong other things, officers reasonably relied on a warrant that was later deemed\\n\\ninvalid for lack of probable cause, see Leon, 468 U.S. at 922, on a warrant that\\n\\nerroneously appeared outstanding due to an error in a court or police database, see\\n\\nArizona v. Evans, 514 U.S. 1, 4 (1995); Herring, 555 U.S. at 137, on a statute that\\n\\nwas later deemed unconstitutional, see Krull, 480 U.S. at 352–53, and on a judicial\\n\\ndecision that was later overruled, Davis, 564 U.S. at 232. The Supreme Court\\n\\nhasn’t, however, directly addressed the particular question before us today—\\n\\nwhether the good-faith exception can be applied to a search conducted in reliance\\n\\non a warrant that was void from the outset.\\n\\n Taylor and Smith insist that the void-voidable distinction is critical.\\n\\nReliance on a voidable warrant—issued in error, perhaps, but by a judge with\\n\\n\\n“whether the exclusionary rule should be applied” when officers act in reasonable reliance on a\\nnegligent police database error (emphasis added)).\\n\\n\\n 22\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 23 of 56\\n\\n\\njurisdiction to act—is different, they contend, from reliance on a warrant that was\\n\\nvoid from the get-go. Because the latter is—as we’ve agreed—“no warrant at all,”\\n\\nTaylor and Smith insist that reliance on it can’t provide an exception to the\\n\\nexclusionary rule. This is so, they continue, because the “heart of the good faith\\n\\nexception is [] officers’ reliance on a neutral third party’s actions within the scope\\n\\nof the third party’s authority.” Br. of Appellant Taylor at 29; Br. of Appellant\\n\\nSmith at 27.\\n\\n There is a certain logic to this argument: In fact, there was never a valid\\n\\nwarrant, so the search was illegal all along. What matters for exclusionary-rule\\n\\nand good-faith purposes, though, isn’t the validity of the warrant “in fact,” but\\n\\nrather the validity of the warrant as it would have reasonably appeared to an officer\\n\\ntasked with executing it. The appropriate question, therefore, is whether, from the\\n\\nperspective of a reasonable officer, there is any difference—for deterrence or\\n\\nculpability purposes—between the warrant issued in this case and the warrants\\n\\nissued in Leon, Evans, and Herring?\\n\\n We don’t think so. The exclusionary rule is concerned with deterring officer\\n\\nmisconduct and punishing officer culpability—not with setting judges straight. See\\n\\nHerring, 555 U.S. at 142 (observing that the “exclusionary rule was crafted to curb\\n\\npolice rather than judicial misconduct”). Viewed from an officer’s perspective,\\n\\nrelying on a facially valid warrant that, as it turns out, was void from the beginning\\n\\n\\n 23\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 24 of 56\\n\\n\\nis no different from relying on a facially valid warrant that, for instance, was later\\n\\ndeemed improper based on a dubious determination of probable cause, see Leon,\\n\\n468 U.S. at 925–26, or appeared outstanding thanks only to a database error, see\\n\\nHerring, 555 U.S. at 136–37. So long as an officer could reasonably have thought\\n\\nthat the warrant was valid, the specific nature of the warrant’s invalidity is\\n\\nimmaterial.\\n\\n In so holding, we join every court of appeals to consider the question, all of\\n\\nwhich have agreed that the good-faith exception applies—and the exclusionary\\n\\nrule doesn’t—in a situation like this. See United States v. Eldred, No. 17-3367-cv,\\n\\n2019 WL 3540415, at *8 (2d Cir. Aug. 5, 2019); United States v. Ganzer, 922 F.3d\\n\\n579, 587–90 (5th Cir.), petition for cert. filed, No. 19-5339 (2019); United States v.\\n\\nMoorehead, 912 F.3d 963, 971 (6th Cir.), petition for cert. filed, No. 19-5444\\n\\n(2019); Werdene, 883 F.3d at 216–17; United States v. McLamb, 880 F.3d 685,\\n\\n691 (4th Cir.), cert. denied, 139 S. Ct. 156 (2018); United States v. Kienast, 907\\n\\nF.3d 522, 527–28 (7th Cir. 2018), cert. denied, 139 S. Ct. 1639 (2019); Henderson,\\n\\n906 F.3d at 1118; United States v. Levin, 874 F.3d 316, 323–24 (1st Cir. 2017);\\n\\nHorton, 863 F.3dat 1050; United States v. Workman, 863 F.3d 1313, 1319 (10th\\n\\nCir. 2017), cert. denied, 138 S. Ct. 1546 (2018). As the Sixth Circuit summarized,\\n\\n“[t]he good-faith exception is not concerned with whether a valid warrant exists,\\n\\nbut instead asks whether a reasonably well-trained officer would have known that a\\n\\n\\n 24\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 25 of 56\\n\\n\\nsearch was illegal.” Moorehead, 912 F.3d at 968. The Third Circuit similarly\\n\\nexplained the “fundamental flaw” in the argument like the one that Taylor and\\n\\nSmith make here: “[I]t does not appreciate the distinction between the validity of\\n\\nthe warrant and the deterrence rationale of the exclusionary rule and the good-faith\\n\\nexception.” Werdene, 883 F.3d at 216.\\n\\n In light of the exclusionary rule’s purpose of deterring culpable police\\n\\nmisconduct, there is no reason to distinguish between good-faith reliance on a void\\n\\nwarrant and any other warrant later deemed defective. We thus hold that the good-\\n\\nfaith exception to the exclusionary rule can apply when police officers reasonably\\n\\nrely on a warrant later determined to have been void ab initio.\\n\\n 2\\n\\n Finally, then, to this particular case: Having determined that the good-faith\\n\\nexception can apply in situations involving void warrants, the question remains\\n\\nwhether the exception should apply to the cases before us today. In Leon, the\\n\\nSupreme Court laid out several situations in which the good-faith exception should\\n\\nnot apply: (1) where the magistrate judge was misled by information in a warrant\\n\\napplication that the applicant knew was false or would have known was false but\\n\\nfor a reckless disregard of the truth; (2) where the magistrate “wholly abandoned”\\n\\nher judicial role; (3) where the affidavit supporting the warrant application was “so\\n\\nlacking in indicia of probable cause as to render official belief in its existence\\n\\n\\n 25\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 26 of 56\\n\\n\\nentirely unreasonable”; or (4) where the warrant was “so facially deficient” that\\n\\nofficers couldn’t have reasonably presumed it to be valid. 468 U.S. at 923.\\n\\n Here, Taylor and Smith contend—and the dissent agrees—that the\\n\\nmagistrate was, within the meaning of Leon, “misled by information” in the\\n\\napplication that the FBI officers knew, or should have known, to be false. The face\\n\\nof the application, they say, prominently represented that the “property to be\\n\\nsearched” was “located in the Eastern District of Virginia” and, more specifically,\\n\\nasserted (in the incorporated Attachment A) that the Playpen server was “located at\\n\\na government facility in the Eastern District of Virginia.” Br. of Appellant Taylor\\n\\nat 42; Br. of Appellant Smith at 41. It wasn’t until page 29 of Agent Macfarlane’s\\n\\n31-page affidavit, Taylor and Smith say, that the application finally acknowledged\\n\\nthat the NIT would search computers “wherever located.” Br. of Appellant Taylor\\n\\nat 42; Br. of Appellant Smith at 41. This approach, they contend, shows that the\\n\\nFBI intentionally misled the magistrate judge and belies any claim to good-faith\\n\\nreliance.\\n\\n In responding that the good-faith exception should apply, the government\\n\\nbegins with the contention that there is no deterrent benefit to exclusion here\\n\\nbecause Rule 41 was recently amended to add a new subsection to cover remote-\\n\\naccess warrants to search electronic storage both within and outside of a magistrate\\n\\n\\n\\n\\n 26\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 27 of 56\\n\\n\\njudge’s district—i.e., precisely the sort of search at issue in this case. 14 But that\\n\\nargument cuts both ways. On the one hand, it indicates that we needn’t necessarily\\n\\ndeter this particular type of search on a going-forward basis. On the other, the\\n\\nrecent amendment of Rule 41 to allow remote-access search warrants underscores\\n\\nthat Rule 41(b) did not permit these warrants at the time the FBI deployed the NIT.\\n\\n Even so, we find no indication that the FBI officers sought to deceive the\\n\\nmagistrate judge or otherwise acted culpably or in a way that necessitates\\n\\ndeterrence—and certainly no indication of the sort of “deliberate[], reckless[], or .\\n\\n. . gross[ly] negligen[t]” conduct that the Supreme Court has recently highlighted\\n\\nas the focus of the exclusionary-rule/good-faith inquiry. Davis, 564 U.S. at 240;\\n\\nsee also Herring, 555 U.S. at 144; Krull, 480 U.S. at 352–53. While the NIT-\\n\\nwarrant application was perhaps not a model of clarity, it seems clear to us that the\\n\\nofficers did the best they could with what they had—a general application form\\n\\nthat was perhaps ill-suited to the complex new technology at issue. 15 It is true, as\\n\\n\\n14Rule 41(b)(6) now states in relevant part: “[A] magistrate judge with authority in any district\\nwhere activities related to a crime may have occurred has authority to issue a warrant to use\\nremote access to search electronic storage media and to seize or copy electronically stored\\ninformation located within or outside that district if . . . the district where the media or\\ninformation is located has been concealed through technological means.”\\n15In concluding that the officers intended to “hoodwink” the magistrate judge, the dissent relies\\nheavily on DOJ’s proposals to amend Rule 41 to better address “remote searches for ‘crimes\\ninvolving Internet anonymizing technology.’” Dissenting Op. at 37–38, 46 (quoting Letter from\\nMythili Raman, Acting Assistant Att’y Gen., to Hon. Reena Raggi, Chair, Advisory Comm. on\\nthe Crim. Rules (Sept. 18, 2013)). Even setting aside the dubious proposition that knowledge of\\ncommunications between the “highest ranking officials in the Criminal Division” and Federal\\n\\n 27\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 28 of 56\\n\\n\\nTaylor and Smith emphasize, that the face of the pre-printed warrant application\\n\\nstated that “the property to be searched” was “located in the Eastern District of\\n\\nVirginia.” It is also true that Attachment A, which described the target property,\\n\\nreported that the Playpen server was “located at a government facility in the\\n\\nEastern District of Virginia.” That being said, there were indications that the FBI\\n\\nwas seeking more broad-ranging search authority. As already noted, the case\\n\\ncaption referred generally to “COMPUTERS THAT ACCESS” Playpen.\\n\\nSomewhat more clearly, Attachment A explained that the NIT would be “deployed\\n\\non” the Playpen-operating server located in the Eastern District of Virginia as a\\n\\nmeans of “obtaining information” from “activating computers,” defined as\\n\\ncomputers “of any user or administrator who logs into” the Playpen site. Finally,\\n\\nand most importantly—if a bit more obscurely than might have been ideal—Agent\\n\\nMacfarlane’s affidavit stated that “the NIT may cause an activating computer—\\n\\nwherever located—to send” identifying information to the FBI.\\n\\n So, was the warrant application here perfect? Not close. But does it\\n\\nevidence “chicanery,” “duplicity,” and “gamesmanship”? See Dissenting Op. at\\n\\n\\nRules Advisory Committee Chairs can be imputed downstream to line-level law-enforcement\\nofficers, see Dissenting Op. at 38–39, these communications in no way demonstrate that the\\nwarrant application here was made in bad faith. We see no benefit to deterring officers from\\nattempting to describe cutting-edge countermeasures using the forms and resources at their\\ndisposal while department heads simultaneously seek to amend the rules to better address\\nadvancing technology. Cf. Eldred, 2019 WL 3540415, at *7; McLamb, 880 F.3d at 691. The\\ndissent’s argument to the contrary is based entirely on speculation about what different\\ngovernment actors could have known.\\n\\n 28\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 29 of 56\\n\\n\\n16, 25. It doesn’t. We conclude that, in their totality, the application and affidavit\\n\\nsufficiently disclosed the bounds of the intended search. In light of the square-\\n\\npeg/round-hole issue that they faced, the officers did what we would hope and\\n\\nexpect—they fully disclosed the mechanics of the intended search, left the\\n\\nconstitutional call to the magistrate judge, and acted in reasonable reliance on the\\n\\nresulting warrant. 16 As already explained, the “exclusionary rule was crafted to\\n\\ncurb police rather than judicial misconduct.” Herring, 555 U.S. at 142. Because\\n\\nwe don’t find the officers’ behavior here culpable and see no deterrent value in\\n\\nsuppressing the evidence found on Taylor’s and Smith’s computers, we find that\\n\\nthe good-faith exception to the exclusionary rule applies in this case.\\n\\n AFFIRMED.\\n\\n\\n\\n\\n16 To the extent that the dissent suggests that officers seeking a search warrant have an\\naffirmative obligation to “flag” potential legal issues in their application, we must respectfully\\ndisagree. See, e.g., Dissenting Op. at 39–40 (stating that the officers here “should have known . .\\n. that the magistrate’s jurisdiction to issue the warrant was in doubt” and that they “had an\\nobligation to flag [this] for the magistrate”). Law-enforcement officers have a duty to lay out\\nfacts—including jurisdictional facts—for reviewing courts, not to anticipate and articulate\\npossible legal hurdles. The warrant application here, particularly when read in conjunction with\\nAgent Macfarlane’s detailed 30-plus-page affidavit, adequately—if imperfectly—lays out the\\nfacts. See, e.g., Levin, 874 F.3d at 323 (determining that there was “no benefit in deterring” the\\ngovernment from “turn[ing] to the courts for guidance” when faced with a novel legal question\\nsuch as whether the NIT warrant could properly issue).\\n\\n\\n 29\\n\\n Case: 17-14915 Date Filed: 08/28/2019 Page: 30 of 56\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4155,"cells":{"id":{"kind":"number","value":4656718,"string":"4,656,718"},"name":{"kind":"string","value":"In RE: Arlene ROTH, Debtor. Arlene Roth, Plaintiff - Appellant, v. Nationstar Mortgage, LLC, Defendant - Appellee."},"name_abbreviation":{"kind":"string","value":"Roth v. Nationstar Mortg., LLC (In Re Roth)"},"decision_date":{"kind":"string","value":"2019-08-28"},"docket_number":{"kind":"string","value":"17-11444"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Wilson, Branch, Anderson\", \"parties\": \"\", \"opinions\": [{\"author\": \"BRANCH, Circuit Judge:\", \"type\": \"010combined\", \"text\": \"Case: 17-11444 Date Filed: 08/28/2019 Page: 1 of 15\\n\\n\\n [PUBLISH]\\n\\n\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No: 17-11444\\n ________________________\\n\\n D.C. Docket Nos. 2:16-cv-00510-SPC; 9:10-bkc-30383-FMD\\n\\n\\nIN RE: ARLENE ROTH,\\n\\n Debtor.\\n\\n ________________________\\n\\n\\nARLENE ROTH,\\n Plaintiff - Appellant,\\n\\n versus\\n\\n\\nNATIONSTAR MORTGAGE, LLC,\\n\\n Defendant - Appellee.\\n\\n ________________________\\n\\n Appeal from the United States District Court\\n for the Middle District of Florida\\n ________________________\\n\\n (August 28, 2019)\\n\\nBefore WILSON, BRANCH, and ANDERSON, Circuit Judges.\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 2 of 15\\n\\n\\nBRANCH, Circuit Judge:\\n\\n Arlene Roth appeals the bankruptcy court’s denial of her second motion for\\n\\nsanctions against Nationstar Mortgage, LLC (“Nationstar”). The basis of the\\n\\nmotion was Nationstar’s conduct of sending a monthly statement to Roth, which\\n\\nshe alleges sought to collect discharged mortgage debt in violation of the\\n\\nbankruptcy code’s prohibition on collection of such debt, 11 U.S.C. § 524. She\\n\\nraises two issues on appeal: (1) whether the bankruptcy court improperly made\\n\\nfactual determinations without the benefit of an evidentiary hearing; and (2)\\n\\nwhether Nationstar’s mortgage statement was an improper attempt to collect a debt\\n\\nin violation of 11 U.S.C. § 524 and justifies sanctions against Nationstar. For the\\n\\nfollowing reasons, we affirm the bankruptcy court’s decision.\\n\\n I.\\n\\n On December 22, 2010, Roth filed a voluntary petition for bankruptcy under\\n\\nChapter 13 of the bankruptcy code. On her bankruptcy schedule, Roth listed a\\n\\nmortgage on non-homestead property, which is at issue here. On this schedule, she\\n\\nindicated she would surrender the property. The Chapter 13 plan provided that\\n\\n“[s]ecured creditors, whether or not dealt with under the Plan, shall retain the liens\\n\\nsecuring such claims.” On October 3, 2011, the bankruptcy court entered an order\\n\\nconfirming her Chapter 13 bankruptcy plan. Subsequently, the mortgage at issue\\n\\nhere (which was subject to the Chapter 13 plan) was transferred to Nationstar. She\\n\\n\\n\\n 2\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 3 of 15\\n\\n\\ncompleted her payments under her Chapter 13 plan, and the bankruptcy court\\n\\ndischarged her debt on June 27, 2014. Nationstar was notified of the discharge.\\n\\n In the discharge order, the bankruptcy court ordered that “the discharge\\n\\nprohibits any attempt to collect from the debtor a debt that has been discharged.\\n\\nFor example, a creditor is not permitted to contact a debtor by mail, phone, or\\n\\notherwise, to file or continue a lawsuit, to attach wages or other property, or to take\\n\\nany other action to collect a discharged debt from the debtor.” The order, however,\\n\\nalso stated that “a creditor may have the right to enforce a valid lien, such as a\\n\\nmortgage or security interest, against the debtor’s property after the bankruptcy, if\\n\\nthat lien was not avoided or eliminated in the bankruptcy case. Also, a debtor may\\n\\nvoluntarily pay any debt that has been discharged.” But even after the discharge of\\n\\nthe mortgage debt, Nationstar did not foreclose on the property; thus, if Roth\\n\\nvoluntarily paid the amount due, she could have retained the property.\\n\\n About four months after the discharge of Roth’s debt, Nationstar started\\n\\nsending Roth monthly statements related to her mortgage. The statements included\\n\\na disclaimer that they were not debt collection, but also included an amount due,\\n\\ndue date, and instructions on how to send payment back to Nationstar. Roth had\\n\\nher attorney send a cease and desist letter, but the statements kept coming. She\\n\\nthen filed her first motion for sanctions in bankruptcy court, alleging that the\\n\\nstatements were improper debt collections in violation of 11 U.S.C. § 524, as well\\n\\n\\n\\n 3\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 4 of 15\\n\\n\\nas a separate civil action claiming violations of the Fair Debt Collection Practices\\n\\nAct (“FDCPA”), 15 U.S.C. § 1692e et seq, and the Florida Consumer Collection\\n\\nPractices Act (FCCPA). Arlene Roth v. Nationstar Mortgage LLC, No. 2:15-cv-\\n\\n508-FtM-38CM (M.D. Fla.). The parties resolved the first sanctions motion and\\n\\nfirst FDCPA action in a settlement.\\n\\n But communication from Nationstar continued in the form of a November\\n\\n18, 2015 “Informational Statement,” which again contained an amount due, due\\n\\ndate, and instructions for how to pay Nationstar, as well as a lengthy disclaimer.\\n\\nThe full text of the disclaimer reads as follows:\\n\\n This statement is sent for informational purposes only and is not\\n intended as an attempt to collect, assess, or recover a discharged debt\\n from you, or as a demand for payment from any individual protected\\n by the United States Bankruptcy Code. If this account is active or has\\n been discharged in a bankruptcy proceeding, be advised this\\n communication is for informational purposes only and is not an attempt\\n to collect a debt. Please note, however Nationstar reserves the right to\\n exercise its legal rights, including but not limited to foreclosure of its\\n lien interest, only against the property securing the original obligation.\\n Consequently, on December 14, 2015, Roth filed a second lawsuit against\\n\\nNationstar alleging the Informational Statement was an improper debt collection\\n\\nunder the FDCPA. Roth v. Nationstar Mortg., LLC, No. 2:15-cv-783-FtM-\\n\\n29MRM, 2016 WL 3570991 at *1 (M.D. Fla. July 1, 2016). Nationstar filed a\\n\\nmotion to dismiss that case, arguing that the Informational Statement was not sent\\n\\nfor debt collection purposes as a matter of law, but the district court denied the\\n\\n\\n\\n 4\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 5 of 15\\n\\n\\nmotion after applying the FDCPA’s “least sophisticated consumer” standard. Id. at\\n\\n*5. On May 30, 2017, the parties notified the district court that they had reached a\\n\\nsettlement in that case. Notice of Settlement, Roth v. Nationstar Mortg., LLC, No.\\n\\n2:15-cv-783-FtM-29MRM (M.D. Fla. May 30, 2017).\\n\\n Based on the same Information Statement, on December 14, 2015—the\\n\\nsame day that she filed her second FDCPA case—Roth filed a second motion for\\n\\nsanctions in her bankruptcy case, which is at issue here. Roth alleged that\\n\\nNationstar’s communication was an attempt to collect a discharged debt in\\n\\nviolation of the discharge order. Nationstar claimed that the statement was only\\n\\ninformational and not a violation of § 524. After briefing, the bankruptcy court\\n\\ndenied Roth’s motion for sanctions at a hearing, and later entered an opinion and\\n\\nan order finding that the “informational statement” was not a debt collection\\n\\nattempt, and therefore was not in violation of the § 524 injunction. Roth appealed\\n\\nthe denial of this second sanctions motion to the district court, which affirmed the\\n\\nbankruptcy court’s opinion and rejected Roth’s request to apply the FDCPA’s\\n\\n“least sophisticated consumer” standard. In re Roth, 568 B.R. 139, 145 (M.D. Fla.\\n\\n2017). This appeal followed.\\n\\n II.\\n\\n In bankruptcy cases, this Court “sits as a second court of review and thus\\n\\nexamines independently the factual and legal determinations of the bankruptcy\\n\\n\\n\\n 5\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 6 of 15\\n\\n\\ncourt and employs the same standards of review as the district court.” In re Ocean\\n\\nWarrior, Inc., 835 F.3d 1310, 1315 (11th Cir. 2016) (quoting In re Fisher Island\\n\\nInvs., Inc., 778 F.3d 1172, 1189 (11th Cir. 2015)); see also In re Gonzalez, 832\\n\\nF.3d 1251, 1253 (11th Cir. 2016), cert. denied sub nom. Fla. Dep’t of Revenue v.\\n\\nGonzalez, 137 S. Ct. 2293 (2017). When the district court affirms the bankruptcy\\n\\ncourt, we review the bankruptcy court’s decision, reviewing its “factual findings\\n\\nfor clear error and its legal conclusions de novo.” In re Ocean Warrior, Inc., 835\\n\\nF.3d at 1315 (internal quotation marks and citations omitted). We review the\\n\\nbankruptcy court’s denial for a motion for sanctions for abuse of discretion. In re\\n\\nDiaz, 647 F.3d 1073, 1082 (11th Cir. 2011).\\n\\n III.\\n\\n Roth challenges (1) the bankruptcy court’s substantive decision to deny her\\n\\nmotion for sanctions, and (2) the bankruptcy court’s decision to dispose of her\\n\\nmotion for sanctions without a separate evidentiary hearing. We first articulate the\\n\\nlegal standard Roth must meet to establish a violation of 11 U.S.C. § 524, and\\n\\napply that standard to review the bankruptcy court’s decision that Nationstar’s\\n\\ncommunications were not a sanctionable violation of § 524. We then determine\\n\\nwhether Roth was entitled to an evidentiary hearing in the bankruptcy court on that\\n\\nsubstantive question.\\n\\n Section 524(a)(2) of the bankruptcy code provides that a discharge of debt in\\n\\n\\n\\n 6\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 7 of 15\\n\\n\\na bankruptcy proceeding “operates as an injunction against the commencement or\\n\\ncontinuation of . . . an act . . . to collect . . . any such [discharged] debt.” 11 U.S.C.\\n\\n§ 524(a)(2).1 This injunction is enforced through section 105, whereby the\\n\\nbankruptcy court “may issue any order, process, or judgment that is necessary or\\n\\nappropriate to carry out the provisions of this title.” 11 U.S.C. § 105(a). Section\\n\\n105 “grants statutory contempt powers in the bankruptcy context.” In re Hardy, 97\\n\\nF.3d 1384, 1389–90 (11th Cir. 1996) (“Therefore, § 105(a) grants courts\\n\\nindependent statutory powers to award monetary and other forms of relief for\\n\\n[violations] to the extent such awards are ‘necessary [or] appropriate’ to carry out\\n\\nthe provisions of the Bankruptcy Code.”) (internal quotation marks and citations\\n\\nomitted). Together, sections 524(a)(2) and 105(a) “authorize a court to impose civil\\n\\n\\n\\n\\n1\\n 11 U.S.C. § 524(a) A discharge in a case under this title--\\n (1) voids any judgment at any time obtained, to the extent that such judgment is a\\n determination of the personal liability of the debtor with respect to any debt discharged\\n under section 727, 944, 1141, 1228, or 1328 of this title, whether or not discharge of such\\n debt is waived;\\n (2) operates as an injunction against the commencement or continuation of an action, the\\n employment of process, or an act, to collect, recover or offset any such debt as a personal\\n liability of the debtor, whether or not discharge of such debt is waived; and\\n (3) operates as an injunction against the commencement or continuation of an action, the\\n employment of process, or an act, to collect or recover from, or offset against, property of\\n the debtor of the kind specified in section 541(a)(2) of this title that is acquired after the\\n commencement of the case, on account of any allowable community claim, except a\\n community claim that is excepted from discharge under section 523, 1228(a)(1), or\\n 1328(a)(1), or that would be so excepted, determined in accordance with the provisions\\n of sections 523(c) and 523(d) of this title, in a case concerning the debtor’s spouse\\n commenced on the date of the filing of the petition in the case concerning the debtor,\\n whether or not discharge of the debt based on such community claim is waived.\\n\\n\\n\\n 7\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 8 of 15\\n\\n\\ncontempt sanctions [for attempting to collect a discharged debt] when there is no\\n\\nobjectively reasonable basis for concluding that the creditor’s conduct might be\\n\\nlawful under the discharge order.” Taggart v. Lorenzen, 139 S. Ct. 1795, 1801\\n\\n(2019).2 In this way, “a court may hold a creditor in civil contempt for violating a\\n\\ndischarge order if there is no fair ground of doubt as to whether the order barred\\n\\nthe creditor’s conduct.” Id. at 1799. 3\\n\\n Accordingly, we first determine whether a communication is a prohibited\\n\\ndebt collection under section 524 by looking to “whether the objective effect of the\\n\\ncreditor’s action is to pressure a debtor to repay a discharged debt.” In re McLean,\\n\\n794 F.3d 1313, 1322 (11th Cir. 2015). If so, we then evaluate whether that\\n\\nviolation of the discharge injunction is sanctionable under section 105, by\\n\\n\\n2\\n After we held oral argument in this case, the Supreme Court clarified in Taggart when\\ncontempt sanctions under § 105 are appropriate for violations of § 524.\\n3\\n In Taggart, the Supreme Court explicitly disavowed our § 105 analysis from In re Hardy in\\nwhich we held that the test “applicable to determining willfulness for violations of the discharge\\ninjunction of § 524” is if the defendant “(1) knew that the [injunction] was invoked and (2)\\nintended the actions which violated the [injunction]. In re Hardy, 97 F.3d 1384, 1390 (11th Cir.\\n1996) (quotations omitted). The Supreme Court reasoned that\\n\\n [b]ecause most creditors are aware of discharge orders and intend the actions they\\n take to collect a debt, this standard would operate much like a strict-liability\\n standard. It would authorize civil contempt sanctions for a violation of a discharge\\n order regardless of the creditor’s subjective beliefs about the scope of the discharge\\n order, and regardless of whether there was a reasonable basis for concluding that\\n the creditor’s conduct did not violate the order.\\n\\nTaggart, 139 S. Ct. at 1803. In place of the “strict-liability standard,” the Court adopted the “fair\\nground of doubt standard.” Id. at 1802.\\n\\n\\n\\n 8\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 9 of 15\\n\\n\\ndetermining if “there is no fair ground of doubt as to whether the order barred the\\n\\ncreditor’s conduct.” Taggart, 139 S. Ct. at 1799.\\n\\n We now turn to the substantive questions presented by this appeal: whether\\n\\nthe Informational Statement sent by Nationstar was an improper attempt at debt\\n\\ncollection in violation of 11 U.S.C. § 524(a)(2), and if so, whether it was a\\n\\nsanctionable under 11 U.S.C. § 105.\\n\\n A. 11 U.S.C. § 524(a)(2)\\n Roth has not met her burden of showing that the Informational Statement\\n\\nwas an unlawful debt collection in violation of § 524; there are several bases for\\n\\nconcluding that “the objective effect” of the Informational Statement was not “to\\n\\npressure [Roth] to repay a discharged debt.” In re McLean, 794 F.3d at 1322. As an\\n\\ninitial matter, the disclaimer is printed in bold on the first page of the statement. It\\n\\ndeclares that it is “for informational purposes only and is not intended as an\\n\\nattempt to collect, assess, or recover a discharged debt from you, or as a demand\\n\\nfor payment from any individual protected by the United States Bankruptcy Code.”\\n\\nIn case the first sentence is not clear enough, the disclaimer then repeats that if the\\n\\naccount “has been discharged in a bankruptcy proceeding” then it is “for\\n\\ninformational purposes only,” and “is not an attempt to collect a debt.” Further, the\\n\\nincluded payment coupon is marked in large lettering as “voluntary.”\\n\\n The fact that the statement includes an “amount due,” “due date,” and\\n\\n\\n\\n 9\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 10 of 15\\n\\n\\nstatements about the negative escrow balance does not diminish the effect of the\\n\\nprominent, clear, and broadly worded disclaimer. Notably, section 524 allows for a\\n\\ndebtor to pay back a discharged debt voluntarily. Under 11 U.S.C. § 524(f),\\n\\n“[n]othing contained in subsection (c) or (d) of this section prevents a debtor from\\n\\nvoluntarily repaying any debt.” Roth had the option to repay the debt (the “amount\\n\\ndue”) and thereby regain the property because Nationstar had not completed a\\n\\nforeclosure on the property. If we accepted Roth’s argument that the Informational\\n\\nStatement is unlawful debt collection under § 524, there would be little daylight\\n\\nbetween (1) a legitimate attempt by Nationstar to inform Roth how she could\\n\\nregain the property and (2) an unlawful attempt at debt collection in violation of\\n\\nthe § 524 injunction. Instead, the statutory scheme clearly allows for Nationstar to\\n\\nsend potentially helpful informational statements to Roth without simultaneously\\n\\ncasting those statements as debt collection. In light of these facts, the Informational\\n\\nStatement sent by Nationstar was not designed to have the “objective effect” of\\n\\n“pressur[ing] the debtor to pay a discharged debt,” In re McLean, 794 F.3d at\\n\\n1322, and so we conclude that it was not an unlawful attempt at debt collection in\\n\\nviolation of § 524.\\n\\n Roth asks us to use a different standard to determine if Nationstar has\\n\\ncommitted a § 524 violation. In Roth’s separate lawsuit, which was brought under\\n\\n\\n\\n\\n 10\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 11 of 15\\n\\n\\nthe FDCPA rather than the bankruptcy code, 4 the district court applied the\\n\\nappropriate FDCPA test using the “least sophisticated consumer” standard. Roth v.\\n\\nNationstar Mortg. LLC, No. 2:15-cv-00783-FtM-29MRM, 2016 WL 3570991, at\\n\\n*3 (M.D. Fla., July 1, 2016). 5 At the motion to dismiss stage, the district court in\\n\\nthe FDCPA case summarized its findings with respect to the Informational\\n\\nStatement:\\n\\n It is true that the Informational Statement does not expressly state that\\n it is “a communication sent for the purpose of collecting a debt.” But,\\n as just discussed, this absence is not dispositive. The Statement lists the\\n total amount due, contains a payment due date, states that a late fee will\\n be charged for an untimely payment, gives six possible payment\\n methods, and separates out from the total amount due the amount of\\n fees and charges previously assessed. That is not all. The statement\\n contains an “Important Messages” box advising [Appellant] that her\\n “escrow account has a negative balance,” and expressly\\n “recommend[ing she] make additional payments” to avoid “an increase\\n in [her] monthly escrow payment.” There is also a detachable “payment\\n coupon,” which states the total amount due and recalculates the amount\\n due for a late payment.\\n\\nRoth, 2016 WL 3570991, at *3. Based on these observations, the court determined\\n\\n\\n\\n4\\n The FDCPA creates a civil cause of action based on certain prohibited debt collection methods,\\nspecifically a “false, deceptive, or misleading representation or means in connection with the\\ncollection of any debt” or an “unfair or unconscionable means” of debt collection. 15 U.S.C.\\n§ 1692e–f.\\n5\\n The FDCPA standard for determining if a communication is a debt collection is whether the\\nstatement would “mislead the least sophisticated consumer regarding the nature of her rights.”\\nCaceres v. McCalla Raymer, LLC, 755 F.3d 1299, 1303 (11th Cir. 2014). “The least\\nsophisticated consumer’ is presumed to have only a “rudimentary amount of information about\\nthe world,” but “it is assumed that the least-sophisticated consumer will be willing to read a\\ncollection notice with some care.” Holzman v. Malcolm S. Gerald & Assocs., Inc., 920 F.3d\\n1264, 1269 (11th Cir. 2019) (internal quotation marks, citations, and alterations omitted).\\n\\n\\n 11\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 12 of 15\\n\\n\\nthat Roth’s FDCPA complaint “plausibly alleges that the Informational Statement\\n\\nwas sent to induce payment on Plaintiff’s mortgage debt,” and was thus “related to\\n\\ndebt collection” under the FDCPA. Id. at *4.\\n\\n Roth argues that the “underlying questions [of § 524 and the FDCPA] are\\n\\ndesigned to protect the same vulnerable parties from the same improper conduct”\\n\\nand so the FDCPA and bankruptcy code should use the same standard. Roth is\\n\\nasking this Court to import the “least sophisticated consumer” standard from the\\n\\nFDCPA into the § 524 context.\\n\\n We decline to follow Roth’s request. 6 This Court has never incorporated the\\n\\n“least sophisticated consumer” test into our § 524 analysis. And for good reason—\\n\\nwhat counts as “debt collection” under one statutory scheme is not necessarily\\n\\n“debt collection” under the other. See Midland Funding, LLC v. Johnson, 137 S.\\n\\nCt. 1407, 1414 (2017) (“The [FDCPA] and the [Bankruptcy] Code have different\\n\\npurposes and structural features. The Act seeks to help consumers . . . . The\\n\\nBankruptcy Code, by way of contrast, creates and maintains what we have called\\n\\nthe delicate balance of a debtor’s protections and obligations.”) (citation and\\n\\ninternal quotations marks omitted). And the Supreme Court reiterated that “civil\\n\\n\\n\\n6\\n The district court in Roth’s FDCPA case was acting in response to Nationstar’s motion to\\ndismiss for failure to state a claim. Thus, the district court in that case did not make a finding on\\nthe merits, but rather it found only that the complaint “state[d] a claim to relief that is plausible\\non its face.” Bell Atl. Corp. v. Twombly, 550 U.S. 544, 547 (2007).\\n\\n\\n\\n 12\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 13 of 15\\n\\n\\ncontempt is a severe remedy,” so it follows that the burden to show contempt\\n\\nshould be a high one. Taggart, 139 S.Ct. at 1802 (citation and internal quotation\\n\\nmarks omitted). There is nothing in either statute that suggests we should evaluate\\n\\npotential § 524 violations using the FDCPA standard, and we find no other reason\\n\\nto do so.\\n\\n B. 11 U.S.C. § 105\\n\\n Because we find that Nationstar’s Informational Statement did not violate\\n\\n§ 524, we do not have to consider whether sanctions would be appropriate under\\n\\n§ 105. But even if we reached a different conclusion based on these facts and\\n\\nfound that the issue of the § 524 violation presented a close call, we would\\n\\nnonetheless find that sanctions under § 105 are unavailable under the Supreme\\n\\nCourt’s recent decision in Taggart, which established the “no fair ground of doubt”\\n\\nstandard. Taggart, 139 S. Ct. at 1799. The Taggart standard is a rigorous one: in\\n\\norder to find that sanctions are appropriate here, we would have to hold that “there\\n\\nis no objectively reasonable basis for concluding that [Nationstar’s] conduct might\\n\\nbe lawful.” Id. With more than a “fair ground of doubt,” Taggart, 139 S. Ct. at\\n\\n1799, as to whether the discharge order barred Nationstar’s conduct, sanctions\\n\\nwould be inappropriate.\\n\\n As we have determined that Nationstar’s Informational Statement did not\\n\\nviolate § 524, the “no fair ground of doubt” standard for § 105 is necessarily not\\n\\n\\n\\n 13\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 14 of 15\\n\\n\\nsatisfied. Accordingly, the bankruptcy court’s denial of Roth’s motion for\\n\\nsanctions was not an abuse of discretion.\\n\\n C. Evidentiary Hearing\\n\\n Finally, Roth argues that the bankruptcy court should have held an\\n\\nevidentiary hearing—despite the fact that she did not ask for one—before it made\\n\\nits findings concerning the Informational Statement.7 The crux of Roth’s argument\\n\\nis that the bankruptcy court concluded that Roth “can hardly have thought that\\n\\nNationstar was trying to collect the discharged debt,” but a determination of her\\n\\nsubjective belief required an evidentiary hearing.\\n\\n In civil contempt proceedings, “when there are no disputed factual matters\\n\\nthat require an evidentiary hearing, the court might properly dispense with the\\n\\nhearing prior to finding the defendant in contempt and sanctioning him.” In re\\n\\nMcLean, 794 F.3d at 1324 (quoting Mercer v. Mitchell, 908 F.2d 763, 769 n.11\\n\\n(11th Cir. 1990)).\\n\\n Neither party in this case requested an evidentiary hearing, and the\\n\\nbankruptcy court was not required to hold an evidentiary hearing sua sponte to\\n\\ninquire about Roth’s subjective beliefs. The bankruptcy court had to look only to\\n\\n\\n7\\n The FDCPA creates a civil cause of action based on certain prohibited debt collection methods,\\nspecifically a “false, deceptive, or misleading representation or means in connection with the\\ncollection of any debt” or an “unfair or unconscionable means” of debt collection. 15 U.S.C.\\n§ 1692e–f.005) (“As we repeatedly have admonished, arguments raised for the first time in a\\nreply brief are not properly before a reviewing court.”) (citation and internal quotation marks\\nomitted).\\n\\n\\n 14\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 15 of 15\\n\\n\\nthe Informational Statement itself to determine “whether the objective effect of [the\\n\\nInformational Statement] is to pressure a debtor to repay a discharged debt,”\\n\\nMcLean, 794 F.3d at 1322 (emphasis added). It did so at length, making numerous\\n\\nfindings about the document and determining that the statement did not constitute\\n\\nan unlawful attempt at debt collection under § 524. Roth’s subjective belief about\\n\\nthe Informational Statement is irrelevant to the analysis, so it was not a “disputed\\n\\nfactual matter[] that require[d] an evidentiary hearing.” In re McLean, 794 F.3d at\\n\\n1324. The bankruptcy court’s statement that Roth “can hardly have thought that\\n\\nNationstar was trying to collect the discharged debt” is simply the court’s\\n\\nconclusion that the communications were not an attempt at debt collection.\\n\\n Accordingly, the bankruptcy court did not err by rendering its decision\\n\\nwithout a hearing.\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 15\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4433971/\", \"author_raw\": \"BRANCH, Circuit Judge:\"}]}"},"judge1":{"kind":"string","value":"WILSON"},"judge2":{"kind":"string","value":"BRANCH"},"judge3":{"kind":"string","value":"ANDERSON"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4656718/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: In RE: Arlene ROTH, Debtor. Arlene Roth, Plaintiff - Appellant, v. Nationstar Mortgage, LLC, Defendant - Appellee.\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-28\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nCase: 17-11444 Date Filed: 08/28/2019 Page: 1 of 15\n\n\n [PUBLISH]\n\n\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n ________________________\n\n No: 17-11444\n ________________________\n\n D.C. Docket Nos. 2:16-cv-00510-SPC; 9:10-bkc-30383-FMD\n\n\nIN RE: ARLENE ROTH,\n\n Debtor.\n\n ________________________\n\n\nARLENE ROTH,\n Plaintiff - Appellant,\n\n versus\n\n\nNATIONSTAR MORTGAGE, LLC,\n\n Defendant - Appellee.\n\n ________________________\n\n Appeal from the United States District Court\n for the Middle District of Florida\n ________________________\n\n (August 28, 2019)\n\nBefore WILSON, BRANCH, and ANDERSON, Circuit Judges.\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 2 of 15\n\n\nBRANCH, Circuit Judge:\n\n Arlene Roth appeals the bankruptcy court’s denial of her second motion for\n\nsanctions against Nationstar Mortgage, LLC (“Nationstar”). The basis of the\n\nmotion was Nationstar’s conduct of sending a monthly statement to Roth, which\n\nshe alleges sought to collect discharged mortgage debt in violation of the\n\nbankruptcy code’s prohibition on collection of such debt, 11 U.S.C. § 524. She\n\nraises two issues on appeal: (1) whether the bankruptcy court improperly made\n\nfactual determinations without the benefit of an evidentiary hearing; and (2)\n\nwhether Nationstar’s mortgage statement was an improper attempt to collect a debt\n\nin violation of 11 U.S.C. § 524 and justifies sanctions against Nationstar. For the\n\nfollowing reasons, we affirm the bankruptcy court’s decision.\n\n I.\n\n On December 22, 2010, Roth filed a voluntary petition for bankruptcy under\n\nChapter 13 of the bankruptcy code. On her bankruptcy schedule, Roth listed a\n\nmortgage on non-homestead property, which is at issue here. On this schedule, she\n\nindicated she would surrender the property. The Chapter 13 plan provided that\n\n“[s]ecured creditors, whether or not dealt with under the Plan, shall retain the liens\n\nsecuring such claims.” On October 3, 2011, the bankruptcy court entered an order\n\nconfirming her Chapter 13 bankruptcy plan. Subsequently, the mortgage at issue\n\nhere (which was subject to the Chapter 13 plan) was transferred to Nationstar. She\n\n\n\n 2\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 3 of 15\n\n\ncompleted her payments under her Chapter 13 plan, and the bankruptcy court\n\ndischarged her debt on June 27, 2014. Nationstar was notified of the discharge.\n\n In the discharge order, the bankruptcy court ordered that “the discharge\n\nprohibits any attempt to collect from the debtor a debt that has been discharged.\n\nFor example, a creditor is not permitted to contact a debtor by mail, phone, or\n\notherwise, to file or continue a lawsuit, to attach wages or other property, or to take\n\nany other action to collect a discharged debt from the debtor.” The order, however,\n\nalso stated that “a creditor may have the right to enforce a valid lien, such as a\n\nmortgage or security interest, against the debtor’s property after the bankruptcy, if\n\nthat lien was not avoided or eliminated in the bankruptcy case. Also, a debtor may\n\nvoluntarily pay any debt that has been discharged.” But even after the discharge of\n\nthe mortgage debt, Nationstar did not foreclose on the property; thus, if Roth\n\nvoluntarily paid the amount due, she could have retained the property.\n\n About four months after the discharge of Roth’s debt, Nationstar started\n\nsending Roth monthly statements related to her mortgage. The statements included\n\na disclaimer that they were not debt collection, but also included an amount due,\n\ndue date, and instructions on how to send payment back to Nationstar. Roth had\n\nher attorney send a cease and desist letter, but the statements kept coming. She\n\nthen filed her first motion for sanctions in bankruptcy court, alleging that the\n\nstatements were improper debt collections in violation of 11 U.S.C. § 524, as well\n\n\n\n 3\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 4 of 15\n\n\nas a separate civil action claiming violations of the Fair Debt Collection Practices\n\nAct (“FDCPA”), 15 U.S.C. § 1692e et seq, and the Florida Consumer Collection\n\nPractices Act (FCCPA). Arlene Roth v. Nationstar Mortgage LLC, No. 2:15-cv-\n\n508-FtM-38CM (M.D. Fla.). The parties resolved the first sanctions motion and\n\nfirst FDCPA action in a settlement.\n\n But communication from Nationstar continued in the form of a November\n\n18, 2015 “Informational Statement,” which again contained an amount due, due\n\ndate, and instructions for how to pay Nationstar, as well as a lengthy disclaimer.\n\nThe full text of the disclaimer reads as follows:\n\n This statement is sent for informational purposes only and is not\n intended as an attempt to collect, assess, or recover a discharged debt\n from you, or as a demand for payment from any individual protected\n by the United States Bankruptcy Code. If this account is active or has\n been discharged in a bankruptcy proceeding, be advised this\n communication is for informational purposes only and is not an attempt\n to collect a debt. Please note, however Nationstar reserves the right to\n exercise its legal rights, including but not limited to foreclosure of its\n lien interest, only against the property securing the original obligation.\n Consequently, on December 14, 2015, Roth filed a second lawsuit against\n\nNationstar alleging the Informational Statement was an improper debt collection\n\nunder the FDCPA. Roth v. Nationstar Mortg., LLC, No. 2:15-cv-783-FtM-\n\n29MRM, 2016 WL 3570991 at *1 (M.D. Fla. July 1, 2016). Nationstar filed a\n\nmotion to dismiss that case, arguing that the Informational Statement was not sent\n\nfor debt collection purposes as a matter of law, but the district court denied the\n\n\n\n 4\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 5 of 15\n\n\nmotion after applying the FDCPA’s “least sophisticated consumer” standard. Id. at\n\n*5. On May 30, 2017, the parties notified the district court that they had reached a\n\nsettlement in that case. Notice of Settlement, Roth v. Nationstar Mortg., LLC, No.\n\n2:15-cv-783-FtM-29MRM (M.D. Fla. May 30, 2017).\n\n Based on the same Information Statement, on December 14, 2015—the\n\nsame day that she filed her second FDCPA case—Roth filed a second motion for\n\nsanctions in her bankruptcy case, which is at issue here. Roth alleged that\n\nNationstar’s communication was an attempt to collect a discharged debt in\n\nviolation of the discharge order. Nationstar claimed that the statement was only\n\ninformational and not a violation of § 524. After briefing, the bankruptcy court\n\ndenied Roth’s motion for sanctions at a hearing, and later entered an opinion and\n\nan order finding that the “informational statement” was not a debt collection\n\nattempt, and therefore was not in violation of the § 524 injunction. Roth appealed\n\nthe denial of this second sanctions motion to the district court, which affirmed the\n\nbankruptcy court’s opinion and rejected Roth’s request to apply the FDCPA’s\n\n“least sophisticated consumer” standard. In re Roth, 568 B.R. 139, 145 (M.D. Fla.\n\n2017). This appeal followed.\n\n II.\n\n In bankruptcy cases, this Court “sits as a second court of review and thus\n\nexamines independently the factual and legal determinations of the bankruptcy\n\n\n\n 5\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 6 of 15\n\n\ncourt and employs the same standards of review as the district court.” In re Ocean\n\nWarrior, Inc., 835 F.3d 1310, 1315 (11th Cir. 2016) (quoting In re Fisher Island\n\nInvs., Inc., 778 F.3d 1172, 1189 (11th Cir. 2015)); see also In re Gonzalez, 832\n\nF.3d 1251, 1253 (11th Cir. 2016), cert. denied sub nom. Fla. Dep’t of Revenue v.\n\nGonzalez, 137 S. Ct. 2293 (2017). When the district court affirms the bankruptcy\n\ncourt, we review the bankruptcy court’s decision, reviewing its “factual findings\n\nfor clear error and its legal conclusions de novo.” In re Ocean Warrior, Inc., 835\n\nF.3d at 1315 (internal quotation marks and citations omitted). We review the\n\nbankruptcy court’s denial for a motion for sanctions for abuse of discretion. In re\n\nDiaz, 647 F.3d 1073, 1082 (11th Cir. 2011).\n\n III.\n\n Roth challenges (1) the bankruptcy court’s substantive decision to deny her\n\nmotion for sanctions, and (2) the bankruptcy court’s decision to dispose of her\n\nmotion for sanctions without a separate evidentiary hearing. We first articulate the\n\nlegal standard Roth must meet to establish a violation of 11 U.S.C. § 524, and\n\napply that standard to review the bankruptcy court’s decision that Nationstar’s\n\ncommunications were not a sanctionable violation of § 524. We then determine\n\nwhether Roth was entitled to an evidentiary hearing in the bankruptcy court on that\n\nsubstantive question.\n\n Section 524(a)(2) of the bankruptcy code provides that a discharge of debt in\n\n\n\n 6\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 7 of 15\n\n\na bankruptcy proceeding “operates as an injunction against the commencement or\n\ncontinuation of . . . an act . . . to collect . . . any such [discharged] debt.” 11 U.S.C.\n\n§ 524(a)(2).1 This injunction is enforced through section 105, whereby the\n\nbankruptcy court “may issue any order, process, or judgment that is necessary or\n\nappropriate to carry out the provisions of this title.” 11 U.S.C. § 105(a). Section\n\n105 “grants statutory contempt powers in the bankruptcy context.” In re Hardy, 97\n\nF.3d 1384, 1389–90 (11th Cir. 1996) (“Therefore, § 105(a) grants courts\n\nindependent statutory powers to award monetary and other forms of relief for\n\n[violations] to the extent such awards are ‘necessary [or] appropriate’ to carry out\n\nthe provisions of the Bankruptcy Code.”) (internal quotation marks and citations\n\nomitted). Together, sections 524(a)(2) and 105(a) “authorize a court to impose civil\n\n\n\n\n1\n 11 U.S.C. § 524(a) A discharge in a case under this title--\n (1) voids any judgment at any time obtained, to the extent that such judgment is a\n determination of the personal liability of the debtor with respect to any debt discharged\n under section 727, 944, 1141, 1228, or 1328 of this title, whether or not discharge of such\n debt is waived;\n (2) operates as an injunction against the commencement or continuation of an action, the\n employment of process, or an act, to collect, recover or offset any such debt as a personal\n liability of the debtor, whether or not discharge of such debt is waived; and\n (3) operates as an injunction against the commencement or continuation of an action, the\n employment of process, or an act, to collect or recover from, or offset against, property of\n the debtor of the kind specified in section 541(a)(2) of this title that is acquired after the\n commencement of the case, on account of any allowable community claim, except a\n community claim that is excepted from discharge under section 523, 1228(a)(1), or\n 1328(a)(1), or that would be so excepted, determined in accordance with the provisions\n of sections 523(c) and 523(d) of this title, in a case concerning the debtor’s spouse\n commenced on the date of the filing of the petition in the case concerning the debtor,\n whether or not discharge of the debt based on such community claim is waived.\n\n\n\n 7\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 8 of 15\n\n\ncontempt sanctions [for attempting to collect a discharged debt] when there is no\n\nobjectively reasonable basis for concluding that the creditor’s conduct might be\n\nlawful under the discharge order.” Taggart v. Lorenzen, 139 S. Ct. 1795, 1801\n\n(2019).2 In this way, “a court may hold a creditor in civil contempt for violating a\n\ndischarge order if there is no fair ground of doubt as to whether the order barred\n\nthe creditor’s conduct.” Id. at 1799. 3\n\n Accordingly, we first determine whether a communication is a prohibited\n\ndebt collection under section 524 by looking to “whether the objective effect of the\n\ncreditor’s action is to pressure a debtor to repay a discharged debt.” In re McLean,\n\n794 F.3d 1313, 1322 (11th Cir. 2015). If so, we then evaluate whether that\n\nviolation of the discharge injunction is sanctionable under section 105, by\n\n\n2\n After we held oral argument in this case, the Supreme Court clarified in Taggart when\ncontempt sanctions under § 105 are appropriate for violations of § 524.\n3\n In Taggart, the Supreme Court explicitly disavowed our § 105 analysis from In re Hardy in\nwhich we held that the test “applicable to determining willfulness for violations of the discharge\ninjunction of § 524” is if the defendant “(1) knew that the [injunction] was invoked and (2)\nintended the actions which violated the [injunction]. In re Hardy, 97 F.3d 1384, 1390 (11th Cir.\n1996) (quotations omitted). The Supreme Court reasoned that\n\n [b]ecause most creditors are aware of discharge orders and intend the actions they\n take to collect a debt, this standard would operate much like a strict-liability\n standard. It would authorize civil contempt sanctions for a violation of a discharge\n order regardless of the creditor’s subjective beliefs about the scope of the discharge\n order, and regardless of whether there was a reasonable basis for concluding that\n the creditor’s conduct did not violate the order.\n\nTaggart, 139 S. Ct. at 1803. In place of the “strict-liability standard,” the Court adopted the “fair\nground of doubt standard.” Id. at 1802.\n\n\n\n 8\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 9 of 15\n\n\ndetermining if “there is no fair ground of doubt as to whether the order barred the\n\ncreditor’s conduct.” Taggart, 139 S. Ct. at 1799.\n\n We now turn to the substantive questions presented by this appeal: whether\n\nthe Informational Statement sent by Nationstar was an improper attempt at debt\n\ncollection in violation of 11 U.S.C. § 524(a)(2), and if so, whether it was a\n\nsanctionable under 11 U.S.C. § 105.\n\n A. 11 U.S.C. § 524(a)(2)\n Roth has not met her burden of showing that the Informational Statement\n\nwas an unlawful debt collection in violation of § 524; there are several bases for\n\nconcluding that “the objective effect” of the Informational Statement was not “to\n\npressure [Roth] to repay a discharged debt.” In re McLean, 794 F.3d at 1322. As an\n\ninitial matter, the disclaimer is printed in bold on the first page of the statement. It\n\ndeclares that it is “for informational purposes only and is not intended as an\n\nattempt to collect, assess, or recover a discharged debt from you, or as a demand\n\nfor payment from any individual protected by the United States Bankruptcy Code.”\n\nIn case the first sentence is not clear enough, the disclaimer then repeats that if the\n\naccount “has been discharged in a bankruptcy proceeding” then it is “for\n\ninformational purposes only,” and “is not an attempt to collect a debt.” Further, the\n\nincluded payment coupon is marked in large lettering as “voluntary.”\n\n The fact that the statement includes an “amount due,” “due date,” and\n\n\n\n 9\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 10 of 15\n\n\nstatements about the negative escrow balance does not diminish the effect of the\n\nprominent, clear, and broadly worded disclaimer. Notably, section 524 allows for a\n\ndebtor to pay back a discharged debt voluntarily. Under 11 U.S.C. § 524(f),\n\n“[n]othing contained in subsection (c) or (d) of this section prevents a debtor from\n\nvoluntarily repaying any debt.” Roth had the option to repay the debt (the “amount\n\ndue”) and thereby regain the property because Nationstar had not completed a\n\nforeclosure on the property. If we accepted Roth’s argument that the Informational\n\nStatement is unlawful debt collection under § 524, there would be little daylight\n\nbetween (1) a legitimate attempt by Nationstar to inform Roth how she could\n\nregain the property and (2) an unlawful attempt at debt collection in violation of\n\nthe § 524 injunction. Instead, the statutory scheme clearly allows for Nationstar to\n\nsend potentially helpful informational statements to Roth without simultaneously\n\ncasting those statements as debt collection. In light of these facts, the Informational\n\nStatement sent by Nationstar was not designed to have the “objective effect” of\n\n“pressur[ing] the debtor to pay a discharged debt,” In re McLean, 794 F.3d at\n\n1322, and so we conclude that it was not an unlawful attempt at debt collection in\n\nviolation of § 524.\n\n Roth asks us to use a different standard to determine if Nationstar has\n\ncommitted a § 524 violation. In Roth’s separate lawsuit, which was brought under\n\n\n\n\n 10\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 11 of 15\n\n\nthe FDCPA rather than the bankruptcy code, 4 the district court applied the\n\nappropriate FDCPA test using the “least sophisticated consumer” standard. Roth v.\n\nNationstar Mortg. LLC, No. 2:15-cv-00783-FtM-29MRM, 2016 WL 3570991, at\n\n*3 (M.D. Fla., July 1, 2016). 5 At the motion to dismiss stage, the district court in\n\nthe FDCPA case summarized its findings with respect to the Informational\n\nStatement:\n\n It is true that the Informational Statement does not expressly state that\n it is “a communication sent for the purpose of collecting a debt.” But,\n as just discussed, this absence is not dispositive. The Statement lists the\n total amount due, contains a payment due date, states that a late fee will\n be charged for an untimely payment, gives six possible payment\n methods, and separates out from the total amount due the amount of\n fees and charges previously assessed. That is not all. The statement\n contains an “Important Messages” box advising [Appellant] that her\n “escrow account has a negative balance,” and expressly\n “recommend[ing she] make additional payments” to avoid “an increase\n in [her] monthly escrow payment.” There is also a detachable “payment\n coupon,” which states the total amount due and recalculates the amount\n due for a late payment.\n\nRoth, 2016 WL 3570991, at *3. Based on these observations, the court determined\n\n\n\n4\n The FDCPA creates a civil cause of action based on certain prohibited debt collection methods,\nspecifically a “false, deceptive, or misleading representation or means in connection with the\ncollection of any debt” or an “unfair or unconscionable means” of debt collection. 15 U.S.C.\n§ 1692e–f.\n5\n The FDCPA standard for determining if a communication is a debt collection is whether the\nstatement would “mislead the least sophisticated consumer regarding the nature of her rights.”\nCaceres v. McCalla Raymer, LLC, 755 F.3d 1299, 1303 (11th Cir. 2014). “The least\nsophisticated consumer’ is presumed to have only a “rudimentary amount of information about\nthe world,” but “it is assumed that the least-sophisticated consumer will be willing to read a\ncollection notice with some care.” Holzman v. Malcolm S. Gerald & Assocs., Inc., 920 F.3d\n1264, 1269 (11th Cir. 2019) (internal quotation marks, citations, and alterations omitted).\n\n\n 11\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 12 of 15\n\n\nthat Roth’s FDCPA complaint “plausibly alleges that the Informational Statement\n\nwas sent to induce payment on Plaintiff’s mortgage debt,” and was thus “related to\n\ndebt collection” under the FDCPA. Id. at *4.\n\n Roth argues that the “underlying questions [of § 524 and the FDCPA] are\n\ndesigned to protect the same vulnerable parties from the same improper conduct”\n\nand so the FDCPA and bankruptcy code should use the same standard. Roth is\n\nasking this Court to import the “least sophisticated consumer” standard from the\n\nFDCPA into the § 524 context.\n\n We decline to follow Roth’s request. 6 This Court has never incorporated the\n\n“least sophisticated consumer” test into our § 524 analysis. And for good reason—\n\nwhat counts as “debt collection” under one statutory scheme is not necessarily\n\n“debt collection” under the other. See Midland Funding, LLC v. Johnson, 137 S.\n\nCt. 1407, 1414 (2017) (“The [FDCPA] and the [Bankruptcy] Code have different\n\npurposes and structural features. The Act seeks to help consumers . . . . The\n\nBankruptcy Code, by way of contrast, creates and maintains what we have called\n\nthe delicate balance of a debtor’s protections and obligations.”) (citation and\n\ninternal quotations marks omitted). And the Supreme Court reiterated that “civil\n\n\n\n6\n The district court in Roth’s FDCPA case was acting in response to Nationstar’s motion to\ndismiss for failure to state a claim. Thus, the district court in that case did not make a finding on\nthe merits, but rather it found only that the complaint “state[d] a claim to relief that is plausible\non its face.” Bell Atl. Corp. v. Twombly, 550 U.S. 544, 547 (2007).\n\n\n\n 12\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 13 of 15\n\n\ncontempt is a severe remedy,” so it follows that the burden to show contempt\n\nshould be a high one. Taggart, 139 S.Ct. at 1802 (citation and internal quotation\n\nmarks omitted). There is nothing in either statute that suggests we should evaluate\n\npotential § 524 violations using the FDCPA standard, and we find no other reason\n\nto do so.\n\n B. 11 U.S.C. § 105\n\n Because we find that Nationstar’s Informational Statement did not violate\n\n§ 524, we do not have to consider whether sanctions would be appropriate under\n\n§ 105. But even if we reached a different conclusion based on these facts and\n\nfound that the issue of the § 524 violation presented a close call, we would\n\nnonetheless find that sanctions under § 105 are unavailable under the Supreme\n\nCourt’s recent decision in Taggart, which established the “no fair ground of doubt”\n\nstandard. Taggart, 139 S. Ct. at 1799. The Taggart standard is a rigorous one: in\n\norder to find that sanctions are appropriate here, we would have to hold that “there\n\nis no objectively reasonable basis for concluding that [Nationstar’s] conduct might\n\nbe lawful.” Id. With more than a “fair ground of doubt,” Taggart, 139 S. Ct. at\n\n1799, as to whether the discharge order barred Nationstar’s conduct, sanctions\n\nwould be inappropriate.\n\n As we have determined that Nationstar’s Informational Statement did not\n\nviolate § 524, the “no fair ground of doubt” standard for § 105 is necessarily not\n\n\n\n 13\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 14 of 15\n\n\nsatisfied. Accordingly, the bankruptcy court’s denial of Roth’s motion for\n\nsanctions was not an abuse of discretion.\n\n C. Evidentiary Hearing\n\n Finally, Roth argues that the bankruptcy court should have held an\n\nevidentiary hearing—despite the fact that she did not ask for one—before it made\n\nits findings concerning the Informational Statement.7 The crux of Roth’s argument\n\nis that the bankruptcy court concluded that Roth “can hardly have thought that\n\nNationstar was trying to collect the discharged debt,” but a determination of her\n\nsubjective belief required an evidentiary hearing.\n\n In civil contempt proceedings, “when there are no disputed factual matters\n\nthat require an evidentiary hearing, the court might properly dispense with the\n\nhearing prior to finding the defendant in contempt and sanctioning him.” In re\n\nMcLean, 794 F.3d at 1324 (quoting Mercer v. Mitchell, 908 F.2d 763, 769 n.11\n\n(11th Cir. 1990)).\n\n Neither party in this case requested an evidentiary hearing, and the\n\nbankruptcy court was not required to hold an evidentiary hearing sua sponte to\n\ninquire about Roth’s subjective beliefs. The bankruptcy court had to look only to\n\n\n7\n The FDCPA creates a civil cause of action based on certain prohibited debt collection methods,\nspecifically a “false, deceptive, or misleading representation or means in connection with the\ncollection of any debt” or an “unfair or unconscionable means” of debt collection. 15 U.S.C.\n§ 1692e–f.005) (“As we repeatedly have admonished, arguments raised for the first time in a\nreply brief are not properly before a reviewing court.”) (citation and internal quotation marks\nomitted).\n\n\n 14\n\f Case: 17-11444 Date Filed: 08/28/2019 Page: 15 of 15\n\n\nthe Informational Statement itself to determine “whether the objective effect of [the\n\nInformational Statement] is to pressure a debtor to repay a discharged debt,”\n\nMcLean, 794 F.3d at 1322 (emphasis added). It did so at length, making numerous\n\nfindings about the document and determining that the statement did not constitute\n\nan unlawful attempt at debt collection under § 524. Roth’s subjective belief about\n\nthe Informational Statement is irrelevant to the analysis, so it was not a “disputed\n\nfactual matter[] that require[d] an evidentiary hearing.” In re McLean, 794 F.3d at\n\n1324. The bankruptcy court’s statement that Roth “can hardly have thought that\n\nNationstar was trying to collect the discharged debt” is simply the court’s\n\nconclusion that the communications were not an attempt at debt collection.\n\n Accordingly, the bankruptcy court did not err by rendering its decision\n\nwithout a hearing.\n\n AFFIRMED.\n\n\n\n\n 15\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: In RE: Arlene ROTH, Debtor. Arlene Roth, Plaintiff - Appellant, v. Nationstar Mortgage, LLC, Defendant - Appellee.\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-28\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 17-11444 Date Filed: 08/28/2019 Page: 1 of 15\\n\\n\\n [PUBLISH]\\n\\n\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No: 17-11444\\n ________________________\\n\\n D.C. Docket Nos. 2:16-cv-00510-SPC; 9:10-bkc-30383-FMD\\n\\n\\nIN RE: ARLENE ROTH,\\n\\n Debtor.\\n\\n ________________________\\n\\n\\nARLENE ROTH,\\n Plaintiff - Appellant,\\n\\n versus\\n\\n\\nNATIONSTAR MORTGAGE, LLC,\\n\\n Defendant - Appellee.\\n\\n ________________________\\n\\n Appeal from the United States District Court\\n for the Middle District of Florida\\n ________________________\\n\\n (August 28, 2019)\\n\\nBefore WILSON, BRANCH, and ANDERSON, Circuit Judges.\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 2 of 15\\n\\n\\nBRANCH, Circuit Judge:\\n\\n Arlene Roth appeals the bankruptcy court’s denial of her second motion for\\n\\nsanctions against Nationstar Mortgage, LLC (“Nationstar”). The basis of the\\n\\nmotion was Nationstar’s conduct of sending a monthly statement to Roth, which\\n\\nshe alleges sought to collect discharged mortgage debt in violation of the\\n\\nbankruptcy code’s prohibition on collection of such debt, 11 U.S.C. § 524. She\\n\\nraises two issues on appeal: (1) whether the bankruptcy court improperly made\\n\\nfactual determinations without the benefit of an evidentiary hearing; and (2)\\n\\nwhether Nationstar’s mortgage statement was an improper attempt to collect a debt\\n\\nin violation of 11 U.S.C. § 524 and justifies sanctions against Nationstar. For the\\n\\nfollowing reasons, we affirm the bankruptcy court’s decision.\\n\\n I.\\n\\n On December 22, 2010, Roth filed a voluntary petition for bankruptcy under\\n\\nChapter 13 of the bankruptcy code. On her bankruptcy schedule, Roth listed a\\n\\nmortgage on non-homestead property, which is at issue here. On this schedule, she\\n\\nindicated she would surrender the property. The Chapter 13 plan provided that\\n\\n“[s]ecured creditors, whether or not dealt with under the Plan, shall retain the liens\\n\\nsecuring such claims.” On October 3, 2011, the bankruptcy court entered an order\\n\\nconfirming her Chapter 13 bankruptcy plan. Subsequently, the mortgage at issue\\n\\nhere (which was subject to the Chapter 13 plan) was transferred to Nationstar. She\\n\\n\\n\\n 2\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 3 of 15\\n\\n\\ncompleted her payments under her Chapter 13 plan, and the bankruptcy court\\n\\ndischarged her debt on June 27, 2014. Nationstar was notified of the discharge.\\n\\n In the discharge order, the bankruptcy court ordered that “the discharge\\n\\nprohibits any attempt to collect from the debtor a debt that has been discharged.\\n\\nFor example, a creditor is not permitted to contact a debtor by mail, phone, or\\n\\notherwise, to file or continue a lawsuit, to attach wages or other property, or to take\\n\\nany other action to collect a discharged debt from the debtor.” The order, however,\\n\\nalso stated that “a creditor may have the right to enforce a valid lien, such as a\\n\\nmortgage or security interest, against the debtor’s property after the bankruptcy, if\\n\\nthat lien was not avoided or eliminated in the bankruptcy case. Also, a debtor may\\n\\nvoluntarily pay any debt that has been discharged.” But even after the discharge of\\n\\nthe mortgage debt, Nationstar did not foreclose on the property; thus, if Roth\\n\\nvoluntarily paid the amount due, she could have retained the property.\\n\\n About four months after the discharge of Roth’s debt, Nationstar started\\n\\nsending Roth monthly statements related to her mortgage. The statements included\\n\\na disclaimer that they were not debt collection, but also included an amount due,\\n\\ndue date, and instructions on how to send payment back to Nationstar. Roth had\\n\\nher attorney send a cease and desist letter, but the statements kept coming. She\\n\\nthen filed her first motion for sanctions in bankruptcy court, alleging that the\\n\\nstatements were improper debt collections in violation of 11 U.S.C. § 524, as well\\n\\n\\n\\n 3\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 4 of 15\\n\\n\\nas a separate civil action claiming violations of the Fair Debt Collection Practices\\n\\nAct (“FDCPA”), 15 U.S.C. § 1692e et seq, and the Florida Consumer Collection\\n\\nPractices Act (FCCPA). Arlene Roth v. Nationstar Mortgage LLC, No. 2:15-cv-\\n\\n508-FtM-38CM (M.D. Fla.). The parties resolved the first sanctions motion and\\n\\nfirst FDCPA action in a settlement.\\n\\n But communication from Nationstar continued in the form of a November\\n\\n18, 2015 “Informational Statement,” which again contained an amount due, due\\n\\ndate, and instructions for how to pay Nationstar, as well as a lengthy disclaimer.\\n\\nThe full text of the disclaimer reads as follows:\\n\\n This statement is sent for informational purposes only and is not\\n intended as an attempt to collect, assess, or recover a discharged debt\\n from you, or as a demand for payment from any individual protected\\n by the United States Bankruptcy Code. If this account is active or has\\n been discharged in a bankruptcy proceeding, be advised this\\n communication is for informational purposes only and is not an attempt\\n to collect a debt. Please note, however Nationstar reserves the right to\\n exercise its legal rights, including but not limited to foreclosure of its\\n lien interest, only against the property securing the original obligation.\\n Consequently, on December 14, 2015, Roth filed a second lawsuit against\\n\\nNationstar alleging the Informational Statement was an improper debt collection\\n\\nunder the FDCPA. Roth v. Nationstar Mortg., LLC, No. 2:15-cv-783-FtM-\\n\\n29MRM, 2016 WL 3570991 at *1 (M.D. Fla. July 1, 2016). Nationstar filed a\\n\\nmotion to dismiss that case, arguing that the Informational Statement was not sent\\n\\nfor debt collection purposes as a matter of law, but the district court denied the\\n\\n\\n\\n 4\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 5 of 15\\n\\n\\nmotion after applying the FDCPA’s “least sophisticated consumer” standard. Id. at\\n\\n*5. On May 30, 2017, the parties notified the district court that they had reached a\\n\\nsettlement in that case. Notice of Settlement, Roth v. Nationstar Mortg., LLC, No.\\n\\n2:15-cv-783-FtM-29MRM (M.D. Fla. May 30, 2017).\\n\\n Based on the same Information Statement, on December 14, 2015—the\\n\\nsame day that she filed her second FDCPA case—Roth filed a second motion for\\n\\nsanctions in her bankruptcy case, which is at issue here. Roth alleged that\\n\\nNationstar’s communication was an attempt to collect a discharged debt in\\n\\nviolation of the discharge order. Nationstar claimed that the statement was only\\n\\ninformational and not a violation of § 524. After briefing, the bankruptcy court\\n\\ndenied Roth’s motion for sanctions at a hearing, and later entered an opinion and\\n\\nan order finding that the “informational statement” was not a debt collection\\n\\nattempt, and therefore was not in violation of the § 524 injunction. Roth appealed\\n\\nthe denial of this second sanctions motion to the district court, which affirmed the\\n\\nbankruptcy court’s opinion and rejected Roth’s request to apply the FDCPA’s\\n\\n“least sophisticated consumer” standard. In re Roth, 568 B.R. 139, 145 (M.D. Fla.\\n\\n2017). This appeal followed.\\n\\n II.\\n\\n In bankruptcy cases, this Court “sits as a second court of review and thus\\n\\nexamines independently the factual and legal determinations of the bankruptcy\\n\\n\\n\\n 5\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 6 of 15\\n\\n\\ncourt and employs the same standards of review as the district court.” In re Ocean\\n\\nWarrior, Inc., 835 F.3d 1310, 1315 (11th Cir. 2016) (quoting In re Fisher Island\\n\\nInvs., Inc., 778 F.3d 1172, 1189 (11th Cir. 2015)); see also In re Gonzalez, 832\\n\\nF.3d 1251, 1253 (11th Cir. 2016), cert. denied sub nom. Fla. Dep’t of Revenue v.\\n\\nGonzalez, 137 S. Ct. 2293 (2017). When the district court affirms the bankruptcy\\n\\ncourt, we review the bankruptcy court’s decision, reviewing its “factual findings\\n\\nfor clear error and its legal conclusions de novo.” In re Ocean Warrior, Inc., 835\\n\\nF.3d at 1315 (internal quotation marks and citations omitted). We review the\\n\\nbankruptcy court’s denial for a motion for sanctions for abuse of discretion. In re\\n\\nDiaz, 647 F.3d 1073, 1082 (11th Cir. 2011).\\n\\n III.\\n\\n Roth challenges (1) the bankruptcy court’s substantive decision to deny her\\n\\nmotion for sanctions, and (2) the bankruptcy court’s decision to dispose of her\\n\\nmotion for sanctions without a separate evidentiary hearing. We first articulate the\\n\\nlegal standard Roth must meet to establish a violation of 11 U.S.C. § 524, and\\n\\napply that standard to review the bankruptcy court’s decision that Nationstar’s\\n\\ncommunications were not a sanctionable violation of § 524. We then determine\\n\\nwhether Roth was entitled to an evidentiary hearing in the bankruptcy court on that\\n\\nsubstantive question.\\n\\n Section 524(a)(2) of the bankruptcy code provides that a discharge of debt in\\n\\n\\n\\n 6\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 7 of 15\\n\\n\\na bankruptcy proceeding “operates as an injunction against the commencement or\\n\\ncontinuation of . . . an act . . . to collect . . . any such [discharged] debt.” 11 U.S.C.\\n\\n§ 524(a)(2).1 This injunction is enforced through section 105, whereby the\\n\\nbankruptcy court “may issue any order, process, or judgment that is necessary or\\n\\nappropriate to carry out the provisions of this title.” 11 U.S.C. § 105(a). Section\\n\\n105 “grants statutory contempt powers in the bankruptcy context.” In re Hardy, 97\\n\\nF.3d 1384, 1389–90 (11th Cir. 1996) (“Therefore, § 105(a) grants courts\\n\\nindependent statutory powers to award monetary and other forms of relief for\\n\\n[violations] to the extent such awards are ‘necessary [or] appropriate’ to carry out\\n\\nthe provisions of the Bankruptcy Code.”) (internal quotation marks and citations\\n\\nomitted). Together, sections 524(a)(2) and 105(a) “authorize a court to impose civil\\n\\n\\n\\n\\n1\\n 11 U.S.C. § 524(a) A discharge in a case under this title--\\n (1) voids any judgment at any time obtained, to the extent that such judgment is a\\n determination of the personal liability of the debtor with respect to any debt discharged\\n under section 727, 944, 1141, 1228, or 1328 of this title, whether or not discharge of such\\n debt is waived;\\n (2) operates as an injunction against the commencement or continuation of an action, the\\n employment of process, or an act, to collect, recover or offset any such debt as a personal\\n liability of the debtor, whether or not discharge of such debt is waived; and\\n (3) operates as an injunction against the commencement or continuation of an action, the\\n employment of process, or an act, to collect or recover from, or offset against, property of\\n the debtor of the kind specified in section 541(a)(2) of this title that is acquired after the\\n commencement of the case, on account of any allowable community claim, except a\\n community claim that is excepted from discharge under section 523, 1228(a)(1), or\\n 1328(a)(1), or that would be so excepted, determined in accordance with the provisions\\n of sections 523(c) and 523(d) of this title, in a case concerning the debtor’s spouse\\n commenced on the date of the filing of the petition in the case concerning the debtor,\\n whether or not discharge of the debt based on such community claim is waived.\\n\\n\\n\\n 7\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 8 of 15\\n\\n\\ncontempt sanctions [for attempting to collect a discharged debt] when there is no\\n\\nobjectively reasonable basis for concluding that the creditor’s conduct might be\\n\\nlawful under the discharge order.” Taggart v. Lorenzen, 139 S. Ct. 1795, 1801\\n\\n(2019).2 In this way, “a court may hold a creditor in civil contempt for violating a\\n\\ndischarge order if there is no fair ground of doubt as to whether the order barred\\n\\nthe creditor’s conduct.” Id. at 1799. 3\\n\\n Accordingly, we first determine whether a communication is a prohibited\\n\\ndebt collection under section 524 by looking to “whether the objective effect of the\\n\\ncreditor’s action is to pressure a debtor to repay a discharged debt.” In re McLean,\\n\\n794 F.3d 1313, 1322 (11th Cir. 2015). If so, we then evaluate whether that\\n\\nviolation of the discharge injunction is sanctionable under section 105, by\\n\\n\\n2\\n After we held oral argument in this case, the Supreme Court clarified in Taggart when\\ncontempt sanctions under § 105 are appropriate for violations of § 524.\\n3\\n In Taggart, the Supreme Court explicitly disavowed our § 105 analysis from In re Hardy in\\nwhich we held that the test “applicable to determining willfulness for violations of the discharge\\ninjunction of § 524” is if the defendant “(1) knew that the [injunction] was invoked and (2)\\nintended the actions which violated the [injunction]. In re Hardy, 97 F.3d 1384, 1390 (11th Cir.\\n1996) (quotations omitted). The Supreme Court reasoned that\\n\\n [b]ecause most creditors are aware of discharge orders and intend the actions they\\n take to collect a debt, this standard would operate much like a strict-liability\\n standard. It would authorize civil contempt sanctions for a violation of a discharge\\n order regardless of the creditor’s subjective beliefs about the scope of the discharge\\n order, and regardless of whether there was a reasonable basis for concluding that\\n the creditor’s conduct did not violate the order.\\n\\nTaggart, 139 S. Ct. at 1803. In place of the “strict-liability standard,” the Court adopted the “fair\\nground of doubt standard.” Id. at 1802.\\n\\n\\n\\n 8\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 9 of 15\\n\\n\\ndetermining if “there is no fair ground of doubt as to whether the order barred the\\n\\ncreditor’s conduct.” Taggart, 139 S. Ct. at 1799.\\n\\n We now turn to the substantive questions presented by this appeal: whether\\n\\nthe Informational Statement sent by Nationstar was an improper attempt at debt\\n\\ncollection in violation of 11 U.S.C. § 524(a)(2), and if so, whether it was a\\n\\nsanctionable under 11 U.S.C. § 105.\\n\\n A. 11 U.S.C. § 524(a)(2)\\n Roth has not met her burden of showing that the Informational Statement\\n\\nwas an unlawful debt collection in violation of § 524; there are several bases for\\n\\nconcluding that “the objective effect” of the Informational Statement was not “to\\n\\npressure [Roth] to repay a discharged debt.” In re McLean, 794 F.3d at 1322. As an\\n\\ninitial matter, the disclaimer is printed in bold on the first page of the statement. It\\n\\ndeclares that it is “for informational purposes only and is not intended as an\\n\\nattempt to collect, assess, or recover a discharged debt from you, or as a demand\\n\\nfor payment from any individual protected by the United States Bankruptcy Code.”\\n\\nIn case the first sentence is not clear enough, the disclaimer then repeats that if the\\n\\naccount “has been discharged in a bankruptcy proceeding” then it is “for\\n\\ninformational purposes only,” and “is not an attempt to collect a debt.” Further, the\\n\\nincluded payment coupon is marked in large lettering as “voluntary.”\\n\\n The fact that the statement includes an “amount due,” “due date,” and\\n\\n\\n\\n 9\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 10 of 15\\n\\n\\nstatements about the negative escrow balance does not diminish the effect of the\\n\\nprominent, clear, and broadly worded disclaimer. Notably, section 524 allows for a\\n\\ndebtor to pay back a discharged debt voluntarily. Under 11 U.S.C. § 524(f),\\n\\n“[n]othing contained in subsection (c) or (d) of this section prevents a debtor from\\n\\nvoluntarily repaying any debt.” Roth had the option to repay the debt (the “amount\\n\\ndue”) and thereby regain the property because Nationstar had not completed a\\n\\nforeclosure on the property. If we accepted Roth’s argument that the Informational\\n\\nStatement is unlawful debt collection under § 524, there would be little daylight\\n\\nbetween (1) a legitimate attempt by Nationstar to inform Roth how she could\\n\\nregain the property and (2) an unlawful attempt at debt collection in violation of\\n\\nthe § 524 injunction. Instead, the statutory scheme clearly allows for Nationstar to\\n\\nsend potentially helpful informational statements to Roth without simultaneously\\n\\ncasting those statements as debt collection. In light of these facts, the Informational\\n\\nStatement sent by Nationstar was not designed to have the “objective effect” of\\n\\n“pressur[ing] the debtor to pay a discharged debt,” In re McLean, 794 F.3d at\\n\\n1322, and so we conclude that it was not an unlawful attempt at debt collection in\\n\\nviolation of § 524.\\n\\n Roth asks us to use a different standard to determine if Nationstar has\\n\\ncommitted a § 524 violation. In Roth’s separate lawsuit, which was brought under\\n\\n\\n\\n\\n 10\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 11 of 15\\n\\n\\nthe FDCPA rather than the bankruptcy code, 4 the district court applied the\\n\\nappropriate FDCPA test using the “least sophisticated consumer” standard. Roth v.\\n\\nNationstar Mortg. LLC, No. 2:15-cv-00783-FtM-29MRM, 2016 WL 3570991, at\\n\\n*3 (M.D. Fla., July 1, 2016). 5 At the motion to dismiss stage, the district court in\\n\\nthe FDCPA case summarized its findings with respect to the Informational\\n\\nStatement:\\n\\n It is true that the Informational Statement does not expressly state that\\n it is “a communication sent for the purpose of collecting a debt.” But,\\n as just discussed, this absence is not dispositive. The Statement lists the\\n total amount due, contains a payment due date, states that a late fee will\\n be charged for an untimely payment, gives six possible payment\\n methods, and separates out from the total amount due the amount of\\n fees and charges previously assessed. That is not all. The statement\\n contains an “Important Messages” box advising [Appellant] that her\\n “escrow account has a negative balance,” and expressly\\n “recommend[ing she] make additional payments” to avoid “an increase\\n in [her] monthly escrow payment.” There is also a detachable “payment\\n coupon,” which states the total amount due and recalculates the amount\\n due for a late payment.\\n\\nRoth, 2016 WL 3570991, at *3. Based on these observations, the court determined\\n\\n\\n\\n4\\n The FDCPA creates a civil cause of action based on certain prohibited debt collection methods,\\nspecifically a “false, deceptive, or misleading representation or means in connection with the\\ncollection of any debt” or an “unfair or unconscionable means” of debt collection. 15 U.S.C.\\n§ 1692e–f.\\n5\\n The FDCPA standard for determining if a communication is a debt collection is whether the\\nstatement would “mislead the least sophisticated consumer regarding the nature of her rights.”\\nCaceres v. McCalla Raymer, LLC, 755 F.3d 1299, 1303 (11th Cir. 2014). “The least\\nsophisticated consumer’ is presumed to have only a “rudimentary amount of information about\\nthe world,” but “it is assumed that the least-sophisticated consumer will be willing to read a\\ncollection notice with some care.” Holzman v. Malcolm S. Gerald & Assocs., Inc., 920 F.3d\\n1264, 1269 (11th Cir. 2019) (internal quotation marks, citations, and alterations omitted).\\n\\n\\n 11\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 12 of 15\\n\\n\\nthat Roth’s FDCPA complaint “plausibly alleges that the Informational Statement\\n\\nwas sent to induce payment on Plaintiff’s mortgage debt,” and was thus “related to\\n\\ndebt collection” under the FDCPA. Id. at *4.\\n\\n Roth argues that the “underlying questions [of § 524 and the FDCPA] are\\n\\ndesigned to protect the same vulnerable parties from the same improper conduct”\\n\\nand so the FDCPA and bankruptcy code should use the same standard. Roth is\\n\\nasking this Court to import the “least sophisticated consumer” standard from the\\n\\nFDCPA into the § 524 context.\\n\\n We decline to follow Roth’s request. 6 This Court has never incorporated the\\n\\n“least sophisticated consumer” test into our § 524 analysis. And for good reason—\\n\\nwhat counts as “debt collection” under one statutory scheme is not necessarily\\n\\n“debt collection” under the other. See Midland Funding, LLC v. Johnson, 137 S.\\n\\nCt. 1407, 1414 (2017) (“The [FDCPA] and the [Bankruptcy] Code have different\\n\\npurposes and structural features. The Act seeks to help consumers . . . . The\\n\\nBankruptcy Code, by way of contrast, creates and maintains what we have called\\n\\nthe delicate balance of a debtor’s protections and obligations.”) (citation and\\n\\ninternal quotations marks omitted). And the Supreme Court reiterated that “civil\\n\\n\\n\\n6\\n The district court in Roth’s FDCPA case was acting in response to Nationstar’s motion to\\ndismiss for failure to state a claim. Thus, the district court in that case did not make a finding on\\nthe merits, but rather it found only that the complaint “state[d] a claim to relief that is plausible\\non its face.” Bell Atl. Corp. v. Twombly, 550 U.S. 544, 547 (2007).\\n\\n\\n\\n 12\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 13 of 15\\n\\n\\ncontempt is a severe remedy,” so it follows that the burden to show contempt\\n\\nshould be a high one. Taggart, 139 S.Ct. at 1802 (citation and internal quotation\\n\\nmarks omitted). There is nothing in either statute that suggests we should evaluate\\n\\npotential § 524 violations using the FDCPA standard, and we find no other reason\\n\\nto do so.\\n\\n B. 11 U.S.C. § 105\\n\\n Because we find that Nationstar’s Informational Statement did not violate\\n\\n§ 524, we do not have to consider whether sanctions would be appropriate under\\n\\n§ 105. But even if we reached a different conclusion based on these facts and\\n\\nfound that the issue of the § 524 violation presented a close call, we would\\n\\nnonetheless find that sanctions under § 105 are unavailable under the Supreme\\n\\nCourt’s recent decision in Taggart, which established the “no fair ground of doubt”\\n\\nstandard. Taggart, 139 S. Ct. at 1799. The Taggart standard is a rigorous one: in\\n\\norder to find that sanctions are appropriate here, we would have to hold that “there\\n\\nis no objectively reasonable basis for concluding that [Nationstar’s] conduct might\\n\\nbe lawful.” Id. With more than a “fair ground of doubt,” Taggart, 139 S. Ct. at\\n\\n1799, as to whether the discharge order barred Nationstar’s conduct, sanctions\\n\\nwould be inappropriate.\\n\\n As we have determined that Nationstar’s Informational Statement did not\\n\\nviolate § 524, the “no fair ground of doubt” standard for § 105 is necessarily not\\n\\n\\n\\n 13\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 14 of 15\\n\\n\\nsatisfied. Accordingly, the bankruptcy court’s denial of Roth’s motion for\\n\\nsanctions was not an abuse of discretion.\\n\\n C. Evidentiary Hearing\\n\\n Finally, Roth argues that the bankruptcy court should have held an\\n\\nevidentiary hearing—despite the fact that she did not ask for one—before it made\\n\\nits findings concerning the Informational Statement.7 The crux of Roth’s argument\\n\\nis that the bankruptcy court concluded that Roth “can hardly have thought that\\n\\nNationstar was trying to collect the discharged debt,” but a determination of her\\n\\nsubjective belief required an evidentiary hearing.\\n\\n In civil contempt proceedings, “when there are no disputed factual matters\\n\\nthat require an evidentiary hearing, the court might properly dispense with the\\n\\nhearing prior to finding the defendant in contempt and sanctioning him.” In re\\n\\nMcLean, 794 F.3d at 1324 (quoting Mercer v. Mitchell, 908 F.2d 763, 769 n.11\\n\\n(11th Cir. 1990)).\\n\\n Neither party in this case requested an evidentiary hearing, and the\\n\\nbankruptcy court was not required to hold an evidentiary hearing sua sponte to\\n\\ninquire about Roth’s subjective beliefs. The bankruptcy court had to look only to\\n\\n\\n7\\n The FDCPA creates a civil cause of action based on certain prohibited debt collection methods,\\nspecifically a “false, deceptive, or misleading representation or means in connection with the\\ncollection of any debt” or an “unfair or unconscionable means” of debt collection. 15 U.S.C.\\n§ 1692e–f.005) (“As we repeatedly have admonished, arguments raised for the first time in a\\nreply brief are not properly before a reviewing court.”) (citation and internal quotation marks\\nomitted).\\n\\n\\n 14\\n\\f Case: 17-11444 Date Filed: 08/28/2019 Page: 15 of 15\\n\\n\\nthe Informational Statement itself to determine “whether the objective effect of [the\\n\\nInformational Statement] is to pressure a debtor to repay a discharged debt,”\\n\\nMcLean, 794 F.3d at 1322 (emphasis added). It did so at length, making numerous\\n\\nfindings about the document and determining that the statement did not constitute\\n\\nan unlawful attempt at debt collection under § 524. Roth’s subjective belief about\\n\\nthe Informational Statement is irrelevant to the analysis, so it was not a “disputed\\n\\nfactual matter[] that require[d] an evidentiary hearing.” In re McLean, 794 F.3d at\\n\\n1324. The bankruptcy court’s statement that Roth “can hardly have thought that\\n\\nNationstar was trying to collect the discharged debt” is simply the court’s\\n\\nconclusion that the communications were not an attempt at debt collection.\\n\\n Accordingly, the bankruptcy court did not err by rendering its decision\\n\\nwithout a hearing.\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 15\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4156,"cells":{"id":{"kind":"number","value":4656793,"string":"4,656,793"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"John Salcedo v. Alex Hanna"},"decision_date":{"kind":"string","value":"2019-08-28"},"docket_number":{"kind":"string","value":"17-14077"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before JILL PRYOR and BRANCH, Circuit Judges, and REEVES, * District Judge.\", \"parties\": \"\", \"opinions\": [{\"author\": \"BRANCH, Circuit Judge:\", \"type\": \"010combined\", \"text\": \"Case: 17-14077 Date Filed: 08/28/2019 Page: 1 of 22\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n\\n\\n\\n No. 17-14077\\n\\n\\n\\n D.C. Docket No. 0:16-cv-62480-DPG\\n\\nJOHN SALCEDO,\\nindividually and on behalf of others similarly situated,\\n\\n Plaintiff - Appellee,\\n\\n versus\\n\\nALEX HANNA,\\nan individual,\\nTHE LAW OFFICES OF ALEX HANNA, P.A.,\\na Florida Professional Association,\\n\\n Defendants - Appellants.\\n\\n\\n\\n Appeal from the United States District Court\\n for the Southern District of Florida\\n\\n\\n\\n (August 28, 2019)\\n\\nBefore JILL PRYOR and BRANCH, Circuit Judges, and REEVES, * District\\nJudge.\\n\\n*\\n Honorable Danny C. Reeves, United States District Judge for the Eastern District of Kentucky,\\nsitting by designation.\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 2 of 22\\n\\n\\nBRANCH, Circuit Judge:\\n\\n Is receiving a single unsolicited text message, sent in violation of a federal\\n\\nstatute, a concrete injury in fact that establishes standing to sue in federal court? To\\n\\nanswer that question, we have examined the statute, our precedent, and—following\\n\\nthe Supreme Court’s guidance—history and the judgment of Congress, and we\\n\\nconclude that the allegations in this suit do not establish standing.\\n\\n I. BACKGROUND\\n\\n At 9:56 a.m. on August 12, 2016, John Salcedo, a former client of Florida\\n\\nattorney Alex Hanna and his law firm, 1 received a multimedia text message from\\n\\nHanna offering a ten percent discount on his services.\\n\\n Salcedo filed suit in the district court as the representative of a putative class\\n\\nof former Hanna clients who received unsolicited text messages from Hanna in the\\n\\npast four years, alleging violations of the Telephone Consumer Protection Act of\\n\\n1991 (“TCPA”), 47 U.S.C. § 227(b)(1)(A)(iii).2 He sought, among other relief,\\n\\nstatutory damages of $500 per text message and treble damages of $1,500 per text\\n\\n\\n\\n1\\n For simplicity, and without implying any view as to Mr. Hanna’s possible personal liability,\\nthroughout this opinion we will refer to both defendants—Mr. Hanna and his law firm—\\ncollectively as “Hanna.”\\n2\\n “It shall be unlawful for any person within the United States . . . to make any call (other than a\\ncall made for emergency purposes or made with the prior express consent of the called party)\\nusing any automatic telephone dialing system or an artificial or prerecorded voice . . . to any\\ntelephone number assigned to a . . . cellular telephone service . . . .” 47 U.S.C.\\n§ 227(b)(1)(A)(iii).\\n 2\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 3 of 22\\n\\n\\nmessage sent willfully or knowingly. See 47 U.S.C. § 227(b)(3).\\n\\n Hanna moved to dismiss the complaint for lack of standing, arguing in the\\n\\nalternative that it should be dismissed as to Mr. Hanna for failure to state a claim\\n\\nagainst him and that certain parts of the complaint should be stricken. The district\\n\\ncourt disagreed, finding in relevant part that Salcedo had standing under Mohamed\\n\\nv. Off Lease Only, Inc., No. 15-23352-Civ-COOKE/TORRES, 2017 WL 1080342\\n\\n(S.D. Fla. Mar. 22, 2017). However, finding that its order “involves a controlling\\n\\nquestion of law as to which there is a substantial ground for difference of opinion,”\\n\\nthe court allowed Salcedo to pursue an interlocutory appeal and stayed its\\n\\nproceedings pending appeal. A panel of our Court granted Hanna’s petition for\\n\\npermission to appeal under 28 U.S.C. § 1292(b). We now consider his appeal.\\n\\n II. STANDARD OF REVIEW\\n\\n “We review standing determinations de novo.” Bochese v. Town of Ponce\\n\\nInlet, 405 F.3d 964, 975 (11th Cir. 2005).\\n\\n III. DISCUSSION\\n\\n Our analysis proceeds as follows. We first introduce the TCPA, the statute\\n\\nunder which Salcedo has filed suit. Next, we discuss the standing requirements of\\n\\nArticle III of the Constitution, which help to define our limited power to resolve\\n\\nonly cases or controversies. We then turn to Salcedo’s particular allegations of\\n\\nharm and analyze them in view of our Circuit precedent, history, and the judgment\\n\\n\\n 3\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 4 of 22\\n\\n\\nof Congress.\\n\\n A. The Telephone Consumer Privacy Act of 1991\\n\\n Because it found that “residential telephone subscribers consider automated\\n\\nor prerecorded telephone calls . . . to be a nuisance and an invasion of privacy,”\\n\\nTelephone Consumer Protection Act of 1991, S. 1462, 102d Cong., Pub. L. No.\\n\\n102-243, § 2, ¶ 10 (1991), in 1991 Congress enacted the TCPA to restrict interstate\\n\\ntelemarketing. The TCPA thus prohibits using automatic telephone dialing systems\\n\\nto call residential or cellular telephone lines without the consent of the called party.\\n\\n47 U.S.C. § 227(b)(1)(A)(iii), (B). It also prohibits sending unsolicited\\n\\nadvertisements via facsimile machine. Id. § 227(b)(1)(C). It authorizes the Federal\\n\\nCommunications Commission (“FCC”) to enact implementing regulations. Id.\\n\\n§ 227(b)(2). Finally for our purposes, the TCPA creates a private right of action\\n\\nwhereby a person or entity may seek compensatory or injunctive relief against\\n\\nviolators. Id. § 227(b)(3).\\n\\n There have been two relevant updates to the TCPA and its enforcement\\n\\nregime since 1991. First, in October 1992, Congress amended the TCPA to allow\\n\\nthe FCC to exempt free-to-receive cellular calls if it so chooses. Id. § 227(b)(2)(C).\\n\\nThe FCC has not done so. Second, the statute has been silent as to text messaging,\\n\\nfor that medium did not exist in 1991. But under its TCPA rulemaking authority,\\n\\nthe FCC has applied the statute’s regulations of voice calls to text messages. 30\\n\\n\\n 4\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 5 of 22\\n\\n\\nFCC Rcd. 7961, 7964 n.3, 7978–79, 8016–22 (2015); 18 FCC Rcd. 14014, 14115\\n\\n(2003); see also Campbell–Ewald Co. v. Gomez, 136 S. Ct. 663, 667 (2016) (“A\\n\\ntext message to a cellular telephone, it is undisputed, qualifies as a ‘call’ within the\\n\\ncompass of § 227(b)(1)(A)(iii).”). Thus, Salcedo’s complaint facially appears to\\n\\nstate a cause of action under the TCPA as interpreted by the FCC.\\n\\n B. Article III Standing\\n\\n Not every right created by Congress or defined by an executive agency is\\n\\nautomatically enforceable in the federal courts. Our tripartite system of\\n\\ngovernment recognizes that “there is no liberty if the power of judging be not\\n\\nseparated from the legislative and executive powers.” The Federalist No. 78, at 465\\n\\n(Alexander Hamilton) (Clinton Rossiter ed. 1961). To protect this separation of\\n\\npowers, we must assure ourselves that our exercise of jurisdiction falls within the\\n\\nConstitution’s grant of judicial power.\\n\\n Article III vests the judicial power in the federal courts and extends that\\n\\npower to “Cases” and “Controversies.” U.S. Const. art. III, §§ 1–2. One tool for\\n\\ndetermining that the matters before us are truly cases or controversies, as\\n\\nunderstood by Article III, is the doctrine of standing. Lujan v. Defenders of\\n\\nWildlife, 504 U.S. 555, 560 (1992). “The law of Article III standing . . . serves to\\n\\nprevent the judicial process from being used to usurp the powers of the political\\n\\nbranches.” Clapper v. Amnesty Int’l USA, 568 U.S. 398, 408 (2013). Even when\\n\\n\\n 5\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 6 of 22\\n\\n\\nthose political branches appear to have granted us jurisdiction by statute and rule,\\n\\nwe are still obliged to examine whether jurisdiction exists under the Constitution.\\n\\n As the Supreme Court has explained, the “irreducible constitutional\\n\\nminimum” to establish Article III standing requires three elements. Lujan, 504 U.S.\\n\\nat 560. “The plaintiff must have (1) suffered an injury in fact, (2) that is fairly\\n\\ntraceable to the challenged conduct of the defendant, and (3) that is likely to be\\n\\nredressed by a favorable judicial decision.” Spokeo, Inc. v. Robins, 136 S. Ct. 1540,\\n\\n1547 (2016) (citing Lujan, 504 U.S. at 560–61). It is the first element—the\\n\\n“foremost” of the three, id. (quoting Steel Co. v. Citizens for a Better Env’t, 523\\n\\nU.S. 83, 103 (1998))—that is disputed in this appeal and to which we now turn.\\n\\n To establish standing, an injury in fact must be concrete.3 Id. at 1548. “A\\n\\n‘concrete’ injury must be ‘de facto’; that is, it must actually exist,” as opposed to\\n\\nbeing hypothetical or speculative. Id. A concrete injury need be only an\\n\\n“identifiable trifle.” United States v. Students Challenging Regulatory Agency\\n\\nProcedures (SCRAP), 412 U.S. 669, 689 n.14 (1973) (noting that sufficiently\\n\\nconcrete injuries have included a fraction of a vote, a $5 fine and costs, and a $1.50\\n\\npoll tax). But sometimes plaintiffs allege intangible injuries that we cannot so\\n\\n\\n\\n3\\n An injury in fact must also be particularized, that is, affecting the plaintiff “in a personal and\\nindividual way.” Spokeo, 136 S. Ct at 1548 (quoting Lujan, 504 U.S. at 560 n.1). It is undisputed\\nthat Salcedo’s allegations are of a personal and individual nature. As the would-be class\\nrepresentative, Salcedo must establish his own personal, concrete injury notwithstanding\\nwhatever injuries may have been suffered by the other members of the class. Id. at 1547 n.6.\\n 6\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 7 of 22\\n\\n\\neasily identify.\\n\\n When the concreteness of an alleged injury is difficult to recognize, we look\\n\\nto “history and the judgment of Congress” for guidance. Spokeo, 136 S. Ct. at\\n\\n1549. But an act of Congress that creates a statutory right and a private right of\\n\\naction to sue does not automatically create standing; “Article III standing requires\\n\\na concrete injury even in the context of a statutory violation.” Id. 4 “[T]he\\n\\nrequirement of injury in fact is a hard floor of Article III jurisdiction that cannot be\\n\\nremoved by statute.” Summers v. Earth Island Inst., 555 U.S. 488, 497 (2009).\\n\\n C. Eleventh Circuit Precedent\\n\\n Because Salcedo bears the burden of establishing federal jurisdiction, Lujan,\\n\\n504 U.S. at 561, we look to the substance of his amended complaint’s allegations\\n\\nto determine if he has standing due to a concrete injury. Salcedo alleged that\\n\\nreceiving the one text message “caused Plaintiff to waste his time answering or\\n\\notherwise addressing the message. While doing so, both Plaintiff and his cellular\\n\\n\\n\\n4\\n Recognizing that “a bare procedural violation” of a statute “may result in no harm,” Spokeo\\nreaffirms the proposition that we must always look for concrete harm when assessing Article III\\nstanding. See 136 S. Ct. at 1550. In some contexts this will mean identifying purely speculative\\n“harm” that never actually materializes as failing to allege an injury in fact. See, e.g., Nicklaw v.\\nCitiMortgage, Inc., 839 F.3d 998, 1003 (11th Cir. 2016), reh’g en banc denied, 855 F.3d 1265\\n(11th Cir. 2017) (holding that a violation of the mortgage satisfaction reporting requirements of a\\nstate law resulted in no concrete harm to the plaintiff).\\n But we should not ignore the Supreme Court’s guidance in Spokeo in cases that purport\\nto allege more than merely technical statutory violations. This appeal presents a close question in\\nwhich we must determine whether Salcedo’s allegations are real and concrete as opposed to\\nfigmentary. Spokeo’s instruction to consider history and the judgment of Congress, id. at 1549,\\nhelpfully guides us in our conclusion that Salcedo has not alleged a concrete injury in fact.\\n 7\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 8 of 22\\n\\n\\nphone were unavailable for otherwise legitimate pursuits.” He further alleged that\\n\\nthe message also “resulted in an invasion of Plaintiff’s privacy and right to enjoy\\n\\nthe full utility of his cellular device.”\\n\\n These allegations are qualitatively different from those in our Circuit\\n\\nprecedent that have been successful in establishing standing to sue over a single\\n\\nviolation of the TCPA. In Palm Beach Golf Center–Boca, Inc. v. John G. Sarris,\\n\\nD.D.S., P.A., 781 F.3d 1245, 1252 (11th Cir. 2015), we found standing for a\\n\\nplaintiff who alleged that receiving a junk fax in violation of the TCPA harmed\\n\\nhim because, during the minute or so that it took to receive and process the fax\\n\\nmessage, his fax machine was unavailable for receiving legitimate business\\n\\nmessages. Accord Florence Endocrine Clinic, PLLC v. Arriva Med., LLC, 858\\n\\nF.3d 1362, 1366 (11th Cir. 2017) (considering also “the cost of printing the\\n\\nunsolicited fax”). To the extent we have relied on tangible costs such as the\\n\\nconsumption of paper and ink or toner to establish injury in fact, Salcedo cannot so\\n\\nrely, since receiving a text message uses no paper, ink, or toner. His complaint\\n\\nalleges generally that some text messages cause recipients to incur costs to their\\n\\nwireless service providers, but he has not alleged specifically that Hanna’s text cost\\n\\nhim any money.\\n\\n Salcedo’s allegations of intangible costs, on the other hand, bear some facial\\n\\nsimilarities to those in Palm Beach Golf. But they differ in kind, rendering Palm\\n\\n\\n 8\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 9 of 22\\n\\n\\nBeach Golf inapplicable. At oral argument, Salcedo asserted that receiving\\n\\nHanna’s message was comparable to using a minute of fax machine time, but his\\n\\ncomplaint does not so allege. Rather, it alleges time wasted only generally. In the\\n\\nabsence of a specific time allegation, we decline to assume an equivalence to the\\n\\nfacts of Palm Beach Golf when receiving a fax message is qualitatively different\\n\\nfrom receiving a text message. A fax message consumes the receiving device\\n\\nentirely, while a text message consumes the receiving device not at all. A cell\\n\\nphone user can continue to use all of the device’s functions, including receiving\\n\\nother messages, while it is receiving a text message.\\n\\n Salcedo also makes an allegation about unavailability, but that too is distinct\\n\\nfrom Palm Beach Golf. There, we were concerned about the fully realized\\n\\nopportunity cost of being unable to receive other faxes for a full minute. By\\n\\ncontrast, Salcedo has alleged no particular loss of opportunity. A fax machine’s\\n\\ninability to receive another message while processing a junk fax has no analogy\\n\\nwith cell phones and text messaging. Salcedo’s assertion that he and his phone\\n\\nwere unavailable appears only to recite language we used in Palm Beach Golf. Cf.\\n\\n781 F.3d at 1252 (quoting H.R. Rep. 102-317, at 10 (1991)). We are entitled to\\n\\nlook past this conclusory recitation to the actual factual substance of Salcedo’s\\n\\nallegations. See Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (“we are not bound to\\n\\naccept as true a legal conclusion couched as a factual allegation”).\\n\\n\\n 9\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 10 of 22\\n\\n\\n Thus, Circuit precedent in Palm Beach Golf does not dictate the outcome of\\n\\nthis appeal. And, for reasons we will discuss below, we find our sister circuit’s\\n\\ndecision involving this precise issue unpersuasive. See Van Patten v. Vertical\\n\\nFitness Group, LLC, 847 F.3d 1037, 1043 (9th Cir. 2017) (holding that the receipt\\n\\nof two unsolicited text messages constituted an injury in fact).5 In the absence of\\n\\ncontrolling authority, we turn our analysis to the framework outlined by the\\n\\nSupreme Court in Spokeo. We look to history and the judgment of Congress to see\\n\\nwhether they support treating Salcedo’s allegations as a concrete injury in fact. Our\\n\\nexamination reveals little support for so doing.\\n\\n D. The Judgment of Congress\\n\\n We first note what Congress has said in the TCPA’s provisions and findings 6\\n\\nabout harms from telemarketing via text message generally: nothing. The TCPA is\\n\\ncompletely silent on the subject of unsolicited text messages. Of course, text\\n\\nmessaging in its current form did not exist in 1991 when the TCPA was enacted,\\n\\n\\n5\\n Nor are we bound by the Supreme Court’s holding in another TCPA text-messaging case that\\nthe case was not mooted by an unaccepted settlement offer. See Campbell–Ewald, 136 S. Ct. at\\n670. The Court did not reach the unraised question of whether the plaintiff had alleged an injury\\nin fact, in part because the defendant apparently never asserted that the plaintiff had failed to do\\nso. See id. at 667–68. “[W]e are not bound by a prior decision’s sub silentio treatment of a\\njurisdictional question.” Okongwu v. Reno, 229 F.3d 1327, 1330 (11th Cir. 2000).\\n6\\n Context matters. We are not suggesting that legislative history should play a role in statutory\\ninterpretation. Salcedo’s allegation is undisputedly a violation of the statute as interpreted by the\\nFCC. Nonetheless, because the Supreme Court has instructed us to consider “the judgment of\\nCongress” in assessing Article III standing, we will consider the congressionally enacted\\nfindings as informative of that judgment. See Spokeo, 136 S. Ct. at 1549; cf. Palm Beach Golf,\\n781 F.3d at 1252 (citing House committee report).\\n 10\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 11 of 22\\n\\n\\nbut Congress has amended the statute several times since then without adding text\\n\\nmessaging to the categories of restricted telemarketing.7 As we have mentioned, it\\n\\nis only through the rulemaking authority of the FCC that the voice call provisions\\n\\nof the TCPA have been extended to text messages. At most, we could take\\n\\nCongress’s silence as tacit approval of that agency action.\\n\\n On the other hand, Congress’s legislative findings about telemarketing\\n\\nsuggest that the receipt of a single text message is qualitatively different from the\\n\\nkinds of things Congress was concerned about when it enacted the TCPA. In\\n\\nparticular, the findings in the TCPA show a concern for privacy within the sanctity\\n\\nof the home that do not necessarily apply to text messaging. “Unrestricted\\n\\ntelemarketing . . . can be an intrusive invasion of privacy,” and “[m]any consumers\\n\\nare outraged over the proliferation of intrusive, nuisance calls to their homes from\\n\\ntelemarketers,” Congress found. Pub. L. No. 102-243, § 2, ¶¶ 5, 6. By contrast, cell\\n\\nphones are often taken outside of the home and often have their ringers silenced,\\n\\npresenting less potential for nuisance and home intrusion. It is thus not surprising\\n\\nthat, after Congress found that the FCC “should have the flexibility to design\\n\\ndifferent rules for those types of automated or prerecorded calls that it finds are not\\n\\n\\n\\n\\n7\\n Following recent amendments, however, the TCPA will expressly include text messaging in its\\nprohibitions on transmitting false caller ID information. Consolidated Appropriations Act, 2018,\\nH.R. 1625, 115th Cong., Pub. L. No. 115-141, div. P, § 503(a) (2018) (to be codified at 47\\nU.S.C. § 227(e)).\\n 11\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 12 of 22\\n\\n\\nconsidered a nuisance or invasion of privacy,” id. ¶ 13, within a year it instructed\\n\\nthe FCC that it may exempt “calls to a telephone number assigned to a cellular\\n\\ntelephone service that are not charged to the called party,” 47 U.S.C.\\n\\n§ 227(b)(2)(C).\\n\\n On text messaging generally, then, the judgment of Congress is ambivalent\\n\\nat best; its privacy and nuisance concerns about residential telemarketing are less\\n\\nclearly applicable to text messaging. Any possible deference to the FCC’s\\n\\ninterpretation of the TCPA 8—the source of its application to text messaging—is\\n\\nnot obviously relevant where the Supreme Court has specifically instructed us to\\n\\nconsider the judgment of Congress. And congressional silence is a poor basis for\\n\\nextending federal jurisdiction to new types of harm. We take seriously the silence\\n\\nof that political branch best positioned to assess and articulate new harms from\\n\\nemerging technologies. See Spokeo, 136 S. Ct. at 1549 (citing Lujan, 504 U.S. at\\n\\n578)). With this point of caution in mind, we now turn to the judgment of Congress\\n\\nabout the specific harms that Salcedo has alleged he suffered when he received\\n\\nHanna’s message.\\n\\n\\n\\n8\\n In this case, we need not reach the issue of whether the agency’s interpretation of the statute is\\nentitled to any deference. See Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837,\\n843 (1984) (requiring deference to agency’s interpretation of silent or ambiguous statute);\\nSkidmore v. Swift & Co., 323 U.S. 134, 140 (1944) (allowing court to determine level of\\ndeference in proportion to agency’s demonstration of persuasive reasoning); cf., e.g., Josendis v.\\nWall to Wall Residence Repairs, Inc., 662 F.3d 1292, 1320 (11th Cir. 2011) (rejecting Chevron\\ndeference where the statutory language was clear and unambiguous).\\n 12\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 13 of 22\\n\\n\\n We consider the judgment of Congress when assessing standing because\\n\\n“Congress is well positioned to identify intangible harms that meet minimum\\n\\nArticle III requirements.” Id. As we have mentioned, Congress was concerned\\n\\nabout “intrusive invasion[s] of privacy” into the home when it enacted the TCPA.\\n\\nPub. L. No. 102-243, § 2, ¶ 5. Salcedo argues that the particular privacy interest\\n\\nCongress has identified is “the freedom from unwanted robocalls,” but that\\n\\nobservation is too general. As we have noted, a single unwelcome text message\\n\\nwill not always involve an intrusion into the privacy of the home in the same way\\n\\nthat a voice call to a residential line necessarily does. Certainly, Salcedo has not\\n\\nalleged that he was in his home when he received Hanna’s message. As we have\\n\\nalso noted, the 1992 amendment allowing the FCC to exempt free-to-receive calls\\n\\nto cell phones, 47 U.S.C. § 227(b)(2)(C), suggests less congressional concern\\n\\nabout calls to cell phones. And by nature of their portability and their ability to be\\n\\nsilenced, cell phone calls may involve less of an intrusion than calls to a home\\n\\nphone. We realize that Congress in 1991 could not have foreseen the explosion in\\n\\npersonal cell phone use, the popularity of text messaging, and the near-extinction\\n\\nof the residential telephone line. But Spokeo instructs us to consider the judgment\\n\\nof Congress about the alleged harm, not to imagine what Congress might say about\\n\\na harm it has not actually addressed.\\n\\n We note that our sister circuit has reached the opposite conclusion in this\\n\\n\\n 13\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 14 of 22\\n\\n\\ncontext. See Van Patten, 847 F.3d at 1043. The Ninth Circuit quoted many of these\\n\\nsame findings, further noting Congress’s purpose of “protect[ing] consumers from\\n\\nthe unwanted intrusion and nuisance of unsolicited telemarketing phone calls and\\n\\nfax advertisements.” See Pub. L. No. 102-243, § 2, ¶ 12. But that court stopped\\n\\nshort of examining whether isolated text messages not received at home come\\n\\nwithin that judgment of Congress. Instead, it concluded that “Congress identified\\n\\nunsolicited contact as a concrete harm.” Van Patten, 847 F.3d at 1043. We disagree\\n\\nwith this broad overgeneralization of the judgment of Congress and have focused\\n\\nour own analysis on text messaging specifically.\\n\\n Other stated concerns behind the TCPA are also inapposite to Salcedo’s\\n\\nallegations. The congressional committee found telemarketing by fax problematic\\n\\nin part because “it occupies the recipient’s facsimile machine so that it is\\n\\nunavailable for legitimate business messages while processing and printing the\\n\\njunk fax.” H.R. Rep. 102-317, at 10 (1991), quoted in Palm Beach Golf, 781 F.3d\\n\\nat 1252. As we have noted, such a concern has little application to the\\n\\ninstantaneous receipt of a text message. The judgment of Congress, then, provides\\n\\nlittle support for finding that Salcedo’s allegations state a concrete injury in fact.9\\n\\n\\n\\n\\n9\\n Congress also stated concerns not raised by either party here: concerns for public safety “when\\nan emergency or medical assistance telephone line is seized,” Pub. L. No. 102-243, § 2, ¶ 5; for\\nthe cost borne by consumers who use technology to avoid unwanted calls, id. ¶ 11; and for\\n“commercial freedoms of speech and trade” that telemarketers enjoy, id. ¶ 9.\\n 14\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 15 of 22\\n\\n\\n E. History\\n\\n We now turn to history for guidance, because the case or controversy\\n\\nrequirement of Article III “is grounded in historical practice.” Spokeo, 136 S. Ct. at\\n\\n1549. Thus, “it is instructive to consider whether an alleged intangible harm has a\\n\\nclose relationship to a harm that has traditionally been regarded as providing a\\n\\nbasis for a lawsuit in English or American courts.” Id.\\n\\n With respect to his allegations of invasion of privacy, we look to the\\n\\ngenerally accepted tort of intrusion upon seclusion,10 which creates liability for\\n\\ninvasions of privacy that would be “highly offensive to a reasonable person.”\\n\\nRestatement (Second) of Torts § 652B. The requirement that the interference be\\n\\n“substantial” and “strongly object[ionable]” instructs us that a plaintiff might be\\n\\nable to establish standing where an intrusion on his privacy is objectively serious\\n\\nand universally condemnable. See id. cmt. d (no liability for one, two, or three\\n\\nphone calls; liability “only when the telephone calls are repeated with such\\n\\npersistence and frequency as to amount to a course of hounding the plaintiff”). By\\n\\ncontrast, Salcedo’s allegations fall short of this degree of harm. We do not see this\\n\\ntype of objectively intense interference where the alleged harm is isolated,\\n\\n\\n\\n\\n10\\n Most of the accepted torts generally known as “invasion of privacy” involve privacy in the\\nrather specific sense of one’s right to be free from unwanted publicity. See Restatement (Second)\\nof Torts § 652A. It is only the privacy tort of intrusion upon seclusion, id. § 652B, that bears any\\npossible relationship to Salcedo’s allegations.\\n 15\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 16 of 22\\n\\n\\nmomentary, and ephemeral.\\n\\n The tort of intrusion upon seclusion also requires an intrusion “upon the\\n\\nsolitude or seclusion of another or his private affairs or concerns.” Id. § 652B.\\n\\nAlthough Salcedo argues that his cell phone is part of his private affairs, the\\n\\nRestatement contemplates a different category of intrusion into private affairs,\\n\\nlisting examples including eavesdropping, wiretapping, and looking through one’s\\n\\npersonal documents. See id. cmt. b. Simply sending one text message to a private\\n\\ncell phone is not closely related to the severe kinds of actively intermeddling\\n\\nintrusions that the traditional tort contemplates. Salcedo’s reasoning would equate\\n\\nopening your private mail—a serious intrusion indeed—with mailing you a\\n\\npostcard.\\n\\n With respect to his allegations of nuisance, Salcedo asks us to compare the\\n\\ntraditional torts of trespass and nuisance, but we find them also to be distinct both\\n\\nin kind and in degree. Trespass requires intentionally “enter[ing] land in the\\n\\npossession of the other,” id. § 158(a), and private nuisance is “a nontrespassory\\n\\ninvasion of another’s interest in the private use and enjoyment of land,” id. § 821D.\\n\\nAlthough, as we have noted, Congress was concerned about intrusions into the\\n\\nhome when it enacted the TCPA, Salcedo has alleged no invasion of any interest in\\n\\nreal property here. Furthermore, even in the context of nuisance to real property, in\\n\\nFlorida, “[m]ere disturbance and annoyance as such do not in themselves\\n\\n\\n 16\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 17 of 22\\n\\n\\nnecessarily give rise to an invasion of a legal right.” A & P Food Stores, Inc. v.\\n\\nKornstein, 121 So. 2d 701, 703 (Fla. 3d Dist. Ct. App. 1960). Hanna’s text\\n\\nmessage is thus not closely related to these traditional harms because it is not\\n\\nalleged to have infringed upon Salcedo’s real property, either directly or indirectly.\\n\\n Salcedo also asks us to consider the personal property torts of conversion\\n\\nand trespass to chattel. Conversion is an interference with chattel “which so\\n\\nseriously interferes with the right of another to control it that the actor may justly\\n\\nbe required to pay the other the full value of the chattel.” Restatement (Second) of\\n\\nTorts § 222A. Salcedo’s allegations are nowhere near a complete and permanent\\n\\ndominion over his phone, so recourse to this serious kind of tort is unhelpful. The\\n\\nsame is true for the tort of trespass to chattel, which involves intentionally “using\\n\\n. . . a chattel in the possession of another.” Id. § 217(b). Traditionally, liability\\n\\narises for this kind of trespass only when “the possessor is deprived of the use of\\n\\nthe chattel for a substantial time” or when the trespass harms “the possessor’s\\n\\nmaterially valuable interest in the physical condition, quality, or value of the\\n\\nchattel.” Id. § 218(c) & cmt. e; cf. United States v. Jones, 565 U.S. 400, 426 (2012)\\n\\n(Alito, J., concurring) (“Trespass to chattels has traditionally required a physical\\n\\ntouching of the property.”). Thus, although Salcedo’s allegations here bear a\\n\\npassing resemblance to this kind of historical harm, they differ so significantly in\\n\\ndegree as to undermine his position. History shows that Salcedo’s allegation is\\n\\n\\n 17\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 18 of 22\\n\\n\\nprecisely the kind of fleeting infraction upon personal property that tort law has\\n\\nresisted addressing.\\n\\n We again note that our sister circuit has reached the opposite conclusion. See\\n\\nVan Patten, 847 F.3d at 1043. We decline to adopt its reasoning and instead\\n\\nembrace more fully the Supreme Court’s instruction to look for a “close\\n\\nrelationship” to a traditionally redressable harm. See Spokeo, 136 S. Ct. at 1549\\n\\n(citing Vt. Agency of Nat. Res., 529 U.S. 765, 775–77 (2000) (discussing traditional\\n\\nqui tam law in a case about qui tam relator Article III standing)). The Ninth\\n\\nCircuit’s one-sentence review of history simply asserted, “Actions to remedy\\n\\ndefendants’ invasions of privacy, intrusion upon seclusion, and nuisance have long\\n\\nbeen heard by American courts, and the right of privacy is recognized by most\\n\\nstates.” Van Patten, 847 F.3d at 1043. But as we have more thoroughly explained,\\n\\nan examination of those torts reveals significant differences in the kind and degree\\n\\nof harm they contemplate providing redress for.\\n\\n In sum, we find that history and the judgment of Congress do not support\\n\\nfinding concrete injury in Salcedo’s allegations. 11 Salcedo has not alleged anything\\n\\n\\n11\\n Salcedo urges us to follow the reasoning that allowed us to find standing in Perry v. Cable\\nNews Network, Inc., 854 F.3d 1336, 1340–41 (11th Cir. 2017). The facts here do not permit us to\\ndo so. In Perry, we found standing under the Video Privacy Protection Act, 18 U.S.C. § 2710,\\nfor a plaintiff suing over privacy violations involving a mobile video app. Perry held that the\\nplaintiff’s allegations and the 1980s-era statute involved precisely the same substantive privacy\\nright. Not so here. As we have discussed, both the judgment of Congress and history here reveal\\nconcerns about intrusions into the privacy of the home and interferences with property that do\\nnot readily transfer to the context of cell phones.\\n 18\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 19 of 22\\n\\n\\nlike enjoying dinner at home with his family and having the domestic peace\\n\\nshattered by the ringing of the telephone. Nor has he alleged that his cell phone\\n\\nwas searched, dispossessed, or seized for any length of time. Salcedo’s allegations\\n\\nof a brief, inconsequential annoyance are categorically distinct from those kinds of\\n\\nreal but intangible harms. The chirp, buzz, or blink of a cell phone receiving a\\n\\nsingle text message is more akin to walking down a busy sidewalk and having a\\n\\nflyer briefly waived in one’s face. Annoying, perhaps, but not a basis for invoking\\n\\nthe jurisdiction of the federal courts. All told, we conclude that Salcedo’s\\n\\nallegations do not state a concrete harm that meets the injury-in-fact requirement of\\n\\nArticle III.\\n\\n F. Quality, Not Quantity\\n\\n To be clear, we are not attempting to measure how small or large Salcedo’s\\n\\nalleged injury is. Article III standing is not a “You must be this tall to ride”\\n\\nmeasuring stick. “There is no minimum quantitative limit required to show injury;\\n\\nrather, the focus is on the qualitative nature of the injury, regardless of how small\\n\\nthe injury may be.” Saladin v. City of Milledgeville, 812 F.2d 687, 691 (11th Cir.\\n\\n1987). Our assessment today is thus qualitative, not quantitative. We have assessed\\n\\nhow concrete and real the alleged harm is, Spokeo, 136 S. Ct. at 1548, and we have\\n\\nconcluded that it is not the kind of harm that constitutes an injury in fact. Some\\n\\nharms that are intangible and ephemeral may do so, but Salcedo’s allegations of\\n\\n\\n 19\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 20 of 22\\n\\n\\nthe harm he suffered from receiving a single text message do not.\\n\\n To be sure, under our precedent, allegations of wasted time can state a\\n\\nconcrete harm for standing purposes. We have found standing where the harm was,\\n\\nfor example, time wasted traveling to the county registrar’s office, Common\\n\\nCause/Ga. v. Billups, 554 F.3d 1340, 1351 (11th Cir. 2009); and correcting credit\\n\\nreporting errors, Pedro v. Equifax, Inc., 868 F.3d 1275, 1280 (11th Cir. 2017).\\n\\nThese precedents strongly suggest that concrete harm from wasted time requires, at\\n\\nthe very least, more than a few seconds. And on this point the judgment of\\n\\nCongress sheds a final ray of light. The TCPA instructs the FCC to establish\\n\\ntelemarketing standards that include releasing the called party’s line within five\\n\\nseconds of a hang-up, 47 U.S.C. § 227(d)(3)(B), demonstrating that, on the margin,\\n\\nCongress does not view tying up a phone line for five seconds as a serious\\n\\nintrusion.\\n\\n Our responsibility to ensure that plaintiffs allege a real injury in fact requires\\n\\nus to look closely at their allegations in light of the statute, our precedent, history,\\n\\nand the judgment of Congress. Such inquiries will, of course, have differing\\n\\noutcomes depending on those inputs. Compare, e.g., Perry, 854 F.3d at 1340–41\\n\\n(finding standing based on intangible harm of statutory violation), and Palm Beach\\n\\nGolf, 781 F.3d at 1252 (same), with Nicklaw v. CitiMortgage, Inc., 839 F.3d 998,\\n\\n1003 (11th Cir. 2016), reh’g denied, 855 F.3d 1265 (11th Cir. 2017) (en banc)\\n\\n\\n 20\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 21 of 22\\n\\n\\n(finding no standing because plaintiff alleged “neither a harm nor a material risk of\\n\\nharm”). We acknowledge that Congress, as a political entity, is well positioned to\\n\\nassess new harms in light of developments in technology and society, and to\\n\\nrespond to the concerns of the American people about novel encroachments on life,\\n\\nliberty, and property. See Spokeo, 136 S. Ct. at 1549. The federal courts are not\\n\\nsimilarly tasked. We have only the power “to say what the law is.” Marbury v.\\n\\nMadison, 5 U.S. (1 Cranch) 137, 177 (1803). And when a plaintiff comes to us\\n\\nwithout alleging a concrete harm, a real injury that states a case or controversy, we\\n\\ncannot do even that much.\\n\\n IV. CONCLUSION\\n\\n We REVERSE the decision of the district court that Salcedo has standing to\\n\\nsue and REMAND with instructions to dismiss without prejudice the amended\\n\\ncomplaint.12\\n\\n\\n\\n\\n12\\n Hanna has asked us to instruct the district court to dismiss Salcedo’s amended complaint with\\nprejudice. But a jurisdictional dismissal is entered without prejudice. Stalley ex rel. United States\\nv. Orlando Reg’l Healthcare Sys., Inc., 524 F.3d 1229, 1232 (11th Cir. 2008). “A dismissal\\n‘without prejudice’ refers to the fact that the dismissal is not on the merits.” Grayson v. K Mart\\nCorp., 79 F.3d 1086, 1094 n.7 (11th Cir. 1996). Although refiling may prove futile (Salcedo has\\nalready amended his complaint once in attempt to shore up his allegations), we and the district\\ncourt presently lack jurisdiction to make that merits determination.\\n 21\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 22 of 22\\n\\n\\nJILL PRYOR, concurring in judgment only:\\n\\n Plaintiff John Salcedo sued defendants Alex Hanna and the Law Offices of\\n\\nAlex Hanna, P.A. (together, “Hanna”) under the Telephone Consumer Protection\\n\\nAct of 1991 (“TCPA”), 47 U.S.C. § 227(b)(1)(A)(iii), after they sent him a single\\n\\nunwanted text message advertisement. I agree with the majority opinion that we\\n\\nlack subject matter jurisdiction because Salcedo has no standing to bring a TCPA\\n\\nclaim. I write separately to emphasize my understanding that the majority’s\\n\\nholding is narrow and the conclusion that Salcedo lacks standing is driven by the\\n\\nallegations in his complaint that Hanna sent him only one text message. The\\n\\nmajority opinion—appropriately, in my view—leaves unaddressed whether a\\n\\nplaintiff who alleged that he had received multiple unwanted and unsolicited text\\n\\nmessages may have standing to sue under the TCPA. With this understanding, I\\n\\nconcur in the majority’s judgment.\\n\\n\\n\\n\\n 22\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434046/\", \"author_raw\": \"BRANCH, Circuit Judge:\"}]}"},"judge1":{"kind":"string","value":"JILL PRYOR"},"judge2":{"kind":"string","value":"BRANCH"},"judge3":{"kind":"string","value":"REEVES"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"JILL PRYOR\": \", Circuit\", \"BRANCH\": \", Circuit\", \"REEVES\": \", District\"}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4656793/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: John Salcedo v. Alex Hanna\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-28\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nCase: 17-14077 Date Filed: 08/28/2019 Page: 1 of 22\n\n\n [PUBLISH]\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n\n\n\n No. 17-14077\n\n\n\n D.C. Docket No. 0:16-cv-62480-DPG\n\nJOHN SALCEDO,\nindividually and on behalf of others similarly situated,\n\n Plaintiff - Appellee,\n\n versus\n\nALEX HANNA,\nan individual,\nTHE LAW OFFICES OF ALEX HANNA, P.A.,\na Florida Professional Association,\n\n Defendants - Appellants.\n\n\n\n Appeal from the United States District Court\n for the Southern District of Florida\n\n\n\n (August 28, 2019)\n\nBefore JILL PRYOR and BRANCH, Circuit Judges, and REEVES, * District\nJudge.\n\n*\n Honorable Danny C. Reeves, United States District Judge for the Eastern District of Kentucky,\nsitting by designation.\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 2 of 22\n\n\nBRANCH, Circuit Judge:\n\n Is receiving a single unsolicited text message, sent in violation of a federal\n\nstatute, a concrete injury in fact that establishes standing to sue in federal court? To\n\nanswer that question, we have examined the statute, our precedent, and—following\n\nthe Supreme Court’s guidance—history and the judgment of Congress, and we\n\nconclude that the allegations in this suit do not establish standing.\n\n I. BACKGROUND\n\n At 9:56 a.m. on August 12, 2016, John Salcedo, a former client of Florida\n\nattorney Alex Hanna and his law firm, 1 received a multimedia text message from\n\nHanna offering a ten percent discount on his services.\n\n Salcedo filed suit in the district court as the representative of a putative class\n\nof former Hanna clients who received unsolicited text messages from Hanna in the\n\npast four years, alleging violations of the Telephone Consumer Protection Act of\n\n1991 (“TCPA”), 47 U.S.C. § 227(b)(1)(A)(iii).2 He sought, among other relief,\n\nstatutory damages of $500 per text message and treble damages of $1,500 per text\n\n\n\n1\n For simplicity, and without implying any view as to Mr. Hanna’s possible personal liability,\nthroughout this opinion we will refer to both defendants—Mr. Hanna and his law firm—\ncollectively as “Hanna.”\n2\n “It shall be unlawful for any person within the United States . . . to make any call (other than a\ncall made for emergency purposes or made with the prior express consent of the called party)\nusing any automatic telephone dialing system or an artificial or prerecorded voice . . . to any\ntelephone number assigned to a . . . cellular telephone service . . . .” 47 U.S.C.\n§ 227(b)(1)(A)(iii).\n 2\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 3 of 22\n\n\nmessage sent willfully or knowingly. See 47 U.S.C. § 227(b)(3).\n\n Hanna moved to dismiss the complaint for lack of standing, arguing in the\n\nalternative that it should be dismissed as to Mr. Hanna for failure to state a claim\n\nagainst him and that certain parts of the complaint should be stricken. The district\n\ncourt disagreed, finding in relevant part that Salcedo had standing under Mohamed\n\nv. Off Lease Only, Inc., No. 15-23352-Civ-COOKE/TORRES, 2017 WL 1080342\n\n(S.D. Fla. Mar. 22, 2017). However, finding that its order “involves a controlling\n\nquestion of law as to which there is a substantial ground for difference of opinion,”\n\nthe court allowed Salcedo to pursue an interlocutory appeal and stayed its\n\nproceedings pending appeal. A panel of our Court granted Hanna’s petition for\n\npermission to appeal under 28 U.S.C. § 1292(b). We now consider his appeal.\n\n II. STANDARD OF REVIEW\n\n “We review standing determinations de novo.” Bochese v. Town of Ponce\n\nInlet, 405 F.3d 964, 975 (11th Cir. 2005).\n\n III. DISCUSSION\n\n Our analysis proceeds as follows. We first introduce the TCPA, the statute\n\nunder which Salcedo has filed suit. Next, we discuss the standing requirements of\n\nArticle III of the Constitution, which help to define our limited power to resolve\n\nonly cases or controversies. We then turn to Salcedo’s particular allegations of\n\nharm and analyze them in view of our Circuit precedent, history, and the judgment\n\n\n 3\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 4 of 22\n\n\nof Congress.\n\n A. The Telephone Consumer Privacy Act of 1991\n\n Because it found that “residential telephone subscribers consider automated\n\nor prerecorded telephone calls . . . to be a nuisance and an invasion of privacy,”\n\nTelephone Consumer Protection Act of 1991, S. 1462, 102d Cong., Pub. L. No.\n\n102-243, § 2, ¶ 10 (1991), in 1991 Congress enacted the TCPA to restrict interstate\n\ntelemarketing. The TCPA thus prohibits using automatic telephone dialing systems\n\nto call residential or cellular telephone lines without the consent of the called party.\n\n47 U.S.C. § 227(b)(1)(A)(iii), (B). It also prohibits sending unsolicited\n\nadvertisements via facsimile machine. Id. § 227(b)(1)(C). It authorizes the Federal\n\nCommunications Commission (“FCC”) to enact implementing regulations. Id.\n\n§ 227(b)(2). Finally for our purposes, the TCPA creates a private right of action\n\nwhereby a person or entity may seek compensatory or injunctive relief against\n\nviolators. Id. § 227(b)(3).\n\n There have been two relevant updates to the TCPA and its enforcement\n\nregime since 1991. First, in October 1992, Congress amended the TCPA to allow\n\nthe FCC to exempt free-to-receive cellular calls if it so chooses. Id. § 227(b)(2)(C).\n\nThe FCC has not done so. Second, the statute has been silent as to text messaging,\n\nfor that medium did not exist in 1991. But under its TCPA rulemaking authority,\n\nthe FCC has applied the statute’s regulations of voice calls to text messages. 30\n\n\n 4\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 5 of 22\n\n\nFCC Rcd. 7961, 7964 n.3, 7978–79, 8016–22 (2015); 18 FCC Rcd. 14014, 14115\n\n(2003); see also Campbell–Ewald Co. v. Gomez, 136 S. Ct. 663, 667 (2016) (“A\n\ntext message to a cellular telephone, it is undisputed, qualifies as a ‘call’ within the\n\ncompass of § 227(b)(1)(A)(iii).”). Thus, Salcedo’s complaint facially appears to\n\nstate a cause of action under the TCPA as interpreted by the FCC.\n\n B. Article III Standing\n\n Not every right created by Congress or defined by an executive agency is\n\nautomatically enforceable in the federal courts. Our tripartite system of\n\ngovernment recognizes that “there is no liberty if the power of judging be not\n\nseparated from the legislative and executive powers.” The Federalist No. 78, at 465\n\n(Alexander Hamilton) (Clinton Rossiter ed. 1961). To protect this separation of\n\npowers, we must assure ourselves that our exercise of jurisdiction falls within the\n\nConstitution’s grant of judicial power.\n\n Article III vests the judicial power in the federal courts and extends that\n\npower to “Cases” and “Controversies.” U.S. Const. art. III, §§ 1–2. One tool for\n\ndetermining that the matters before us are truly cases or controversies, as\n\nunderstood by Article III, is the doctrine of standing. Lujan v. Defenders of\n\nWildlife, 504 U.S. 555, 560 (1992). “The law of Article III standing . . . serves to\n\nprevent the judicial process from being used to usurp the powers of the political\n\nbranches.” Clapper v. Amnesty Int’l USA, 568 U.S. 398, 408 (2013). Even when\n\n\n 5\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 6 of 22\n\n\nthose political branches appear to have granted us jurisdiction by statute and rule,\n\nwe are still obliged to examine whether jurisdiction exists under the Constitution.\n\n As the Supreme Court has explained, the “irreducible constitutional\n\nminimum” to establish Article III standing requires three elements. Lujan, 504 U.S.\n\nat 560. “The plaintiff must have (1) suffered an injury in fact, (2) that is fairly\n\ntraceable to the challenged conduct of the defendant, and (3) that is likely to be\n\nredressed by a favorable judicial decision.” Spokeo, Inc. v. Robins, 136 S. Ct. 1540,\n\n1547 (2016) (citing Lujan, 504 U.S. at 560–61). It is the first element—the\n\n“foremost” of the three, id. (quoting Steel Co. v. Citizens for a Better Env’t, 523\n\nU.S. 83, 103 (1998))—that is disputed in this appeal and to which we now turn.\n\n To establish standing, an injury in fact must be concrete.3 Id. at 1548. “A\n\n‘concrete’ injury must be ‘de facto’; that is, it must actually exist,” as opposed to\n\nbeing hypothetical or speculative. Id. A concrete injury need be only an\n\n“identifiable trifle.” United States v. Students Challenging Regulatory Agency\n\nProcedures (SCRAP), 412 U.S. 669, 689 n.14 (1973) (noting that sufficiently\n\nconcrete injuries have included a fraction of a vote, a $5 fine and costs, and a $1.50\n\npoll tax). But sometimes plaintiffs allege intangible injuries that we cannot so\n\n\n\n3\n An injury in fact must also be particularized, that is, affecting the plaintiff “in a personal and\nindividual way.” Spokeo, 136 S. Ct at 1548 (quoting Lujan, 504 U.S. at 560 n.1). It is undisputed\nthat Salcedo’s allegations are of a personal and individual nature. As the would-be class\nrepresentative, Salcedo must establish his own personal, concrete injury notwithstanding\nwhatever injuries may have been suffered by the other members of the class. Id. at 1547 n.6.\n 6\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 7 of 22\n\n\neasily identify.\n\n When the concreteness of an alleged injury is difficult to recognize, we look\n\nto “history and the judgment of Congress” for guidance. Spokeo, 136 S. Ct. at\n\n1549. But an act of Congress that creates a statutory right and a private right of\n\naction to sue does not automatically create standing; “Article III standing requires\n\na concrete injury even in the context of a statutory violation.” Id. 4 “[T]he\n\nrequirement of injury in fact is a hard floor of Article III jurisdiction that cannot be\n\nremoved by statute.” Summers v. Earth Island Inst., 555 U.S. 488, 497 (2009).\n\n C. Eleventh Circuit Precedent\n\n Because Salcedo bears the burden of establishing federal jurisdiction, Lujan,\n\n504 U.S. at 561, we look to the substance of his amended complaint’s allegations\n\nto determine if he has standing due to a concrete injury. Salcedo alleged that\n\nreceiving the one text message “caused Plaintiff to waste his time answering or\n\notherwise addressing the message. While doing so, both Plaintiff and his cellular\n\n\n\n4\n Recognizing that “a bare procedural violation” of a statute “may result in no harm,” Spokeo\nreaffirms the proposition that we must always look for concrete harm when assessing Article III\nstanding. See 136 S. Ct. at 1550. In some contexts this will mean identifying purely speculative\n“harm” that never actually materializes as failing to allege an injury in fact. See, e.g., Nicklaw v.\nCitiMortgage, Inc., 839 F.3d 998, 1003 (11th Cir. 2016), reh’g en banc denied, 855 F.3d 1265\n(11th Cir. 2017) (holding that a violation of the mortgage satisfaction reporting requirements of a\nstate law resulted in no concrete harm to the plaintiff).\n But we should not ignore the Supreme Court’s guidance in Spokeo in cases that purport\nto allege more than merely technical statutory violations. This appeal presents a close question in\nwhich we must determine whether Salcedo’s allegations are real and concrete as opposed to\nfigmentary. Spokeo’s instruction to consider history and the judgment of Congress, id. at 1549,\nhelpfully guides us in our conclusion that Salcedo has not alleged a concrete injury in fact.\n 7\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 8 of 22\n\n\nphone were unavailable for otherwise legitimate pursuits.” He further alleged that\n\nthe message also “resulted in an invasion of Plaintiff’s privacy and right to enjoy\n\nthe full utility of his cellular device.”\n\n These allegations are qualitatively different from those in our Circuit\n\nprecedent that have been successful in establishing standing to sue over a single\n\nviolation of the TCPA. In Palm Beach Golf Center–Boca, Inc. v. John G. Sarris,\n\nD.D.S., P.A., 781 F.3d 1245, 1252 (11th Cir. 2015), we found standing for a\n\nplaintiff who alleged that receiving a junk fax in violation of the TCPA harmed\n\nhim because, during the minute or so that it took to receive and process the fax\n\nmessage, his fax machine was unavailable for receiving legitimate business\n\nmessages. Accord Florence Endocrine Clinic, PLLC v. Arriva Med., LLC, 858\n\nF.3d 1362, 1366 (11th Cir. 2017) (considering also “the cost of printing the\n\nunsolicited fax”). To the extent we have relied on tangible costs such as the\n\nconsumption of paper and ink or toner to establish injury in fact, Salcedo cannot so\n\nrely, since receiving a text message uses no paper, ink, or toner. His complaint\n\nalleges generally that some text messages cause recipients to incur costs to their\n\nwireless service providers, but he has not alleged specifically that Hanna’s text cost\n\nhim any money.\n\n Salcedo’s allegations of intangible costs, on the other hand, bear some facial\n\nsimilarities to those in Palm Beach Golf. But they differ in kind, rendering Palm\n\n\n 8\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 9 of 22\n\n\nBeach Golf inapplicable. At oral argument, Salcedo asserted that receiving\n\nHanna’s message was comparable to using a minute of fax machine time, but his\n\ncomplaint does not so allege. Rather, it alleges time wasted only generally. In the\n\nabsence of a specific time allegation, we decline to assume an equivalence to the\n\nfacts of Palm Beach Golf when receiving a fax message is qualitatively different\n\nfrom receiving a text message. A fax message consumes the receiving device\n\nentirely, while a text message consumes the receiving device not at all. A cell\n\nphone user can continue to use all of the device’s functions, including receiving\n\nother messages, while it is receiving a text message.\n\n Salcedo also makes an allegation about unavailability, but that too is distinct\n\nfrom Palm Beach Golf. There, we were concerned about the fully realized\n\nopportunity cost of being unable to receive other faxes for a full minute. By\n\ncontrast, Salcedo has alleged no particular loss of opportunity. A fax machine’s\n\ninability to receive another message while processing a junk fax has no analogy\n\nwith cell phones and text messaging. Salcedo’s assertion that he and his phone\n\nwere unavailable appears only to recite language we used in Palm Beach Golf. Cf.\n\n781 F.3d at 1252 (quoting H.R. Rep. 102-317, at 10 (1991)). We are entitled to\n\nlook past this conclusory recitation to the actual factual substance of Salcedo’s\n\nallegations. See Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (“we are not bound to\n\naccept as true a legal conclusion couched as a factual allegation”).\n\n\n 9\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 10 of 22\n\n\n Thus, Circuit precedent in Palm Beach Golf does not dictate the outcome of\n\nthis appeal. And, for reasons we will discuss below, we find our sister circuit’s\n\ndecision involving this precise issue unpersuasive. See Van Patten v. Vertical\n\nFitness Group, LLC, 847 F.3d 1037, 1043 (9th Cir. 2017) (holding that the receipt\n\nof two unsolicited text messages constituted an injury in fact).5 In the absence of\n\ncontrolling authority, we turn our analysis to the framework outlined by the\n\nSupreme Court in Spokeo. We look to history and the judgment of Congress to see\n\nwhether they support treating Salcedo’s allegations as a concrete injury in fact. Our\n\nexamination reveals little support for so doing.\n\n D. The Judgment of Congress\n\n We first note what Congress has said in the TCPA’s provisions and findings 6\n\nabout harms from telemarketing via text message generally: nothing. The TCPA is\n\ncompletely silent on the subject of unsolicited text messages. Of course, text\n\nmessaging in its current form did not exist in 1991 when the TCPA was enacted,\n\n\n5\n Nor are we bound by the Supreme Court’s holding in another TCPA text-messaging case that\nthe case was not mooted by an unaccepted settlement offer. See Campbell–Ewald, 136 S. Ct. at\n670. The Court did not reach the unraised question of whether the plaintiff had alleged an injury\nin fact, in part because the defendant apparently never asserted that the plaintiff had failed to do\nso. See id. at 667–68. “[W]e are not bound by a prior decision’s sub silentio treatment of a\njurisdictional question.” Okongwu v. Reno, 229 F.3d 1327, 1330 (11th Cir. 2000).\n6\n Context matters. We are not suggesting that legislative history should play a role in statutory\ninterpretation. Salcedo’s allegation is undisputedly a violation of the statute as interpreted by the\nFCC. Nonetheless, because the Supreme Court has instructed us to consider “the judgment of\nCongress” in assessing Article III standing, we will consider the congressionally enacted\nfindings as informative of that judgment. See Spokeo, 136 S. Ct. at 1549; cf. Palm Beach Golf,\n781 F.3d at 1252 (citing House committee report).\n 10\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 11 of 22\n\n\nbut Congress has amended the statute several times since then without adding text\n\nmessaging to the categories of restricted telemarketing.7 As we have mentioned, it\n\nis only through the rulemaking authority of the FCC that the voice call provisions\n\nof the TCPA have been extended to text messages. At most, we could take\n\nCongress’s silence as tacit approval of that agency action.\n\n On the other hand, Congress’s legislative findings about telemarketing\n\nsuggest that the receipt of a single text message is qualitatively different from the\n\nkinds of things Congress was concerned about when it enacted the TCPA. In\n\nparticular, the findings in the TCPA show a concern for privacy within the sanctity\n\nof the home that do not necessarily apply to text messaging. “Unrestricted\n\ntelemarketing . . . can be an intrusive invasion of privacy,” and “[m]any consumers\n\nare outraged over the proliferation of intrusive, nuisance calls to their homes from\n\ntelemarketers,” Congress found. Pub. L. No. 102-243, § 2, ¶¶ 5, 6. By contrast, cell\n\nphones are often taken outside of the home and often have their ringers silenced,\n\npresenting less potential for nuisance and home intrusion. It is thus not surprising\n\nthat, after Congress found that the FCC “should have the flexibility to design\n\ndifferent rules for those types of automated or prerecorded calls that it finds are not\n\n\n\n\n7\n Following recent amendments, however, the TCPA will expressly include text messaging in its\nprohibitions on transmitting false caller ID information. Consolidated Appropriations Act, 2018,\nH.R. 1625, 115th Cong., Pub. L. No. 115-141, div. P, § 503(a) (2018) (to be codified at 47\nU.S.C. § 227(e)).\n 11\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 12 of 22\n\n\nconsidered a nuisance or invasion of privacy,” id. ¶ 13, within a year it instructed\n\nthe FCC that it may exempt “calls to a telephone number assigned to a cellular\n\ntelephone service that are not charged to the called party,” 47 U.S.C.\n\n§ 227(b)(2)(C).\n\n On text messaging generally, then, the judgment of Congress is ambivalent\n\nat best; its privacy and nuisance concerns about residential telemarketing are less\n\nclearly applicable to text messaging. Any possible deference to the FCC’s\n\ninterpretation of the TCPA 8—the source of its application to text messaging—is\n\nnot obviously relevant where the Supreme Court has specifically instructed us to\n\nconsider the judgment of Congress. And congressional silence is a poor basis for\n\nextending federal jurisdiction to new types of harm. We take seriously the silence\n\nof that political branch best positioned to assess and articulate new harms from\n\nemerging technologies. See Spokeo, 136 S. Ct. at 1549 (citing Lujan, 504 U.S. at\n\n578)). With this point of caution in mind, we now turn to the judgment of Congress\n\nabout the specific harms that Salcedo has alleged he suffered when he received\n\nHanna’s message.\n\n\n\n8\n In this case, we need not reach the issue of whether the agency’s interpretation of the statute is\nentitled to any deference. See Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837,\n843 (1984) (requiring deference to agency’s interpretation of silent or ambiguous statute);\nSkidmore v. Swift & Co., 323 U.S. 134, 140 (1944) (allowing court to determine level of\ndeference in proportion to agency’s demonstration of persuasive reasoning); cf., e.g., Josendis v.\nWall to Wall Residence Repairs, Inc., 662 F.3d 1292, 1320 (11th Cir. 2011) (rejecting Chevron\ndeference where the statutory language was clear and unambiguous).\n 12\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 13 of 22\n\n\n We consider the judgment of Congress when assessing standing because\n\n“Congress is well positioned to identify intangible harms that meet minimum\n\nArticle III requirements.” Id. As we have mentioned, Congress was concerned\n\nabout “intrusive invasion[s] of privacy” into the home when it enacted the TCPA.\n\nPub. L. No. 102-243, § 2, ¶ 5. Salcedo argues that the particular privacy interest\n\nCongress has identified is “the freedom from unwanted robocalls,” but that\n\nobservation is too general. As we have noted, a single unwelcome text message\n\nwill not always involve an intrusion into the privacy of the home in the same way\n\nthat a voice call to a residential line necessarily does. Certainly, Salcedo has not\n\nalleged that he was in his home when he received Hanna’s message. As we have\n\nalso noted, the 1992 amendment allowing the FCC to exempt free-to-receive calls\n\nto cell phones, 47 U.S.C. § 227(b)(2)(C), suggests less congressional concern\n\nabout calls to cell phones. And by nature of their portability and their ability to be\n\nsilenced, cell phone calls may involve less of an intrusion than calls to a home\n\nphone. We realize that Congress in 1991 could not have foreseen the explosion in\n\npersonal cell phone use, the popularity of text messaging, and the near-extinction\n\nof the residential telephone line. But Spokeo instructs us to consider the judgment\n\nof Congress about the alleged harm, not to imagine what Congress might say about\n\na harm it has not actually addressed.\n\n We note that our sister circuit has reached the opposite conclusion in this\n\n\n 13\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 14 of 22\n\n\ncontext. See Van Patten, 847 F.3d at 1043. The Ninth Circuit quoted many of these\n\nsame findings, further noting Congress’s purpose of “protect[ing] consumers from\n\nthe unwanted intrusion and nuisance of unsolicited telemarketing phone calls and\n\nfax advertisements.” See Pub. L. No. 102-243, § 2, ¶ 12. But that court stopped\n\nshort of examining whether isolated text messages not received at home come\n\nwithin that judgment of Congress. Instead, it concluded that “Congress identified\n\nunsolicited contact as a concrete harm.” Van Patten, 847 F.3d at 1043. We disagree\n\nwith this broad overgeneralization of the judgment of Congress and have focused\n\nour own analysis on text messaging specifically.\n\n Other stated concerns behind the TCPA are also inapposite to Salcedo’s\n\nallegations. The congressional committee found telemarketing by fax problematic\n\nin part because “it occupies the recipient’s facsimile machine so that it is\n\nunavailable for legitimate business messages while processing and printing the\n\njunk fax.” H.R. Rep. 102-317, at 10 (1991), quoted in Palm Beach Golf, 781 F.3d\n\nat 1252. As we have noted, such a concern has little application to the\n\ninstantaneous receipt of a text message. The judgment of Congress, then, provides\n\nlittle support for finding that Salcedo’s allegations state a concrete injury in fact.9\n\n\n\n\n9\n Congress also stated concerns not raised by either party here: concerns for public safety “when\nan emergency or medical assistance telephone line is seized,” Pub. L. No. 102-243, § 2, ¶ 5; for\nthe cost borne by consumers who use technology to avoid unwanted calls, id. ¶ 11; and for\n“commercial freedoms of speech and trade” that telemarketers enjoy, id. ¶ 9.\n 14\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 15 of 22\n\n\n E. History\n\n We now turn to history for guidance, because the case or controversy\n\nrequirement of Article III “is grounded in historical practice.” Spokeo, 136 S. Ct. at\n\n1549. Thus, “it is instructive to consider whether an alleged intangible harm has a\n\nclose relationship to a harm that has traditionally been regarded as providing a\n\nbasis for a lawsuit in English or American courts.” Id.\n\n With respect to his allegations of invasion of privacy, we look to the\n\ngenerally accepted tort of intrusion upon seclusion,10 which creates liability for\n\ninvasions of privacy that would be “highly offensive to a reasonable person.”\n\nRestatement (Second) of Torts § 652B. The requirement that the interference be\n\n“substantial” and “strongly object[ionable]” instructs us that a plaintiff might be\n\nable to establish standing where an intrusion on his privacy is objectively serious\n\nand universally condemnable. See id. cmt. d (no liability for one, two, or three\n\nphone calls; liability “only when the telephone calls are repeated with such\n\npersistence and frequency as to amount to a course of hounding the plaintiff”). By\n\ncontrast, Salcedo’s allegations fall short of this degree of harm. We do not see this\n\ntype of objectively intense interference where the alleged harm is isolated,\n\n\n\n\n10\n Most of the accepted torts generally known as “invasion of privacy” involve privacy in the\nrather specific sense of one’s right to be free from unwanted publicity. See Restatement (Second)\nof Torts § 652A. It is only the privacy tort of intrusion upon seclusion, id. § 652B, that bears any\npossible relationship to Salcedo’s allegations.\n 15\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 16 of 22\n\n\nmomentary, and ephemeral.\n\n The tort of intrusion upon seclusion also requires an intrusion “upon the\n\nsolitude or seclusion of another or his private affairs or concerns.” Id. § 652B.\n\nAlthough Salcedo argues that his cell phone is part of his private affairs, the\n\nRestatement contemplates a different category of intrusion into private affairs,\n\nlisting examples including eavesdropping, wiretapping, and looking through one’s\n\npersonal documents. See id. cmt. b. Simply sending one text message to a private\n\ncell phone is not closely related to the severe kinds of actively intermeddling\n\nintrusions that the traditional tort contemplates. Salcedo’s reasoning would equate\n\nopening your private mail—a serious intrusion indeed—with mailing you a\n\npostcard.\n\n With respect to his allegations of nuisance, Salcedo asks us to compare the\n\ntraditional torts of trespass and nuisance, but we find them also to be distinct both\n\nin kind and in degree. Trespass requires intentionally “enter[ing] land in the\n\npossession of the other,” id. § 158(a), and private nuisance is “a nontrespassory\n\ninvasion of another’s interest in the private use and enjoyment of land,” id. § 821D.\n\nAlthough, as we have noted, Congress was concerned about intrusions into the\n\nhome when it enacted the TCPA, Salcedo has alleged no invasion of any interest in\n\nreal property here. Furthermore, even in the context of nuisance to real property, in\n\nFlorida, “[m]ere disturbance and annoyance as such do not in themselves\n\n\n 16\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 17 of 22\n\n\nnecessarily give rise to an invasion of a legal right.” A & P Food Stores, Inc. v.\n\nKornstein, 121 So. 2d 701, 703 (Fla. 3d Dist. Ct. App. 1960). Hanna’s text\n\nmessage is thus not closely related to these traditional harms because it is not\n\nalleged to have infringed upon Salcedo’s real property, either directly or indirectly.\n\n Salcedo also asks us to consider the personal property torts of conversion\n\nand trespass to chattel. Conversion is an interference with chattel “which so\n\nseriously interferes with the right of another to control it that the actor may justly\n\nbe required to pay the other the full value of the chattel.” Restatement (Second) of\n\nTorts § 222A. Salcedo’s allegations are nowhere near a complete and permanent\n\ndominion over his phone, so recourse to this serious kind of tort is unhelpful. The\n\nsame is true for the tort of trespass to chattel, which involves intentionally “using\n\n. . . a chattel in the possession of another.” Id. § 217(b). Traditionally, liability\n\narises for this kind of trespass only when “the possessor is deprived of the use of\n\nthe chattel for a substantial time” or when the trespass harms “the possessor’s\n\nmaterially valuable interest in the physical condition, quality, or value of the\n\nchattel.” Id. § 218(c) & cmt. e; cf. United States v. Jones, 565 U.S. 400, 426 (2012)\n\n(Alito, J., concurring) (“Trespass to chattels has traditionally required a physical\n\ntouching of the property.”). Thus, although Salcedo’s allegations here bear a\n\npassing resemblance to this kind of historical harm, they differ so significantly in\n\ndegree as to undermine his position. History shows that Salcedo’s allegation is\n\n\n 17\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 18 of 22\n\n\nprecisely the kind of fleeting infraction upon personal property that tort law has\n\nresisted addressing.\n\n We again note that our sister circuit has reached the opposite conclusion. See\n\nVan Patten, 847 F.3d at 1043. We decline to adopt its reasoning and instead\n\nembrace more fully the Supreme Court’s instruction to look for a “close\n\nrelationship” to a traditionally redressable harm. See Spokeo, 136 S. Ct. at 1549\n\n(citing Vt. Agency of Nat. Res., 529 U.S. 765, 775–77 (2000) (discussing traditional\n\nqui tam law in a case about qui tam relator Article III standing)). The Ninth\n\nCircuit’s one-sentence review of history simply asserted, “Actions to remedy\n\ndefendants’ invasions of privacy, intrusion upon seclusion, and nuisance have long\n\nbeen heard by American courts, and the right of privacy is recognized by most\n\nstates.” Van Patten, 847 F.3d at 1043. But as we have more thoroughly explained,\n\nan examination of those torts reveals significant differences in the kind and degree\n\nof harm they contemplate providing redress for.\n\n In sum, we find that history and the judgment of Congress do not support\n\nfinding concrete injury in Salcedo’s allegations. 11 Salcedo has not alleged anything\n\n\n11\n Salcedo urges us to follow the reasoning that allowed us to find standing in Perry v. Cable\nNews Network, Inc., 854 F.3d 1336, 1340–41 (11th Cir. 2017). The facts here do not permit us to\ndo so. In Perry, we found standing under the Video Privacy Protection Act, 18 U.S.C. § 2710,\nfor a plaintiff suing over privacy violations involving a mobile video app. Perry held that the\nplaintiff’s allegations and the 1980s-era statute involved precisely the same substantive privacy\nright. Not so here. As we have discussed, both the judgment of Congress and history here reveal\nconcerns about intrusions into the privacy of the home and interferences with property that do\nnot readily transfer to the context of cell phones.\n 18\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 19 of 22\n\n\nlike enjoying dinner at home with his family and having the domestic peace\n\nshattered by the ringing of the telephone. Nor has he alleged that his cell phone\n\nwas searched, dispossessed, or seized for any length of time. Salcedo’s allegations\n\nof a brief, inconsequential annoyance are categorically distinct from those kinds of\n\nreal but intangible harms. The chirp, buzz, or blink of a cell phone receiving a\n\nsingle text message is more akin to walking down a busy sidewalk and having a\n\nflyer briefly waived in one’s face. Annoying, perhaps, but not a basis for invoking\n\nthe jurisdiction of the federal courts. All told, we conclude that Salcedo’s\n\nallegations do not state a concrete harm that meets the injury-in-fact requirement of\n\nArticle III.\n\n F. Quality, Not Quantity\n\n To be clear, we are not attempting to measure how small or large Salcedo’s\n\nalleged injury is. Article III standing is not a “You must be this tall to ride”\n\nmeasuring stick. “There is no minimum quantitative limit required to show injury;\n\nrather, the focus is on the qualitative nature of the injury, regardless of how small\n\nthe injury may be.” Saladin v. City of Milledgeville, 812 F.2d 687, 691 (11th Cir.\n\n1987). Our assessment today is thus qualitative, not quantitative. We have assessed\n\nhow concrete and real the alleged harm is, Spokeo, 136 S. Ct. at 1548, and we have\n\nconcluded that it is not the kind of harm that constitutes an injury in fact. Some\n\nharms that are intangible and ephemeral may do so, but Salcedo’s allegations of\n\n\n 19\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 20 of 22\n\n\nthe harm he suffered from receiving a single text message do not.\n\n To be sure, under our precedent, allegations of wasted time can state a\n\nconcrete harm for standing purposes. We have found standing where the harm was,\n\nfor example, time wasted traveling to the county registrar’s office, Common\n\nCause/Ga. v. Billups, 554 F.3d 1340, 1351 (11th Cir. 2009); and correcting credit\n\nreporting errors, Pedro v. Equifax, Inc., 868 F.3d 1275, 1280 (11th Cir. 2017).\n\nThese precedents strongly suggest that concrete harm from wasted time requires, at\n\nthe very least, more than a few seconds. And on this point the judgment of\n\nCongress sheds a final ray of light. The TCPA instructs the FCC to establish\n\ntelemarketing standards that include releasing the called party’s line within five\n\nseconds of a hang-up, 47 U.S.C. § 227(d)(3)(B), demonstrating that, on the margin,\n\nCongress does not view tying up a phone line for five seconds as a serious\n\nintrusion.\n\n Our responsibility to ensure that plaintiffs allege a real injury in fact requires\n\nus to look closely at their allegations in light of the statute, our precedent, history,\n\nand the judgment of Congress. Such inquiries will, of course, have differing\n\noutcomes depending on those inputs. Compare, e.g., Perry, 854 F.3d at 1340–41\n\n(finding standing based on intangible harm of statutory violation), and Palm Beach\n\nGolf, 781 F.3d at 1252 (same), with Nicklaw v. CitiMortgage, Inc., 839 F.3d 998,\n\n1003 (11th Cir. 2016), reh’g denied, 855 F.3d 1265 (11th Cir. 2017) (en banc)\n\n\n 20\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 21 of 22\n\n\n(finding no standing because plaintiff alleged “neither a harm nor a material risk of\n\nharm”). We acknowledge that Congress, as a political entity, is well positioned to\n\nassess new harms in light of developments in technology and society, and to\n\nrespond to the concerns of the American people about novel encroachments on life,\n\nliberty, and property. See Spokeo, 136 S. Ct. at 1549. The federal courts are not\n\nsimilarly tasked. We have only the power “to say what the law is.” Marbury v.\n\nMadison, 5 U.S. (1 Cranch) 137, 177 (1803). And when a plaintiff comes to us\n\nwithout alleging a concrete harm, a real injury that states a case or controversy, we\n\ncannot do even that much.\n\n IV. CONCLUSION\n\n We REVERSE the decision of the district court that Salcedo has standing to\n\nsue and REMAND with instructions to dismiss without prejudice the amended\n\ncomplaint.12\n\n\n\n\n12\n Hanna has asked us to instruct the district court to dismiss Salcedo’s amended complaint with\nprejudice. But a jurisdictional dismissal is entered without prejudice. Stalley ex rel. United States\nv. Orlando Reg’l Healthcare Sys., Inc., 524 F.3d 1229, 1232 (11th Cir. 2008). “A dismissal\n‘without prejudice’ refers to the fact that the dismissal is not on the merits.” Grayson v. K Mart\nCorp., 79 F.3d 1086, 1094 n.7 (11th Cir. 1996). Although refiling may prove futile (Salcedo has\nalready amended his complaint once in attempt to shore up his allegations), we and the district\ncourt presently lack jurisdiction to make that merits determination.\n 21\n\f Case: 17-14077 Date Filed: 08/28/2019 Page: 22 of 22\n\n\nJILL PRYOR, concurring in judgment only:\n\n Plaintiff John Salcedo sued defendants Alex Hanna and the Law Offices of\n\nAlex Hanna, P.A. (together, “Hanna”) under the Telephone Consumer Protection\n\nAct of 1991 (“TCPA”), 47 U.S.C. § 227(b)(1)(A)(iii), after they sent him a single\n\nunwanted text message advertisement. I agree with the majority opinion that we\n\nlack subject matter jurisdiction because Salcedo has no standing to bring a TCPA\n\nclaim. I write separately to emphasize my understanding that the majority’s\n\nholding is narrow and the conclusion that Salcedo lacks standing is driven by the\n\nallegations in his complaint that Hanna sent him only one text message. The\n\nmajority opinion—appropriately, in my view—leaves unaddressed whether a\n\nplaintiff who alleged that he had received multiple unwanted and unsolicited text\n\nmessages may have standing to sue under the TCPA. With this understanding, I\n\nconcur in the majority’s judgment.\n\n\n\n\n 22\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: John Salcedo v. Alex Hanna\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-28\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 17-14077 Date Filed: 08/28/2019 Page: 1 of 22\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n\\n\\n\\n No. 17-14077\\n\\n\\n\\n D.C. Docket No. 0:16-cv-62480-DPG\\n\\nJOHN SALCEDO,\\nindividually and on behalf of others similarly situated,\\n\\n Plaintiff - Appellee,\\n\\n versus\\n\\nALEX HANNA,\\nan individual,\\nTHE LAW OFFICES OF ALEX HANNA, P.A.,\\na Florida Professional Association,\\n\\n Defendants - Appellants.\\n\\n\\n\\n Appeal from the United States District Court\\n for the Southern District of Florida\\n\\n\\n\\n (August 28, 2019)\\n\\nBefore JILL PRYOR and BRANCH, Circuit Judges, and REEVES, * District\\nJudge.\\n\\n*\\n Honorable Danny C. Reeves, United States District Judge for the Eastern District of Kentucky,\\nsitting by designation.\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 2 of 22\\n\\n\\nBRANCH, Circuit Judge:\\n\\n Is receiving a single unsolicited text message, sent in violation of a federal\\n\\nstatute, a concrete injury in fact that establishes standing to sue in federal court? To\\n\\nanswer that question, we have examined the statute, our precedent, and—following\\n\\nthe Supreme Court’s guidance—history and the judgment of Congress, and we\\n\\nconclude that the allegations in this suit do not establish standing.\\n\\n I. BACKGROUND\\n\\n At 9:56 a.m. on August 12, 2016, John Salcedo, a former client of Florida\\n\\nattorney Alex Hanna and his law firm, 1 received a multimedia text message from\\n\\nHanna offering a ten percent discount on his services.\\n\\n Salcedo filed suit in the district court as the representative of a putative class\\n\\nof former Hanna clients who received unsolicited text messages from Hanna in the\\n\\npast four years, alleging violations of the Telephone Consumer Protection Act of\\n\\n1991 (“TCPA”), 47 U.S.C. § 227(b)(1)(A)(iii).2 He sought, among other relief,\\n\\nstatutory damages of $500 per text message and treble damages of $1,500 per text\\n\\n\\n\\n1\\n For simplicity, and without implying any view as to Mr. Hanna’s possible personal liability,\\nthroughout this opinion we will refer to both defendants—Mr. Hanna and his law firm—\\ncollectively as “Hanna.”\\n2\\n “It shall be unlawful for any person within the United States . . . to make any call (other than a\\ncall made for emergency purposes or made with the prior express consent of the called party)\\nusing any automatic telephone dialing system or an artificial or prerecorded voice . . . to any\\ntelephone number assigned to a . . . cellular telephone service . . . .” 47 U.S.C.\\n§ 227(b)(1)(A)(iii).\\n 2\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 3 of 22\\n\\n\\nmessage sent willfully or knowingly. See 47 U.S.C. § 227(b)(3).\\n\\n Hanna moved to dismiss the complaint for lack of standing, arguing in the\\n\\nalternative that it should be dismissed as to Mr. Hanna for failure to state a claim\\n\\nagainst him and that certain parts of the complaint should be stricken. The district\\n\\ncourt disagreed, finding in relevant part that Salcedo had standing under Mohamed\\n\\nv. Off Lease Only, Inc., No. 15-23352-Civ-COOKE/TORRES, 2017 WL 1080342\\n\\n(S.D. Fla. Mar. 22, 2017). However, finding that its order “involves a controlling\\n\\nquestion of law as to which there is a substantial ground for difference of opinion,”\\n\\nthe court allowed Salcedo to pursue an interlocutory appeal and stayed its\\n\\nproceedings pending appeal. A panel of our Court granted Hanna’s petition for\\n\\npermission to appeal under 28 U.S.C. § 1292(b). We now consider his appeal.\\n\\n II. STANDARD OF REVIEW\\n\\n “We review standing determinations de novo.” Bochese v. Town of Ponce\\n\\nInlet, 405 F.3d 964, 975 (11th Cir. 2005).\\n\\n III. DISCUSSION\\n\\n Our analysis proceeds as follows. We first introduce the TCPA, the statute\\n\\nunder which Salcedo has filed suit. Next, we discuss the standing requirements of\\n\\nArticle III of the Constitution, which help to define our limited power to resolve\\n\\nonly cases or controversies. We then turn to Salcedo’s particular allegations of\\n\\nharm and analyze them in view of our Circuit precedent, history, and the judgment\\n\\n\\n 3\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 4 of 22\\n\\n\\nof Congress.\\n\\n A. The Telephone Consumer Privacy Act of 1991\\n\\n Because it found that “residential telephone subscribers consider automated\\n\\nor prerecorded telephone calls . . . to be a nuisance and an invasion of privacy,”\\n\\nTelephone Consumer Protection Act of 1991, S. 1462, 102d Cong., Pub. L. No.\\n\\n102-243, § 2, ¶ 10 (1991), in 1991 Congress enacted the TCPA to restrict interstate\\n\\ntelemarketing. The TCPA thus prohibits using automatic telephone dialing systems\\n\\nto call residential or cellular telephone lines without the consent of the called party.\\n\\n47 U.S.C. § 227(b)(1)(A)(iii), (B). It also prohibits sending unsolicited\\n\\nadvertisements via facsimile machine. Id. § 227(b)(1)(C). It authorizes the Federal\\n\\nCommunications Commission (“FCC”) to enact implementing regulations. Id.\\n\\n§ 227(b)(2). Finally for our purposes, the TCPA creates a private right of action\\n\\nwhereby a person or entity may seek compensatory or injunctive relief against\\n\\nviolators. Id. § 227(b)(3).\\n\\n There have been two relevant updates to the TCPA and its enforcement\\n\\nregime since 1991. First, in October 1992, Congress amended the TCPA to allow\\n\\nthe FCC to exempt free-to-receive cellular calls if it so chooses. Id. § 227(b)(2)(C).\\n\\nThe FCC has not done so. Second, the statute has been silent as to text messaging,\\n\\nfor that medium did not exist in 1991. But under its TCPA rulemaking authority,\\n\\nthe FCC has applied the statute’s regulations of voice calls to text messages. 30\\n\\n\\n 4\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 5 of 22\\n\\n\\nFCC Rcd. 7961, 7964 n.3, 7978–79, 8016–22 (2015); 18 FCC Rcd. 14014, 14115\\n\\n(2003); see also Campbell–Ewald Co. v. Gomez, 136 S. Ct. 663, 667 (2016) (“A\\n\\ntext message to a cellular telephone, it is undisputed, qualifies as a ‘call’ within the\\n\\ncompass of § 227(b)(1)(A)(iii).”). Thus, Salcedo’s complaint facially appears to\\n\\nstate a cause of action under the TCPA as interpreted by the FCC.\\n\\n B. Article III Standing\\n\\n Not every right created by Congress or defined by an executive agency is\\n\\nautomatically enforceable in the federal courts. Our tripartite system of\\n\\ngovernment recognizes that “there is no liberty if the power of judging be not\\n\\nseparated from the legislative and executive powers.” The Federalist No. 78, at 465\\n\\n(Alexander Hamilton) (Clinton Rossiter ed. 1961). To protect this separation of\\n\\npowers, we must assure ourselves that our exercise of jurisdiction falls within the\\n\\nConstitution’s grant of judicial power.\\n\\n Article III vests the judicial power in the federal courts and extends that\\n\\npower to “Cases” and “Controversies.” U.S. Const. art. III, §§ 1–2. One tool for\\n\\ndetermining that the matters before us are truly cases or controversies, as\\n\\nunderstood by Article III, is the doctrine of standing. Lujan v. Defenders of\\n\\nWildlife, 504 U.S. 555, 560 (1992). “The law of Article III standing . . . serves to\\n\\nprevent the judicial process from being used to usurp the powers of the political\\n\\nbranches.” Clapper v. Amnesty Int’l USA, 568 U.S. 398, 408 (2013). Even when\\n\\n\\n 5\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 6 of 22\\n\\n\\nthose political branches appear to have granted us jurisdiction by statute and rule,\\n\\nwe are still obliged to examine whether jurisdiction exists under the Constitution.\\n\\n As the Supreme Court has explained, the “irreducible constitutional\\n\\nminimum” to establish Article III standing requires three elements. Lujan, 504 U.S.\\n\\nat 560. “The plaintiff must have (1) suffered an injury in fact, (2) that is fairly\\n\\ntraceable to the challenged conduct of the defendant, and (3) that is likely to be\\n\\nredressed by a favorable judicial decision.” Spokeo, Inc. v. Robins, 136 S. Ct. 1540,\\n\\n1547 (2016) (citing Lujan, 504 U.S. at 560–61). It is the first element—the\\n\\n“foremost” of the three, id. (quoting Steel Co. v. Citizens for a Better Env’t, 523\\n\\nU.S. 83, 103 (1998))—that is disputed in this appeal and to which we now turn.\\n\\n To establish standing, an injury in fact must be concrete.3 Id. at 1548. “A\\n\\n‘concrete’ injury must be ‘de facto’; that is, it must actually exist,” as opposed to\\n\\nbeing hypothetical or speculative. Id. A concrete injury need be only an\\n\\n“identifiable trifle.” United States v. Students Challenging Regulatory Agency\\n\\nProcedures (SCRAP), 412 U.S. 669, 689 n.14 (1973) (noting that sufficiently\\n\\nconcrete injuries have included a fraction of a vote, a $5 fine and costs, and a $1.50\\n\\npoll tax). But sometimes plaintiffs allege intangible injuries that we cannot so\\n\\n\\n\\n3\\n An injury in fact must also be particularized, that is, affecting the plaintiff “in a personal and\\nindividual way.” Spokeo, 136 S. Ct at 1548 (quoting Lujan, 504 U.S. at 560 n.1). It is undisputed\\nthat Salcedo’s allegations are of a personal and individual nature. As the would-be class\\nrepresentative, Salcedo must establish his own personal, concrete injury notwithstanding\\nwhatever injuries may have been suffered by the other members of the class. Id. at 1547 n.6.\\n 6\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 7 of 22\\n\\n\\neasily identify.\\n\\n When the concreteness of an alleged injury is difficult to recognize, we look\\n\\nto “history and the judgment of Congress” for guidance. Spokeo, 136 S. Ct. at\\n\\n1549. But an act of Congress that creates a statutory right and a private right of\\n\\naction to sue does not automatically create standing; “Article III standing requires\\n\\na concrete injury even in the context of a statutory violation.” Id. 4 “[T]he\\n\\nrequirement of injury in fact is a hard floor of Article III jurisdiction that cannot be\\n\\nremoved by statute.” Summers v. Earth Island Inst., 555 U.S. 488, 497 (2009).\\n\\n C. Eleventh Circuit Precedent\\n\\n Because Salcedo bears the burden of establishing federal jurisdiction, Lujan,\\n\\n504 U.S. at 561, we look to the substance of his amended complaint’s allegations\\n\\nto determine if he has standing due to a concrete injury. Salcedo alleged that\\n\\nreceiving the one text message “caused Plaintiff to waste his time answering or\\n\\notherwise addressing the message. While doing so, both Plaintiff and his cellular\\n\\n\\n\\n4\\n Recognizing that “a bare procedural violation” of a statute “may result in no harm,” Spokeo\\nreaffirms the proposition that we must always look for concrete harm when assessing Article III\\nstanding. See 136 S. Ct. at 1550. In some contexts this will mean identifying purely speculative\\n“harm” that never actually materializes as failing to allege an injury in fact. See, e.g., Nicklaw v.\\nCitiMortgage, Inc., 839 F.3d 998, 1003 (11th Cir. 2016), reh’g en banc denied, 855 F.3d 1265\\n(11th Cir. 2017) (holding that a violation of the mortgage satisfaction reporting requirements of a\\nstate law resulted in no concrete harm to the plaintiff).\\n But we should not ignore the Supreme Court’s guidance in Spokeo in cases that purport\\nto allege more than merely technical statutory violations. This appeal presents a close question in\\nwhich we must determine whether Salcedo’s allegations are real and concrete as opposed to\\nfigmentary. Spokeo’s instruction to consider history and the judgment of Congress, id. at 1549,\\nhelpfully guides us in our conclusion that Salcedo has not alleged a concrete injury in fact.\\n 7\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 8 of 22\\n\\n\\nphone were unavailable for otherwise legitimate pursuits.” He further alleged that\\n\\nthe message also “resulted in an invasion of Plaintiff’s privacy and right to enjoy\\n\\nthe full utility of his cellular device.”\\n\\n These allegations are qualitatively different from those in our Circuit\\n\\nprecedent that have been successful in establishing standing to sue over a single\\n\\nviolation of the TCPA. In Palm Beach Golf Center–Boca, Inc. v. John G. Sarris,\\n\\nD.D.S., P.A., 781 F.3d 1245, 1252 (11th Cir. 2015), we found standing for a\\n\\nplaintiff who alleged that receiving a junk fax in violation of the TCPA harmed\\n\\nhim because, during the minute or so that it took to receive and process the fax\\n\\nmessage, his fax machine was unavailable for receiving legitimate business\\n\\nmessages. Accord Florence Endocrine Clinic, PLLC v. Arriva Med., LLC, 858\\n\\nF.3d 1362, 1366 (11th Cir. 2017) (considering also “the cost of printing the\\n\\nunsolicited fax”). To the extent we have relied on tangible costs such as the\\n\\nconsumption of paper and ink or toner to establish injury in fact, Salcedo cannot so\\n\\nrely, since receiving a text message uses no paper, ink, or toner. His complaint\\n\\nalleges generally that some text messages cause recipients to incur costs to their\\n\\nwireless service providers, but he has not alleged specifically that Hanna’s text cost\\n\\nhim any money.\\n\\n Salcedo’s allegations of intangible costs, on the other hand, bear some facial\\n\\nsimilarities to those in Palm Beach Golf. But they differ in kind, rendering Palm\\n\\n\\n 8\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 9 of 22\\n\\n\\nBeach Golf inapplicable. At oral argument, Salcedo asserted that receiving\\n\\nHanna’s message was comparable to using a minute of fax machine time, but his\\n\\ncomplaint does not so allege. Rather, it alleges time wasted only generally. In the\\n\\nabsence of a specific time allegation, we decline to assume an equivalence to the\\n\\nfacts of Palm Beach Golf when receiving a fax message is qualitatively different\\n\\nfrom receiving a text message. A fax message consumes the receiving device\\n\\nentirely, while a text message consumes the receiving device not at all. A cell\\n\\nphone user can continue to use all of the device’s functions, including receiving\\n\\nother messages, while it is receiving a text message.\\n\\n Salcedo also makes an allegation about unavailability, but that too is distinct\\n\\nfrom Palm Beach Golf. There, we were concerned about the fully realized\\n\\nopportunity cost of being unable to receive other faxes for a full minute. By\\n\\ncontrast, Salcedo has alleged no particular loss of opportunity. A fax machine’s\\n\\ninability to receive another message while processing a junk fax has no analogy\\n\\nwith cell phones and text messaging. Salcedo’s assertion that he and his phone\\n\\nwere unavailable appears only to recite language we used in Palm Beach Golf. Cf.\\n\\n781 F.3d at 1252 (quoting H.R. Rep. 102-317, at 10 (1991)). We are entitled to\\n\\nlook past this conclusory recitation to the actual factual substance of Salcedo’s\\n\\nallegations. See Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (“we are not bound to\\n\\naccept as true a legal conclusion couched as a factual allegation”).\\n\\n\\n 9\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 10 of 22\\n\\n\\n Thus, Circuit precedent in Palm Beach Golf does not dictate the outcome of\\n\\nthis appeal. And, for reasons we will discuss below, we find our sister circuit’s\\n\\ndecision involving this precise issue unpersuasive. See Van Patten v. Vertical\\n\\nFitness Group, LLC, 847 F.3d 1037, 1043 (9th Cir. 2017) (holding that the receipt\\n\\nof two unsolicited text messages constituted an injury in fact).5 In the absence of\\n\\ncontrolling authority, we turn our analysis to the framework outlined by the\\n\\nSupreme Court in Spokeo. We look to history and the judgment of Congress to see\\n\\nwhether they support treating Salcedo’s allegations as a concrete injury in fact. Our\\n\\nexamination reveals little support for so doing.\\n\\n D. The Judgment of Congress\\n\\n We first note what Congress has said in the TCPA’s provisions and findings 6\\n\\nabout harms from telemarketing via text message generally: nothing. The TCPA is\\n\\ncompletely silent on the subject of unsolicited text messages. Of course, text\\n\\nmessaging in its current form did not exist in 1991 when the TCPA was enacted,\\n\\n\\n5\\n Nor are we bound by the Supreme Court’s holding in another TCPA text-messaging case that\\nthe case was not mooted by an unaccepted settlement offer. See Campbell–Ewald, 136 S. Ct. at\\n670. The Court did not reach the unraised question of whether the plaintiff had alleged an injury\\nin fact, in part because the defendant apparently never asserted that the plaintiff had failed to do\\nso. See id. at 667–68. “[W]e are not bound by a prior decision’s sub silentio treatment of a\\njurisdictional question.” Okongwu v. Reno, 229 F.3d 1327, 1330 (11th Cir. 2000).\\n6\\n Context matters. We are not suggesting that legislative history should play a role in statutory\\ninterpretation. Salcedo’s allegation is undisputedly a violation of the statute as interpreted by the\\nFCC. Nonetheless, because the Supreme Court has instructed us to consider “the judgment of\\nCongress” in assessing Article III standing, we will consider the congressionally enacted\\nfindings as informative of that judgment. See Spokeo, 136 S. Ct. at 1549; cf. Palm Beach Golf,\\n781 F.3d at 1252 (citing House committee report).\\n 10\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 11 of 22\\n\\n\\nbut Congress has amended the statute several times since then without adding text\\n\\nmessaging to the categories of restricted telemarketing.7 As we have mentioned, it\\n\\nis only through the rulemaking authority of the FCC that the voice call provisions\\n\\nof the TCPA have been extended to text messages. At most, we could take\\n\\nCongress’s silence as tacit approval of that agency action.\\n\\n On the other hand, Congress’s legislative findings about telemarketing\\n\\nsuggest that the receipt of a single text message is qualitatively different from the\\n\\nkinds of things Congress was concerned about when it enacted the TCPA. In\\n\\nparticular, the findings in the TCPA show a concern for privacy within the sanctity\\n\\nof the home that do not necessarily apply to text messaging. “Unrestricted\\n\\ntelemarketing . . . can be an intrusive invasion of privacy,” and “[m]any consumers\\n\\nare outraged over the proliferation of intrusive, nuisance calls to their homes from\\n\\ntelemarketers,” Congress found. Pub. L. No. 102-243, § 2, ¶¶ 5, 6. By contrast, cell\\n\\nphones are often taken outside of the home and often have their ringers silenced,\\n\\npresenting less potential for nuisance and home intrusion. It is thus not surprising\\n\\nthat, after Congress found that the FCC “should have the flexibility to design\\n\\ndifferent rules for those types of automated or prerecorded calls that it finds are not\\n\\n\\n\\n\\n7\\n Following recent amendments, however, the TCPA will expressly include text messaging in its\\nprohibitions on transmitting false caller ID information. Consolidated Appropriations Act, 2018,\\nH.R. 1625, 115th Cong., Pub. L. No. 115-141, div. P, § 503(a) (2018) (to be codified at 47\\nU.S.C. § 227(e)).\\n 11\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 12 of 22\\n\\n\\nconsidered a nuisance or invasion of privacy,” id. ¶ 13, within a year it instructed\\n\\nthe FCC that it may exempt “calls to a telephone number assigned to a cellular\\n\\ntelephone service that are not charged to the called party,” 47 U.S.C.\\n\\n§ 227(b)(2)(C).\\n\\n On text messaging generally, then, the judgment of Congress is ambivalent\\n\\nat best; its privacy and nuisance concerns about residential telemarketing are less\\n\\nclearly applicable to text messaging. Any possible deference to the FCC’s\\n\\ninterpretation of the TCPA 8—the source of its application to text messaging—is\\n\\nnot obviously relevant where the Supreme Court has specifically instructed us to\\n\\nconsider the judgment of Congress. And congressional silence is a poor basis for\\n\\nextending federal jurisdiction to new types of harm. We take seriously the silence\\n\\nof that political branch best positioned to assess and articulate new harms from\\n\\nemerging technologies. See Spokeo, 136 S. Ct. at 1549 (citing Lujan, 504 U.S. at\\n\\n578)). With this point of caution in mind, we now turn to the judgment of Congress\\n\\nabout the specific harms that Salcedo has alleged he suffered when he received\\n\\nHanna’s message.\\n\\n\\n\\n8\\n In this case, we need not reach the issue of whether the agency’s interpretation of the statute is\\nentitled to any deference. See Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837,\\n843 (1984) (requiring deference to agency’s interpretation of silent or ambiguous statute);\\nSkidmore v. Swift & Co., 323 U.S. 134, 140 (1944) (allowing court to determine level of\\ndeference in proportion to agency’s demonstration of persuasive reasoning); cf., e.g., Josendis v.\\nWall to Wall Residence Repairs, Inc., 662 F.3d 1292, 1320 (11th Cir. 2011) (rejecting Chevron\\ndeference where the statutory language was clear and unambiguous).\\n 12\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 13 of 22\\n\\n\\n We consider the judgment of Congress when assessing standing because\\n\\n“Congress is well positioned to identify intangible harms that meet minimum\\n\\nArticle III requirements.” Id. As we have mentioned, Congress was concerned\\n\\nabout “intrusive invasion[s] of privacy” into the home when it enacted the TCPA.\\n\\nPub. L. No. 102-243, § 2, ¶ 5. Salcedo argues that the particular privacy interest\\n\\nCongress has identified is “the freedom from unwanted robocalls,” but that\\n\\nobservation is too general. As we have noted, a single unwelcome text message\\n\\nwill not always involve an intrusion into the privacy of the home in the same way\\n\\nthat a voice call to a residential line necessarily does. Certainly, Salcedo has not\\n\\nalleged that he was in his home when he received Hanna’s message. As we have\\n\\nalso noted, the 1992 amendment allowing the FCC to exempt free-to-receive calls\\n\\nto cell phones, 47 U.S.C. § 227(b)(2)(C), suggests less congressional concern\\n\\nabout calls to cell phones. And by nature of their portability and their ability to be\\n\\nsilenced, cell phone calls may involve less of an intrusion than calls to a home\\n\\nphone. We realize that Congress in 1991 could not have foreseen the explosion in\\n\\npersonal cell phone use, the popularity of text messaging, and the near-extinction\\n\\nof the residential telephone line. But Spokeo instructs us to consider the judgment\\n\\nof Congress about the alleged harm, not to imagine what Congress might say about\\n\\na harm it has not actually addressed.\\n\\n We note that our sister circuit has reached the opposite conclusion in this\\n\\n\\n 13\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 14 of 22\\n\\n\\ncontext. See Van Patten, 847 F.3d at 1043. The Ninth Circuit quoted many of these\\n\\nsame findings, further noting Congress’s purpose of “protect[ing] consumers from\\n\\nthe unwanted intrusion and nuisance of unsolicited telemarketing phone calls and\\n\\nfax advertisements.” See Pub. L. No. 102-243, § 2, ¶ 12. But that court stopped\\n\\nshort of examining whether isolated text messages not received at home come\\n\\nwithin that judgment of Congress. Instead, it concluded that “Congress identified\\n\\nunsolicited contact as a concrete harm.” Van Patten, 847 F.3d at 1043. We disagree\\n\\nwith this broad overgeneralization of the judgment of Congress and have focused\\n\\nour own analysis on text messaging specifically.\\n\\n Other stated concerns behind the TCPA are also inapposite to Salcedo’s\\n\\nallegations. The congressional committee found telemarketing by fax problematic\\n\\nin part because “it occupies the recipient’s facsimile machine so that it is\\n\\nunavailable for legitimate business messages while processing and printing the\\n\\njunk fax.” H.R. Rep. 102-317, at 10 (1991), quoted in Palm Beach Golf, 781 F.3d\\n\\nat 1252. As we have noted, such a concern has little application to the\\n\\ninstantaneous receipt of a text message. The judgment of Congress, then, provides\\n\\nlittle support for finding that Salcedo’s allegations state a concrete injury in fact.9\\n\\n\\n\\n\\n9\\n Congress also stated concerns not raised by either party here: concerns for public safety “when\\nan emergency or medical assistance telephone line is seized,” Pub. L. No. 102-243, § 2, ¶ 5; for\\nthe cost borne by consumers who use technology to avoid unwanted calls, id. ¶ 11; and for\\n“commercial freedoms of speech and trade” that telemarketers enjoy, id. ¶ 9.\\n 14\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 15 of 22\\n\\n\\n E. History\\n\\n We now turn to history for guidance, because the case or controversy\\n\\nrequirement of Article III “is grounded in historical practice.” Spokeo, 136 S. Ct. at\\n\\n1549. Thus, “it is instructive to consider whether an alleged intangible harm has a\\n\\nclose relationship to a harm that has traditionally been regarded as providing a\\n\\nbasis for a lawsuit in English or American courts.” Id.\\n\\n With respect to his allegations of invasion of privacy, we look to the\\n\\ngenerally accepted tort of intrusion upon seclusion,10 which creates liability for\\n\\ninvasions of privacy that would be “highly offensive to a reasonable person.”\\n\\nRestatement (Second) of Torts § 652B. The requirement that the interference be\\n\\n“substantial” and “strongly object[ionable]” instructs us that a plaintiff might be\\n\\nable to establish standing where an intrusion on his privacy is objectively serious\\n\\nand universally condemnable. See id. cmt. d (no liability for one, two, or three\\n\\nphone calls; liability “only when the telephone calls are repeated with such\\n\\npersistence and frequency as to amount to a course of hounding the plaintiff”). By\\n\\ncontrast, Salcedo’s allegations fall short of this degree of harm. We do not see this\\n\\ntype of objectively intense interference where the alleged harm is isolated,\\n\\n\\n\\n\\n10\\n Most of the accepted torts generally known as “invasion of privacy” involve privacy in the\\nrather specific sense of one’s right to be free from unwanted publicity. See Restatement (Second)\\nof Torts § 652A. It is only the privacy tort of intrusion upon seclusion, id. § 652B, that bears any\\npossible relationship to Salcedo’s allegations.\\n 15\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 16 of 22\\n\\n\\nmomentary, and ephemeral.\\n\\n The tort of intrusion upon seclusion also requires an intrusion “upon the\\n\\nsolitude or seclusion of another or his private affairs or concerns.” Id. § 652B.\\n\\nAlthough Salcedo argues that his cell phone is part of his private affairs, the\\n\\nRestatement contemplates a different category of intrusion into private affairs,\\n\\nlisting examples including eavesdropping, wiretapping, and looking through one’s\\n\\npersonal documents. See id. cmt. b. Simply sending one text message to a private\\n\\ncell phone is not closely related to the severe kinds of actively intermeddling\\n\\nintrusions that the traditional tort contemplates. Salcedo’s reasoning would equate\\n\\nopening your private mail—a serious intrusion indeed—with mailing you a\\n\\npostcard.\\n\\n With respect to his allegations of nuisance, Salcedo asks us to compare the\\n\\ntraditional torts of trespass and nuisance, but we find them also to be distinct both\\n\\nin kind and in degree. Trespass requires intentionally “enter[ing] land in the\\n\\npossession of the other,” id. § 158(a), and private nuisance is “a nontrespassory\\n\\ninvasion of another’s interest in the private use and enjoyment of land,” id. § 821D.\\n\\nAlthough, as we have noted, Congress was concerned about intrusions into the\\n\\nhome when it enacted the TCPA, Salcedo has alleged no invasion of any interest in\\n\\nreal property here. Furthermore, even in the context of nuisance to real property, in\\n\\nFlorida, “[m]ere disturbance and annoyance as such do not in themselves\\n\\n\\n 16\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 17 of 22\\n\\n\\nnecessarily give rise to an invasion of a legal right.” A & P Food Stores, Inc. v.\\n\\nKornstein, 121 So. 2d 701, 703 (Fla. 3d Dist. Ct. App. 1960). Hanna’s text\\n\\nmessage is thus not closely related to these traditional harms because it is not\\n\\nalleged to have infringed upon Salcedo’s real property, either directly or indirectly.\\n\\n Salcedo also asks us to consider the personal property torts of conversion\\n\\nand trespass to chattel. Conversion is an interference with chattel “which so\\n\\nseriously interferes with the right of another to control it that the actor may justly\\n\\nbe required to pay the other the full value of the chattel.” Restatement (Second) of\\n\\nTorts § 222A. Salcedo’s allegations are nowhere near a complete and permanent\\n\\ndominion over his phone, so recourse to this serious kind of tort is unhelpful. The\\n\\nsame is true for the tort of trespass to chattel, which involves intentionally “using\\n\\n. . . a chattel in the possession of another.” Id. § 217(b). Traditionally, liability\\n\\narises for this kind of trespass only when “the possessor is deprived of the use of\\n\\nthe chattel for a substantial time” or when the trespass harms “the possessor’s\\n\\nmaterially valuable interest in the physical condition, quality, or value of the\\n\\nchattel.” Id. § 218(c) & cmt. e; cf. United States v. Jones, 565 U.S. 400, 426 (2012)\\n\\n(Alito, J., concurring) (“Trespass to chattels has traditionally required a physical\\n\\ntouching of the property.”). Thus, although Salcedo’s allegations here bear a\\n\\npassing resemblance to this kind of historical harm, they differ so significantly in\\n\\ndegree as to undermine his position. History shows that Salcedo’s allegation is\\n\\n\\n 17\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 18 of 22\\n\\n\\nprecisely the kind of fleeting infraction upon personal property that tort law has\\n\\nresisted addressing.\\n\\n We again note that our sister circuit has reached the opposite conclusion. See\\n\\nVan Patten, 847 F.3d at 1043. We decline to adopt its reasoning and instead\\n\\nembrace more fully the Supreme Court’s instruction to look for a “close\\n\\nrelationship” to a traditionally redressable harm. See Spokeo, 136 S. Ct. at 1549\\n\\n(citing Vt. Agency of Nat. Res., 529 U.S. 765, 775–77 (2000) (discussing traditional\\n\\nqui tam law in a case about qui tam relator Article III standing)). The Ninth\\n\\nCircuit’s one-sentence review of history simply asserted, “Actions to remedy\\n\\ndefendants’ invasions of privacy, intrusion upon seclusion, and nuisance have long\\n\\nbeen heard by American courts, and the right of privacy is recognized by most\\n\\nstates.” Van Patten, 847 F.3d at 1043. But as we have more thoroughly explained,\\n\\nan examination of those torts reveals significant differences in the kind and degree\\n\\nof harm they contemplate providing redress for.\\n\\n In sum, we find that history and the judgment of Congress do not support\\n\\nfinding concrete injury in Salcedo’s allegations. 11 Salcedo has not alleged anything\\n\\n\\n11\\n Salcedo urges us to follow the reasoning that allowed us to find standing in Perry v. Cable\\nNews Network, Inc., 854 F.3d 1336, 1340–41 (11th Cir. 2017). The facts here do not permit us to\\ndo so. In Perry, we found standing under the Video Privacy Protection Act, 18 U.S.C. § 2710,\\nfor a plaintiff suing over privacy violations involving a mobile video app. Perry held that the\\nplaintiff’s allegations and the 1980s-era statute involved precisely the same substantive privacy\\nright. Not so here. As we have discussed, both the judgment of Congress and history here reveal\\nconcerns about intrusions into the privacy of the home and interferences with property that do\\nnot readily transfer to the context of cell phones.\\n 18\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 19 of 22\\n\\n\\nlike enjoying dinner at home with his family and having the domestic peace\\n\\nshattered by the ringing of the telephone. Nor has he alleged that his cell phone\\n\\nwas searched, dispossessed, or seized for any length of time. Salcedo’s allegations\\n\\nof a brief, inconsequential annoyance are categorically distinct from those kinds of\\n\\nreal but intangible harms. The chirp, buzz, or blink of a cell phone receiving a\\n\\nsingle text message is more akin to walking down a busy sidewalk and having a\\n\\nflyer briefly waived in one’s face. Annoying, perhaps, but not a basis for invoking\\n\\nthe jurisdiction of the federal courts. All told, we conclude that Salcedo’s\\n\\nallegations do not state a concrete harm that meets the injury-in-fact requirement of\\n\\nArticle III.\\n\\n F. Quality, Not Quantity\\n\\n To be clear, we are not attempting to measure how small or large Salcedo’s\\n\\nalleged injury is. Article III standing is not a “You must be this tall to ride”\\n\\nmeasuring stick. “There is no minimum quantitative limit required to show injury;\\n\\nrather, the focus is on the qualitative nature of the injury, regardless of how small\\n\\nthe injury may be.” Saladin v. City of Milledgeville, 812 F.2d 687, 691 (11th Cir.\\n\\n1987). Our assessment today is thus qualitative, not quantitative. We have assessed\\n\\nhow concrete and real the alleged harm is, Spokeo, 136 S. Ct. at 1548, and we have\\n\\nconcluded that it is not the kind of harm that constitutes an injury in fact. Some\\n\\nharms that are intangible and ephemeral may do so, but Salcedo’s allegations of\\n\\n\\n 19\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 20 of 22\\n\\n\\nthe harm he suffered from receiving a single text message do not.\\n\\n To be sure, under our precedent, allegations of wasted time can state a\\n\\nconcrete harm for standing purposes. We have found standing where the harm was,\\n\\nfor example, time wasted traveling to the county registrar’s office, Common\\n\\nCause/Ga. v. Billups, 554 F.3d 1340, 1351 (11th Cir. 2009); and correcting credit\\n\\nreporting errors, Pedro v. Equifax, Inc., 868 F.3d 1275, 1280 (11th Cir. 2017).\\n\\nThese precedents strongly suggest that concrete harm from wasted time requires, at\\n\\nthe very least, more than a few seconds. And on this point the judgment of\\n\\nCongress sheds a final ray of light. The TCPA instructs the FCC to establish\\n\\ntelemarketing standards that include releasing the called party’s line within five\\n\\nseconds of a hang-up, 47 U.S.C. § 227(d)(3)(B), demonstrating that, on the margin,\\n\\nCongress does not view tying up a phone line for five seconds as a serious\\n\\nintrusion.\\n\\n Our responsibility to ensure that plaintiffs allege a real injury in fact requires\\n\\nus to look closely at their allegations in light of the statute, our precedent, history,\\n\\nand the judgment of Congress. Such inquiries will, of course, have differing\\n\\noutcomes depending on those inputs. Compare, e.g., Perry, 854 F.3d at 1340–41\\n\\n(finding standing based on intangible harm of statutory violation), and Palm Beach\\n\\nGolf, 781 F.3d at 1252 (same), with Nicklaw v. CitiMortgage, Inc., 839 F.3d 998,\\n\\n1003 (11th Cir. 2016), reh’g denied, 855 F.3d 1265 (11th Cir. 2017) (en banc)\\n\\n\\n 20\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 21 of 22\\n\\n\\n(finding no standing because plaintiff alleged “neither a harm nor a material risk of\\n\\nharm”). We acknowledge that Congress, as a political entity, is well positioned to\\n\\nassess new harms in light of developments in technology and society, and to\\n\\nrespond to the concerns of the American people about novel encroachments on life,\\n\\nliberty, and property. See Spokeo, 136 S. Ct. at 1549. The federal courts are not\\n\\nsimilarly tasked. We have only the power “to say what the law is.” Marbury v.\\n\\nMadison, 5 U.S. (1 Cranch) 137, 177 (1803). And when a plaintiff comes to us\\n\\nwithout alleging a concrete harm, a real injury that states a case or controversy, we\\n\\ncannot do even that much.\\n\\n IV. CONCLUSION\\n\\n We REVERSE the decision of the district court that Salcedo has standing to\\n\\nsue and REMAND with instructions to dismiss without prejudice the amended\\n\\ncomplaint.12\\n\\n\\n\\n\\n12\\n Hanna has asked us to instruct the district court to dismiss Salcedo’s amended complaint with\\nprejudice. But a jurisdictional dismissal is entered without prejudice. Stalley ex rel. United States\\nv. Orlando Reg’l Healthcare Sys., Inc., 524 F.3d 1229, 1232 (11th Cir. 2008). “A dismissal\\n‘without prejudice’ refers to the fact that the dismissal is not on the merits.” Grayson v. K Mart\\nCorp., 79 F.3d 1086, 1094 n.7 (11th Cir. 1996). Although refiling may prove futile (Salcedo has\\nalready amended his complaint once in attempt to shore up his allegations), we and the district\\ncourt presently lack jurisdiction to make that merits determination.\\n 21\\n\\f Case: 17-14077 Date Filed: 08/28/2019 Page: 22 of 22\\n\\n\\nJILL PRYOR, concurring in judgment only:\\n\\n Plaintiff John Salcedo sued defendants Alex Hanna and the Law Offices of\\n\\nAlex Hanna, P.A. (together, “Hanna”) under the Telephone Consumer Protection\\n\\nAct of 1991 (“TCPA”), 47 U.S.C. § 227(b)(1)(A)(iii), after they sent him a single\\n\\nunwanted text message advertisement. I agree with the majority opinion that we\\n\\nlack subject matter jurisdiction because Salcedo has no standing to bring a TCPA\\n\\nclaim. I write separately to emphasize my understanding that the majority’s\\n\\nholding is narrow and the conclusion that Salcedo lacks standing is driven by the\\n\\nallegations in his complaint that Hanna sent him only one text message. The\\n\\nmajority opinion—appropriately, in my view—leaves unaddressed whether a\\n\\nplaintiff who alleged that he had received multiple unwanted and unsolicited text\\n\\nmessages may have standing to sue under the TCPA. With this understanding, I\\n\\nconcur in the majority’s judgment.\\n\\n\\n\\n\\n 22\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4157,"cells":{"id":{"kind":"number","value":4656938,"string":"4,656,938"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"Regions Bank v. Legal Outsource PA"},"decision_date":{"kind":"string","value":"2019-08-28"},"docket_number":{"kind":"string","value":"17-11736"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before WILLIAM PRYOR, and ROSENBAUM, Circuit Judges, and MOORE, * District Judge.\", \"parties\": \"\", \"opinions\": [{\"author\": \"WILLIAM PRYOR, Circuit Judge:\", \"type\": \"010combined\", \"text\": \"Case: 17-11736 Date Filed: 08/28/2019 Page: 1 of 88\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 17-11736\\n ________________________\\n\\n D.C. Docket No. 2:14-cv-000476-PAM-MRM\\n\\n\\n\\nREGIONS BANK,\\nan Alabama state-chartered bank,\\n\\n Plaintiff–Counter Defendant–Appellee,\\n\\nversus\\n\\nLEGAL OUTSOURCE PA,\\na Florida professional association,\\nPERIWINKLE PARTNERS, LLC,\\na Florida limited liability company,\\nCHARLES PAUL-THOMAS PHOENIX,\\nindividually, a.k.a. Charles PT Phoenix,\\nLISA M. PHOENIX,\\nindividually,\\n\\n Defendants–Counter Claimants–Appellants.\\n\\n ________________________\\n\\n Appeal from the United States District Court\\n for the Middle District of Florida\\n ________________________\\n\\n (August 28, 2019)\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 2 of 88\\n\\n\\nBefore WILLIAM PRYOR, and ROSENBAUM, Circuit Judges, and MOORE, *\\nDistrict Judge.\\n\\nWILLIAM PRYOR, Circuit Judge:\\n\\n The main issue presented by this appeal has divided our sister circuits:\\n\\nwhether a guarantor constitutes an “applicant” under the Equal Credit Opportunity\\n\\nAct. See 15 U.S.C. §§ 1691(a), 1691a(b). Compare Hawkins v. Cmty. Bank of\\n\\nRaymore, 761 F.3d 937 (8th Cir. 2014), aff’d by an equally divided Court, 136 S.\\n\\nCt. 1072 (2016) (holding that a guarantor unambiguously is not an “applicant”\\n\\nunder the Act), and Moran Foods, Inc. v. Mid-Atl. Mkt. Dev. Co., 476 F.3d 436\\n\\n(7th Cir. 2007) (opining the same), with RL BB Acquisition, LLC v. Bridgemill\\n\\nCommons Dev. Grp., 754 F.3d 380 (6th Cir. 2014) (holding that the term\\n\\n“applicant” is ambiguous and applying Chevron deference to an agency\\n\\ninterpretation that a guarantor is an “applicant”). Legal Outsource PA, a law firm\\n\\nwholly owned by Charles Phoenix, defaulted on a loan from Regions Bank, which\\n\\ntriggered the default of a loan and mortgage that Regions issued to Periwinkle\\n\\nPartners, LLC, an entity wholly owned by Charles’s wife, Lisa Phoenix. After the\\n\\nobligors refused to cure the defaults, Regions sued to enforce its rights under the\\n\\nloans and mortgage. The obligors filed several counterclaims asserting that\\n\\nRegions violated the Equal Credit Opportunity Act by discriminating against Lisa\\n\\n\\n *\\n Honorable K. Michael Moore, United States District Chief Judge for the Southern District\\nof Florida, sitting by designation.\\n 2\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 3 of 88\\n\\n\\nand Charles based on their marital status when it demanded that they and Legal\\n\\nOutsource guarantee the Periwinkle loan. The district court granted summary\\n\\njudgment in favor of Regions. The district court ruled that Lisa Phoenix’s\\n\\ncounterclaims failed because she lacked standing as an “applicant” when she was\\n\\ninstead a guarantor. Because we conclude that a guarantor is not an “applicant”\\n\\nunder the Equal Credit Opportunity Act, we affirm the summary judgment in favor\\n\\nof Regions. But the parties agree that we must remand to correct an error in the\\n\\njudgment.\\n\\n I. BACKGROUND\\n\\n Beginning in 2005, Regions Bank extended a $450,000 line of credit to\\n\\nLegal Outsource PA, a law firm owned by Charles Phoenix. Legal Outsource\\n\\nrenewed the loan on a yearly or semi-yearly basis, and it was last renewed in May\\n\\n2013 with a maturity date in February 2014. Charles Phoenix also guaranteed the\\n\\n2013 Outsource loan.\\n\\n In 2011, Regions lent nearly $1.7 million to Periwinkle Partners, LLC, for\\n\\nthe purchase of a shopping center on Sanibel Island, Florida. At that time, the sole\\n\\nmember of Periwinkle Partners was a company owned by Charles Phoenix’s wife,\\n\\nLisa Phoenix. Charles Phoenix, Lisa Phoenix, and Legal Outsource all guaranteed\\n\\nthe Periwinkle loan. Under the Periwinkle loan, a default by any of the parties,\\n\\n\\n\\n\\n 3\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 4 of 88\\n\\n\\nincluding the guarantors, on any other loans that they had with Regions constitutes\\n\\na default under the Periwinkle loan.\\n\\n In August 2013, Regions concluded that the Outsource and Periwinkle loans\\n\\nwere in default based on the obligors’ failure to provide requested financial\\n\\ninformation and based on Periwinkle’s failure to pay its property taxes. Regions\\n\\nthen warned the obligors several times that it would accelerate the loans if the\\n\\nobligors failed to cure the default. In February 2014, the Outsource loan matured\\n\\nand Legal Outsource, which was no longer in operation, failed to pay it. Two\\n\\nmonths later, Regions declared the Outsource loan in default and demanded its full\\n\\nand immediate payment. According to the obligors, this declaration was a bad-faith\\n\\nattempt by Regions to coerce Lisa Phoenix into securing the Outsource loan with\\n\\nPeriwinkle as collateral, but she refused to do so. After the Outsource loan default,\\n\\nRegions also declared the Periwinkle loan in default and demanded its full and\\n\\nimmediate payment. The obligors never cured any of the defaults.\\n\\n In August 2014, Regions filed a complaint against Charles and Lisa Phoenix,\\n\\nLegal Outsource, and Periwinkle Partners for breach of the Legal Outsource\\n\\npromissory note and guaranty, breach of the Periwinkle promissory note and\\n\\nguaranties, foreclosure of the Periwinkle mortgage, and receivership. The obligors\\n\\nanswered the complaint and interposed 73 affirmative defenses and eight\\n\\ncounterclaims.\\n\\n\\n 4\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 5 of 88\\n\\n\\n The obligors twice amended the answer and added four new counterclaims\\n\\nthat each asserted a violation of the Equal Credit Opportunity Act. The\\n\\ncounterclaims—three of which were individually brought by Charles Phoenix, Lisa\\n\\nPhoenix, and Legal Outsource respectively, and one of which was brought by Lisa\\n\\nPhoenix and Periwinkle Partners—alleged that Regions discriminated on the basis\\n\\nof marital status when it required the Phoenixes and Legal Outsource to guarantee\\n\\nthe Periwinkle loan. Regions then moved to dismiss the newly added\\n\\ncounterclaims, and the district count granted that motion in part. The district court\\n\\nruled that the guarantors of the Periwinkle loan all lacked statutory standing\\n\\nbecause they were not “applicant[s]” under the Equal Credit Opportunity Act. But\\n\\nthe court also ruled that one of the counterclaims, which was brought on behalf of\\n\\nLisa Phoenix and Periwinkle Partners, had sufficiently alleged that Lisa Phoenix\\n\\nand Periwinkle Partners were “applicants” under the Act, so it denied the motion as\\n\\nto that count.\\n\\n After Regions moved for summary judgment, the district court granted\\n\\nsummary judgment in favor of Regions both for its claims for breach of the\\n\\npromissory notes and guaranties and against the obligors’ counterclaims. The\\n\\ndistrict court ruled that the obligors “do not dispute that they were in default under\\n\\nthe relevant notes and guaranties,” and it ruled that the counterclaims had “no\\n\\nmerit.” With respect to the remaining counterclaim under the Equal Credit\\n\\n\\n 5\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 6 of 88\\n\\n\\nOpportunity Act—the joint claim by Lisa Phoenix and Periwinkle—the district\\n\\ncourt ruled that Periwinkle’s claim of discrimination was “frivolous” because, as\\n\\nan entity, it had no marital status. And the district court ruled that “[t]he claim fails\\n\\nas to Lisa Phoenix as well because, aside from the lack of any evidence to establish\\n\\nany alleged discrimination on the basis of marital status, she was not an ‘applicant’\\n\\nfor the Periwinkle loan[;] she was a guarantor.” The district court referred to its\\n\\nearlier order ruling that guarantors were not “applicants.”\\n\\n The district court later issued a second summary judgment order granting\\n\\nforeclosure on the Periwinkle mortgage. The court then dismissed the matter with\\n\\nprejudice and directed the clerk to enter the judgment. The clerk entered the\\n\\njudgment, and the obligors filed their notice of appeal.\\n\\n Regions moved to amend the judgment to state, among other things, the\\n\\namounts due to Regions from the obligors. The district court granted Regions’\\n\\nmotion in part, instructing the clerk to enter an amended judgment providing for\\n\\nthe following relief:\\n\\n [T]he Court will order the Clerk to amend the Judgment to provide that\\n Regions Bank prevails on its claims against Defendants. The Judgment\\n will further provide that Regions Bank is entitled to recover\\n $540,054.24 from Defendants for the Legal Outsource loan . . . .\\n\\nThe clerk then entered the amended judgment, and the obligors amended their\\n\\nnotice of appeal to include the order granting Regions’ motion to clarify and the\\n\\namended judgment among the items subject to their appeal.\\n 6\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 7 of 88\\n\\n\\n II. STANDARD OF REVIEW\\n\\n We review a grant of summary judgment de novo. Moore ex rel. Moore v.\\n\\nReese, 637 F.3d 1220, 1231 (11th Cir. 2011).\\n\\n III. DISCUSSION\\n\\n This appeal presents several issues about whether the obligors are liable for\\n\\nthe default of the Legal Outsource loan and the Periwinkle loan and mortgage.\\n\\nAlthough the obligors raise a host of issues that seek to obscure the nature of their\\n\\ndefaults, all but one of them lack any merit, and some border on being frivolous.\\n\\nWe decline to address them any further.\\n\\n We divide our discussion of the remaining issues in two parts. First, we\\n\\nexplain that Lisa Phoenix’s counterclaims under the Equal Credit Opportunity Act\\n\\nfail because a guarantor does not qualify as an “applicant” under the Act. Second,\\n\\nwe explain that a limited remand to correct erroneous language from the amended\\n\\njudgment is warranted.\\n\\n A. The District Court Correctly Granted Summary Judgment Against the\\n Equal Credit Opportunity Act Counterclaims by Lisa Phoenix.\\n\\n The district court did not err when it granted summary judgment against the\\n\\ncounterclaims by Lisa Phoenix under the Equal Credit Opportunity Act. As an\\n\\ninitial matter, although the obligors briefly mention Periwinkle’s counterclaim in\\n\\ntheir argument about the Equal Credit Opportunity Act, they have failed to argue\\n\\nor cite caselaw in either the district court or on appeal to rebut the conclusion that\\n\\n 7\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 8 of 88\\n\\n\\nits status as an entity defeats its claim, as the district court ruled, so we consider\\n\\nthat issue abandoned. See Sapuppo v. Allstate Floridian Ins. Co., 739 F.3d 678,\\n\\n680 (11th Cir. 2014) (“[A]n appellant must convince us that every stated ground\\n\\nfor the judgment against him is incorrect.”). We discuss only our reasons for\\n\\nconcluding that the district court correctly granted summary judgment against Lisa\\n\\nPhoenix’s counterclaims on the ground that a guarantor is not an “applicant” for\\n\\ncredit within the meaning of the Act, see 15 U.S.C. § 1691a(b).\\n\\n The Equal Credit Opportunity Act makes it unlawful for “any creditor to\\n\\ndiscriminate against any applicant, with respect to any aspect of a credit transaction\\n\\n. . . on the basis of . . . marital status.” Id. § 1691(a)–(a)(1). The Act defines an\\n\\n“applicant” as “any person who applies to a creditor directly for . . . credit, or\\n\\napplies to a creditor indirectly by use of an existing credit plan for an amount\\n\\nexceeding a previously established credit limit.” Id. § 1691a(b) (emphases added).\\n\\nThe Act initially required the Federal Reserve Board to promulgate regulations to\\n\\nenforce the Act. See 15 U.S.C. § 1691b (1974). And the Federal Reserve Board\\n\\npromulgated Regulation B, which defines an applicant as “any person who\\n\\nrequests or who has received an extension of credit from a creditor,” which\\n\\nincludes “any person who is or may become contractually liable regarding an\\n\\nextension of credit.” 12 C.F.R. § 202.2(e). That regulation further provides that the\\n\\nterm “applicant” includes “guarantors, sureties, endorsers, and similar parties.” Id.\\n\\n\\n 8\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 9 of 88\\n\\n\\n(emphasis added). Regulation B also prohibits a creditor from requiring the\\n\\nsignature of an applicant’s spouse, other than a joint applicant, on any credit\\n\\ninstrument if the applicant independently qualifies as creditworthy. Id.\\n\\n§ 202.7(d)(1).\\n\\n The obligors rely on the definition of “applicant” in Regulation B to argue\\n\\nthat Lisa Phoenix has statutory standing under the Act, so we must determine\\n\\nwhether we should defer to this regulation under the two-step framework\\n\\nannounced in Chevron, U.S.A., Inc. v. Nat’l Res. Def. Council, Inc., 467 U.S. 837\\n\\n(1984). See Arevalo v. U.S. Att’y Gen., 872 F.3d 1184, 1187–88 (11th Cir. 2017).\\n\\nFirst, we ask whether, after applying the “traditional tools of statutory\\n\\nconstruction,” we can determine whether Congress has spoken clearly on the issue.\\n\\nBarton v. U.S. Att’y Gen., 904 F.3d 1294, 1298 (11th Cir. 2018) (quoting Fajardo\\n\\nv. U.S. Att’y Gen., 659 F.3d 1303, 1307 (11th Cir. 2011)). If the statute is\\n\\nunambiguous, we apply it according to its terms and give no deference to the\\n\\nadministrative interpretation. Arevalo, 872 F.3d at 1188. Second, “if the statute is\\n\\nsilent or ambiguous with respect to the specific issue presented, we must then\\n\\ndetermine whether the agency’s interpretation is reasonable or based on a\\n\\npermissible construction of the statute.” Id. An interpretation is reasonable if it is\\n\\n“rational and consistent with the statute.” Id. (quoting Sullivan v. Everhart, 494\\n\\nU.S. 83, 89 (1990)).\\n\\n\\n 9\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 10 of 88\\n\\n\\n In applying the “traditional tools of statutory construction,” we begin “with\\n\\nthe statutory text, and proceed from the understanding that unless otherwise\\n\\ndefined, statutory terms are generally interpreted in accordance with their ordinary\\n\\nmeaning.” Barton, 904 F.3d at 1298 (quoting Sebelius v. Cloer, 569 U.S. 369, 376\\n\\n(2013)); see also Antonin Scalia & Bryan A. Garner, Reading Law: The\\n\\nInterpretation of Legal Texts § 6, at 69 (2012) (“Words are to be understood in\\n\\ntheir ordinary, everyday meaning—unless the context indicates that they bear a\\n\\ntechnical sense.”). And we interpret the words of a statute based on their meaning\\n\\nat the time of enactment. See New Prime Inc. v. Oliveira, 139 S. Ct. 532, 539\\n\\n(2019); Scalia & Garner, Reading Law § 7, at 78 (“Words must be given the\\n\\nmeaning they had when the text was adopted.”).\\n\\n The Act, which was adopted in 1974, defines an applicant as “any person\\n\\nwho applies to a creditor directly for . . . credit, or applies to a creditor indirectly\\n\\nby use of an existing credit plan for an amount exceeding a previously established\\n\\ncredit limit.” 15 U.S.C. § 1691a(b) (emphases added). English-language\\n\\ndictionaries both before and after the enactment define the term “apply” to refer to\\n\\na request for something. See Apply, 1 The Oxford English Dictionary 407 (corr.\\n\\nreprint 1961) (1933) (“To address oneself for information or aid, to have recourse,\\n\\nto make application to”); Apply, 1 The Oxford English Dictionary 577 (2d ed.\\n\\n1989) (same); Apply, Webster’s New International Dictionary of the English\\n\\n\\n 10\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 11 of 88\\n\\n\\nLanguage (Webster’s Second) 132 (2d ed. 1961) (“To make request; to have\\n\\nrecourse with a view to gain something; to solicit”); Apply, Webster’s New\\n\\nCollegiate Dictionary 55 (1977) (“[T]o make an appeal or request esp[ecially] in\\n\\nthe form of a written application.”). So too do legal dictionaries. See Apply, Black’s\\n\\nLaw Dictionary 128 (rev. 4th ed. 1968) (“To make a formal request or petition,\\n\\nusually in writing, to a court, officer, board, or company, for the granting of some\\n\\nfavor, or of some rule or order, which is within his or their power or discretion”);\\n\\nApplication, id. at 127 (“The act of making a request for something”). The Sixth\\n\\nand Eighth Circuits have also both cited a definition of the term “apply” as\\n\\nmeaning “a request . . . usually for something of benefit to oneself.” See Hawkins,\\n\\n761 F.3d at 941 (alteration adopted) (emphasis added) (quoting Webster’s Third\\n\\nNew Int’l Dictionary 105 (2002)); RL BB, 754 F.3d at 385 (quoting Webster’s\\n\\nThird New Int’l Dictionary 105 (1993)); see also Apply, Webster’s Third New Int’l\\n\\nDictionary 105 (1971) (defining “apply” as “to make an appeal or a request\\n\\nesp[ecially] formally and often in writing and usu[ally] for something of benefit to\\n\\noneself”). So, taken together, these definitions suggest that the ordinary meaning of\\n\\nthe term “applicant” is one who requests credit to benefit himself.\\n\\n A guarantor does not fit within this definition. At the time of enactment,\\n\\nEnglish-language dictionaries defined “guaranty” to mean a promise by a\\n\\nguarantor to answer for the payment of some debt if the person liable in the first\\n\\n\\n 11\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 12 of 88\\n\\n\\ninstance is unable to pay. See Guaranty, 4 The Oxford English Dictionary 477\\n\\n(corr. reprint 1961) (1933) (“a written undertaking made by a person (called the\\n\\nguarantor) to be answerable for the payment of a debt or the performance of an\\n\\nobligation by another person, who is in the first instance liable to such payment or\\n\\nobligation”); Guaranty, 6 The Oxford English Dictionary 912 (2d ed. 1989)\\n\\n(same); Guaranty, Webster’s Second 1110 (“An undertaking to answer for the\\n\\npayment of some debt, or the performance of some duty, of another, in case of the\\n\\nfailure of such other to pay or perform”); Guaranty, Webster’s New Collegiate\\n\\nDictionary 509 (“[A]n undertaking to answer for the payment of a debt or the\\n\\nperformance of a duty of another in case of the other’s default or miscarriage”).\\n\\nLegal dictionaries defined “guaranty” the same way. See Guaranty, Black’s Law\\n\\nDictionary 833 (rev. 4th ed.) (“A promise to answer for payment of debt or the\\n\\nperformance of obligation if person liable in the first instance fails to make\\n\\npayment or perform obligation”); Guarantor, Black’s Law Dictionary 634 (5th ed.\\n\\n1979) (“One who becomes secondarily liable for another’s debt or performance”).\\n\\nAlthough a guarantor makes a promise related to an applicant’s request for credit,\\n\\nthe guaranty is not itself a request for credit, and certainly not a request for credit\\n\\nfor the guarantor. See Hawkins, 761 F.3d at 942 (“We find it to be unambiguous\\n\\nthat assuming a secondary, contingent liability does not amount to a request for\\n\\ncredit.”).\\n\\n\\n 12\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 13 of 88\\n\\n\\n To be sure, as the dissent points out, a guarantor’s promise supports a\\n\\nwould-be debtor’s loan application and ordinarily stems from the guarantor’s\\n\\ndesire that the application be granted. See Dissenting Op. at 62–63. But to say that\\n\\na guarantor requests credit by supporting another’s request for credit is to push the\\n\\nbounds of ordinary usage—at the very least, it is to use one word in two obviously\\n\\ndifferent senses. And to say that the guarantor applies for credit by supporting\\n\\nanother’s application is to leave ordinary usage behind entirely.\\n\\n An example should make this point clear. Suppose a high-school senior is\\n\\napplying to her parents’ alma mater, and her parents—who happen to be wealthy\\n\\ndonors—promise the school that they will make a large gift if their daughter is\\n\\nadmitted. The parents’ promise supports the daughter’s application for admission,\\n\\njust as a guarantor’s promise supports a loan applicant’s application for credit. The\\n\\nparents will be grateful if their daughter is admitted, as a guarantor ordinarily is\\n\\ngrateful when the debtor’s application for credit is granted. But it would be\\n\\nunnatural to say that the parents have “applied” for their daughter’s admission or to\\n\\ncall them “applicants” for admission. Under any ordinary use of the word, the\\n\\nstudent is the only “applicant” in this scenario.\\n\\n Applying the whole-text and consistent-usage canons to the Act further\\n\\nconfirms that the term “applicant” excludes guarantors. The whole-text canon\\n\\nrefers to the principle that a “judicial interpreter [should] consider the entire text, in\\n\\n\\n 13\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 14 of 88\\n\\n\\nview of its structure and of the physical and logical relation of its many parts,”\\n\\nwhen interpreting any particular part of the text. Scalia & Garner, Reading Law\\n\\n§ 24, at 167. “Properly applied, it typically establishes that only one of the possible\\n\\nmeanings that a word or phrase can bear is compatible with use of the same word\\n\\nor phrase elsewhere in the statute . . . .” Id. at 168. Closely related to the whole-text\\n\\ncanon is the principle that “[a] word or phrase is presumed to bear the same\\n\\nmeaning throughout a text” unless context requires otherwise. Id. § 25, at 170;\\n\\naccord Atl. Cleaners & Dyers v. United States, 286 U.S. 427, 433 (1932)\\n\\n(“Undoubtedly, there is a natural presumption that identical words used in different\\n\\nparts of the same act are intended to have the same meaning.”).\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434191/\", \"author_raw\": \"WILLIAM PRYOR, Circuit Judge:\"}, {\"author\": \"As Judge Colloton explained in his concurring\", \"type\": \"concurrence\", \"text\": \"As Judge Colloton explained in his concurring opinion in Hawkins, three\\n\\naspects of the statutory text strongly suggest that the term “applicant” is only\\n\\ncompatible with “a first-party applicant who requests credit to benefit herself.”\\n\\nHawkins, 761 F.3d at 943 (Colloton, J., concurring). First, the Act uses the term\\n\\n“applicant” in several provisions that can only refer to a first-party applicant. For\\n\\nexample, section 1691 speaks of a “completed application for credit,” 15 U.S.C.\\n\\n§ 1691(d)(1), and of a creditor taking action in connection with the “applicant’s\\n\\napplication for a loan,” id.§ 1691(e)(1). We agree that “it would be unnatural to\\n\\nconclude that a third party who offers a promise in support of an applicant thereby\\n\\nsubmits what the statute describes as an ‘application for a loan,’ and a ‘completed\\n\\n\\n 14\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 15 of 88\\n\\n\\napplication for credit.’” Hawkins, 761 F.3d at 943–44 (Colloton, J., concurring)\\n\\n(citations omitted). Similarly, the Act provides that within thirty days “after receipt\\n\\nof a completed application for credit, a creditor shall notify the applicant of its\\n\\nactions on the application.” 15 U.S.C. § 1691(d)(1) (emphasis added). “The\\n\\nstatute’s use of the definite article shows that applicant is the single person to\\n\\nwhom credit would be extended, not a third party asking on behalf of the putative\\n\\ndebtor.” Hawkins, 761 F.3d at 943 (Colloton, J., concurring). In contrast with these\\n\\nprovisions, we are aware of no instance in which the Act refers to an “applicant” in\\n\\na context that would naturally suggest that a third-party guarantor could qualify.\\n\\n Second, the statutory definition of “adverse action” on a credit application\\n\\nexcludes from that phrase “a refusal to extend additional credit under an existing\\n\\ncredit arrangement where the applicant is delinquent or otherwise in default.” 15\\n\\nU.S.C. § 1691(d)(6) (emphasis added). This provision suggests that “the applicant”\\n\\nhas received credit and is responsible for making payments on an existing loan. “A\\n\\nguarantor or other third-party requestor does not in ordinary usage become\\n\\n‘delinquent’ or ‘in default’ on a loan or other existing credit arrangement.”\\n\\nHawkins, 761 F.3d at 944 (Colloton, J., concurring). And “if a guarantor could be\\n\\nan ‘applicant,’ then the creditor’s refusal to extend additional credit to a delinquent\\n\\nborrower would be an ‘adverse action’ on the guarantor’s ‘application,’ thus\\n\\nentitling the third-party guarantor to a statement of reasons that the creditor need\\n\\n\\n 15\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 16 of 88\\n\\n\\nnot furnish to the first-party applicant. [15 U.S.C.] § 1691(d)(2).” Id. As Judge\\n\\nColloton concluded, “This is not a natural reading of the text.” Id.\\n\\n Third, the Act recognizes that third parties can be involved in requesting an\\n\\nextension of credit to a first-party applicant, but it “distinguishes between the third-\\n\\nparty requestor and the ‘applicant.’” Id. The Act provides that “[w]here a creditor\\n\\nhas been requested by a third party to make a specific extension of credit directly\\n\\nor indirectly to an applicant, the notification and statement of reasons required by\\n\\nthis subsection may be made directly by such creditor, or indirectly through the\\n\\nthird party, provided in either case that the identity of the creditor is disclosed.” Id.\\n\\n(alterations adopted) (quoting 15 U.S.C. § 1691(d)(4)). That Congress chose to use\\n\\nthe term “applicant” to refer to the party receiving the credit and “third party” to\\n\\nrefer to a separate party who requests an extension of credit for the “applicant” is\\n\\ntelling. In short, after examining the term “applicant” in the context of the statute\\n\\nas a whole, we conclude that there is ample evidence that the term bears the\\n\\nordinary meaning of a person who requests a benefit for himself.\\n\\n Two of the three of our sister circuits that have considered whether the\\n\\nadministrative interpretation of the term “applicant” deserves Chevron deference\\n\\nhave also concluded that the Act unambiguously excludes guarantors. See\\n\\nHawkins, 761 F.3d at 942 (rejecting Chevron deference and explaining “[w]e find\\n\\nit to be unambiguous that assuming a secondary, contingent liability does not\\n\\n\\n 16\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 17 of 88\\n\\n\\namount to a request for credit”); Moran, 476 F.3d at 441 (“[T]here is nothing\\n\\nambiguous about ‘applicant’ and no way to confuse an applicant with a\\n\\nguarantor.”). But see RL BB, 754 F.3d at 385 (applying Chevron deference to\\n\\nsection 202.2(e) on the theory that the term “applies” is ambiguous and that\\n\\nguarantors qualify as requesting credit because they “make formal requests for aid\\n\\nin the form of credit for a third party”). Because we agree with the Seventh and\\n\\nEighth Circuits that the ordinary meaning of “applicant” does not encompass a\\n\\nguarantor, we hold that no deference is due section 202.2(e). So the district court\\n\\ncorrectly granted summary judgment against the counterclaim by Lisa Phoenix’s\\n\\nbecause she was not an “applicant” under the Act.\\n\\n The dissenting opinion disagrees with our analysis of the meaning of the\\n\\nterm “applicant” under the Act on three grounds. First, the dissent argues that the\\n\\nordinary meaning of the word “applicant” reasonably includes guarantors.\\n\\nDissenting Op. at 60–63. Second, the dissent contends that our analysis fails to\\n\\nreflect the “overriding national policy against discrimination that underlies the\\n\\n[ECOA].” Id. at 76 (alteration in original) (internal quotation marks omitted).\\n\\nAlong the way to that conclusion, the dissent relies on two favorites of\\n\\npurposivists: the notion that the Act, as a remedial statute, must be construed\\n\\nbroadly, id. at 49–51, 61, 65–71, 76, and the notion that the words of the Act must\\n\\nbe construed in the light of the Act’s overall purpose, id. at 76. Finally, the dissent\\n\\n\\n 17\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 18 of 88\\n\\n\\ncontends that Congress acquiesced to the Board’s definition of “applicant” by\\n\\nfailing to amend the Act to expressly preclude the Board’s definition. Id. at 79–82.\\n\\nNone of these reasons is persuasive.\\n\\n The dissent’s analysis begins by focusing on how the word “any” appears\\n\\nfour times in two relevant sentences of the Act. Id. at 60–61 (quoting 15 U.S.C.\\n\\n§ 1691(a)(1) (making it “unlawful for any creditor to discriminate against any\\n\\napplication, with respect to any aspect of a credit transaction . . . .” (emphases\\n\\nadded)); id. § 1691a(b) (defining “applicant” to mean “any person who applies to a\\n\\ncreditor directly for . . . credit” (emphasis added))). The dissent argues that the\\n\\nrepeated use of the word “any” suggests that Congress intended for the statute to\\n\\nhave expansive reach and that we should not “engraft artificial limitations” to curb\\n\\nthe “expansive remedial purposes” of the Act. Id. at 61 (quoting Blue Shield of Va.\\n\\nv. McCready, 457 U.S. 465, 472 (1982)). If all the dissent is arguing is that we\\n\\nshould not go beyond the ordinary meaning of the term “applicant” to narrow it\\n\\nartificially, we agree entirely.\\n\\n But the use of the word “any” does not change the meaning of the term\\n\\n“applicant.” We have repeatedly explained that when Congress uses the word\\n\\n“any” without “language limiting the breadth of that word, ‘any’ means all.” CBS\\n\\nInc. v. PrimeTime 24 Joint Venture, 245 F.3d 1217, 1223 (11th Cir. 2001)\\n\\n(alteration adopted) (quoting Merritt v. Dillard Paper Co., 120 F.3d 1181, 1186\\n\\n\\n 18\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 19 of 88\\n\\n\\n(11th Cir. 1997)). So when the Act speaks of an applicant as “any person who\\n\\napplies to a creditor directly for . . . credit,” the word “any” signifies only that all\\n\\npersons who apply to a creditor directly for credit qualify. The word or term that is\\n\\nmodified by “any” is still defined by its ordinary meaning. See, e.g., id. (using\\n\\ndictionary definitions to interpret the word “termination” from “any termination”).\\n\\nSo the relevant question remains what the ordinary meaning of the term “person\\n\\nwho applies to a creditor directly for . . . . credit” includes.\\n\\n The dissent’s attempt to answer this question leans heavily on the definition\\n\\nof “apply” as “to make an appeal or request . . . usually for something of benefit to\\n\\noneself.” Dissenting Op. at 62 (emphasis altered) (quoting Hawkins, 761 F.3d at\\n\\n941 (quoting Webster’s Third New Int’l Dictionary at 105 (1993)). The dissent\\n\\nreasons, “[o]bviously, if something is ‘usually for something of benefit to oneself,’\\n\\nit must sometimes be for something of benefit to another.” Id. (emphasis altered).\\n\\nSo, the dissent concludes, the word “applicant” can fairly be interpreted to include\\n\\na guarantor.\\n\\n Yet the definition the dissent relies on proves the opposite. As Judge\\n\\nColloton pointed out in his concurrence in Hawkins, “under th[e] usual meaning,\\n\\nan ‘applicant’ who ‘applies for credit’ is one who requests credit to benefit herself,\\n\\nnot credit to benefit a third party. That there are unusual meanings of ‘apply’ that\\n\\nencompass making a request on behalf of another is not sufficient to make a term\\n\\n\\n 19\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 20 of 88\\n\\n\\nambiguous for purposes of Chevron.” 761 F.3d at 943 (Colloton, J., concurring).\\n\\nThe only circumstance in which it is reasonable to construe a term according to an\\n\\nunusual meaning is when the context makes the unusual meaning a natural one.\\n\\nBut, as we have explained, there is nothing natural about calling a guarantor an\\n\\napplicant for credit, and the whole text of the Act makes that usage even less\\n\\nplausible.\\n\\n The dissent charges that our reading of the Act fails to apply the whole-text\\n\\ncanon, Dissenting Op. at 65–66, 66 n.20, but the dissent’s assertion is notably\\n\\nlacking in references to the text. According to the dissent, if we viewed the Act as\\n\\na whole, we would see that “‘[t]he overriding national policy against\\n\\ndiscrimination that underlies the [Act]’ means that ‘we cannot give’ words in that\\n\\nstatute a ‘narrow interpretation.’” Id. at 76 (quoting Bros. v. First Leasing, 724\\n\\nF.2d 789, 793 (9th Cir. 1984)). But apart from its logically irrelevant reliance on\\n\\nthe word “any,” the dissent fails to point to any other provisions of the Act that\\n\\nsuggest that the term “applicant” includes a third party who requests a benefit for\\n\\nthe first-party applicant. And it is hornbook abuse of the whole-text canon to argue\\n\\n“that since the overall purpose of the statute is to achieve x, any interpretation of\\n\\nthe text that limits the achieving of x must be disfavored.” Scalia & Garner,\\n\\nReading Law § 24, at 168.\\n\\n\\n\\n\\n 20\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 21 of 88\\n\\n\\n The dissent also contends that we fail to reconcile our reading of the text\\n\\nwith the “familiar canon of statutory construction that remedial legislation should\\n\\nbe construed broadly to effectuate its purposes.” Dissenting Op. at 65 (internal\\n\\nquotation marks omitted) (quoting Tcherepnin v. Knight, 389 U.S. 332, 336\\n\\n(1967)). Indeed, we do not apply this so-called canon because it is of dubious\\n\\nvalue. An eight-member majority of the Supreme Court has ridiculed it as “th[e]\\n\\nlast redoubt of losing causes,” Dir., Office of Workers’ Comp. Programs, Dep’t of\\n\\nLabor v. Newport News Shipbuilding & Dry Dock Co., 514 U.S. 122, 135 (1995),\\n\\nand the Court has rejected applying it in a number of other decisions since 1995.\\n\\nSee, e.g., CTS Corp. v. Waldburger, 573 U.S. 1, 12 (2014) (“The Court of Appeals\\n\\nsupported its interpretation of [section] 9658 by invoking the proposition that\\n\\nremedial statutes should be interpreted in a liberal manner. The Court of Appeals\\n\\nwas in error when it treated this as a substitute for a conclusion grounded in the\\n\\nstatute’s text and structure.”); Norfolk S. Ry. Co. v. Sorrell, 549 U.S. 158, 171\\n\\n(2007) (rejecting argument based on remedial-purpose canon and explaining that\\n\\nalthough a statute’s remedial purpose was to benefit employees, “this remedial\\n\\npurpose [does not] require[] us to interpret every uncertainty in the Act in favor of\\n\\nemployees”); Household Credit Servs., Inc. v. Pfennig, 541 U.S. 232, 237, 244\\n\\n(2004) (reversing court of appeals that relied on remedial-purpose canon to broadly\\n\\ninterpret the term “finance charge” from the Truth in Lending Act); Inyo Cty. v.\\n\\n\\n 21\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 22 of 88\\n\\n\\nPaiute-Shoshone Indians of the Bishop Cmty. of the Bishop Colony, 538 U.S. 701,\\n\\n710–12 (2003) (rejecting argument based on the remedial-purpose canon that the\\n\\nterm “person” should be construed broadly under a statute to include Native\\n\\nAmerican tribes). As the Supreme Court has recently explained, the fundamental\\n\\nproblem with this so-called canon is its indeterminate coverage, as “almost every\\n\\nstatute might be described as remedial in the sense that all statutes are designed to\\n\\nremedy some problem.” CTS Corp., 573 U.S. at 12. And the “canon” too is of\\n\\nindeterminate effect because, “even if the Court [has] identified some subset of\\n\\nstatutes as especially remedial, [it] has emphasized that no legislation pursues its\\n\\npurposes at all costs.” Id. (citation and internal quotation marks omitted).\\n\\n Indeed, it is hard to imagine a more widely criticized “canon” of\\n\\ninterpretation. A leading treatise has labeled it a “false” canon and has explained\\n\\nthat it is “an open invitation to engage in ‘purposive’ rather than textual\\n\\ninterpretation, and generally to engage in judicial improvisation.” Scalia & Garner,\\n\\nReading Law § 64, at 364–66. And jurists as varied as Antonin Scalia and Richard\\n\\nPosner share the same view. See Antonin Scalia, Assorted Canards of\\n\\nContemporary Legal Analysis, 40 Case W. Res. L. Rev. 581, 581 (1990) (calling\\n\\nthe canon one of “the prime examples of lego-babble”); Richard A. Posner,\\n\\nStatutory Interpretation—in the Classroom and in the Courtroom, 50 U. Chi. L.\\n\\nRev. 800, 809 (1983) (explaining that the canon is “unrealistic about legislative\\n\\n\\n 22\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 23 of 88\\n\\n\\nobjectives” because it assumes that legislatures pursue a single remedial purpose,\\n\\nwhen in reality a “statute [often] is a compromise between one group of legislators\\n\\nthat holds a simple remedial objective but lacks a majority and another group that\\n\\nhas reservations about the objective”). We agree with these authorities that we\\n\\nshould not employ this false canon to contravene the text of the Act.\\n\\n The dissent also insists that we must “construe the literal language of the\\n\\n[Act]” in the light of the “overriding national policy against discrimination that\\n\\nunderlies the [Act].” Dissenting Op. at 76 (quoting First Leasing, 724 F.2d at 793).\\n\\nWe take no issue with the dissent’s explanation of the vital role that women have\\n\\nplayed in our nation’s history, of the discrimination they have faced in obtaining\\n\\ncredit, and that one of the purposes of the Act was to remedy this discrimination.\\n\\nSee Dissenting Op. at 51–60. But the dissent ignores that “[n]o legislation pursues\\n\\nits purposes at all costs,” and that “it frustrates rather than effectuates legislative\\n\\nintent simplistically to assume that whatever furthers the statute’s primary\\n\\nobjective must be the law.” Pension Benefit Guar. Corp. v. LTV Corp., 496 U.S.\\n\\n633, 646–47 (1990) (quoting Rodriguez v. United States, 480 U.S. 522, 525–26\\n\\n(1987)). When a statute includes limiting provisions, those provisions “are no less\\n\\na reflection of the genuine ‘purpose’ of the statute than the operative provisions,\\n\\nand it is not the court’s function to alter the legislative compromise.” Scalia &\\n\\nGarner, Reading Law at 21. Here, Congress created a right that runs only to\\n\\n\\n 23\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 24 of 88\\n\\n\\n“applicants.” And as the Seventh Circuit explained in Moran, because the term\\n\\n“applicant” unambiguously does not encompass “guarantors,” reading the statute\\n\\nin the way the dissent does would frustrate the limitations that Congress imposed\\n\\non statutory standing and “opens vistas of liability” that Congress did not envision.\\n\\n476 F.3d at 441.\\n\\n The dissent’s final substantive argument is that Congress has impliedly\\n\\nadopted the Board’s definition of “applicant” by amending the Act without\\n\\nchanging the statutory definition during the 30 years since Regulation B was first\\n\\npromulgated. Dissenting Op. at 79–82. Although the Supreme Court has\\n\\nrecognized that congressional inaction can, in limited circumstances, support an\\n\\ninference that Congress has acquiesced to an agency or judicial interpretation, it\\n\\nhas explained that “[l]egislative silence is a poor beacon to follow” in construing a\\n\\nstatute. Zuber v. Allen, 396 U.S. 168, 185 (1969). And it has repeatedly warned\\n\\nthat congressional silence alone is ordinarily not enough to infer acquiescence. See\\n\\nSolid Waste Agency of N. Cook Cty. v. U.S. Army Corps of Eng’rs, 531 U.S. 159,\\n\\n169 (2001) (“Although we have recognized congressional acquiescence to\\n\\nadministrative interpretations of a statute in some situations, we have done so with\\n\\nextreme care.”); United States v. Riverside Bayview Homes, Inc., 474 U.S. 121,\\n\\n137 (1985) (“[W]e are chary of attributing significance to Congress’ failure to\\n\\n\\n\\n\\n 24\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 25 of 88\\n\\n\\nact . . . .”); Bob Jones Univ. v. United States, 461 U.S. 574, 600 (1983)\\n\\n(“Nonaction by Congress is not often a useful guide . . . .”).\\n\\n The dissent cites Texas Department of Housing & Community Affairs v.\\n\\nInclusive Communities Project, Inc., 135 S. Ct. 2507 (2015), in arguing that we\\n\\nshould infer congressional acquiescence, Dissenting Op. at 79–80, but this reliance\\n\\nis misplaced. There, the Supreme Court held that when Congress amended the Fair\\n\\nHousing Act in 1988, it was aware of and adopted the unanimous view of nine\\n\\ncircuits that had considered the matter and concluded that the statute authorized\\n\\ndisparate-impact claims. See Texas Dep’t of Hous., 135 S. Ct. at 2519.\\n\\n The dissenting opinion contends that until 2014, the “vast majority of\\n\\ncourts” that had “examined” the issue held that guarantors had standing under the\\n\\nEqual Credit Opportunity Act, Dissenting Op. at 80 (quoting RL BB, 754 F.3d at\\n\\n386), but this argument is misleading. Before 2007, several federal and state courts\\n\\napplied Regulation B, but they did so without discussing whether it was entitled to\\n\\ndeference. See, e.g., Silverman v. Eastrich Multiple Inv’r Fund, L.P., 51 F.3d 28,\\n\\n30–31 (3d Cir. 1995) (assuming without discussion that the Board’s interpretation\\n\\nwas valid); Fed. Deposit Ins. Corp. v. Medmark, Inc., 897 F. Supp. 511, 514 (D.\\n\\nKan. 1995) (same); see also United States v. L.A. Tucker Truck Lines, Inc., 344\\n\\nU.S. 33, 38 (1952) (explaining that a decision is not precedential with respect to an\\n\\nissue “not there raised in briefs or argument nor discussed in the opinion of the\\n\\n\\n 25\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 26 of 88\\n\\n\\nCourt”). To our knowledge, the first court to “examine[]” whether the Board’s\\n\\ndefinition deserved deference was the Seventh Circuit in Moran, in 2007, and it\\n\\nconcluded that it did not. See 476 F.3d. at 441. By the time Congress next amended\\n\\nthe Act in July 2010, the only other courts to opine on the issue were a federal\\n\\ndistrict court, which agreed with Moran and ruled that guarantors lack statutory\\n\\nstanding, see Champion Bank v. Reg’l Dev., LLC, No. 4:08-cv-1807-CDP, 2009\\n\\nWL 1351122, at *3 (E.D. Mo. May 13, 2009) (reasoning that “a guarantor does\\n\\nnot, by definition, apply for anything”), and the Supreme Court of Iowa, which\\n\\napplied Regulation B without considering Moran or whether it was entitled to\\n\\ndeference, see Bank of the West v. Kline, 782 N.W.2d 453, 458 (Iowa 2010). So\\n\\nwhen Congress amended the Act in 2010, the weight of reasoned authority was\\n\\nagainst the Board’s definition of applicant. And Congress’s prior amendments to\\n\\nthe Act took place before any court had considered the validity of Regulation B.\\n\\nThis situation obviously is nothing like the unanimous, reasoned precedent of nine\\n\\ncircuits in favor of an agency interpretation featured in Texas Department of\\n\\nHousing. We can hardly infer congressional acquiescence in this circumstance.\\n\\n The dissent advances one other objection to our analysis: that it “opine[s] on\\n\\nan issue that . . . Lisa Phoenix and Periwinkle . . . never raised on appeal.”\\n\\nDissenting Op. at 35. As the dissent sees it, Lisa Phoenix has abandoned both of\\n\\n\\n\\n\\n 26\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 27 of 88\\n\\n\\nher counterclaims by failing to raise them “plainly and prominently” enough in the\\n\\nobligors’ initial brief. Id. at 38 (quoting Sapuppo, 739 F.3d at 681). We disagree.\\n\\n To be sure, the obligors’ briefing could most charitably be described as\\n\\nclumsy. We emphatically “do not condone the unartful way in which [the obligors]\\n\\nha[ve] stated and argued the issues on this appeal.” Fed. Sav. & Loan Ins. Corp. v.\\n\\nHaralson, 813 F.2d 370, 373 n.3. (11th Cir. 1987). Even so, reading their initial\\n\\nbrief in the light of the record, the other briefs in this appeal, and the principle that\\n\\n“briefs should be read liberally to ascertain the issues raised on appeal,” Allstate\\n\\nIns. Co. v. Swann, 27 F.3d 1539, 1542 (11th Cir. 1994), we have no doubt that Lisa\\n\\nPhoenix has fairly presented the argument that the district court erred when it\\n\\ndismissed at least one of her counterclaims relating to the Periwinkle loan based on\\n\\nher status as a guarantor.\\n\\n Consider what happened in the district court. The obligors pleaded four\\n\\ncounterclaims under the Equal Credit Opportunity Act based on the Periwinkle\\n\\nloan: one by Charles Phoenix, one by Legal Outsource, one by Lisa Phoenix and\\n\\nPeriwinkle Partners jointly, and one by Lisa Phoenix individually. Citing Hawkins\\n\\nfor the proposition that guarantors are not applicants, the district court dismissed\\n\\nthe claims by Charles Phoenix, Legal Outsource, and Lisa Phoenix individually on\\n\\nthe ground that those claims rested solely on guarantor standing. But the district\\n\\ncourt withheld judgment on the joint claim by Lisa Phoenix and Periwinkle\\n\\n\\n 27\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 28 of 88\\n\\n\\nbecause that claim—counterclaim 11—appeared to allege that both Lisa Phoenix\\n\\nand Periwinkle were applicants. After further considering the matter, the district\\n\\ncourt granted summary judgment against counterclaim 11 on the grounds that Lisa\\n\\nPhoenix lacked standing and that Periwinkle had no marital status. The district\\n\\ncourt also mentioned “the lack of any evidence to establish any alleged\\n\\ndiscrimination on the basis of marital status,” which was a sufficient alternative\\n\\nbasis for the summary judgment, although the district court did not clearly\\n\\ndesignate it as such.\\n\\n Now consider the briefs. In a discrete section of their initial brief, the\\n\\nobligors protest the district court’s “reli[ance] . . . on the Eighth Circuit’s ruling in\\n\\nHawkins” “to substantiate dismissing Lisa Phoenix’s and Periwinkle’s claims”\\n\\nunder the Equal Credit Opportunity Act. Initial Br. of Appellants at 30. And they\\n\\nargue that the regulation defining the term “applicant” to include guarantors, see 12\\n\\nC.F.R. § 202.2(e), is a valid exercise of regulatory power to implement the Act.\\n\\nInitial Br. of Appellants at 31–32. They conclude that the claims by “Lisa Phoenix\\n\\nand Periwinkle” “fall squarely within ECOA’s protections.” Id. at 32.\\n\\n Logically, this argument can relate only to counterclaim 11—the joint claim\\n\\nby Lisa Phoenix and Periwinkle—or counterclaim 12—Lisa Phoenix’s individual\\n\\nclaim as a guarantor of the Periwinkle loan. That the brief refers to both Lisa\\n\\nPhoenix and Periwinkle suggests that the obligors might have counterclaim 11 in\\n\\n\\n 28\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 29 of 88\\n\\n\\nmind, but the district court did not rely on Hawkins to dismiss any claim by\\n\\nPeriwinkle. But, with respect to Lisa Phoenix, the argument responds squarely to\\n\\nthe primary basis on which the district court dismissed her share of counterclaim\\n\\n11 and the sole basis on which it dismissed her counterclaim 12. In this\\n\\ncircumstance, we cannot agree with the dissent that “no claim is properly before us\\n\\non appeal.” Dissenting Op. at 35.\\n\\n Consider the consequences if we accepted the obligors’ argument—that is, if\\n\\nthe obligors convinced us that the district court erred when it “relied . . . on the\\n\\nEighth Circuit’s ruling in Hawkins to substantiate dismissing Lisa Phoenix’s and\\n\\nPeriwinkle’s claims” because, contrary to Hawkins, guarantors are applicants under\\n\\nthe Act. Initial Br. of Appellants at 30. In that case, they necessarily would have\\n\\nconvinced us that “every stated ground for the judgment against [counterclaim 12]\\n\\nis incorrect,” which is exactly what we have said an appellant must do “[t]o obtain\\n\\nreversal of a district court judgment.” Sapuppo, 739 F.3d at 680. They also would\\n\\nhave convinced us that the primary and arguably the only “stated ground for the\\n\\njudgment against [counterclaim 11] is incorrect.” Id. So, unless some alternative\\n\\nground for affirmance appeared from Regions’ brief or from the record, accepting\\n\\nthe obligors’ argument would require us to reverse the dismissal of counterclaim\\n\\n12 at least and perhaps counterclaim 11 as well.\\n\\n\\n\\n\\n 29\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 30 of 88\\n\\n\\n We consider it telling that Regions agrees with us, at least in part, that the\\n\\nobligors have raised the issue of guarantor standing. In its brief, Regions contends\\n\\nthat the obligors “have waived or abandoned any issue or argument with respect to\\n\\nthe dismissal of [counterclaims] IX, X, and XII.” Br. of Appellee Regions Bank at\\n\\n34. But Regions does not contend that the obligors have abandoned counterclaim\\n\\n11 or the general issue of guarantor standing. On the contrary, it reads the obligors’\\n\\nbrief to “contend . . . that summary judgment for Regions on Counterclaim count\\n\\nXI was in error because, under the definition of ‘applicant’ supplied in Regulation\\n\\nB [the governing regulation], Mrs. Phoenix possessed standing to sue in her\\n\\ncapacity as a guarantor of the Periwinkle loan.” Id. Regions proceeds to argue on\\n\\nthe merits that the district court correctly granted summary judgment against\\n\\ncounterclaim 11 because the obligors have produced no evidence of discrimination\\n\\nand because guarantors are not applicants. See id. at 35–42.\\n\\n That Regions does not share the dissent’s view that Lisa Phoenix\\n\\n“indisputably abandoned” counterclaim 11 is relevant in two respects. Dissenting\\n\\nOp. at 41. First, even if Lisa Phoenix waived or forfeited counterclaim 11 in her\\n\\ninitial brief, Regions has waived or forfeited the waiver or forfeiture by conceding\\n\\nin its own brief that she raised it. And nothing Regions said at oral argument can\\n\\nundo that concession. See APA Excelsior III L.P. v. Premiere Techs., Inc., 476 F.3d\\n\\n\\n\\n\\n 30\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 31 of 88\\n\\n\\n1261, 1269 (11th Cir. 2007) (“[W]e do not consider claims not raised in a party’s\\n\\ninitial brief and made for the first time at oral argument.”).\\n\\n The second way in which Regions’ concession matters is that the main\\n\\nprinciple that animates the abandonment rule is fair notice. As we have explained,\\n\\n“an appellee is entitled to rely on the content of an appellant’s brief for the scope\\n\\nof the issues appealed.” Access Now, Inc. v. Sw. Airlines Co., 385 F.3d 1324, 1330\\n\\n(11th Cir. 2004) (quoting Pignons S.A. de Mecanique v. Polaroid Corp., 701 F.2d\\n\\n1, 3 (1st Cir. 1983)); accord Haralson, 813 F.2d at 373 n.3. That Regions thought\\n\\nthe obligors raised counterclaim 11 in their brief—even as it concluded that they\\n\\nhad abandoned other claims—suggests that the obligors’ brief gave Regions fair\\n\\nnotice that counterclaim 11 was an issue.\\n\\n To be sure, an appellee’s assertions about what is or is not abandoned do not\\n\\nbind us. Claims of abandonment may often be overstated; in this appeal, we are not\\n\\nsure we agree with Regions that Lisa Phoenix has abandoned counterclaim 12. And\\n\\nthere are many reasons why appellees may fail to raise meritorious abandonment\\n\\narguments. But the point remains: that Regions understood the obligors’ brief to\\n\\npresent a live argument about counterclaim 11 and the general issue of guarantor\\n\\nstanding is surely evidence that it is reasonably understood to do so.\\n\\n Consider also that both amicus briefs—one by the Consumer Financial\\n\\nProtection Bureau and one by five bankers associations—fully briefed the issue of\\n\\n\\n 31\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 32 of 88\\n\\n\\nLisa Phoenix’s guarantor standing under the Act. Neither brief ever hinted at the\\n\\npossibility that Lisa Phoenix abandoned her counterclaims. That all of the amici,\\n\\nlike Regions, consider Lisa Phoenix to have adequately raised the issue of her\\n\\nstatutory standing reinforces our conclusion that Lisa Phoenix did not abandon her\\n\\nargument about counterclaims 11 and 12.\\n\\n Our dissenting colleague’s contention that Lisa Phoenix has abandoned her\\n\\ncounterclaims turns on her reading—which we respectfully submit is an\\n\\noverreading—of three sentences in the obligors’ brief, each of which seems to\\n\\nsuggest that the Equal Credit Opportunity Act claim on appeal is that Regions\\n\\nviolated the Act when it allegedly demanded that Lisa Phoenix guarantee the\\n\\nOutsource loan. See Dissenting Op. at 38–39. The obligors’ statement of the issues\\n\\npresents the question, “Does a wife who refuses to collateralize loans or guaranty\\n\\nher husband’s unsecured business debt have Equal Credit Opportunity Act\\n\\n(ECOA) standing as an ‘applicant’ . . . ?” Initial Br. of Appellants at 1 (emphasis\\n\\nadded). Then, in their argument section, the obligors protest that excluding\\n\\nguarantors from the definition of “applicant” “gives lenders an untethered license\\n\\nto require spouses to collateralize and sign for their husbands’ loans with\\n\\nimpunity,” and they contend that Lisa Phoenix was required “to sign for and\\n\\ncollateralize her husband’s business debts.” Id. at 30, 32 (emphases added). As the\\n\\ndissent sees it, these phrases can mean only that Lisa Phoenix has replaced her\\n\\n\\n 32\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 33 of 88\\n\\n\\ncounterclaims based on the Periwinkle claim with a wholly novel claim based on\\n\\nRegions’ alleged demand that she guarantee the Outsource loan, even though the\\n\\nobligors never pleaded such a claim in the district court and no such claim is\\n\\nconcerned in the judgment on appeal, and even though it would be nonsensical for\\n\\nLisa Phoenix to assert guarantor standing with respect to a loan she never\\n\\nguaranteed.\\n\\n Although we certainly agree that the obligors’ brief is no model of clarity,\\n\\nwe do not think that its references to the Outsource loan nullify Lisa Phoenix’s\\n\\nchallenge to the ground on which her counterclaims were dismissed. In the\\n\\nobligors’ telling of the facts, Regions demanded that Lisa Phoenix collateralize the\\n\\nOutsource loan, she refused, and Regions punished her refusal by declaring falsely\\n\\nthat the Periwinkle loan was in default. We read the phrases on which the dissent\\n\\nleans as highlighting the crucial narrative role of Regions’ demand and Lisa\\n\\nPhoenix’s refusal in the obligors’ version of the facts, not as obliterating her legal\\n\\ntheory. And we do not see how we can read them any other way without violating\\n\\nthe rule that “briefs should be read liberally to ascertain the issues raised on\\n\\nappeal.” Swann, 27 F.3d at 1542. The abandonment rule would swell to a scope\\n\\npreviously unimagined if we were to hold that an appellant abandons her legal\\n\\ntheory whenever she places undue emphasis on one part of her factual narrative\\n\\nover another. We conclude that Lisa Phoenix has preserved at least one of her\\n\\n\\n 33\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 34 of 88\\n\\n\\ncounterclaims under the Equal Credit Opportunity Act and that the issue of\\n\\nguarantor standing is before us. And, for the reasons we have explained, we hold\\n\\nthat the district court correctly granted summary judgment against those\\n\\ncounterclaims because a guarantor is not an “applicant” for credit under the Act.\\n\\n B. The Amended Judgment Must Be Corrected on Remand.\\n\\n The amended judgment states “that Regions Bank is entitled to recover\\n\\n$540,054.24 from Defendants for the Legal Outsource loan.” As both parties\\n\\nagreed at oral argument, the counts regarding the Legal Outsource loan name only\\n\\nCharles Phoenix and Legal Outsource as defendants, so the judgment erroneously\\n\\nstates that Lisa Phoenix and Periwinkle Partners are also liable for the Outsource\\n\\nloan. We remand with instructions to correct the judgment to state that only\\n\\nCharles Phoenix and Legal Outsource are liable for the damages owed for the\\n\\ndefault of the Outsource loan.\\n\\n IV. CONCLUSION\\n\\n We AFFIRM the summary judgment in favor of Regions, and we\\n\\nREMAND with instructions to correct the judgment.\\n\\n\\n\\n\\n 34\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 35 of 88\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434191/\", \"author_raw\": \"As Judge Colloton explained in his concurring\"}, {\"author\": \"ROSENBAUM, Circuit Judge, concurring\", \"type\": \"concurrence\", \"text\": \"ROSENBAUM, Circuit Judge, concurring in part and dissenting in part:\\n\\n Today we opine on an issue that Appellants Lisa Phoenix and Periwinkle\\n\\nPartners LLC (“Periwinkle”) never raised on appeal. Courts have noted that\\n\\ndeciding an issue no appellant raised is generally unwise. But the Majority Opinion\\n\\nnevertheless insists that we do so. And in following this course, it purports to\\n\\nconstrict the Equal Credit Opportunity Act (“ECOA”) to preclude it from\\n\\naccomplishing one of its primary remedial goals: disentangling spouses’ financial\\n\\nintertwinement when such intertwinement is not necessary. I therefore feel it\\n\\nnecessary to explain the problems with the Majority Opinion’s analysis.\\n\\n Section I of this dissent demonstrates that no claim is properly before us on\\n\\nappeal. And Section II responds to the Majority Opinion’s incorrect conclusion that\\n\\nguarantors lack standing as “applicants” under the ECOA.\\n\\n I.\\n\\n At different points in this litigation, Defendants-Counterclaimants-Appellants\\n\\nLisa Phoenix and Periwinkle have arguably raised two ECOA claims possibly\\n\\nrelevant to this appeal. 1 Each claim involves a separate Regions Bank loan.\\n\\n\\n\\n\\n 1\\n The other Defendants-Counterclaimants-Appellants’ claims are not relevant to the ECOA\\nanalysis the Majority Opinion has chosen to conduct, so I do not address them here.\\n\\n 35\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 36 of 88\\n\\n\\n The Majority Opinion construes the claim on appeal to concern a loan Regions\\n\\nmade to Periwinkle, a company Lisa Phoenix indirectly owns. As the Majority\\n\\nOpinion interprets this claim, Lisa Phoenix invokes the ECOA to contest Regions’s\\n\\ndemand that her husband, Charles Phoenix, and his law firm, Legal Outsource PA,\\n\\nguaranty the Periwinkle loan (“Periwinkle Loan Claim”).\\n\\n The second potential ECOA claim possibly relevant on this appeal involves a\\n\\nloan Regions made to Legal Outsource. Under this potential claim, Lisa 2 asserts that\\n\\nRegions violated the ECOA when it required her (through her company Periwinkle)\\n\\nto guaranty a loan it had made to her husband Charles’s business, Legal Outsource\\n\\n(“Legal Outsource Loan Claim”).\\n\\n But neither the Periwinkle Loan Claim nor the Legal Outsource Loan Claim\\n\\nis properly presented on appeal.\\n\\n To understand why, it makes sense to start by looking at the sole ECOA claim\\n\\nLisa identified in the issues she presented on appeal: “Does a wife who refuses to\\n\\ncollateralize loans or guaranty her husband’s unsecured business debt have\\n\\n[ECOA] standing as an ‘applicant’ to sue if the lender then forces over a dozen\\n\\ntechnical defaults as pretense to falsely accelerate and foreclose on her separate\\n\\nsecured loan?” Appellants’ Br. at 1 (emphasis added). This language\\n\\nunambiguously seeks to assert Lisa’s status as an “applicant” on the Legal Outsource\\n\\n\\n 2\\n To avoid confusion, I refer to Charles and Lisa Phoenix by their first names.\\n 36\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 37 of 88\\n\\n\\nLoan Claim, as a result of Periwinkle’s legal status as a guarantor for that loan, since\\n\\nonly the Legal Outsource Loan Claim involved an alleged demand by Regions to\\n\\nguaranty Lisa’s husband’s business debt. The language of the issue as Lisa has\\n\\nphrased it on appeal does not implicate the Periwinkle Loan Claim, since that was a\\n\\nloan involving Lisa’s own business, not her husband’s business.\\n\\n Lisa’s actual argument in support of her ECOA claim also leaves no doubt\\n\\nthat she challenges only Regions’s demand that she (through her company\\n\\nPeriwinkle) guaranty her husband’s Legal Outsource Loan. Her argument is short;\\n\\nit’s less than three pages. But in that space she repeatedly argues that Regions\\n\\nviolated the ECOA by trying to make her guaranty her husband’s loans. Appellants’\\n\\nBr. at 30, 32. Unfortunately for Lisa, though, she did not raise the Legal Outsource\\n\\nLoan Claim in the district court.\\n\\n As for the Periwinkle Loan Claim, Lisa never once argues it in this appeal.\\n\\nTrue, the loan to Periwinkle comes up in her briefs, but only in the context of\\n\\narguments that the Majority Opinion (correctly) decides are meritless, if not\\n\\n“frivolous.” Maj. Op. at 7. Significantly, she never argues that the district court was\\n\\nwrong in finding she lacked standing as a guarantor to contest the Periwinkle Loan\\n\\nClaim.\\n\\n So what exactly is properly on appeal?\\n\\n Nothing.\\n\\n\\n 37\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 38 of 88\\n\\n\\n We have repeatedly explained that “[a]ny issue that an appellant wants the\\n\\nCourt to address should be specifically and clearly identified in the brief.” Access\\n\\nNow, Inc. v. Sw. Airlines Co., 385 F.3d 1324, 1330 (11th Cir. 2004).3 Not only that,\\n\\nbut “[a] party fails to adequately brief a claim when he does not plainly and\\n\\nprominently raise it, for instance by devoting a discrete section of his argument to\\n\\nthose claims.” Sapuppo v. Allstate Floridian Ins. Co., 739 F.3d 678, 681 (11th Cir.\\n\\n2014) (citation and internal quotation marks omitted). When that happens, “the\\n\\nissue—even if properly preserved at trial—will be considered abandoned.” Access\\n\\nNow, 385 F.3d at 1330 (citation and internal quotation marks omitted).\\n\\n Here, Lisa’s briefs certainly do not “plainly and prominently raise” her\\n\\nPeriwinkle Loan Claim. For starters, her opening brief does not “devot[e] a discrete\\n\\nsection of [the] argument” to it. In fact, as I have noted, it does not even identify the\\n\\nclaim in an issue on appeal or ever expressly mention the Periwinkle Loan Claim in\\n\\nits ECOA arguments. And her reply brief does not address the issue at all.\\n\\n Nor, contrary to the Majority Opinion’s suggestion, does Lisa’s mention of\\n\\nthe district court’s reliance on Hawkins v. Community Bank of Raymore, 761 F.3d\\n\\n937 (8th Cir. 2014), aff’d by an equally divided Court, 136 S. Ct. 1072 (2016), justify\\n\\nthe Majority Opinion’s determination of an issue that Lisa did not raise on appeal.\\n\\n\\n 3\\n The Federal Rules of Appellate Procedure similarly require an appellant’s brief to\\n“contain, under appropriate headings and in the order indicated . . . a statement of the issues\\npresented for review.” Fed. R. App. P. 28(a)(5).\\n 38\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 39 of 88\\n\\n\\nTo be sure, the district court did invoke Hawkins to dismiss counterclaims relating\\n\\nto the Periwinkle Loan Claim. Maj. Op. at 27. But Lisa’s brief leaves no doubt that\\n\\nshe raised the district court’s reliance on Hawkins for only an altogether different\\n\\nand unrelated point: she cited the district court’s mention of Hawkins solely to\\n\\nsupport her argument concerning the Legal Outsource Claim. 4\\n\\n In fact, in the very next paragraph following the general reference to Hawkins\\n\\nthe Majority Opinion cites, see Maj. Op. at 27, Lisa characterized Hawkins as\\n\\n“grant[ing] lenders an untethered license to require spouses to collateralize and sign\\n\\nfor their husbands’ loans with impunity.” Appellants’ Br. at 30 (emphasis added).\\n\\nAnd while Lisa argued that she and Periwinkle “fall squarely within ECOA’s\\n\\nprotections,” Maj. Op. at 28 (quoting Appellants’ Br. at 32), she made clear that she\\n\\nmeant that the ECOA protected her (through her ownership of Periwinkle) and\\n\\nPeriwinkle as guarantors of the Legal Outsource Loan: she said that “Regions\\n\\nrequired [her through Periwinkle] to sign for and collateralize her husband’s\\n\\nbusiness debts hence violating ECOA prohibitions.” Appellants’ Br. at 32 (emphasis\\n\\nadded).\\n\\n There’s no doubt that Lisa argued that guarantors have standing under the\\n\\nECOA, but her brief makes clear that she made this argument in the context of the\\n\\n\\n 4\\n Lest there be any question about what Lisa argued on appeal, I have included as an\\nappendix to this opinion the statement of issues from Lisa’s opening brief, as well as the entirety\\nof her ECOA argument.\\n 39\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 40 of 88\\n\\n\\nLegal Outsource Loan Claim—that is, that under the ECOA, Regions could not\\n\\nrequire her individually or through her company to guarantee her husband’s\\n\\ncompany Legal Outsource’s loan. But as I have noted, the problem with that\\n\\nargument is Lisa never raised it in the district court.\\n\\n As for the Periwinkle Loan Claim she did raise in the district court, the most\\n\\nwe can say about that is that Lisa’s brief on appeal nakedly cites her Counterclaim\\n\\n11, which in turn, alleged the Periwinkle Loan Claim in the district court. But\\n\\nsignificantly, Lisa’s brief cites Counterclaim 11 in the context of arguing the Legal\\n\\nOutsource Loan Claim, in an apparent effort to suggest that the Legal Outsource\\n\\nLoan Claim was properly raised in the district court. (It was not.)\\n\\n And even if we ignore the context of the citations to Counterclaim 11,\\n\\n(literally) bare citations alone do not preserve an argument on appeal. We have\\n\\nfound abandonment when litigants have done more to preserve their claims, such as\\n\\n“when [at least] passing references appear in the argument section of an opening\\n\\nbrief,” and when “they are buried within . . . arguments.” Sapuppo, 739 F.3d at 682.\\n\\nAs this Court has recently “remind[ed] the . . . bar[,] . . . specific factual and legal\\n\\nargumentation at every stage of . . . proceedings” is “importan[t]” to preserving\\n\\nissues on appeal. United States v. Corbett, 921 F.3d 1032, 1043 (11th Cir. 2019)\\n\\n\\n\\n\\n 40\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 41 of 88\\n\\n\\n(W. Pryor, Newsom, Vratil, JJ.).5 So under our unequivocal precedent, Lisa’s\\n\\nbriefing indisputably abandoned the Periwinkle Loan Claim.\\n\\n Lisa’s counsel likewise did not raise the Periwinkle Loan Claim issue at all\\n\\nduring his initial oral argument. And during his rebuttal, while Lisa’s counsel briefly\\n\\naddressed the “ECOA issue,” he argued only that Lisa had raised a triable issue of\\n\\nfact that Regions had required Lisa “to provide collateral for her husband’s loan.”\\n\\nOral Argument at 24:04, 24:25, Regions Bank v. Legal Outsource PA, et al. (No. 17-\\n\\n11736), http://www.ca11.uscourts.gov/oral-argument-recordings?title=17-11736&\\n\\nfield_oar_case_name_value=&field_oral_argument_date_value%5Bvalue%5D%\\n\\n\\n\\n\\n 5\\n Indeed, this Court has consistently enforced the abandonment rule. See, e.g., Quality\\nAuto Painting Ctr. of Roselle, Inc. v. State Farm Indem. Co., 917 F.3d 1249, 1259 n.9 (11th Cir.\\n2019) (en banc); Nance v. Warden, Ga. Diagnostic Prison, 922 F.3d 1298, 1302 n.2 (11th Cir.\\n2019) (Tjoflat, E. Carnes, W. Pryor, JJ.); Hornsby-Culpepper v. Ware, 906 F.3d 1302, 1306 n.1\\n(11th Cir. 2018) (W. Pryor, Branch, Anderson, JJ.); United States v. Wenxia Man, 891 F.3d 1253,\\n1275 (11th Cir. 2018) (W. Pryor, J. Pryor, Black, JJ.); Feldman v. Am. Dawn, Inc., 849 F.3d 1333,\\n1345 (11th Cir. 2017) (W. Pryor, Jordan, Baldock, JJ.); United States v. Stein, 846 F.3d 1135, 1151\\nn.15 (11th Cir. 2017) (W. Pryor, J. Pryor, Story, JJ.); Jeune v. U.S. Att’y Gen., 810 F.3d 792, 797\\nn.2 (11th Cir. 2016) (W. Pryor, J. Carnes, Siler, JJ.); Sorrels v. NCL (Bah.) Ltd., 796 F.3d 1275,\\n1283 (11th Cir. 2015) (W. Pryor, Jordan, Jones, JJ.); In re McFarland, 790 F.3d 1182, 1191 n.10\\n(11th Cir. 2015) (Tjoflat, W. Pryor, Baldock, JJ.); Perez v. U.S. Bureau of Citizenship &\\nImmigration Servs., 774 F.3d 960, 964 (11th Cir. 2014) (per curiam) (Marcus, W. Pryor, Fay, JJ.).\\nBut puzzlingly, today, the Majority Opinion selectively applies the abandonment rule to correctly\\ndecline ruling on whether a corporation can be an “applicant” under the ECOA, since “the obligors\\nbriefly mention Periwinkle’s counterclaim in their argument about the Equal Credit Opportunity\\nAct, [but] they have failed to argue or cite caselaw in either the district court or on appeal to rebut\\nthe conclusion that its status as an entity defeats its claim, as the district court ruled . . .” Maj. Op.\\nat 7–8. Yet the Majority Opinion incorrectly stretches to “resolve” a specific claim Lisa does not\\neven raise at all in her briefs on appeal—whether Lisa may rely on the ECOA to contest Regions’s\\ndemand that her husband, Charles, and his law firm, Legal Outsource PA, guaranty the loan to\\nPeriwinkle.\\n\\n\\n 41\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 42 of 88\\n\\n\\n5Byear%5D=&field_oral_argument_date_value%5Bvalue%5D%5Bmonth%5D=\\n\\n(“Oral Argument”).\\n\\n The only time during this appeal that either party raised Lisa’s role as a\\n\\nguarantor of the Periwinkle Loan was in Regions’s opposition brief. Appellee’s Br.\\n\\nat 39–42. According to the Majority Opinion, this makes the argument fair game.\\n\\nMaj. Op. at 30–31. But here’s the problem: Lisa never took the bait. She didn’t\\n\\nadopt Regions’s characterization of her argument. She didn’t respond to Regions’s\\n\\nargument in her reply brief. Her counsel did not agree with it during oral argument.\\n\\nAnd by the time Regions’s counsel got to oral argument, Regions argued—over no\\n\\nclaim to the contrary—that Lisa had abandoned the Periwinkle Loan Claim. 6 So on\\n\\nappeal, no one ever argued in favor of Lisa and Periwinkle’s position on the\\n\\nPeriwinkle Loan Claim.\\n\\n To recap, Lisa never argued in her opening brief that she had standing as a\\n\\nguarantor to challenge the loan made to Periwinkle. Regions raised that argument\\n\\nin its opposition brief, but Lisa did not respond to it in her reply brief. Lisa’s counsel\\n\\ndid not raise the argument during oral argument. Regions’s counsel then argued that\\n\\n\\n 6\\n See Oral Argument at 13:48 (“If you look at the counterclaim that was brought before the\\ndistrict court and what has been argued here in the briefs, it’s [a] completely different factual\\nsituation. They haven’t even advanced the right argument in our position. . . .”) (emphasis added),\\n18:02 (“[T]hey’re arguing that there was discrimination because when the Legal Outsource loan\\nmatured and Regions Bank said we can’t extend the maturity on this because Legal Outsource is\\ndefunct, it’s not credit-worthy, you need to provide an additional source of repayment collateral—\\nsomething to that effect—they’re saying in the brief, which they never raised in front of Judge\\nMagnuson, that that is the ECOA violation.”) (emphasis added).\\n 42\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 43 of 88\\n\\n\\nLisa had abandoned this argument. And Lisa’s counsel never disputed this. It’s not\\n\\non appeal.\\n\\n Perhaps for this reason, the Majority Opinion introduces the novel suggestion\\n\\nthat amici can decide what issues are before us. Maj. Op. at 31–32. Amici frequently\\n\\nprovide valuable insight to the court by “supplementing the efforts of private counsel\\n\\nand by drawing the court’s attention to law that might otherwise escape\\n\\nconsideration[.]” Shoemaker v. City of Howell, 795 F.3d 553, 562 (6th Cir. 2015)\\n\\n(quoting 3-28 Moore’s Manual—Federal Practice and Procedure § 28.84 (2014)).\\n\\nWe are grateful for their contribution. But we don’t let them select the issues. 7 See\\n\\nCellnet Commc’ns, Inc. v. F.C.C., 149 F.3d 429, 443 (6th Cir. 1998) (“To the extent\\n\\nthat the amicus raises issues or make[s] arguments that exceed those properly raised\\n\\nby the parties, we may not consider such issues.”).\\n\\n\\n 7\\n And besides, the Majority Opinion is wrong when it suggests that the amicus briefs\\nsupport its conclusion that Appellants raised the ECOA standing issue in the context of the\\nPeriwinkle Loan claim. See Maj. Op. at 31–32. On the contrary, the banking associations’ amicus\\nbrief plainly understood Appellants to have raised the ECOA standing issue as it relates to only\\nthe Legal Outsource Loan Claim on appeal. Their brief describes the issue Appellants raise as\\nfollows: “Appellants incorrectly argue that the Equal Credit Opportunity Act affords ‘Lisa\\nPhoenix and Perwinkle [sic] [Partners, LLC]’ a claim that Regions [Bank] required them to sign\\nfor and collateralize her husband’s business debts[.]” Bankers’ Br. at 10 (emphasis added)\\n(alterations in original). Clearly, this pertains to only the Legal Outsource Loan Claim on appeal.\\nIt does not reference the Periwinkle Loan Claim. To be sure, the Bankers briefed the issue of\\nwhether a guarantor has standing, but according to their own words, they did so in the context of\\nthe Legal Outsource Loan Claim because that’s what they understood Appellants to argue in their\\nbrief. The Bankers obviously did not understand Appellants to raise the guarantor-standing issue\\nas it relates to the Periwinkle Loan Claim. For good reason: on appeal, Appellants never argued\\nthat issue in the context of the Periwinkle Loan Claim. As for the other amicus brief, submitted\\nby the Consumer Financial Protection Bureau, it focused on what it disagreed with in the district\\ncourt’s opinion, as opposed to characterizing Appellants’ argument on appeal.\\n 43\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 44 of 88\\n\\n\\n Other courts have identified only very limited circumstances warranting a\\n\\ncourt’s decision to overlook abandonment of an argument or issue and decide it,\\n\\nanyway. None applies here.\\n\\n In Silber v. United States, 370 U.S. 717, 717–18 (1962), for example, the\\n\\nSupreme Court explained that when a party fails to raise an issue on appeal, courts\\n\\n“[i]n exceptional circumstances, especially in criminal cases, . . .in the public\\n\\ninterest, may, of their own motion, notice errors to which no exception has been\\n\\ntaken, if the errors are obvious, or if they otherwise seriously affect the fairness,\\n\\nintegrity, or public reputation of judicial proceedings.” (emphasis added) (citation\\n\\nand quotation marks omitted). The Majority Opinion has offered no reason why this\\n\\ncase presents “exceptional circumstances,” and I am aware of no such reason. This\\n\\nis also not a criminal case. Nor, in light of the fact that the Supreme Court recently\\n\\nsplit 4-4 on the issue the Majority insists on addressing today, Hawkins, 136 S. Ct.\\n\\nat 1072, and other circuits are similarly split two to one, can we say the answer to\\n\\nthe unraised issue is “obvious.” And finally, since the Majority Opinion’s answer\\n\\nto the unraised issue results in precisely the same outcome as declaring the issue\\n\\nabandoned—either way, the district court is affirmed—I can see no way that the\\n\\nunraised issue “seriously affect[s]”—or, for that matter, affects at all—“the fairness,\\n\\nintegrity, or public reputation of judicial proceedings.”\\n\\n\\n\\n\\n 44\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 45 of 88\\n\\n\\n Other Circuits have also identified extremely limited circumstances in which\\n\\nit might be appropriate for an appellate court to address an issue that no party raised.\\n\\nThese other Circuits have reasoned that a court may determine an unraised issue\\n\\nwhere doing so would “enhanc[e] the efficiency of the decisionmaking process and\\n\\nthe conservation of scarce judicial resources,” and “remand ‘would further postpone\\n\\nthe ultimate resolution of’ the petitioner’s underlying claim for relief.” United States\\n\\nv. Holness, 706 F.3d 579, 592 (4th Cir. 2013) (quoting LaBruna v. U.S. Marshal,\\n\\n665 F.2d 439, 442 (2d Cir. 1981)). Similarly, the Seventh Circuit has concluded it\\n\\nis appropriate to review an unraised issue only when these controlling factors\\n\\ndecidedly favor review: the size and complexity of the record, the certainty of the\\n\\nresolution of the issue, and the need for “protracted, costly, and ultimately futile\\n\\nproceedings in the district court,” in the absence of addressing the forfeited issue.\\n\\nUnited States v. Giovannetti, 928 F.2d 225, 227 (7th Cir. 1991).\\n\\n Of course, not one of these circumstances describes this case. The record here\\n\\nis not particularly lengthy or complex; resolution of the issue here, as I have noted,\\n\\nis not cut and dry; and declining to reach the issue would not “result in protracted,\\n\\ncostly, and ultimately futile proceedings in the district court.” Id. Whether we\\n\\naddress the unraised issue or not, the result is the same: affirmance of the district\\n\\ncourt’s order granting summary judgment to Regions. In fact, ironically, deciding\\n\\nthe unraised issue here makes today’s opinion unnecessarily lengthy and complex.\\n\\n\\n 45\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 46 of 88\\n\\n\\nSo it should go without saying that this case does not present an appropriate vehicle\\n\\nto justify a wholly unnecessary journey into the deeply debated question of whether\\n\\n“guarantors” can be “applicants,” when the claim as resolved by the Majority\\n\\nOpinion should be dismissed, regardless, since it was clearly abandoned.\\n\\n Apparently recognizing the inadvisability of determining an issue no\\n\\nAppellant invoked and failing to convince even itself that Appellants have raised the\\n\\nPeriwinkle Loan Claim on appeal, the Majority Opinion tries a different tack and\\n\\ninsists that Appellants’ Legal Outsource Loan Claim arguments “logically . . . can\\n\\nrelate only” to Counterclaims 11 or 12 (Lisa’s individual claim as a guarantor of the\\n\\nPeriwinkle loan). Maj. Op. at 28. At the risk of repetition but for the avoidance of\\n\\ndoubt, I offer a simpler explanation: on appeal, Lisa, for the first time in this case,\\n\\nhas raised a new claim about Regions’s attempt to make her (through Periwinkle)\\n\\nguaranty the Legal Outsource Loan. She did not plead the claim in her\\n\\nCounterclaims—including Counterclaims 11 or 12—and she did not otherwise\\n\\nsufficiently develop it in the district court.\\n\\n Perhaps it was “nonsensical” for Lisa to raise the Legal Outsource Loan Claim\\n\\nat this point, Maj. Op. at 33, but that was her prerogative, not ours. Our task here is\\n\\nto “decide . . . [the] questions presented by the parties.” Greenlaw v. United States,\\n\\n554 U.S. 237, 244 (2008) (citation and quotation marks omitted). Our responsibility\\n\\nto construe briefs “liberally” does not grant us permission to recast what was\\n\\n\\n 46\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 47 of 88\\n\\n\\nappealed simply because we want to reach an issue that the parties left behind. See\\n\\nNat’l Aeronautics and Space Admin. v. Nelson, 562 U.S. 134, 147 n.10 (2011) (“The\\n\\npremise of our adversarial system is that appellate courts do not sit as self-directed\\n\\nboards of legal inquiry and research, but essentially as arbiters of legal questions\\n\\npresented and argued by the parties before them.”) (quoting Carducci v. Regan, 714\\n\\nF.2d 171, 177 (D.C. Cir. 1983) (opinion for the court by Scalia, J.)); Greenlaw, 554\\n\\nU.S. at 243 (“In our adversary system . . . we rely on the parties to frame the issues\\n\\nfor decision and assign to courts the role of neutral arbiter of matters the parties\\n\\npresent.”).\\n\\n As for the Legal Outsource Loan Claim, this is hardly the first time a litigant\\n\\nhas tried to raise a “wholly novel claim” for the first time on appeal. Maj. Op. at 33.\\n\\nWhen that happens, it has long been this Court’s practice to refuse to consider the\\n\\nissue. See In re Holywell Corp., 874 F.2d 780, 782 (11th Cir. 1989) (refusing to\\n\\nconsider claim that was not presented to the district court); Troxler v. Owens-Illinois,\\n\\nInc., 717 F.2d 530, 532 (11th Cir. 1983) (refusing to consider affirmative defense\\n\\nraised for the first time on appeal); see also Access Now, 385 F.3d at 1331 (“[W]e\\n\\nreview claims of judicial error in the trial courts. If we were to regularly address\\n\\nquestions—particularly fact-bound issues—that districts court never had a chance to\\n\\nexamine, we would not only waste our resources, but also deviate from the essential\\n\\nnature, purpose, and competence of an appellate court.”).\\n\\n\\n 47\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 48 of 88\\n\\n\\n Because Lisa raises the Legal Outsource Loan Claim for the first time on\\n\\nappeal, the better course would be to forgo considering it at this late hour. Access\\n\\nNow, 385 F.3d at 1330–35. We can, however, make exceptions to our general\\n\\nprohibition against considering new claims for the first time on appeal if one of our\\n\\nlimited exceptions to the rule applies. 8 See id. at 1332. The only exception that\\n\\ncould even possibly pertain here allows us to evaluate a new issue for the first time\\n\\non appeal “if that issue presents significant questions of general impact or of great\\n\\npublic concern.” Id. at 1332.\\n\\n Perhaps we could rely on that exception to consider whether guarantors are\\n\\nconsidered “applicants” under the ECOA in the context of Lisa’s contentions arising\\n\\nout of the Legal Outsource Loan Claim. But if we are going to do that, we should\\n\\nsay so and explain why the exception applies.\\n\\n Unfortunately, though, that’s not what the Majority Opinion does. Instead,\\n\\ninexplicably, it simply considers whether guarantors have standing under the ECOA,\\n\\nin the context of “resolving” the Periwinkle Loan Claim, ignoring Lisa’s\\n\\nabandonment of that claim on appeal and Regions’s argument that Lisa has forsaken\\n\\nthat claim for good by not appealing it. Because Lisa opted not to present the\\n\\n\\n 8\\n We have explained that we may consider an issue raised for the first time on appeal (1)\\n“if it involves a pure question of law, and if refusal to consider it would result in a miscarriage of\\njustice”; (2) if “the appellant raises an objection to an order which he had no opportunity to raise\\nat the district court level”; (3) if “the interest of substantial justice is at stake”; (4) if “the proper\\nresolution is beyond any doubt”; or (5) “if that issue presents significant questions of general\\nimpact or of great public concern.” Access Now, 385 F.3d at 1332.\\n 48\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 49 of 88\\n\\n\\nPeriwinkle Loan Claim to us on appeal, it should be clear that that claim leaves\\n\\nnothing to be “resolved.”\\n\\n II.\\n\\n Nevertheless, because the Majority Opinion purports to improperly narrow\\n\\nthe ECOA, I must explain why its reasoning is wrong. I begin by setting forth the\\n\\nframework we apply to the question the Majority Opinion has chosen to address:\\n\\nwhether guarantors are included within the meaning of “applicants” under the\\n\\nECOA.\\n\\n Since we are construing the meaning of a statute, we must start with the text.\\n\\nRobinson v. Shell Oil Co., 519 U.S. 337, 340 (1997). We consider whether the text\\n\\nplainly and unambiguously answers the particular question at issue. Id. In assessing\\n\\nthe text, we keep in mind that “[a] word in a statute may or may not extend to the\\n\\nouter limits of its definitional possibilities.” Dolan v. U.S. Postal Serv., 546 U.S.\\n\\n481, 486 (2006). To determine the proper meaning, we evaluate “the whole statutory\\n\\ntext, considering the purpose and context of the statute, and consult[] any precedents\\n\\nor authorities that inform the analysis.” Id.; see also Crandon v. United States, 494\\n\\nU.S. 152, 158 (1990) (“In determining the meaning of the statute, we look not only\\n\\nto the particular statutory language, but to the design of the statute as a whole and to\\n\\nits object and policy.”). And when a statute is remedial in nature, we apply the well-\\n\\nestablished rule that requires us to construe the text “broadly to effectuate its\\n\\n\\n 49\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 50 of 88\\n\\n\\npurposes.” Tcherepnin v. Knight, 389 U.S. 332, 336 (1967); Stewart v. Kahn, 78\\n\\nU.S. 493, 504 (1870); SEC v. Levin, 849 F.3d 995, 1001 (11th Cir. 2017) (observing\\n\\nthat remedial legislation “is entitled to a broad construction”) (citations omitted). If,\\n\\nafter we apply these rules, the text is unambiguous, yielding but a single possible\\n\\nanswer to the question at issue, our analysis ends. For we “must give effect to the\\n\\nunambiguously expressed intent of Congress.” Cadet v. Bulger, 377 F.3d 1173,\\n\\n1185 (11th Cir. 2004) (citation and quotation marks omitted).\\n\\n But if the text does not unambiguously answer the precise question we are\\n\\nexamining, we must conduct an analysis under Chevron, U.S.A., Inc. v. Nat. Res.\\n\\nDef. Council, Inc., 467 U.S. 837 (1984). Under Chevron, we determine whether the\\n\\nagency charged with filling in the ECOA’s gaps—originally the Federal Reserve\\n\\nBoard but beginning in 2010, the Consumer Financial Protection Bureau 9—has\\n\\nestablished a permissible answer to our question. The agency’s interpretation is\\n\\npermissible if it is “reasonable.” Id. at 844. And we must defer to an agency’s\\n\\nreasonable construction of a statute for which that agency enjoys rulemaking\\n\\nauthority. Id. If, on the other hand, the agency’s interpretation of the statute as it\\n\\n\\n\\n 9\\n In 2010, Congress transferred the Federal Reserve’s rulemaking authority to the\\nConsumer Financial Protection Bureau. 15 U.S.C. 1691b(a). The Bureau has since repromulgated\\nRegulation B without making material changes. 12 C.F.R. Pt. 1002 & Supp. I; see Equal Credit\\nOpportunity (Regulation B), 76 Fed. Reg. 79,442 (Dec. 21, 2011). Because Regulation B has\\nremained materially the same since its original promulgation by the Federal Reserve, to avoid\\nconfusion, for the remainder of this opinion, I refer solely to the Federal Reserve as the\\nadministering body.\\n 50\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 51 of 88\\n\\n\\nconcerns the precise issue in controversy is “arbitrary, capricious, or manifestly\\n\\ncontrary to the statute,” the agency’s answer is entitled to no weight. Id.\\n\\n With this framework in mind, I consider whether a guarantor qualifies as an\\n\\n“applicant” under the ECOA. I divide my discussion into three parts. In accordance\\n\\nwith the Supreme Court’s guidance for construing statutory text, see Dolan, 546 U.S.\\n\\nat 486, and because the ECOA has “broad remedial goals,” Barney v. Holzer Clinic,\\n\\nLtd., 110 F.3d 1207, 1211 n.6 (6th Cir. 1997), Section A reviews the background,\\n\\npurpose, and broader context of the ECOA as a whole. Cf. Zuni Pub. Sch. Dist.\\n\\nNo.89 v. Dep’t of Educ., 550 U.S. 81, 89–90 (2007) (departing from usual order of\\n\\nanalysis for purposes of exposition). Section B examines the statutory text. And\\n\\nSection C considers the Federal Reserve’s approach to answering the question of\\n\\nwhether a guarantor qualifies as an “applicant” under the ECOA.\\n\\n A.\\n\\n Women were integral to the United States’s World War II triumph through,\\n\\namong other ways, their efforts in the workforce. 10 Buoyed by their experiences\\n\\nduring the War, women helped detonate America’s historic Post-War economic\\n\\nboom by flocking to workplaces in unprecedented droves.11\\n\\n\\n 10\\n Jone Johnson Lewis, Women and World War II: Women at Work, ThoughtCo, Mar. 5,\\n2019, available at https://www.thoughtco.com/world-war-ii-women-at-work-3530690 (last\\nvisited Aug. 27, 2019).\\n 11\\n Howard N. Fullerton, Jr., Labor Force Participation: 75 Years of Change, 1950-98 and\\n1998-2025, Monthly Lab. Rev. 3 (Dec. 1999), available at http://www.bls.gov/mlr/1999/12/\\nart1full.pdf (last visited Aug. 27, 2019).\\n 51\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 52 of 88\\n\\n\\n But significant impediments to women’s economic parity nonetheless\\n\\npersisted. A particularly important obstacle involved credit availability: creditors\\n\\nwould not lend to women so they could start their own businesses or buy their own\\n\\nhomes because creditors refused to consider women’s applications on par with\\n\\nmen’s. And the increasing necessity of credit to achieve economic security made\\n\\nthis impediment particularly restrictive.12\\n\\n That women faced rampant discrimination in accessing credit was no secret.\\n\\nIn fact, in the early 1970s, the then-President of the American Bankers Association\\n\\nopenly admitted that banks “do in fact discriminate against women when it comes\\n\\nto granting credit.”13 And in 1972, the National Commission on Consumer Finance\\n\\nfound five systemic patterns of gender-based credit discrimination: (1) single\\n\\nwomen had more trouble obtaining credit than single men; (2) creditors required\\n\\nwomen to reapply for credit upon marriage, usually in the husband’s name; (3)\\n\\ncreditors declined to extend credit to a married woman in her own name; (4) creditors\\n\\noften refused to count the wife’s income when the couple applied for credit; and (5)\\n\\nwomen who divorced or widowed had trouble reestablishing credit since the\\n\\n\\n\\n\\n 12\\n Dubravka Ritter, Do We Still Need the Equal Credit Opportunity Act?, Fed. Res. Bank\\nof Phila. (Sept. 2012), available at https://www.philadelphiafed.org/-/media/consumer-finance-\\ninstitute/payment-cards-center/publications/discussion-papers/2012/d-2012-equal-credit-\\nopportunity-act.pdf?la=en (last visited Aug. 27, 2019).\\n 13\\n Susan Smith Blakely, Credit Opportunity for Women: The ECOA and Its Effects, 1981\\nWis. L. Rev. 655, 657 (1981) (citation omitted).\\n 52\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 53 of 88\\n\\n\\naccounts were in the husband’s name. Nat’l Comm. On Consumer Fin., Consumer\\n\\nCredit in the United States, 152–53 (1972).\\n\\n So before marriage, regardless of a woman’s wealth accumulation, creditors\\n\\nviewed women as poor credits risks because they considered them “inherently\\n\\nunstable and incapable of handling [their] own affairs . . . and likely to change [their]\\n\\nmarital status.” Gail R. Reizenstein, A Fresh Look at the Equal Credit Opportunity\\n\\nAct, 14 Akron L. Rev. 215, 226 (1980) (citation omitted) (“Fresh Look”). Many\\n\\ncreditors even demanded that women disclose the type of birth control they relied\\n\\nupon. Credit Equality Comes to Women: An Analysis of the Equal Credit\\n\\nOpportunity Act, 13 San Diego L. Rev. 960, 965 (1976). Some creditors took it yet\\n\\na step further by making women choose between babies and credit by conditioning\\n\\na woman’s credit approval on her “swear . . . that [she] would not endanger [her]\\n\\nability to repay [her] debt[] by having children.” Id. (citations omitted).\\n\\n Getting married usually didn’t improve women’s lots in credit markets\\n\\nbecause lenders often folded whatever credit women had before marriage into their\\n\\nhusbands’ credit. For instance, after winning multiple Wimbledon championships\\n\\nand supporting her household with those earnings, Billie Jean King could not get a\\n\\ncredit card in her own name; rather, she had to apply in the name of her husband,\\n\\nwho at that time, was an unemployed law student. Gail Collins, When Everything\\n\\nChanged: The Amazing Journey of American Women from 1960 to the Present, 182\\n\\n\\n 53\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 54 of 88\\n\\n\\n(Little, Brown & Co. ed., 2014). And bankers told the mayor of Davenport, Iowa,\\n\\nthat they would give her a credit card only if her husband would sign for it. Id.\\n\\n This state of affairs created a paradox for women: they couldn’t have their\\n\\nown credit during marriage, and if they divorced or became widowed, lenders would\\n\\ndeny their applications because they lacked established credit of their own.\\n\\nEconomic Problems of Women: Hearings Before the Joint Economic Committee,\\n\\n93d Cong. 152, 1st Sess. (1973) (“The irony of these credit practices is that when a\\n\\nwoman is divorced, separated, or widowed she often is denied credit by these same\\n\\ncredit companies on the grounds that she has no established credit record.”)\\n\\n(Statement of U.S. Rep. Griffiths, Member, Joint Econ. Comm.) (“Economic\\n\\nProblems”). And many women started post-marriage life even worse off than\\n\\nhaving no credit because lenders insisted on tying women’s financial fortunes to\\n\\nthose of their ex-husbands, and their ex-husbands’ delinquent accounts often\\n\\ndetracted from the women’s credit scores even after the marriage ended. See Fresh\\n\\nLook at 225.\\n\\n This atmosphere evolved out of the wrong 14 but “widely-held” presumptions\\n\\nthat “probability of pregnancy, the subsequent termination of employment upon\\n\\n\\n 14\\n Despite higher levels of unemployment than men at the time, women were, in fact, better\\ncredit risks than men, studies showed. A study commissioned in 1964 examined the possible\\ncorrelation between consumer credit risk and sex. It found that out of 8,795 credit accounts\\nestablished for single men, 176 defaulted (2%); while among 4,337 accounts established for single\\nwomen, only 33 defaulted (0.75%). See Sharon Thornton, The Not-So-Equal Credit Opportunity\\nAct, 5 Orange Cty. B.J. 363, 366 (1978).\\n 54\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 55 of 88\\n\\n\\nchildbirth, and the general instability and inability of women to control their personal\\n\\naffairs (especially single and divorced women)” made women bad bets. Fresh Look\\n\\nat 216 (citation omitted).\\n\\n Eventually, Congress began to recognize the inequities of tethering women’s\\n\\neconomic prospects to their husbands and the risks this discrimination posed to the\\n\\nnew economy. So in the early 1970s, Congress addressed the problem.\\n\\n Testimony before Congress established that at that time, to obtain credit,\\n\\nwomen needed a higher salary and more stable employment than men. See Senate\\n\\nComm. On Banking, Housing & Urban Affairs, S. Rep. No. 93–278, 93rd Cong., 1st\\n\\nSess. 16–17 (1973) (“Senate Rpt.”); see also Economic Problems at 197; Credit\\n\\nDiscrimination: Hearings Before the Subcomm. on Consumer Affairs of the House\\n\\nComm. on Banking and Currency on H.R. 14856 and H.R. 14908, 93rd Cong., 2d\\n\\nSess. 315, 636 (1974).15 Creditors operated under the assumption that women were\\n\\neconomically dependent upon their husbands, even if in reality, the wife’s earnings\\n\\noutpaced her husband’s. See Senate Rpt. at 17–18. Creditors also usually altered a\\n\\nwife’s credit rating to match her husband’s and frequently refused to give married\\n\\nwomen separate accounts. See id. at 17.\\n\\n\\n\\n\\n 15\\n During the hearings, one person relayed that an employee of a credit institution had stated\\nthat it was “un-American to count a woman’s income and that the only way a woman’s income\\ncould be counted would be if she were to have a hysterectomy.” Economic Problems at 192\\n(quotation marks omitted).\\n 55\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 56 of 88\\n\\n\\n The congressional hearings culminated in the Senate Subcommittee’s report\\n\\nthat found thirteen different “widespread” ways creditors discriminated on the basis\\n\\nof sex and marital status:\\n\\n (1) Women were held to different standards than men in obtaining\\n credit;\\n\\n (2) Creditors generally required a woman upon marriage to reapply for\\n credit;\\n\\n (3) Creditors often refused to extend credit to a married woman in her\\n own name;\\n\\n (4) Creditors were usually unwilling to consider the wife’s income\\n when a married couple applied for credit;\\n\\n (5) Women who had separated had a particularly difficult time, since\\n the accounts may still have been in the husband’s name;\\n\\n (6) Creditors arbitrarily refused to consider alimony and child support\\n as a valid source of income;\\n\\n (7) Creditors applied stricter standards to married applicants where the\\n wife was the high earner;\\n\\n (8) Creditors used information about women’s birth-control practices\\n in evaluating credit applications;\\n\\n (9) Creditors used information concerning the creditworthiness of a\\n spouse where an otherwise-creditworthy married person applied\\n for credit;\\n\\n (10) Creditors refused to issue separate accounts to married persons;\\n\\n (11) Creditors considered women dependent upon their husbands even\\n if the women were employed;\\n\\n\\n\\n 56\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 57 of 88\\n\\n\\n (12) Creditors used credit-scoring systems that applied different values\\n depending on sex or marital status; and\\n\\n (13) Creditors altered women’s credit ratings to match their husbands’\\n credit ratings.\\nSenate Rpt. at 16–17.\\n\\n Against this background and in light of these findings, Congress enacted the\\n\\nECOA in 1974. In doing so, Congress concluded sex and marital status “are, and\\n\\nmust be, irrelevant to a credit judgment.” Senate Rep. (Banking Hous. and Urban\\n\\nAffairs Comm.), S. Rep. No. 94 – 589, 94th Cong., 2d Sess. 3 (1976), reprinted in\\n\\n1976 U.S.C.C.A.N. 403, 405; see also 15 U.S.C. § 1691 (articulating the ECOA’s\\n\\nprohibition of credit discrimination based on gender or marital status).\\n\\n So it is not surprising that the ECOA makes it “unlawful for any creditor to\\n\\ndiscriminate against any applicant, with respect to any aspect of a credit transaction\\n\\n[] on the basis of . . . sex or marital status. . . .” 15 U.S.C. § 1691(a)(1). 16 Indeed,\\n\\nthe ECOA represents Congress’s resolve to “eradicate credit discrimination waged\\n\\nagainst women, especially married women whom creditors traditionally refused to\\n\\nconsider for individual credit,” Anderson v. United Fin. Co., 666 F.2d 1274, 1277\\n\\n(9th Cir. 1982), and “to prevent loans from being conditioned automatically on the\\n\\n\\n\\n\\n 16\\n In 1976, Congress extended the ECOA to bar creditors from discriminating on the basis\\nof race, color, religion and national origin. See 15 U.S.C. § 1691(a); Ritter, supra, at 2–3.\\n 57\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 58 of 88\\n\\n\\nsecuring of the signature of the non-borrowing spouse,” Mayes v. Chrysler Credit\\n\\nCorp., 167 F.3d 675, 676 (1st Cir. 1999).\\n\\n Against the crazy quilt of discriminatory practices creditors were using at that\\n\\ntime, Congress declined to spell out each methodology or action that amounts to\\n\\nimpermissible discrimination. Instead, Congress rested on the ECOA’s broad\\n\\nlanguage and “entrust[ed] its construction to an agency with the necessary\\n\\nexperience and resources to monitor its operation.” Mourning v. Family Publ’ns\\n\\nServ., Inc., 411 U.S. 356, 365 (1973). Specifically, Congress delegated to the\\n\\nFederal Reserve the task of adopting “classifications, differentiation, or other\\n\\nprovision[s],” that were, in the Federal Reserve’s judgment, “necessary or proper to\\n\\neffectuate the [ECOA’s] purposes. . . .” 15 U.S.C. § 1691b(a).\\n\\n Shortly after Congress enacted the ECOA, the Federal Reserve took the baton\\n\\nfrom Congress and promulgated Regulation B to implement Congress’s directive to\\n\\ncombat sex- and marital-status-based credit discrimination. 12 C.F.R. § 202.1(b)\\n\\n(stating that the purpose of Regulation B is “to promote the availability of credit to\\n\\nall creditworthy applicants without regard to . . . sex [or] marital status. . . .”).\\n\\nIncluded within Regulation B is the Spousal Guaranty Rule, which generally\\n\\nprohibits creditors from treating married people differently by requiring spouses to\\n\\nassume liability for each other’s debt obligations. 12 C.F.R. 202.7(d); Equal Credit\\n\\nOpportunity Final Rulemaking, 40 Fed. Reg. 49,308–09 (Oct. 22, 1975). Therefore,\\n\\n\\n 58\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 59 of 88\\n\\n\\never since 1977, Regulation B has prohibited creditors from obligating an individual\\n\\nto guaranty her spouse’s debts.\\n\\n Some courts, though, interpreted Regulation B to mean that when a creditor\\n\\nviolated the Spousal Guaranty Rule, the only “applicant” who suffered\\n\\ndiscrimination was the primary borrower, not the spouse who guaranteed the loan.\\n\\nSee, e.g., Morse v. Mut. Fed. Sav. & Loan Ass’n of Whitman, 536 F. Supp. 1271,\\n\\n1278 (D. Mass. 1982). The Federal Reserve recognized that this restrictive reading\\n\\nthwarted one of the purposes of the ECOA—ensuring access to an independent\\n\\ncredit profile by detangling spouses’ credit—since the guarantor spouse could be\\n\\nsaddled with her husband’s bad debt with no recourse under the ECOA.\\n\\n So in 1985, the Federal Reserve expressly construed the ECOA’s definition\\n\\nof “applicant”—“any person who applies to a creditor directly for an extension,\\n\\nrenewal, or continuation of credit, or applies to a creditor indirectly by use of an\\n\\nexisting credit plan for an amount exceeding a previously established credit limit,”\\n\\n15 U.S.C. § 1691a(b)—to include “guarantors, sureties, endorsers, and similar\\n\\nparties” “[f]or purposes of” the Spousal Guaranty Rule. 12 C.F.R. 202.2(e). In\\n\\ndoing so, the Federal Reserve emphasized that it was not imposing any new\\n\\nobligations, since in accordance with the ECOA, Regulation B had long prohibited\\n\\ncreditors from requiring a guaranty from a borrower’s spouse. Rather, the Federal\\n\\nReserve explained, it was just recognizing that the guarantor spouse also has\\n\\n\\n 59\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 60 of 88\\n\\n\\nstanding when the creditor straps the spouses’ credit fortunes together. Equal Credit\\n\\nOpportunity; Revision of Regulation B; Official Staff Commentary, 50 Fed. Reg.\\n\\n48,018 (Nov. 20, 1985). 17\\n\\n B.\\n\\n We begin our analysis of whether the ECOA includes guarantors within the\\n\\ndefinition of “applicants” by evaluating the statutory text. The ECOA makes it\\n\\n“unlawful for any creditor to discriminate against any applicant, with respect to any\\n\\naspect of a credit transaction [] on the basis of . . . marital status. . . .” 15 U.S.C. §\\n\\n1691(a)(1) (emphasis added). In turn, the ECOA defines “applicant” as “any person\\n\\nwho applies to a creditor directly for an extension, renewal, or continuation of credit,\\n\\nor applies to a creditor indirectly by use of an existing credit plan for an amount\\n\\nexceeding a previously established credit limit.” 15 U.S.C. § 1691a(b) (emphasis\\n\\nadded).\\n\\n Two aspects of this text immediately stand out.\\n\\n First, Congress employed the word “any” four times in these two sentences.\\n\\nThe Supreme Court has explained that “the word ‘any’ has an expansive meaning.”\\n\\nUnited States v. Gonzales, 520 U.S. 1, 5 (1997) (citation omitted). The Majority\\n\\n\\n\\n\\n 17\\n The Federal Reserve made Regulation B firm but flexible. If a party is needed to support\\nthe loan request because the potential debtor husband’s credit is insufficient, the wife can guaranty\\nthe loan. Regulation B just bars lenders from “requir[ing] that the spouse be the additional party.”\\n12 C.F.R. 202.7(d)(5).\\n 60\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 61 of 88\\n\\n\\nOpinion is correct that the word “any” does not change the definition of the word it\\n\\nmodifies. Maj. Op. at 18–19. But when that word appears repeatedly over just a\\n\\ncouple of sentences—here, four times in two sentences—“it begins to seem that\\n\\nCongress meant the statute to have expansive reach.” See United States v. Clintwood\\n\\nElkhorn Mining Co., 553 U.S. 1, 7 (2008). 18\\n\\n This would be true for any statutory provision that abundantly uses the word\\n\\n“any,” but it is especially so when dealing with remedial legislation. Where\\n\\nCongress uses the term “any” with a “lack of [accompanying] restrictive language,”\\n\\nthe Supreme Court has instructed us to “refuse[] to engraft artificial limitations” that\\n\\ncurb the “expansive remedial purpose[s]” connoted by such language. Blue Shield\\n\\nof Va. v. McCready, 457 U.S. 465, 472 (1982); see id. (broadly construing “[a]ny\\n\\nperson who shall be injured in his business or property by reason of anything\\n\\nforbidden in the antitrust laws”). So, for example, the text’s prohibition against\\n\\nmarital-status discrimination “with respect to any aspect of a credit transaction,” 15\\n\\nU.S.C. § 1691(a)(1) (emphasis added), purports to preclude marital-status\\n\\ndiscrimination in all aspects of credit transactions, including against guarantors.\\n\\n\\n\\n\\n 18\\n See also Dennis v. Higgins, 498 U.S. 439, 443 (1991) (“A broad construction of § 1983\\nis compelled by the statutory language, which speaks of deprivations of ‘any rights, privileges, or\\nimmunities secured by the Constitution and laws.’ Accordingly, we have repeatedly held that the\\ncoverage of [§ 1983] must be broadly construed.”) (citations and quotation marks omitted).\\n 61\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 62 of 88\\n\\n\\n Nevertheless, and second, the statutory provisions do not expressly answer the\\n\\nkey question at issue here: must the applicant request credit for herself or may she\\n\\nrequest it for somebody else? As the Sixth Circuit has noted, “the applicant and the\\n\\ndebtor are not always the same person.” RL BB Acquisition, LLC v. Bridgemill\\n\\nCommons Dev. Grp., LLC, 754 F.3d 380, 385 (6th Cir. 2014). And the Majority\\n\\nOpinion’s listed common dictionary definitions of “apply” do not limit applicants to\\n\\nonly the proposed debtor, because they do not foreclose the scenario where an\\n\\napplicant makes a request for credit on someone else’s behalf.\\n\\n Similarly, both the Sixth and Eighth Circuits, in construing the text at issue,\\n\\nhave relied on common dictionary definitions of “apply” as meaning “to make an\\n\\nappeal or request especially formally and often in writing and usually for something\\n\\nof benefit to oneself.” Hawkins, 761 F.3d at 941 (quoting Webster’s Third New Int’l\\n\\nDictionary at 105 (2002) (quotation marks and alterations omitted) (emphasis\\n\\nadded); RL BB Acquisition, 754 F.3d at 385 (quoting Webster’s Third New Int’l\\n\\nDictionary at 105 (1993) (cleaned up)); see also Webster’s Third New Int’l\\n\\nDictionary at 105 (1961) (same definition, cleaned up). Obviously, if something is\\n\\n“usually for something of benefit to oneself,” it must sometimes be for something\\n\\nof benefit to another.\\n\\n So we must consider whether a guarantor may ever reasonably and naturally\\n\\nbe viewed as requesting “something of benefit” to another—that is, credit for the\\n\\n\\n 62\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 63 of 88\\n\\n\\nbenefit of the proposed debtor. As it turns out, under the plain meaning of\\n\\n“guarantor,” she may.\\n\\n As Corbin on Contracts explains, “In most cases of guaranty contracts, the\\n\\noffer comes from the guarantor requesting the giving of credit to a principal debtor.\\n\\n. . .” See Timothy Murray, Corbin on Contracts § 3.14, at 467 (rev. ed. 2018). 19 In\\n\\nfact, a guaranty is typically enforceable only because, in exchange for the creditor’s\\n\\npromise to extend credit to the debtor, the guarantor promises to repay the loan if the\\n\\nprimary debtor defaults. See Restatement (Second) of Contracts § 71(2) (1981).\\n\\nThus, guaranty contracts would be unenforceable absent the exchange of\\n\\nconsideration by the guarantor—a promise to repay—and the creditor—the\\n\\nfulfillment of the guarantor’s “application” that it lends to the debtor. See\\n\\nRestatement (Third) of Suretyship and Guaranty § 9, cmt. a (1996); Joseph M.\\n\\nPerillo, Corbin on Corbin § 9.4, at 252–253 (rev. ed. 1996); see also United States\\n\\nv. Burgreen, 591 F.2d 291, 294 (5th Cir. 1979) (finding that guaranty was supported\\n\\nby consideration through the loan to the primary borrower).\\n\\n The Majority Opinion charges that this definition of “apply” falls outside the\\n\\nword’s natural meaning. Maj. Op. at 20. But there are three problems with the\\n\\nMajority Opinion’s position.\\n\\n\\n\\n 19\\n Additionally, other banking regulations provide that guarantors request and receive\\nextensions of credit. See, e.g., 12 C.F.R. 215.3(a)(4).\\n 63\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 64 of 88\\n\\n\\n First, in support of its argument, the Majority Opinion inexplicably invites\\n\\nreaders to consider an incongruous hypothetical, arguing that parents who seek to\\n\\nbribe their daughter’s way into college by offering to “make a large gift if their\\n\\ndaughter is admitted” are not “applicants.” Maj. Op. at 13. I agree. But nor are they\\n\\nanalogous to guarantors. Unlike the parents in the Majority Opinion’s hypothetical,\\n\\nwho can obtain admission for their daughter only by paying a bribe to the college, a\\n\\nguarantor does not bribe or otherwise pay the creditor for credit to be extended to\\n\\nanother. And if the eventual debtor pays back the credit, the guarantor will never\\n\\npay anything. So the Majority Opinion’s efforts to explain in tangible terms why a\\n\\nguarantor cannot fall within the meaning of “applicant” fail.\\n\\n Second, none of the many definitions of the word “apply” that the Majority\\n\\nOpinion cites excludes the interpretation that the application was for someone else.\\n\\nMaj. Op. at 10–11. And that is significant to our analysis. In Smith v. United States,\\n\\n508 U.S. 223, 225 (1993), for example, the Supreme Court had to determine whether\\n\\nthe exchange of a gun for narcotics amounted to the “use” of a firearm “during and\\n\\nin relation to . . . [a] drug trafficking crime,” within the meaning of 18 U.S.C. §\\n\\n924(c)(1). The Court rejected the petitioner’s request to apply only the limited and\\n\\nmost common meaning of the phrase. Id. at 229. Rather, the Court determined that\\n\\ntrading a firearm for drugs also qualified as “use” of the firearm under the statute,\\n\\nexplaining, “It is one thing to say that the ordinary meaning of ‘uses a firearm’\\n\\n\\n 64\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 65 of 88\\n\\n\\nincludes using a firearm as a weapon, since that is the intended purpose of a firearm\\n\\nand the example of ‘use’ that most immediately comes to mind. But it is quite\\n\\nanother to conclude that, as a result, the phrase also excludes any other use.” Id.\\n\\n Applying Smith, we must conclude that the availability of a frequently used\\n\\ndefinition of “applicant”—one who seeks credit for herself—does not necessarily\\n\\nnegate a less-common definition—one who seeks credit for others. That is so\\n\\nbecause the definitions of “applicant” do not exclude use of the word to refer to\\n\\nguarantors.\\n\\n And third, the ECOA is a remedial statute that we must construe “broad[ly]”\\n\\nto effectuate its remedial goals. Barney, 110 F.3d at 1211 n.6; see also Levin, 849\\n\\nF.3d at 1001 (observing that remedial legislation “is entitled to a broad\\n\\nconstruction”); Morante-Navarro v. T&Y Pine Straw, Inc., 350 F.3d 1163, 1166\\n\\n(11th Cir. 2003) (construing the remedial Migrant and Seasonal Agricultural\\n\\nWorkers Protection Act “broadly to effect its . . . purpose”). Our task cannot begin\\n\\nand end with a definitional popularity contest. If it did, there would be no way to\\n\\nconstrue remedial legislation “broadly” to effectuate its purpose, because all words\\n\\nin all statutes would be read as conveying only a single potential meaning.\\n\\n Notably, the Majority Opinion makes no attempt whatsoever to reconcile the\\n\\n“familiar canon of statutory construction that remedial legislation should be\\n\\nconstrued broadly to effectuate its purposes,” Tcherepnin, 389 U.S. at 336, or the\\n\\n\\n 65\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 66 of 88\\n\\n\\ncaveat that “[a] word in a statute may or may not extend to the outer limits of its\\n\\ndefinitional possibilities,” Dolan, 546 U.S. at 486, with the Majority Opinion’s the-\\n\\nwords-can-have-only-a-single-meaning mantra. 20 Of course, if either or both of\\n\\nthese doctrines apply, the Majority Opinion cannot be right.\\n\\n So instead of responding substantively, the Majority Opinion asserts the\\n\\nSupreme Court has “rejected applying” the remedial-purpose doctrine. Maj. Op. at\\n\\n21. And it has no answer at all to the Supreme Court’s instruction in Dolan.\\n\\n But while the Majority Opinion pulls colorful language from context to\\n\\nsuggest that the Supreme Court has determined that the remedial-purpose doctrine\\n\\nis no longer good law, there’s a problem with this response: the Supreme Court has\\n\\nnot rejected the doctrine. Rather, the Court concluded the doctrine was simply\\n\\ninapplicable in the cases the Majority Opinion cites.\\n\\n For example, there were multiple reasons why the doctrine would not apply\\n\\nto the Comprehensive Environmental Response, Compensation, and Liability Act,\\n\\nthe statute at issue in CTS Corp. v. Waldburger, 573 U.S. 1 (2014): (1) the Court\\n\\n\\n\\n 20\\n As Judge William Pryor explained last year, “the text must be construed as a whole. . . .\\nStrict construction sequesters the words of a text from their context. That is one of the reasons why\\nstrict construction is foolish.” Pictet Overseas Inc. v. Helvetia Trust, 905 F.3d 1183, 1190 (11th\\nCir. 2018) (William Pryor, J., concurring) (cleaned up); see also Deal v. United States, 508 U.S.\\n129, 132 (1993) superseded by statute as stated in United States v. Davis, 139 S. Ct. 2319, 2324\\nn.1 (2019) (it is a “fundamental principle of statutory construction (and indeed, of language itself),\\nthat the meaning of a word cannot be determined in isolation, but must be drawn from the context\\nin which it is used.”). So we ought not to pluck words out of their context and define them narrowly\\njust because a dictionary indicates that the narrower meaning is the more common one.\\n 66\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 67 of 88\\n\\n\\nheld that the statute “d[id] not provide a complete remedial framework,” and so it\\n\\nwas not necessarily a remedial statute, see id. at 18; (2) even if it was a remedial\\n\\nstatute, the statute’s legislative history supported a construction of the statutory text\\n\\nat odds with the remedial purpose, see id. at 14–17; and (3) the statute, which would\\n\\ntrump state law if the Court adopted the respondents’ argument, was subject to a\\n\\ncompeting presumption against preemption, see id. at 12.\\n\\n Likewise, the respondents in Inyo County v. Paiute-Shoshone Indians of the\\n\\nBishop Community of the Bishop Colony, 538 U.S. 701 (2003), argued that the\\n\\nBishop Paiute Tribe should be considered a “person” under 42 U.S.C. § 1983, a\\n\\nconstruction—according to respondents—that would be consistent with the law’s\\n\\nremedial purpose. See Paiute-Shoshone Indians, 538 U.S. at 710. The Court\\n\\nrejected this bid, holding that “Section 1983 was designed to secure private rights\\n\\nagainst government encroachment, not to advance a sovereign’s prerogative to\\n\\nwithhold evidence relevant to a criminal investigation.” See id. at 711–12 (citation\\n\\nomitted). So rather than rejecting the doctrine, the Court disagreed that the statute\\n\\nwas meant to remedy the harms that the respondents addressed.\\n\\n Norfolk Southern Railway Co. v. Sorrell, 549 U.S. 158 (2007), is similarly\\n\\nuninstructive here. Norfolk involved the Federal Employers’ Liability Act\\n\\n(“FELA”), which makes railroads liable to their employees for injuries “resulting in\\n\\nwhole or in part from the negligence” of the railroad, 45 U.S.C. § 51. 549 U.S. at\\n\\n\\n 67\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 68 of 88\\n\\n\\n160. In that case, the Court considered whether the statute permitted different\\n\\nstandards of causation for railroad and employee contributory negligence. Id.\\n\\nAlthough the Court recognized that FELA was “enacted to benefit railroad\\n\\nemployees” and that a more lenient standard of causation for employee contributory\\n\\nnegligence would favor employees, it held that FELA required application of a\\n\\nsingle negligence standard, stating that FELA’s “remedial purpose [does not]\\n\\nrequire[] us to interpret every uncertainty in the Act in favor of employees.” Id at\\n\\n171. But what’s important for this case is that the Court decided that the “system of\\n\\ncomparative fault” would not “work” if “the basis of comparison [meaning the\\n\\nstandard for contributory negligence] [were not] the same.” Id. at 170, 171 (citations\\n\\nomitted). In other words, the Court left open the possibility that the remedial-\\n\\npurpose doctrine would apply in cases where the object of interpretation makes the\\n\\ndifference between accomplishing the statute’s central purposes at all and not.\\n\\n This is just such a case. First, unlike the situation with FELA in Norfolk,\\n\\nconstruing “applicant” to include guarantors under § 1691(a)(1) does not prevent the\\n\\nECOA’s mechanisms from working. And unlike whether the contributory\\n\\nnegligence standard is the same for employees and employers under the FELA—a\\n\\nquestion that impacts only the amount of the recovery but not entitlement to\\n\\nrecovery—whether “applicant” encompasses guarantors under § 1691(a)(1) does not\\n\\ninvolve just some random “uncertainty in the Act,” Norfolk, 549 U.S. at 171. Rather,\\n\\n\\n 68\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 69 of 88\\n\\n\\nthe answer to that question determines whether the ECOA will effect one of its\\n\\ncentral purposes at all—that is, whether the ECOA will allow spouses to disentangle\\n\\ntheir credit, or whether it will permit lenders to require such entanglement.\\n\\n The Majority Opinion’s reliance on Household Credit Servs., Inc. v. Pfennig,\\n\\n541 U.S. 232 (2004), fares no better. In that case, the Supreme Court concluded that\\n\\nthe statutory language of the Truth in Lending Act was ambiguous on the question\\n\\nat issue there, so it noted that the agency’s regulation answering the question was\\n\\n“binding” unless, among other things, “manifestly contrary to the statute.” Id. at 242\\n\\n(citation omitted). The Court then concluded that the regulation was “in no way\\n\\nmanifestly contrary” to the statute. Id.\\n\\n As was the case with the statute in Household Credit, here, the ECOA is\\n\\nambiguous on the precise question at issue. And when we turn to the agency’s\\n\\ninterpretation—Regulation B—as in Household Credit, that regulation is not\\n\\n“manifestly contrary to the statute.” Id. Rather, as in Household Credit, it furthers\\n\\none of the “primary goals” of the statute, see id. at 243, as I have explained.\\n\\n Finally, turning to Director, Office of Workers’ Compensation Programs,\\n\\nDepartment of Labor v. Newport News Shipbuilding & Dry Dock Co., 514 U.S. 122\\n\\n(1995), which the Majority Opinion suggests “ridiculed” the remedial-purpose\\n\\ndoctrine and rendered it no longer good law, Maj. Op. at 21 (citing Newport News,\\n\\n514 U.S. at 135), that’s not an accurate description of that case. Rather, the Court\\n\\n\\n 69\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 70 of 88\\n\\n\\nridiculed the petitioner’s attempt to rely on the doctrine in that case, when the Court\\n\\nwas not even analyzing a remedial statute. In fact, the Court discussed—at some\\n\\nlength—that the statute’s “goal” was not clear. See id. at 131. Not only did the\\n\\nCourt not invalidate the remedial-purpose doctrine, the Court recognized that the\\n\\nprinciple of construing any statute broadly to achieve its purposes could be invoked\\n\\n“in case of ambiguity,” id. at 135, which I submit is the case here. And finally, the\\n\\nCourt in Newport News went on to construe the statute at issue “as liberally as can\\n\\nbe.” Id. at 136.\\n\\n But you don’t have to take it from me; you can take it from the Court itself.\\n\\nSeven years after issuing Newport News, the Supreme Court unanimously applied\\n\\nthe remedial-purpose doctrine in S.E.C. v. Zandford, 535 U.S. 813, 819 (2002)\\n\\n(“[W]e have explained that [Section 10(b) of the Securities Exchange Act] should\\n\\nbe construed not technically and restrictively, but flexibly to effectuate its remedial\\n\\npurposes.”) (citation and internal quotation marks omitted).\\n\\n Instead of addressing the substance of this argument on the remedial-purpose\\n\\ndoctrine and Dolan’s instruction that “[a] word in a statute may or may not extend\\n\\nto the outer limits of its definitional possibilities,” 546 U.S. at 486, the Majority\\n\\nOpinion accuses this dissent of engaging in “hornbook abuse of the whole-text\\n\\ncanon” and mischaracterizes my argument as “since the overall purpose of the statute\\n\\n\\n\\n\\n 70\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 71 of 88\\n\\n\\nis to achieve x, any interpretation of the text that limits the achieving of x must be\\n\\ndisfavored.” Maj. Op. at 20.\\n\\n But it’s not that the Majority Opinion’s interpretation of “applicant” “limits\\n\\nthe achieving” of the ECOA’s purpose of disentangling spouses’ financial fortunes,\\n\\nsee Senate Rpt. at 16–17; 21 it’s that it is antithetical to that reason for the ECOA.\\n\\n The guarantor spouse often suffers a unique economic injury that the primary\\n\\ndebtor spouse does not. Say the creditor refuses to extend the loan unless the wife\\n\\nguarantees it, so the wife agrees to do so. But the husband—the direct debtor—may\\n\\nhave no complaints, since he received the loan he was after. And even if he did, he\\n\\nmay well not have economic damages to assert in a lawsuit. See, e.g., Mayes, 167\\n\\nF.3d at 678 (“[The husband] has no claim for damages or injunctive relief under\\n\\nECOA for harm done to his wife. If anyone was injured by requiring [the wife] to\\n\\nsign the guarantee, it was she and not [the husband], who after all received the loan\\n\\nhe had sought.”). Without the potential for redress, he would lack standing.\\n\\n By contrast, the wife who guaranteed the loan may have a separate injury\\n\\nstemming from the fact that her credit score is likely lower because she agreed to be\\n\\nsecondarily liable for the loan. See Mechel Glass, Equifax, Should I Co-Sign On a\\n\\nLoan for a Family Member?, available https://blog.equifax.com/tag/credit-\\n\\n\\n\\n 21\\n See also Mayes, 167 F.3d at 676 (Congress also set out “to prevent loans from being\\nconditioned automatically on the securing of the signature of the non-borrowing spouse.”).\\n 71\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 72 of 88\\n\\n\\nscore/page/4/ (last visited Aug. 27, 2019). So she may have suffered an injury if she\\n\\nsought credit and was unable to obtain it, or if any credit she did receive either was\\n\\nmore limited or bore a higher interest rate than it would have had she not been\\n\\nrequired to guaranty her husband’s loan. Yet construing “applicants” to exclude\\n\\nguarantors would mean nobody in this scenario would have standing to pursue the\\n\\ncreditor’s flagrant violation of the ECOA.\\n\\n And worse yet, allowing lenders to, with impunity, require wives to guaranty\\n\\ntheir husbands’ debts actually creates the same financial intertwinement problems\\n\\nthat arose when lenders demanded that women like Billie Jean King obtain their\\n\\nhusbands’ guaranties before lenders would extend credit to them. In both cases, the\\n\\nwives’ credit is forever tied to the husbands’ credit fortunes.\\n\\n As a result, guaranteeing the husband’s loan can “negatively impact [the\\n\\nwife’s] credit report and creditworthiness,” since guaranteeing a loan shows up on\\n\\ncredit reports immediately. Glass, supra, at 26. And if the husband “miss[es]\\n\\npayments or default[s] on the loan, [the wife’s] credit reports will show the\\n\\ndelinquencies,” ruining her credit history and ability to secure future credit.\\n\\nTransUnion, The Benefits & Issues of Co-Signing a Loan, available at\\n\\nhttp://www.transunion.com/personal-credit/credit-issues-bad-credit/cosigning-a-\\n\\nloan.page (last visited Aug. 27, 2019).\\n\\n\\n\\n\\n 72\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 73 of 88\\n\\n\\n So far from “‘open[ing] vistas of liability’ that Congress did not envision,”\\n\\nMaj. Op. at 24 (quoting Moran Foods, Inc. v. Mid-Atl. Market Dev. Co., LLC, 476\\n\\nF.3d 436, 441 (7th Cir. 2007)), construing “applicants” to include guarantors\\n\\nactually effectuates Congress’s goal of disentangling spouses’ credit. In contrast,\\n\\nthe interpretation the Majority Opinion and the Eighth Circuit in Hawkins offer\\n\\ncreates the very entanglement of spouses’ credit Congress sought to eradicate. That\\n\\nCongress would have prohibited one form of enforced credit entanglement and\\n\\nunworthiness only to allow another to fully replace it makes about as much sense as\\n\\nice-hockey cleats. See W. Air Lines, Inc. v. Bd. of Equalization, 480 U.S. 123, 133\\n\\n(1987) (“[I]llogical results . . . argue strongly against the conclusion that Congress\\n\\nintended” a particular statutory construction).\\n\\n The Majority Opinion’s reliance on Judge Colloton’s Hawkins concurrence—\\n\\nwhere Judge Colloton argued that the word “application” should have only one\\n\\nmeaning throughout the ECOA—is similarly misplaced. Maj. Op. at 14–16. I\\n\\nunderstand Judge Colloton’s concern. After all, we “ordinarily assume[] that\\n\\nidentical words used in different parts of the same act are intended to have the same\\n\\nmeaning.” Utility Air Reg. Grp. v. E.P.A., 573 U.S. 302, 319-20 (2014) (citations\\n\\nomitted). But that assumption is not absolute. It “readily yields whenever there is\\n\\nsuch variation in the connection in which the words are used as reasonably to warrant\\n\\nthe conclusion that they were employed in different parts of the act with different\\n\\n\\n 73\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 74 of 88\\n\\n\\nintent.” Atl. Cleaners & Dyers v. United States, 286 U.S. 427, 433 (1932); see also\\n\\nYates v. United States, 135 S. Ct. 1074, 1082 (2015) (“In law as in life . . . the same\\n\\nwords, placed in different contexts, sometimes mean different things.”). This is one\\n\\nof those times, or else the ECOA ensures the existence of the very problem it seeks\\n\\nto eliminate.\\n\\n Even assuming some parts of the ECOA suggest—or even require—a narrow\\n\\ninterpretation of the word “applicant,” that does not mean that the term as used in 15\\n\\nU.S.C. § 1691(a)(1) must be similarly construed, since “a characterization fitting in\\n\\ncertain contexts may be unsuitable in others.” See NationsBank of N.C., N.A. v.\\n\\nVariable Annuity Life Ins. Co., 513 U.S. 251, 262 (1995). For instance, to make his\\n\\npoint, Judge Colloton primarily invokes 15 U.S.C. § 1691(d), titled, “Reason for\\n\\nadverse action; procedure applicable; ‘adverse action’ defined.” That subsection\\n\\narticulates what happens when a lender denies or revokes credit to a debtor. It\\n\\ntherefore makes sense that it applies to the person who would have received or\\n\\nactually received the loan.\\n\\n Section 1691(a), on the other hand, is the statute’s vehicle for proscribing\\n\\ndiscrimination on the basis of marital status. By both its terms and its function, it\\n\\nhas broader application than the provisions to which Judge Colloton points. Cf. Atl.\\n\\nCleaners, 286 U.S. at 435 (“A consideration of the history [of the ECOA] . . .\\n\\nsanctions the conclusion that Congress meant to deal comprehensively and\\n\\n\\n 74\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 75 of 88\\n\\n\\neffectively with the evils resulting” from discrimination). Indeed, it directly effects\\n\\nCongress’s purposes in enacting the ECOA in the first place—preventing credit\\n\\ndiscrimination on the basis of marital status and allowing spouses to disentangle their\\n\\ncredit from one another. And its text, as I have noted, makes it applicable to “any\\n\\naspect” of a credit transaction.\\n\\n That Judge Colton’s citations are exclusively to other provisions of 15 U.S.C.\\n\\n§ 1691 does not change this fact. See Hawkins, 761 F.3d at 943–44 (Colloton, J.,\\n\\nconcurring). “Most words have different shades of meaning and consequently may\\n\\nbe variously construed, not only when they occur in different statutes, but when used\\n\\nmore than once in the same statute or even in the same section.” Envtl. Def. v. Duke\\n\\nEnergy Corp., 549 U.S. 561, 574 (2007) (emphasis added) (citation and alteration\\n\\nomitted); accord Gen. Dynamics, 540 U.S. at 595-97 (“age” has different meanings\\n\\nin 29 U.S.C. § 623(a) and § 623(f) of the Age Discrimination in Employment Act of\\n\\n1967); Robinson, 519 U.S. at 342–43 (“employee” has different meanings in\\n\\ndifferent parts of 42 U.S.C. § 2000e–16(b) of Title VII of the Civil Rights Act of\\n\\n1964); see also Wachovia Bank, N.A. v. Schmidt, 546 U.S. 303, 313–14 (2006)\\n\\n(“located” has different meanings in different sections of the National Bank Act).\\n\\n But even if we were to hold that the term “applicant” as used throughout the\\n\\nstatute includes guarantors, that is a better answer than the Majority Opinion’s\\n\\nconclusion. Judge Colloton’s examples suggest that such an interpretation could\\n\\n\\n 75\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 76 of 88\\n\\n\\ncreate unexpected benefits for guarantors. That is preferable to gutting the statute’s\\n\\nreason for being. The Majority Opinion’s interpretation does not simply “limit[] the\\n\\nachieving” of what Congress sought with the ECOA, Maj. Op. at 20 (citation\\n\\nomitted); as I have explained, it “would be destructive of [the law’s] purpose. . . .”\\n\\nRobinson, 519 U.S. at 346.\\n\\n So the Majority Opinion’s analysis cannot carry the day. “[T]he overriding\\n\\nnational policy against discrimination that underlies the [ECOA]” means that “we\\n\\ncannot give” words in that statute a “narrow interpretation.” Bros. v. First Leasing,\\n\\n724 F.2d 789, 793 (9th Cir. 1984). Instead, we must “construe the literal language\\n\\nof the ECOA in light of the clear, strong purpose evidenced by the Act and adopt an\\n\\ninterpretation that will serve to effectuate that purpose.” Id.; see also United States\\n\\nv. Am. Trucking Ass’ns, 310 U.S. 534, 543 (1940) (“[W]hen the plain meaning did\\n\\nnot produce absurd results but merely an unreasonable one ‘plainly at variance with\\n\\nthe policy of the legislation as a whole[,]’ this Court has followed that purpose, rather\\n\\nthan the literal words.”) (citation omitted). Considering Congress’s remedial\\n\\npurposes in enacting the ECOA, as well as that the many definitions of “applicant”\\n\\ndo not exclude the possibility that an “applicant” includes a guarantor, we must\\n\\nconclude that Congress’s employment of the term “applicant” is ambiguous with\\n\\nrespect to whether it covers guarantors. For this reason, we must conduct a Chevron\\n\\n\\n\\n\\n 76\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 77 of 88\\n\\n\\nanalysis of the Federal Reserve’s answer to our question of whether the ECOA\\n\\nincludes guarantors within the definition of “applicants.”\\n\\n C.\\n\\n Under Chevron, “we may not disturb an agency rule unless it is ‘arbitrary or\\n\\ncapricious in substance, or manifestly contrary to the statute.’” Mayo Found. for\\n\\nMed. Educ. & Res. v. United States, 562 U.S. 44, 53 (2011) (quoting Pfennig, 541\\n\\nU.S. at 242). At bottom, then, we must ask whether the Federal Reserve’s inclusion\\n\\nof guarantors in the ECOA’s definition of “applicant” is a “reasonable\\n\\ninterpretation” of the ECOA’s text. Chevron, 467 U.S. at 844. The Federal\\n\\nReserve’s interpretation is reasonable for at least three reasons: it accords with the\\n\\nECOA’s text; it furthers a primary purpose of the ECOA; and, Congress has tacitly\\n\\napproved of the Federal Reserve’s interpretation.\\n\\n 1. The Federal Reserve’s inclusion of guarantors within the meaning of\\n “applicants” is consistent with the ECOA’s text.\\n\\n First, guarantors fit within the plain meaning of “applicants.” As I have noted,\\n\\nleading authorities recognize that guarantors apply to the creditor by asking it to\\n\\nextend credit to the primary debtor. See supra at 63.\\n\\n The Majority Opinion’s conclusion to the contrary rests on its premise that\\n\\n“[a]lthough a guarantor makes a promise related to an applicant’s request for credit,\\n\\nthe guaranty is not itself a request for credit.” Maj. Op. at 12. But the Majority\\n\\nOpinion offers no support (beyond the Eighth Circuit’s erroneous conclusion in\\n 77\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 78 of 88\\n\\n\\nHawkins) for the notion that the guarantor is not requesting that the creditor extend\\n\\ncredit to the primary borrower. Not a treatise. Not an industry source. Not another\\n\\nstatute. Not a dictionary. Not even something from the remote reaches of the\\n\\ninternet. Instead, the Majority Opinion relies on its own understanding of credit\\n\\nmarkets.\\n\\n But the Majority Opinion’s own view—regardless of how learned and\\n\\nreasonable—of how terms are used in credit markets, does not control if the agency’s\\n\\ninterpretation is also reasonable. Indeed, “[t]he reviewing court may not substitute\\n\\nits judgment for that of the agency but must, instead, defer to the agency’s technical\\n\\nexpertise” if its interpretation is within the range of reasonableness. Miami-Dade\\n\\nCty. v. E.P.A., 529 F.3d 1049, 1065 (11th Cir. 2008) (citation and quotation marks\\n\\nomitted). And here, as I have noted, leading treatises and industry sources recognize\\n\\nthat guarantors request credit. See supra at 63. Since the Federal Reserve’s\\n\\ninterpretation of the text is reasonable—not to mention well-supported—we must\\n\\ndefer to it.\\n\\n 2. The Federal Reserve’s inclusion of guarantors within the meaning of\\n “applicants” is consistent with primary purposes of the ECOA.\\n\\n Second, the Federal Reserve’s interpretation is reasonable because covering\\n\\n“guarantors” fits squarely within the wheelhouse of the ECOA’s aims. As I have\\n\\nnoted, in enacting the ECOA, Congress sought to disentangle sex and marital status\\n\\nfrom credit. See Senate Rpt. at 16–17. Congress also set out “to prevent loans from\\n 78\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 79 of 88\\n\\n\\nbeing conditioned automatically on the securing of the signature of the non-\\n\\nborrowing spouse.” Mayes, 167 F.3d at 676.\\n\\n The Federal Reserve’s interpretation accounts for and respects these primary\\n\\ngoals of the ECOA. But not including guarantors within the definition of\\n\\n“applicants” yanks the teeth out of the ECOA. See supra at 71–73.\\n\\n 3. The Federal Reserve’s inclusion of guarantors within the meaning of\\n “applicants” is consistent with congressional abstention from amending the\\n ECOA to preclude that interpretation, despite Congress’s other amendments\\n to the ECOA.\\n\\n Third, the Federal Reserve’s interpretation of “applicants” to include\\n\\nguarantors is reasonable, in view of congressional action on the ECOA since its\\n\\nenactment in 1974. The Supreme Court has instructed us to tread lightly where\\n\\nCongress has amended a statute but declined to disturb the agency’s interpretation.\\n\\n For instance, in Texas Department of Housing & Community Affairs v.\\n\\nInclusive Communities Project, Inc., 135 S. Ct. 2507 (2015), the Court considered\\n\\nwhether the statutory language authorized disparate-impact claims under the Fair\\n\\nHousing Act. In conducting its analysis, the Court found it significant that Congress\\n\\ndeclined to alter the language at issue to preclude such claims, even though Congress\\n\\nhad amended the statute after several Circuits had determined that the statute\\n\\nauthorized such claims. Id. at 2519. Congress’s inaction, the Court explained, was\\n\\n“convincing support for the conclusion that Congress accepted and ratified the\\n\\nunanimous holdings of the Courts of Appeals finding disparate-impact liability.” Id.\\n 79\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 80 of 88\\n\\n\\nat 2520; see also Gen. Dynamics, 540 U.S. at 593–94 & n.6 (noting that the two\\n\\nfederal appellate courts and nearly all the district courts to consider the issue held\\n\\nthat the Age Discrimination in Employment Act of 1967 did not prohibit favoring\\n\\nolder employees over younger ones: “The very strength of this consensus is enough\\n\\nto rule out any serious claim of ambiguity, and congressional silence after years of\\n\\njudicial interpretation supports adherence to the traditional view.”).\\n\\n Here, Congress has tinkered with the ECOA no fewer than three times since\\n\\n1985, when the Federal Reserve began construing the ECOA to include guarantors\\n\\nwithin the term “applicants.” See, e.g., Dodd-Frank Wall Street Reform and\\n\\nConsumer Protection Act, Pub. L. 111-203, §§ 1071, 1474, 124 Stat. 1376, 2056–\\n\\n57, 2199–2200 (2010); Omnibus Consolidated Appropriation Act, 1997, Pub. L.\\n\\n104-208, § 2302, 110 Stat. 3009 (1996); Federal Deposit Insurance Corporation\\n\\nImprovement Act of 1991, Pub. L. 102-242, §§ 212(d), 223, 105 Stat. 2236 (1991).\\n\\n And up until 2014, when the Eighth Circuit issued Hawkins, the “vast majority\\n\\nof courts” that examined the issue found that guarantors had standing under the\\n\\nECOA. See RL BB, 754 F.3d at 386. True, as the Majority Opinion notes, many of\\n\\nthe cases deciding this issue offered little reasoning for their decisions. Maj. Op. at\\n\\n25–26. But the Supreme Court does not require otherwise. While the Court\\n\\nacknowledged the judicial consensus of the interpretation of the Fair Housing Act in\\n\\nTexas Department of Housing, contrary to the Majority Opinion’s characterization,\\n\\n\\n 80\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 81 of 88\\n\\n\\nit did not require the lower courts’ decisions to be “reasoned.” Maj. Op. at 26. It\\n\\nnoted simply that the courts that had “addressed” the issue had consistently come\\n\\nout the same way. Tex. Dep’t of Hous., 135 S. Ct. at 2519. And most of the cases\\n\\nthe Court referenced were not particularly well-reasoned; they just adopted what\\n\\nother circuits had done. See, e.g., Hanson v. Veterans Admin., 800 F.2d 1381, 1386\\n\\n(5th Cir. 1986) (noting and adopting analysis from other circuits that Fair Housing\\n\\nAct violations can be proven with a showing of discriminatory effect); Arthur v. City\\n\\nof Toledo, Ohio, 782 F.2d 565, 574–75 (6th Cir. 1986) (same); United States v.\\n\\nMarengo Cty. Comm’n, 731 F.2d 1546, 1559, n.20 (11th Cir. 1984) (same); Smith v.\\n\\nTown of Clarkton, N.C., 682 F.2d 1055, 1065 (4th Cir. 1982) (same); Halet v. Wend\\n\\nInv. Co., 672 F.2d 1305, 1311 (9th Cir. 1982) (same).\\n\\n The more important point, though, is that Congress has long been on notice\\n\\nas to how federal courts have interpreted the ECOA and Regulation B. And with\\n\\nthat knowledge in hand, Congress amended the law multiple times and never\\n\\nconsidered tweaking the ECOA to abrogate these courts’ approval of the Federal\\n\\nReserve’s interpretation. Even after the Seventh Circuit issued Moran, in which a\\n\\nCourt of Appeals—for the first time—voiced doubts in dicta about the Federal\\n\\nReserve’s interpretation, Congress left the ECOA’s definition of “applicant” the\\n\\nsame, even though it substantively amended the ECOA three years later, in 2010.\\n\\nSo Congress declined to take the Seventh Circuit up on its invitation to cap the\\n\\n\\n 81\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 82 of 88\\n\\n\\nsupposed “vistas of liability” that the Seventh Circuit concluded Congress was\\n\\n“unlikely to accept.” Moran, 476 F.3d at 441. Congressional inaction on this point\\n\\nin the face of thirty years of uniformity “supports adherence to the traditional view,”\\n\\nGen. Dynamics, 540 U.S. at 594, that the Federal Reserve’s inclusion of guarantors\\n\\nwithin the term “applicants” is valid.\\n\\n III.\\n\\n We should dismiss this case because it is not properly before us. The Majority\\n\\nOpinion’s insistence on “resolving” a claim that no Appellant has presented to us\\n\\nrenders its discussion of the ECOA unnecessary. But because that discussion\\n\\nincorrectly purports to truncate the ECOA’s broad reach, it leaves me no choice but\\n\\nto respond. Unfortunately, the Majority Opinion’s analysis artificially limits the\\n\\ndefinitions of “applicant” and “guaranty”; it conflicts with congressional aims in\\n\\nenacting the ECOA; and it is rebutted by Congress’s failure to amend the ECOA to\\n\\nmake guarantors fall outside the meaning of “applicants,” in the 34 years since the\\n\\nFederal Reserve has construed the term “applicants” to include guarantors.\\n\\nUltimately, the Majority Opinion’s interpretation of the ECOA ironically allows\\n\\nlenders to get away with discriminating on the basis of sex or marital status, the very\\n\\nthing the ECOA meant to eliminate. I therefore respectfully dissent.\\n\\n\\n\\n\\n 82\\n\\nCase: 17-11736 Date Filed: 08/28/2019 Page: 83 of 88\\n\\nCase: 17-11736 Date Filed: 08/28/2019 Page: 84 of 88\\n\\nCase: 17-11736 Date Filed: 08/28/2019 Page: 85 of 88\\n\\nCase: 17-11736 Date Filed: 08/28/2019 Page: 86 of 88\\n\\nCase: 17-11736 Date Filed: 08/28/2019 Page: 87 of 88\\n\\nCase: 17-11736 Date Filed: 08/28/2019 Page: 88 of 88\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434191/\", \"author_raw\": \"ROSENBAUM, Circuit Judge, concurring\"}]}"},"judge1":{"kind":"string","value":"WILLIAM PRYOR"},"judge2":{"kind":"string","value":"ROSENBAUM"},"judge3":{"kind":"string","value":"MOORE"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"WILLIAM PRYOR\": \", Circuit\", \"ROSENBAUM\": \", Circuit\", \"MOORE\": \", District\"}"},"opinions_count":{"kind":"number","value":3,"string":"3"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":2,"string":"2"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4656938/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Regions Bank v. Legal Outsource PA\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-28\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: Yes\nAdditional notes: \n\n[CASE_TEXT]\nCase: 17-11736 Date Filed: 08/28/2019 Page: 1 of 88\n\n\n [PUBLISH]\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n ________________________\n\n No. 17-11736\n ________________________\n\n D.C. Docket No. 2:14-cv-000476-PAM-MRM\n\n\n\nREGIONS BANK,\nan Alabama state-chartered bank,\n\n Plaintiff–Counter Defendant–Appellee,\n\nversus\n\nLEGAL OUTSOURCE PA,\na Florida professional association,\nPERIWINKLE PARTNERS, LLC,\na Florida limited liability company,\nCHARLES PAUL-THOMAS PHOENIX,\nindividually, a.k.a. Charles PT Phoenix,\nLISA M. PHOENIX,\nindividually,\n\n Defendants–Counter Claimants–Appellants.\n\n ________________________\n\n Appeal from the United States District Court\n for the Middle District of Florida\n ________________________\n\n (August 28, 2019)\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 2 of 88\n\n\nBefore WILLIAM PRYOR, and ROSENBAUM, Circuit Judges, and MOORE, *\nDistrict Judge.\n\nWILLIAM PRYOR, Circuit Judge:\n\n The main issue presented by this appeal has divided our sister circuits:\n\nwhether a guarantor constitutes an “applicant” under the Equal Credit Opportunity\n\nAct. See 15 U.S.C. §§ 1691(a), 1691a(b). Compare Hawkins v. Cmty. Bank of\n\nRaymore, 761 F.3d 937 (8th Cir. 2014), aff’d by an equally divided Court, 136 S.\n\nCt. 1072 (2016) (holding that a guarantor unambiguously is not an “applicant”\n\nunder the Act), and Moran Foods, Inc. v. Mid-Atl. Mkt. Dev. Co., 476 F.3d 436\n\n(7th Cir. 2007) (opining the same), with RL BB Acquisition, LLC v. Bridgemill\n\nCommons Dev. Grp., 754 F.3d 380 (6th Cir. 2014) (holding that the term\n\n“applicant” is ambiguous and applying Chevron deference to an agency\n\ninterpretation that a guarantor is an “applicant”). Legal Outsource PA, a law firm\n\nwholly owned by Charles Phoenix, defaulted on a loan from Regions Bank, which\n\ntriggered the default of a loan and mortgage that Regions issued to Periwinkle\n\nPartners, LLC, an entity wholly owned by Charles’s wife, Lisa Phoenix. After the\n\nobligors refused to cure the defaults, Regions sued to enforce its rights under the\n\nloans and mortgage. The obligors filed several counterclaims asserting that\n\nRegions violated the Equal Credit Opportunity Act by discriminating against Lisa\n\n\n *\n Honorable K. Michael Moore, United States District Chief Judge for the Southern District\nof Florida, sitting by designation.\n 2\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 3 of 88\n\n\nand Charles based on their marital status when it demanded that they and Legal\n\nOutsource guarantee the Periwinkle loan. The district court granted summary\n\njudgment in favor of Regions. The district court ruled that Lisa Phoenix’s\n\ncounterclaims failed because she lacked standing as an “applicant” when she was\n\ninstead a guarantor. Because we conclude that a guarantor is not an “applicant”\n\nunder the Equal Credit Opportunity Act, we affirm the summary judgment in favor\n\nof Regions. But the parties agree that we must remand to correct an error in the\n\njudgment.\n\n I. BACKGROUND\n\n Beginning in 2005, Regions Bank extended a $450,000 line of credit to\n\nLegal Outsource PA, a law firm owned by Charles Phoenix. Legal Outsource\n\nrenewed the loan on a yearly or semi-yearly basis, and it was last renewed in May\n\n2013 with a maturity date in February 2014. Charles Phoenix also guaranteed the\n\n2013 Outsource loan.\n\n In 2011, Regions lent nearly $1.7 million to Periwinkle Partners, LLC, for\n\nthe purchase of a shopping center on Sanibel Island, Florida. At that time, the sole\n\nmember of Periwinkle Partners was a company owned by Charles Phoenix’s wife,\n\nLisa Phoenix. Charles Phoenix, Lisa Phoenix, and Legal Outsource all guaranteed\n\nthe Periwinkle loan. Under the Periwinkle loan, a default by any of the parties,\n\n\n\n\n 3\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 4 of 88\n\n\nincluding the guarantors, on any other loans that they had with Regions constitutes\n\na default under the Periwinkle loan.\n\n In August 2013, Regions concluded that the Outsource and Periwinkle loans\n\nwere in default based on the obligors’ failure to provide requested financial\n\ninformation and based on Periwinkle’s failure to pay its property taxes. Regions\n\nthen warned the obligors several times that it would accelerate the loans if the\n\nobligors failed to cure the default. In February 2014, the Outsource loan matured\n\nand Legal Outsource, which was no longer in operation, failed to pay it. Two\n\nmonths later, Regions declared the Outsource loan in default and demanded its full\n\nand immediate payment. According to the obligors, this declaration was a bad-faith\n\nattempt by Regions to coerce Lisa Phoenix into securing the Outsource loan with\n\nPeriwinkle as collateral, but she refused to do so. After the Outsource loan default,\n\nRegions also declared the Periwinkle loan in default and demanded its full and\n\nimmediate payment. The obligors never cured any of the defaults.\n\n In August 2014, Regions filed a complaint against Charles and Lisa Phoenix,\n\nLegal Outsource, and Periwinkle Partners for breach of the Legal Outsource\n\npromissory note and guaranty, breach of the Periwinkle promissory note and\n\nguaranties, foreclosure of the Periwinkle mortgage, and receivership. The obligors\n\nanswered the complaint and interposed 73 affirmative defenses and eight\n\ncounterclaims.\n\n\n 4\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 5 of 88\n\n\n The obligors twice amended the answer and added four new counterclaims\n\nthat each asserted a violation of the Equal Credit Opportunity Act. The\n\ncounterclaims—three of which were individually brought by Charles Phoenix, Lisa\n\nPhoenix, and Legal Outsource respectively, and one of which was brought by Lisa\n\nPhoenix and Periwinkle Partners—alleged that Regions discriminated on the basis\n\nof marital status when it required the Phoenixes and Legal Outsource to guarantee\n\nthe Periwinkle loan. Regions then moved to dismiss the newly added\n\ncounterclaims, and the district count granted that motion in part. The district court\n\nruled that the guarantors of the Periwinkle loan all lacked statutory standing\n\nbecause they were not “applicant[s]” under the Equal Credit Opportunity Act. But\n\nthe court also ruled that one of the counterclaims, which was brought on behalf of\n\nLisa Phoenix and Periwinkle Partners, had sufficiently alleged that Lisa Phoenix\n\nand Periwinkle Partners were “applicants” under the Act, so it denied the motion as\n\nto that count.\n\n After Regions moved for summary judgment, the district court granted\n\nsummary judgment in favor of Regions both for its claims for breach of the\n\npromissory notes and guaranties and against the obligors’ counterclaims. The\n\ndistrict court ruled that the obligors “do not dispute that they were in default under\n\nthe relevant notes and guaranties,” and it ruled that the counterclaims had “no\n\nmerit.” With respect to the remaining counterclaim under the Equal Credit\n\n\n 5\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 6 of 88\n\n\nOpportunity Act—the joint claim by Lisa Phoenix and Periwinkle—the district\n\ncourt ruled that Periwinkle’s claim of discrimination was “frivolous” because, as\n\nan entity, it had no marital status. And the district court ruled that “[t]he claim fails\n\nas to Lisa Phoenix as well because, aside from the lack of any evidence to establish\n\nany alleged discrimination on the basis of marital status, she was not an ‘applicant’\n\nfor the Periwinkle loan[;] she was a guarantor.” The district court referred to its\n\nearlier order ruling that guarantors were not “applicants.”\n\n The district court later issued a second summary judgment order granting\n\nforeclosure on the Periwinkle mortgage. The court then dismissed the matter with\n\nprejudice and directed the clerk to enter the judgment. The clerk entered the\n\njudgment, and the obligors filed their notice of appeal.\n\n Regions moved to amend the judgment to state, among other things, the\n\namounts due to Regions from the obligors. The district court granted Regions’\n\nmotion in part, instructing the clerk to enter an amended judgment providing for\n\nthe following relief:\n\n [T]he Court will order the Clerk to amend the Judgment to provide that\n Regions Bank prevails on its claims against Defendants. The Judgment\n will further provide that Regions Bank is entitled to recover\n $540,054.24 from Defendants for the Legal Outsource loan . . . .\n\nThe clerk then entered the amended judgment, and the obligors amended their\n\nnotice of appeal to include the order granting Regions’ motion to clarify and the\n\namended judgment among the items subject to their appeal.\n 6\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 7 of 88\n\n\n II. STANDARD OF REVIEW\n\n We review a grant of summary judgment de novo. Moore ex rel. Moore v.\n\nReese, 637 F.3d 1220, 1231 (11th Cir. 2011).\n\n III. DISCUSSION\n\n This appeal presents several issues about whether the obligors are liable for\n\nthe default of the Legal Outsource loan and the Periwinkle loan and mortgage.\n\nAlthough the obligors raise a host of issues that seek to obscure the nature of their\n\ndefaults, all but one of them lack any merit, and some border on being frivolous.\n\nWe decline to address them any further.\n\n We divide our discussion of the remaining issues in two parts. First, we\n\nexplain that Lisa Phoenix’s counterclaims under the Equal Credit Opportunity Act\n\nfail because a guarantor does not qualify as an “applicant” under the Act. Second,\n\nwe explain that a limited remand to correct erroneous language from the amended\n\njudgment is warranted.\n\n A. The District Court Correctly Granted Summary Judgment Against the\n Equal Credit Opportunity Act Counterclaims by Lisa Phoenix.\n\n The district court did not err when it granted summary judgment against the\n\ncounterclaims by Lisa Phoenix under the Equal Credit Opportunity Act. As an\n\ninitial matter, although the obligors briefly mention Periwinkle’s counterclaim in\n\ntheir argument about the Equal Credit Opportunity Act, they have failed to argue\n\nor cite caselaw in either the district court or on appeal to rebut the conclusion that\n\n 7\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 8 of 88\n\n\nits status as an entity defeats its claim, as the district court ruled, so we consider\n\nthat issue abandoned. See Sapuppo v. Allstate Floridian Ins. Co., 739 F.3d 678,\n\n680 (11th Cir. 2014) (“[A]n appellant must convince us that every stated ground\n\nfor the judgment against him is incorrect.”). We discuss only our reasons for\n\nconcluding that the district court correctly granted summary judgment against Lisa\n\nPhoenix’s counterclaims on the ground that a guarantor is not an “applicant” for\n\ncredit within the meaning of the Act, see 15 U.S.C. § 1691a(b).\n\n The Equal Credit Opportunity Act makes it unlawful for “any creditor to\n\ndiscriminate against any applicant, with respect to any aspect of a credit transaction\n\n. . . on the basis of . . . marital status.” Id. § 1691(a)–(a)(1). The Act defines an\n\n“applicant” as “any person who applies to a creditor directly for . . . credit, or\n\napplies to a creditor indirectly by use of an existing credit plan for an amount\n\nexceeding a previously established credit limit.” Id. § 1691a(b) (emphases added).\n\nThe Act initially required the Federal Reserve Board to promulgate regulations to\n\nenforce the Act. See 15 U.S.C. § 1691b (1974). And the Federal Reserve Board\n\npromulgated Regulation B, which defines an applicant as “any person who\n\nrequests or who has received an extension of credit from a creditor,” which\n\nincludes “any person who is or may become contractually liable regarding an\n\nextension of credit.” 12 C.F.R. § 202.2(e). That regulation further provides that the\n\nterm “applicant” includes “guarantors, sureties, endorsers, and similar parties.” Id.\n\n\n 8\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 9 of 88\n\n\n(emphasis added). Regulation B also prohibits a creditor from requiring the\n\nsignature of an applicant’s spouse, other than a joint applicant, on any credit\n\ninstrument if the applicant independently qualifies as creditworthy. Id.\n\n§ 202.7(d)(1).\n\n The obligors rely on the definition of “applicant” in Regulation B to argue\n\nthat Lisa Phoenix has statutory standing under the Act, so we must determine\n\nwhether we should defer to this regulation under the two-step framework\n\nannounced in Chevron, U.S.A., Inc. v. Nat’l Res. Def. Council, Inc., 467 U.S. 837\n\n(1984). See Arevalo v. U.S. Att’y Gen., 872 F.3d 1184, 1187–88 (11th Cir. 2017).\n\nFirst, we ask whether, after applying the “traditional tools of statutory\n\nconstruction,” we can determine whether Congress has spoken clearly on the issue.\n\nBarton v. U.S. Att’y Gen., 904 F.3d 1294, 1298 (11th Cir. 2018) (quoting Fajardo\n\nv. U.S. Att’y Gen., 659 F.3d 1303, 1307 (11th Cir. 2011)). If the statute is\n\nunambiguous, we apply it according to its terms and give no deference to the\n\nadministrative interpretation. Arevalo, 872 F.3d at 1188. Second, “if the statute is\n\nsilent or ambiguous with respect to the specific issue presented, we must then\n\ndetermine whether the agency’s interpretation is reasonable or based on a\n\npermissible construction of the statute.” Id. An interpretation is reasonable if it is\n\n“rational and consistent with the statute.” Id. (quoting Sullivan v. Everhart, 494\n\nU.S. 83, 89 (1990)).\n\n\n 9\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 10 of 88\n\n\n In applying the “traditional tools of statutory construction,” we begin “with\n\nthe statutory text, and proceed from the understanding that unless otherwise\n\ndefined, statutory terms are generally interpreted in accordance with their ordinary\n\nmeaning.” Barton, 904 F.3d at 1298 (quoting Sebelius v. Cloer, 569 U.S. 369, 376\n\n(2013)); see also Antonin Scalia & Bryan A. Garner, Reading Law: The\n\nInterpretation of Legal Texts § 6, at 69 (2012) (“Words are to be understood in\n\ntheir ordinary, everyday meaning—unless the context indicates that they bear a\n\ntechnical sense.”). And we interpret the words of a statute based on their meaning\n\nat the time of enactment. See New Prime Inc. v. Oliveira, 139 S. Ct. 532, 539\n\n(2019); Scalia & Garner, Reading Law § 7, at 78 (“Words must be given the\n\nmeaning they had when the text was adopted.”).\n\n The Act, which was adopted in 1974, defines an applicant as “any person\n\nwho applies to a creditor directly for . . . credit, or applies to a creditor indirectly\n\nby use of an existing credit plan for an amount exceeding a previously established\n\ncredit limit.” 15 U.S.C. § 1691a(b) (emphases added). English-language\n\ndictionaries both before and after the enactment define the term “apply” to refer to\n\na request for something. See Apply, 1 The Oxford English Dictionary 407 (corr.\n\nreprint 1961) (1933) (“To address oneself for information or aid, to have recourse,\n\nto make application to”); Apply, 1 The Oxford English Dictionary 577 (2d ed.\n\n1989) (same); Apply, Webster’s New International Dictionary of the English\n\n\n 10\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 11 of 88\n\n\nLanguage (Webster’s Second) 132 (2d ed. 1961) (“To make request; to have\n\nrecourse with a view to gain something; to solicit”); Apply, Webster’s New\n\nCollegiate Dictionary 55 (1977) (“[T]o make an appeal or request esp[ecially] in\n\nthe form of a written application.”). So too do legal dictionaries. See Apply, Black’s\n\nLaw Dictionary 128 (rev. 4th ed. 1968) (“To make a formal request or petition,\n\nusually in writing, to a court, officer, board, or company, for the granting of some\n\nfavor, or of some rule or order, which is within his or their power or discretion”);\n\nApplication, id. at 127 (“The act of making a request for something”). The Sixth\n\nand Eighth Circuits have also both cited a definition of the term “apply” as\n\nmeaning “a request . . . usually for something of benefit to oneself.” See Hawkins,\n\n761 F.3d at 941 (alteration adopted) (emphasis added) (quoting Webster’s Third\n\nNew Int’l Dictionary 105 (2002)); RL BB, 754 F.3d at 385 (quoting Webster’s\n\nThird New Int’l Dictionary 105 (1993)); see also Apply, Webster’s Third New Int’l\n\nDictionary 105 (1971) (defining “apply” as “to make an appeal or a request\n\nesp[ecially] formally and often in writing and usu[ally] for something of benefit to\n\noneself”). So, taken together, these definitions suggest that the ordinary meaning of\n\nthe term “applicant” is one who requests credit to benefit himself.\n\n A guarantor does not fit within this definition. At the time of enactment,\n\nEnglish-language dictionaries defined “guaranty” to mean a promise by a\n\nguarantor to answer for the payment of some debt if the person liable in the first\n\n\n 11\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 12 of 88\n\n\ninstance is unable to pay. See Guaranty, 4 The Oxford English Dictionary 477\n\n(corr. reprint 1961) (1933) (“a written undertaking made by a person (called the\n\nguarantor) to be answerable for the payment of a debt or the performance of an\n\nobligation by another person, who is in the first instance liable to such payment or\n\nobligation”); Guaranty, 6 The Oxford English Dictionary 912 (2d ed. 1989)\n\n(same); Guaranty, Webster’s Second 1110 (“An undertaking to answer for the\n\npayment of some debt, or the performance of some duty, of another, in case of the\n\nfailure of such other to pay or perform”); Guaranty, Webster’s New Collegiate\n\nDictionary 509 (“[A]n undertaking to answer for the payment of a debt or the\n\nperformance of a duty of another in case of the other’s default or miscarriage”).\n\nLegal dictionaries defined “guaranty” the same way. See Guaranty, Black’s Law\n\nDictionary 833 (rev. 4th ed.) (“A promise to answer for payment of debt or the\n\nperformance of obligation if person liable in the first instance fails to make\n\npayment or perform obligation”); Guarantor, Black’s Law Dictionary 634 (5th ed.\n\n1979) (“One who becomes secondarily liable for another’s debt or performance”).\n\nAlthough a guarantor makes a promise related to an applicant’s request for credit,\n\nthe guaranty is not itself a request for credit, and certainly not a request for credit\n\nfor the guarantor. See Hawkins, 761 F.3d at 942 (“We find it to be unambiguous\n\nthat assuming a secondary, contingent liability does not amount to a request for\n\ncredit.”).\n\n\n 12\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 13 of 88\n\n\n To be sure, as the dissent points out, a guarantor’s promise supports a\n\nwould-be debtor’s loan application and ordinarily stems from the guarantor’s\n\ndesire that the application be granted. See Dissenting Op. at 62–63. But to say that\n\na guarantor requests credit by supporting another’s request for credit is to push the\n\nbounds of ordinary usage—at the very least, it is to use one word in two obviously\n\ndifferent senses. And to say that the guarantor applies for credit by supporting\n\nanother’s application is to leave ordinary usage behind entirely.\n\n An example should make this point clear. Suppose a high-school senior is\n\napplying to her parents’ alma mater, and her parents—who happen to be wealthy\n\ndonors—promise the school that they will make a large gift if their daughter is\n\nadmitted. The parents’ promise supports the daughter’s application for admission,\n\njust as a guarantor’s promise supports a loan applicant’s application for credit. The\n\nparents will be grateful if their daughter is admitted, as a guarantor ordinarily is\n\ngrateful when the debtor’s application for credit is granted. But it would be\n\nunnatural to say that the parents have “applied” for their daughter’s admission or to\n\ncall them “applicants” for admission. Under any ordinary use of the word, the\n\nstudent is the only “applicant” in this scenario.\n\n Applying the whole-text and consistent-usage canons to the Act further\n\nconfirms that the term “applicant” excludes guarantors. The whole-text canon\n\nrefers to the principle that a “judicial interpreter [should] consider the entire text, in\n\n\n 13\n\n Case: 17-11736 Date Filed: 08/28/2019 Page: 14 of 88\n\n\nview of its structure and of the physical and logical relation of its many parts,”\n\nwhen interpreting any particular part of the text. Scalia & Garner, Reading Law\n\n§ 24, at 167. “Properly applied, it typically establishes that only one of the possible\n\nmeanings that a word or phrase can bear is compatible with use of the same word\n\nor phrase elsewhere in the statute . . . .” Id. at 168. Closely related to the whole-text\n\ncanon is the principle that “[a] word or phrase is presumed to bear the same\n\nmeaning throughout a text” unless context requires otherwise. Id. § 25, at 170;\n\naccord Atl. Cleaners & Dyers v. United States, 286 U.S. 427, 433 (1932)\n\n(“Undoubtedly, there is a natural presumption that identical words used in different\n\nparts of the same act are intended to have the same meaning.”).\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Regions Bank v. Legal Outsource PA\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-28\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: Yes\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 17-11736 Date Filed: 08/28/2019 Page: 1 of 88\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 17-11736\\n ________________________\\n\\n D.C. Docket No. 2:14-cv-000476-PAM-MRM\\n\\n\\n\\nREGIONS BANK,\\nan Alabama state-chartered bank,\\n\\n Plaintiff–Counter Defendant–Appellee,\\n\\nversus\\n\\nLEGAL OUTSOURCE PA,\\na Florida professional association,\\nPERIWINKLE PARTNERS, LLC,\\na Florida limited liability company,\\nCHARLES PAUL-THOMAS PHOENIX,\\nindividually, a.k.a. Charles PT Phoenix,\\nLISA M. PHOENIX,\\nindividually,\\n\\n Defendants–Counter Claimants–Appellants.\\n\\n ________________________\\n\\n Appeal from the United States District Court\\n for the Middle District of Florida\\n ________________________\\n\\n (August 28, 2019)\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 2 of 88\\n\\n\\nBefore WILLIAM PRYOR, and ROSENBAUM, Circuit Judges, and MOORE, *\\nDistrict Judge.\\n\\nWILLIAM PRYOR, Circuit Judge:\\n\\n The main issue presented by this appeal has divided our sister circuits:\\n\\nwhether a guarantor constitutes an “applicant” under the Equal Credit Opportunity\\n\\nAct. See 15 U.S.C. §§ 1691(a), 1691a(b). Compare Hawkins v. Cmty. Bank of\\n\\nRaymore, 761 F.3d 937 (8th Cir. 2014), aff’d by an equally divided Court, 136 S.\\n\\nCt. 1072 (2016) (holding that a guarantor unambiguously is not an “applicant”\\n\\nunder the Act), and Moran Foods, Inc. v. Mid-Atl. Mkt. Dev. Co., 476 F.3d 436\\n\\n(7th Cir. 2007) (opining the same), with RL BB Acquisition, LLC v. Bridgemill\\n\\nCommons Dev. Grp., 754 F.3d 380 (6th Cir. 2014) (holding that the term\\n\\n“applicant” is ambiguous and applying Chevron deference to an agency\\n\\ninterpretation that a guarantor is an “applicant”). Legal Outsource PA, a law firm\\n\\nwholly owned by Charles Phoenix, defaulted on a loan from Regions Bank, which\\n\\ntriggered the default of a loan and mortgage that Regions issued to Periwinkle\\n\\nPartners, LLC, an entity wholly owned by Charles’s wife, Lisa Phoenix. After the\\n\\nobligors refused to cure the defaults, Regions sued to enforce its rights under the\\n\\nloans and mortgage. The obligors filed several counterclaims asserting that\\n\\nRegions violated the Equal Credit Opportunity Act by discriminating against Lisa\\n\\n\\n *\\n Honorable K. Michael Moore, United States District Chief Judge for the Southern District\\nof Florida, sitting by designation.\\n 2\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 3 of 88\\n\\n\\nand Charles based on their marital status when it demanded that they and Legal\\n\\nOutsource guarantee the Periwinkle loan. The district court granted summary\\n\\njudgment in favor of Regions. The district court ruled that Lisa Phoenix’s\\n\\ncounterclaims failed because she lacked standing as an “applicant” when she was\\n\\ninstead a guarantor. Because we conclude that a guarantor is not an “applicant”\\n\\nunder the Equal Credit Opportunity Act, we affirm the summary judgment in favor\\n\\nof Regions. But the parties agree that we must remand to correct an error in the\\n\\njudgment.\\n\\n I. BACKGROUND\\n\\n Beginning in 2005, Regions Bank extended a $450,000 line of credit to\\n\\nLegal Outsource PA, a law firm owned by Charles Phoenix. Legal Outsource\\n\\nrenewed the loan on a yearly or semi-yearly basis, and it was last renewed in May\\n\\n2013 with a maturity date in February 2014. Charles Phoenix also guaranteed the\\n\\n2013 Outsource loan.\\n\\n In 2011, Regions lent nearly $1.7 million to Periwinkle Partners, LLC, for\\n\\nthe purchase of a shopping center on Sanibel Island, Florida. At that time, the sole\\n\\nmember of Periwinkle Partners was a company owned by Charles Phoenix’s wife,\\n\\nLisa Phoenix. Charles Phoenix, Lisa Phoenix, and Legal Outsource all guaranteed\\n\\nthe Periwinkle loan. Under the Periwinkle loan, a default by any of the parties,\\n\\n\\n\\n\\n 3\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 4 of 88\\n\\n\\nincluding the guarantors, on any other loans that they had with Regions constitutes\\n\\na default under the Periwinkle loan.\\n\\n In August 2013, Regions concluded that the Outsource and Periwinkle loans\\n\\nwere in default based on the obligors’ failure to provide requested financial\\n\\ninformation and based on Periwinkle’s failure to pay its property taxes. Regions\\n\\nthen warned the obligors several times that it would accelerate the loans if the\\n\\nobligors failed to cure the default. In February 2014, the Outsource loan matured\\n\\nand Legal Outsource, which was no longer in operation, failed to pay it. Two\\n\\nmonths later, Regions declared the Outsource loan in default and demanded its full\\n\\nand immediate payment. According to the obligors, this declaration was a bad-faith\\n\\nattempt by Regions to coerce Lisa Phoenix into securing the Outsource loan with\\n\\nPeriwinkle as collateral, but she refused to do so. After the Outsource loan default,\\n\\nRegions also declared the Periwinkle loan in default and demanded its full and\\n\\nimmediate payment. The obligors never cured any of the defaults.\\n\\n In August 2014, Regions filed a complaint against Charles and Lisa Phoenix,\\n\\nLegal Outsource, and Periwinkle Partners for breach of the Legal Outsource\\n\\npromissory note and guaranty, breach of the Periwinkle promissory note and\\n\\nguaranties, foreclosure of the Periwinkle mortgage, and receivership. The obligors\\n\\nanswered the complaint and interposed 73 affirmative defenses and eight\\n\\ncounterclaims.\\n\\n\\n 4\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 5 of 88\\n\\n\\n The obligors twice amended the answer and added four new counterclaims\\n\\nthat each asserted a violation of the Equal Credit Opportunity Act. The\\n\\ncounterclaims—three of which were individually brought by Charles Phoenix, Lisa\\n\\nPhoenix, and Legal Outsource respectively, and one of which was brought by Lisa\\n\\nPhoenix and Periwinkle Partners—alleged that Regions discriminated on the basis\\n\\nof marital status when it required the Phoenixes and Legal Outsource to guarantee\\n\\nthe Periwinkle loan. Regions then moved to dismiss the newly added\\n\\ncounterclaims, and the district count granted that motion in part. The district court\\n\\nruled that the guarantors of the Periwinkle loan all lacked statutory standing\\n\\nbecause they were not “applicant[s]” under the Equal Credit Opportunity Act. But\\n\\nthe court also ruled that one of the counterclaims, which was brought on behalf of\\n\\nLisa Phoenix and Periwinkle Partners, had sufficiently alleged that Lisa Phoenix\\n\\nand Periwinkle Partners were “applicants” under the Act, so it denied the motion as\\n\\nto that count.\\n\\n After Regions moved for summary judgment, the district court granted\\n\\nsummary judgment in favor of Regions both for its claims for breach of the\\n\\npromissory notes and guaranties and against the obligors’ counterclaims. The\\n\\ndistrict court ruled that the obligors “do not dispute that they were in default under\\n\\nthe relevant notes and guaranties,” and it ruled that the counterclaims had “no\\n\\nmerit.” With respect to the remaining counterclaim under the Equal Credit\\n\\n\\n 5\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 6 of 88\\n\\n\\nOpportunity Act—the joint claim by Lisa Phoenix and Periwinkle—the district\\n\\ncourt ruled that Periwinkle’s claim of discrimination was “frivolous” because, as\\n\\nan entity, it had no marital status. And the district court ruled that “[t]he claim fails\\n\\nas to Lisa Phoenix as well because, aside from the lack of any evidence to establish\\n\\nany alleged discrimination on the basis of marital status, she was not an ‘applicant’\\n\\nfor the Periwinkle loan[;] she was a guarantor.” The district court referred to its\\n\\nearlier order ruling that guarantors were not “applicants.”\\n\\n The district court later issued a second summary judgment order granting\\n\\nforeclosure on the Periwinkle mortgage. The court then dismissed the matter with\\n\\nprejudice and directed the clerk to enter the judgment. The clerk entered the\\n\\njudgment, and the obligors filed their notice of appeal.\\n\\n Regions moved to amend the judgment to state, among other things, the\\n\\namounts due to Regions from the obligors. The district court granted Regions’\\n\\nmotion in part, instructing the clerk to enter an amended judgment providing for\\n\\nthe following relief:\\n\\n [T]he Court will order the Clerk to amend the Judgment to provide that\\n Regions Bank prevails on its claims against Defendants. The Judgment\\n will further provide that Regions Bank is entitled to recover\\n $540,054.24 from Defendants for the Legal Outsource loan . . . .\\n\\nThe clerk then entered the amended judgment, and the obligors amended their\\n\\nnotice of appeal to include the order granting Regions’ motion to clarify and the\\n\\namended judgment among the items subject to their appeal.\\n 6\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 7 of 88\\n\\n\\n II. STANDARD OF REVIEW\\n\\n We review a grant of summary judgment de novo. Moore ex rel. Moore v.\\n\\nReese, 637 F.3d 1220, 1231 (11th Cir. 2011).\\n\\n III. DISCUSSION\\n\\n This appeal presents several issues about whether the obligors are liable for\\n\\nthe default of the Legal Outsource loan and the Periwinkle loan and mortgage.\\n\\nAlthough the obligors raise a host of issues that seek to obscure the nature of their\\n\\ndefaults, all but one of them lack any merit, and some border on being frivolous.\\n\\nWe decline to address them any further.\\n\\n We divide our discussion of the remaining issues in two parts. First, we\\n\\nexplain that Lisa Phoenix’s counterclaims under the Equal Credit Opportunity Act\\n\\nfail because a guarantor does not qualify as an “applicant” under the Act. Second,\\n\\nwe explain that a limited remand to correct erroneous language from the amended\\n\\njudgment is warranted.\\n\\n A. The District Court Correctly Granted Summary Judgment Against the\\n Equal Credit Opportunity Act Counterclaims by Lisa Phoenix.\\n\\n The district court did not err when it granted summary judgment against the\\n\\ncounterclaims by Lisa Phoenix under the Equal Credit Opportunity Act. As an\\n\\ninitial matter, although the obligors briefly mention Periwinkle’s counterclaim in\\n\\ntheir argument about the Equal Credit Opportunity Act, they have failed to argue\\n\\nor cite caselaw in either the district court or on appeal to rebut the conclusion that\\n\\n 7\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 8 of 88\\n\\n\\nits status as an entity defeats its claim, as the district court ruled, so we consider\\n\\nthat issue abandoned. See Sapuppo v. Allstate Floridian Ins. Co., 739 F.3d 678,\\n\\n680 (11th Cir. 2014) (“[A]n appellant must convince us that every stated ground\\n\\nfor the judgment against him is incorrect.”). We discuss only our reasons for\\n\\nconcluding that the district court correctly granted summary judgment against Lisa\\n\\nPhoenix’s counterclaims on the ground that a guarantor is not an “applicant” for\\n\\ncredit within the meaning of the Act, see 15 U.S.C. § 1691a(b).\\n\\n The Equal Credit Opportunity Act makes it unlawful for “any creditor to\\n\\ndiscriminate against any applicant, with respect to any aspect of a credit transaction\\n\\n. . . on the basis of . . . marital status.” Id. § 1691(a)–(a)(1). The Act defines an\\n\\n“applicant” as “any person who applies to a creditor directly for . . . credit, or\\n\\napplies to a creditor indirectly by use of an existing credit plan for an amount\\n\\nexceeding a previously established credit limit.” Id. § 1691a(b) (emphases added).\\n\\nThe Act initially required the Federal Reserve Board to promulgate regulations to\\n\\nenforce the Act. See 15 U.S.C. § 1691b (1974). And the Federal Reserve Board\\n\\npromulgated Regulation B, which defines an applicant as “any person who\\n\\nrequests or who has received an extension of credit from a creditor,” which\\n\\nincludes “any person who is or may become contractually liable regarding an\\n\\nextension of credit.” 12 C.F.R. § 202.2(e). That regulation further provides that the\\n\\nterm “applicant” includes “guarantors, sureties, endorsers, and similar parties.” Id.\\n\\n\\n 8\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 9 of 88\\n\\n\\n(emphasis added). Regulation B also prohibits a creditor from requiring the\\n\\nsignature of an applicant’s spouse, other than a joint applicant, on any credit\\n\\ninstrument if the applicant independently qualifies as creditworthy. Id.\\n\\n§ 202.7(d)(1).\\n\\n The obligors rely on the definition of “applicant” in Regulation B to argue\\n\\nthat Lisa Phoenix has statutory standing under the Act, so we must determine\\n\\nwhether we should defer to this regulation under the two-step framework\\n\\nannounced in Chevron, U.S.A., Inc. v. Nat’l Res. Def. Council, Inc., 467 U.S. 837\\n\\n(1984). See Arevalo v. U.S. Att’y Gen., 872 F.3d 1184, 1187–88 (11th Cir. 2017).\\n\\nFirst, we ask whether, after applying the “traditional tools of statutory\\n\\nconstruction,” we can determine whether Congress has spoken clearly on the issue.\\n\\nBarton v. U.S. Att’y Gen., 904 F.3d 1294, 1298 (11th Cir. 2018) (quoting Fajardo\\n\\nv. U.S. Att’y Gen., 659 F.3d 1303, 1307 (11th Cir. 2011)). If the statute is\\n\\nunambiguous, we apply it according to its terms and give no deference to the\\n\\nadministrative interpretation. Arevalo, 872 F.3d at 1188. Second, “if the statute is\\n\\nsilent or ambiguous with respect to the specific issue presented, we must then\\n\\ndetermine whether the agency’s interpretation is reasonable or based on a\\n\\npermissible construction of the statute.” Id. An interpretation is reasonable if it is\\n\\n“rational and consistent with the statute.” Id. (quoting Sullivan v. Everhart, 494\\n\\nU.S. 83, 89 (1990)).\\n\\n\\n 9\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 10 of 88\\n\\n\\n In applying the “traditional tools of statutory construction,” we begin “with\\n\\nthe statutory text, and proceed from the understanding that unless otherwise\\n\\ndefined, statutory terms are generally interpreted in accordance with their ordinary\\n\\nmeaning.” Barton, 904 F.3d at 1298 (quoting Sebelius v. Cloer, 569 U.S. 369, 376\\n\\n(2013)); see also Antonin Scalia & Bryan A. Garner, Reading Law: The\\n\\nInterpretation of Legal Texts § 6, at 69 (2012) (“Words are to be understood in\\n\\ntheir ordinary, everyday meaning—unless the context indicates that they bear a\\n\\ntechnical sense.”). And we interpret the words of a statute based on their meaning\\n\\nat the time of enactment. See New Prime Inc. v. Oliveira, 139 S. Ct. 532, 539\\n\\n(2019); Scalia & Garner, Reading Law § 7, at 78 (“Words must be given the\\n\\nmeaning they had when the text was adopted.”).\\n\\n The Act, which was adopted in 1974, defines an applicant as “any person\\n\\nwho applies to a creditor directly for . . . credit, or applies to a creditor indirectly\\n\\nby use of an existing credit plan for an amount exceeding a previously established\\n\\ncredit limit.” 15 U.S.C. § 1691a(b) (emphases added). English-language\\n\\ndictionaries both before and after the enactment define the term “apply” to refer to\\n\\na request for something. See Apply, 1 The Oxford English Dictionary 407 (corr.\\n\\nreprint 1961) (1933) (“To address oneself for information or aid, to have recourse,\\n\\nto make application to”); Apply, 1 The Oxford English Dictionary 577 (2d ed.\\n\\n1989) (same); Apply, Webster’s New International Dictionary of the English\\n\\n\\n 10\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 11 of 88\\n\\n\\nLanguage (Webster’s Second) 132 (2d ed. 1961) (“To make request; to have\\n\\nrecourse with a view to gain something; to solicit”); Apply, Webster’s New\\n\\nCollegiate Dictionary 55 (1977) (“[T]o make an appeal or request esp[ecially] in\\n\\nthe form of a written application.”). So too do legal dictionaries. See Apply, Black’s\\n\\nLaw Dictionary 128 (rev. 4th ed. 1968) (“To make a formal request or petition,\\n\\nusually in writing, to a court, officer, board, or company, for the granting of some\\n\\nfavor, or of some rule or order, which is within his or their power or discretion”);\\n\\nApplication, id. at 127 (“The act of making a request for something”). The Sixth\\n\\nand Eighth Circuits have also both cited a definition of the term “apply” as\\n\\nmeaning “a request . . . usually for something of benefit to oneself.” See Hawkins,\\n\\n761 F.3d at 941 (alteration adopted) (emphasis added) (quoting Webster’s Third\\n\\nNew Int’l Dictionary 105 (2002)); RL BB, 754 F.3d at 385 (quoting Webster’s\\n\\nThird New Int’l Dictionary 105 (1993)); see also Apply, Webster’s Third New Int’l\\n\\nDictionary 105 (1971) (defining “apply” as “to make an appeal or a request\\n\\nesp[ecially] formally and often in writing and usu[ally] for something of benefit to\\n\\noneself”). So, taken together, these definitions suggest that the ordinary meaning of\\n\\nthe term “applicant” is one who requests credit to benefit himself.\\n\\n A guarantor does not fit within this definition. At the time of enactment,\\n\\nEnglish-language dictionaries defined “guaranty” to mean a promise by a\\n\\nguarantor to answer for the payment of some debt if the person liable in the first\\n\\n\\n 11\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 12 of 88\\n\\n\\ninstance is unable to pay. See Guaranty, 4 The Oxford English Dictionary 477\\n\\n(corr. reprint 1961) (1933) (“a written undertaking made by a person (called the\\n\\nguarantor) to be answerable for the payment of a debt or the performance of an\\n\\nobligation by another person, who is in the first instance liable to such payment or\\n\\nobligation”); Guaranty, 6 The Oxford English Dictionary 912 (2d ed. 1989)\\n\\n(same); Guaranty, Webster’s Second 1110 (“An undertaking to answer for the\\n\\npayment of some debt, or the performance of some duty, of another, in case of the\\n\\nfailure of such other to pay or perform”); Guaranty, Webster’s New Collegiate\\n\\nDictionary 509 (“[A]n undertaking to answer for the payment of a debt or the\\n\\nperformance of a duty of another in case of the other’s default or miscarriage”).\\n\\nLegal dictionaries defined “guaranty” the same way. See Guaranty, Black’s Law\\n\\nDictionary 833 (rev. 4th ed.) (“A promise to answer for payment of debt or the\\n\\nperformance of obligation if person liable in the first instance fails to make\\n\\npayment or perform obligation”); Guarantor, Black’s Law Dictionary 634 (5th ed.\\n\\n1979) (“One who becomes secondarily liable for another’s debt or performance”).\\n\\nAlthough a guarantor makes a promise related to an applicant’s request for credit,\\n\\nthe guaranty is not itself a request for credit, and certainly not a request for credit\\n\\nfor the guarantor. See Hawkins, 761 F.3d at 942 (“We find it to be unambiguous\\n\\nthat assuming a secondary, contingent liability does not amount to a request for\\n\\ncredit.”).\\n\\n\\n 12\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 13 of 88\\n\\n\\n To be sure, as the dissent points out, a guarantor’s promise supports a\\n\\nwould-be debtor’s loan application and ordinarily stems from the guarantor’s\\n\\ndesire that the application be granted. See Dissenting Op. at 62–63. But to say that\\n\\na guarantor requests credit by supporting another’s request for credit is to push the\\n\\nbounds of ordinary usage—at the very least, it is to use one word in two obviously\\n\\ndifferent senses. And to say that the guarantor applies for credit by supporting\\n\\nanother’s application is to leave ordinary usage behind entirely.\\n\\n An example should make this point clear. Suppose a high-school senior is\\n\\napplying to her parents’ alma mater, and her parents—who happen to be wealthy\\n\\ndonors—promise the school that they will make a large gift if their daughter is\\n\\nadmitted. The parents’ promise supports the daughter’s application for admission,\\n\\njust as a guarantor’s promise supports a loan applicant’s application for credit. The\\n\\nparents will be grateful if their daughter is admitted, as a guarantor ordinarily is\\n\\ngrateful when the debtor’s application for credit is granted. But it would be\\n\\nunnatural to say that the parents have “applied” for their daughter’s admission or to\\n\\ncall them “applicants” for admission. Under any ordinary use of the word, the\\n\\nstudent is the only “applicant” in this scenario.\\n\\n Applying the whole-text and consistent-usage canons to the Act further\\n\\nconfirms that the term “applicant” excludes guarantors. The whole-text canon\\n\\nrefers to the principle that a “judicial interpreter [should] consider the entire text, in\\n\\n\\n 13\\n\\n Case: 17-11736 Date Filed: 08/28/2019 Page: 14 of 88\\n\\n\\nview of its structure and of the physical and logical relation of its many parts,”\\n\\nwhen interpreting any particular part of the text. Scalia & Garner, Reading Law\\n\\n§ 24, at 167. “Properly applied, it typically establishes that only one of the possible\\n\\nmeanings that a word or phrase can bear is compatible with use of the same word\\n\\nor phrase elsewhere in the statute . . . .” Id. at 168. Closely related to the whole-text\\n\\ncanon is the principle that “[a] word or phrase is presumed to bear the same\\n\\nmeaning throughout a text” unless context requires otherwise. Id. § 25, at 170;\\n\\naccord Atl. Cleaners & Dyers v. United States, 286 U.S. 427, 433 (1932)\\n\\n(“Undoubtedly, there is a natural presumption that identical words used in different\\n\\nparts of the same act are intended to have the same meaning.”).\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4158,"cells":{"id":{"kind":"number","value":4656967,"string":"4,656,967"},"name":{"kind":"string","value":"UNITED STATES of America, Plaintiff - Appellee, v. Junior Jean BAPTISTE, Defendant - Appellant."},"name_abbreviation":{"kind":"string","value":"United States v. Junior Jean Baptiste"},"decision_date":{"kind":"string","value":"2019-08-28"},"docket_number":{"kind":"string","value":"16-17175; 16-17595"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Pryor, Newsom, Branch\", \"parties\": \"\", \"opinions\": [{\"author\": \"NEWSOM, Circuit Judge:\", \"type\": \"010combined\", \"text\": \"Case: 16-17175 Date Filed: 08/28/2019 Page: 1 of 28\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n Nos. 16-17175, 16-17595\\n ________________________\\n\\n\\n D.C. Docket No. 1:15-cr-20777-JEM-1\\n\\n\\nUNITED STATES OF AMERICA,\\n\\n Plaintiff - Appellee,\\n versus\\n\\nJUNIOR JEAN BAPTISTE,\\n\\n Defendant - Appellant.\\n\\n\\n ________________________\\n\\n Appeals from the United States District Court\\n for the Southern District of Florida\\n ________________________\\n\\n (August 28, 2019)\\n\\n\\n\\nBefore WILLIAM PRYOR, NEWSOM, and BRANCH, Circuit Judges.\\n\\nNEWSOM, Circuit Judge:\\n\\n Jean Junior Baptiste was a businessman. In addition to a portfolio of media\\n\\nand entertainment companies, Baptiste owned and operated Surveillance Master\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 2 of 28\\n\\n\\nCCTV. Surveillance Master installed security cameras, repaired computers, and—\\n\\none of these things is not like the others—cashed checks. In 2016, a jury found\\n\\nthat Baptiste had taken part in an $11 million scheme that involved cashing tax-\\n\\nrefund checks that he had fraudulently obtained in the names of inmates, minors,\\n\\ndead people, and other insentient or otherwise unsuspecting “clients.” Baptiste\\n\\nnow appeals his conviction and 212-month sentence.\\n\\n Although Baptiste raises a number of issues on appeal, we focus primarily\\n\\non two questions related to the hearsay testimony of a government witness. The\\n\\nabridged version of the story: Francesse Chery was one of Baptiste’s key\\n\\nwitnesses. The government countered with her brother, Anael Chery, who testified\\n\\n(among other things) that Francesse had told him that, in exchange for her (false)\\n\\ntestimony supporting Baptiste’s narrative, Baptiste would give her a Mercedes.\\n\\nBaptiste argues that Anael’s testimony was inadmissible hearsay and that the\\n\\ndistrict court’s error in allowing the jury to hear it tainted both his conviction and\\n\\nhis sentence.\\n\\n Baptiste’s challenge presents two questions. First, was Anael’s testimony\\n\\nindeed inadmissible hearsay? The district court admitted the testimony pursuant to\\n\\nthe statement-against-interest exception to the general prohibition on hearsay\\n\\nevidence, and on appeal the government has offered a smattering of additional\\n\\ntheories of admissibility. We conclude that we needn’t decide whether Anael’s\\n\\n\\n 2\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 3 of 28\\n\\n\\ntestimony was inadmissible hearsay because even if the district court did err in\\n\\nallowing it, the error was harmless. There was more than enough compelling—and\\n\\nundoubtedly admissible—evidence to support Baptiste’s conviction.\\n\\n Second, and (sort of) relatedly, did the district court err in relying on Anael’s\\n\\ntestimony when it imposed a sentencing enhancement for obstructing justice? If\\n\\nyou’re saying, “Didn’t they just say they weren’t going to decide whether the\\n\\ntestimony was admissible?”—we hear you. As it turns out, though, thanks to a\\n\\ndoctrine called (somewhat oxymoronically) “reliable hearsay” we can answer the\\n\\nsecond question without deciding the first. Under the reliable-hearsay doctrine, so\\n\\nlong as certain preconditions are met, a sentencing court can rely on evidence that\\n\\nwould be off-limits in the guilt phase. For Baptiste, this means that even if Anael’s\\n\\ndescription of his sister’s supposed deal was inadmissible hearsay (and we aren’t\\n\\nsaying either way) the district court might not have erred in relying on that\\n\\ntestimony for the obstruction enhancement—again, so long as the preconditions are\\n\\nmet.\\n\\n So, what are they? Well, our case law has arguably sent mixed signals about\\n\\nthat. There is, though, a synthesis. We hold (and clarify) today that the Sentencing\\n\\nGuidelines permit use of hearsay testimony so long as the overall record provides\\n\\n“sufficient indicia of reliability”—and we conclude that the indicia of reliability\\n\\nhere are sufficient.\\n\\n\\n 3\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 4 of 28\\n\\n\\n Accordingly—and because we find that none of Baptiste’s other arguments\\n\\nhas merit—we affirm Baptiste’s conviction and, with one asterisk, his sentence.\\n\\nThe asterisk? Pursuant to the parties’ joint request, we remand the case to the\\n\\ndistrict court for the limited purpose of allowing Baptiste to allocute, a right that he\\n\\nwas denied at sentencing.\\n\\n I\\n\\n During Baptiste’s trial, two cooperating witnesses, Andy Louissaint and Karl\\n\\nMoltimer, described the tax-refund scheme. Baptiste would first provide tax\\n\\npreparers such as Louissaint and Moltimer with a list of individuals—many of\\n\\nthem young, incarcerated, dead, or otherwise unlikely to discover the use of their\\n\\nidentities—along with accompanying personal information such as birthdates and\\n\\nsocial security numbers. The tax preparers used that information to (fraudulently)\\n\\nobtain refund checks from the IRS, which they then brought to Baptiste’s\\n\\nSurveillance Master for cashing.\\n\\n Baptiste contracted with a third party to process the checks. He would\\n\\nfabricate the necessary documents—work permits, passports, green cards, driver’s\\n\\nlicenses—to avoid raising red flags. Baptiste and the tax preparers would divide\\n\\nthe proceeds, with Baptiste often taking a cut of between 25% and 50%. Not bad,\\n\\ngiven that typical check-cashing fees range from 1.5% to 3%.\\n\\n\\n\\n\\n 4\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 5 of 28\\n\\n\\n When the IRS reviewed Surveillance Master’s records in conjunction with a\\n\\nroutine inspection, an agent noticed that several refund checks were for identical\\n\\n(or nearly identical) amounts. During a subsequent investigation, the IRS obtained\\n\\ncopies of identification documents, Treasury checks, and lists containing personal\\n\\ninformation for both the quick and the dead. Customs and Immigration Services\\n\\nconducted an evaluation of approximately 630 permanent-resident and\\n\\nemployment-authorization cards that Baptiste had used to verify the checks; all but\\n\\none—belonging to Baptiste’s wife—were fraudulent.\\n\\n The ensuing indictment charged Baptiste with money laundering, conspiracy\\n\\nto commit money laundering, possession of false identification documents with\\n\\nunlawful intent, theft of government money, and aggravated identity theft. At trial,\\n\\nBaptiste’s story was that he was an unwitting participant in a broader scheme\\n\\nconcocted by a now-dead business partner, Marvin Pagnon. Pagnon’s ex-\\n\\ngirlfriend, the aforementioned Francesse Chery, echoed Baptiste’s account in her\\n\\ntestimony and explained how Pagnon operated. In particular, she explained that\\n\\nbecause Pagnon had poor credit, he would ask would-be associates such as\\n\\nBaptiste to partner with him. Pagnon, she said, laid out a collaboration that was\\n\\nstraightforward and above board: Pagnon would obtain (real) refund checks from\\n\\ntax preparers; Baptiste would cash whatever Pagnon handed him. Baptiste signed\\n\\nup, and he seems to have been well compensated for his efforts. During the\\n\\n\\n 5\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 6 of 28\\n\\n\\nscheme, Baptiste purchased a Mercedes Benz CL 63, a Mercedes Benz E Class, a\\n\\nRange Rover, and—again, one of these things is not like the others—a $345,000\\n\\nshare in a 200-foot long cargo ship.\\n\\n The government painted a different picture. Moltimer and Louissaint both\\n\\ntestified, for instance, that Pagnon played no role in Baptiste’s check-cashing\\n\\noperation. Louissaint added that when Baptiste first became aware that an\\n\\nindictment could be coming down the pike, he had told Louissaint that he planned\\n\\nto “basically blame everything on Marvin [Pagnon].” The government also offered\\n\\ntestimony from Francesse Chery’s brother, Anael. Anael first opined, as a general\\n\\nmatter, that his sister just wasn’t a very truthful person. He went on to testify,\\n\\nmuch more specifically, that Francesse had told him that Baptiste was going to\\n\\ngive her a Mercedes CLA 45 so long as she “h[e]ld up her end of the bargain” by\\n\\nbacking Baptiste’s it-was-all-Pagnon’s-idea narrative. Baptiste objected that\\n\\nAnael’s testimony—relaying what his sister had allegedly told him—was\\n\\ninadmissible hearsay. The district court ultimately admitted Anael’s statement,\\n\\nciting the statement-against-interest exception to the general prohibition on hearsay\\n\\nevidence. See Fed. R. Evid. 804(b)(3).\\n\\n The jury convicted Baptiste on all counts. At sentencing, the district court\\n\\napplied enhancements for the amount of money that Baptiste had fraudulently\\n\\nobtained from the government (exceeding $11 million), for the number of victims\\n\\n\\n 6\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 7 of 28\\n\\n\\n(perhaps hundreds), for the victims’ vulnerable status (many incapacitated in some\\n\\nform or fashion), and for obstructing justice (for the perjury-for-a-Mercedes deal\\n\\nwith Francesse). The district court sentenced Baptiste to 212 months’\\n\\nimprisonment and restitution of more than $11 million. 1 This is Baptiste’s appeal.\\n\\n II\\n\\n As already noted, our focus here will be the related contentions that the\\n\\ndistrict court (1) abused its discretion by admitting Anael’s hearsay testimony as\\n\\nevidence against Baptiste and (2) violated the Sentencing Guidelines by imposing\\n\\nan obstruction-based enhancement that relied on the same. Before getting there,\\n\\nthough, we’ll address a few other guilt-phase matters—namely, whether the district\\n\\ncourt erroneously refused to admit Rule 404(b) evidence, whether the government\\n\\nfailed to reveal (alleged) deals with its witnesses, and whether the prosecutor made\\n\\nimproper statements to the jury during his closing argument. Then, after dealing\\n\\nwith the Anael-related issues, which span the trial and sentencing phases, we’ll\\n\\nconclude with a few additional, sentencing-specific arguments—namely, that the\\n\\ndistrict court erred in increasing Baptiste’s sentence due to the amount of money\\n\\nthat the government lost as well as the number and vulnerability of the scheme’s\\n\\nvictims, as well as in declining to permit Baptiste to allocute personally.\\n\\n A\\n\\n1\\n The court ordered that $70,000 of that total be paid jointly and severally with Moltimer.\\n\\n 7\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 8 of 28\\n\\n\\n 1\\n\\n During the trial, the district court refused to allow Baptiste to put on\\n\\nevidence that Marvin Pagnon had duped others into participating in similar\\n\\nschemes. On appeal, Baptiste contends that, in so doing, the district court\\n\\nmisapplied Federal Rule of Evidence 404(b), which states, in relevant part—\\n\\n (1) Prohibited Uses. Evidence of a crime, wrong, or other act is not\\n admissible to prove a person’s character in order to show that on a\\n particular occasion the person acted in accordance with the character.\\n (2) Permitted Uses; Notice in a Criminal Case. This evidence may\\n be admissible for another purpose, such as proving motive,\\n opportunity, intent, preparation, plan, knowledge, identity, absence of\\n mistake, or lack of accident.\\n\\nFed. R. Evid. 404(b). In particular, Baptiste argues that his evidence would have\\n\\ndemonstrated only Pagnon’s capacity to implement his scheme—he had done so\\n\\nbefore, and so could do so again—rather than Pagnon’s character or propensity\\n\\nmore generally, and was therefore admissible as “reverse” 404(b) evidence—i.e.,\\n\\nbad-acts evidence that “tends to negate the defendant’s guilt of the crime charged\\n\\nagainst him.” United States v. Seals, 419 F.3d 600, 606 (7th Cir. 2005); see also,\\n\\ne.g., United States v. Cohen, 888 F.2d 770, 776 (11th Cir. 1989) (admitting\\n\\nevidence showing that the government’s witness “was capable of concocting and\\n\\nmanaging [a] fraudulent scheme”).\\n\\n The government doesn’t really dispute Baptiste’s premise—that the\\n\\nevidence bore on whether Pagnon was competent of pulling off such a scheme—\\n\\n\\n 8\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 9 of 28\\n\\n\\nand, indeed, it quotes Baptiste’s own brief back at him to make its point. In his\\n\\nbrief, the government says, Baptiste admits that “[t]he purpose of the proffered\\n\\nevidence was to show that Pagnon was capable of implementing his fraudulent tax-\\n\\nrefund scheme without the knowledge and involvement of the nominal owners.”\\n\\nBr. of Appellant at 41. In so stating, the government contends, Baptiste has\\n\\neffectively conceded that the evidence was of the “pure propensity” variety,\\n\\nprobative only of Pagnon’s general mendacity.\\n\\n The parties thus read the same bit of testimony in two very different ways.\\n\\nAnd not implausibly so. The evidence demonstrating that Pagnon had previously\\n\\ndeceived others could be construed as indicating either Pagnon’s capacity to do so\\n\\nagain and/or his propensity to do so again—either “Pagnon is well-equipped to lie\\n\\nin these circumstances” or “Pagnon is a liar.” The two interpretations are\\n\\ninterrelated—mutually reinforcing, you might say—but only the former is\\n\\ngenerally admissible under Rule 404(b).\\n\\n The trouble for Baptiste is that even scales won’t get the job done.\\n\\n“Determinations of admissibility of evidence rest largely within the discretion of\\n\\nthe trial judge and will not be disturbed on appeal absent a clear showing of an\\n\\nabuse of discretion.” United States v. Russell, 703 F.2d 1243, 1249 (11th Cir.\\n\\n1983). That this evidence can reasonably be read as (permissible) capacity\\n\\n\\n\\n\\n 9\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 10 of 28\\n\\n\\nevidence or (impermissible) character evidence demonstrates that Baptiste hasn’t\\n\\nmade the required “clear showing.”\\n\\n 2\\n\\n Baptiste also objects to how the government prosecuted his case. He makes\\n\\ntwo claims: first, that the government failed to disclose material issues involving\\n\\nkey witnesses; and second, that the prosecutor’s closing argument mischaracterized\\n\\nand exaggerated the evidence.\\n\\n a\\n\\n Baptiste alleges three disclosure violations. First, he argues that the\\n\\ngovernment failed to correct false testimony when one of its key witnesses,\\n\\nMoltimer, stated that he had derivative U.S. citizenship. Baptiste contends that, in\\n\\nfact, Moltimer was granted citizenship in exchange for testifying against Baptiste.\\n\\nSecond, Baptiste claims that Moltimer lied—and the government failed to\\n\\ncorrect—when he testified that he only met Pagnon in 2013. Pagnon died in 2011,\\n\\nand according to Baptiste, the government’s failure to set the record straight\\n\\nundermined Baptiste’s defense that “Pagnon masterminded the scheme to dupe the\\n\\nunsuspecting Baptiste.” Br. of Appellant at 19. Finally, Baptiste asserts that the\\n\\ngovernment failed to disclose a cooperation agreement with Louissaint, preventing\\n\\nthe jury from taking the true measure of Louissaint’s incentives.\\n\\n\\n\\n\\n 10\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 11 of 28\\n\\n\\n All three claims fall short. Baptiste’s first argument—about the\\n\\ngovernment’s alleged citizenship deal with Moltimer—rests on misleadingly\\n\\nselective quotations of the record. Baptiste explains that Moltimer’s attorney\\n\\nindicated that Moltimer had been “granted American citizenship,” encouraging us\\n\\nto infer that this “grant[]” was in exchange for Moltimer’s testimony against\\n\\nBaptiste. But just before the snippet that Baptiste highlights, Moltimer’s attorney\\n\\nstated that Moltimer “has been granted American citizenship because of his being\\n\\nbrought here when he was very young” and that Moltimer’s “mother was an\\n\\nAmerican citizen.” (emphasis added). And just after the passage excerpted by\\n\\nBaptiste, Moltimer’s attorney continued, “ICE came to see him and went forward,\\n\\nand he has been granted American citizenship, which he should have had from the\\n\\nstart.” (emphasis added). Come on. Baptiste’s assertion that the government\\n\\nfailed to disclose a citizenship-for-testimony deal with Moltimer is worse than\\n\\nconjectural—it appears to be flatly contradicted by the record.\\n\\n As for Moltimer’s testimony that he first met Pagnon three years after\\n\\nPagnon’s death, Baptiste is of course correct that—barring either time travel or\\n\\nnecromancy—Moltimer was wrong. But the focus for our purposes is not the\\n\\nshakiness of Moltimer’s chronology but rather the government’s failure to correct\\n\\nit. See Giglio v. United States, 405 U.S. 150, 153 (1972). To prove that the\\n\\ngovernment’s inaction violated his rights, Baptiste must show not only that the\\n\\n\\n 11\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 12 of 28\\n\\n\\ngovernment “[1] knowingly used perjured testimony, or failed to correct what [it]\\n\\nsubsequently learned was false testimony, and [2] that the falsehood was material,”\\n\\nbut also that the testimony was “[3] given with the willful intent to provide false\\n\\ntestimony and not as a result of a mistake, confusion, or faulty memory.” United\\n\\nStates v. Horner, 853 F.3d 1201, 1206 (11th Cir. 2017) (quoting another source),\\n\\ncert. denied, 138 S. Ct. 674 (2018). Even setting aside that the jury was repeatedly\\n\\nreminded—including by Baptiste’s attorney—that Pagnon was long dead by 2014,\\n\\nand that the falsehood was thus unlikely to have been material, Moltimer’s\\n\\ntestimony clearly indicates “mistake, confusion, or faulty memory.” See id. He\\n\\nsaid, for instance: “I’ve been in prison for almost two years. I think met Marvin\\n\\n[Pagnon] in 2014. I don’t remember the exact date . . . .” That testimony, we\\n\\nthink, hardly suggests a “willful intent” to deceive on the government’s part.\\n\\n Finally, Baptiste alleges that the government failed to disclose a cooperation\\n\\nagreement with Louissaint. Baptiste seizes on Louissaint’s testimony that he was\\n\\n“still going to get charged.” Baptiste points out that, in fact, Louissaint was never\\n\\ncharged, and he asks us to infer that the government cut a(nother) secret deal.\\n\\nAgain, we aren’t persuaded—in part because the claim is highly speculative, and in\\n\\npart because, in order to prove materiality, Baptiste must show that the jury made\\n\\ncredibility inferences from Louissaint’s comment that don’t necessarily follow.\\n\\nFor Louissaint’s comment to be material, the jurors would have had to follow\\n\\n\\n 12\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 13 of 28\\n\\n\\nsomething like the following chain of logic: (1) Louissaint testified that he was\\n\\n“still going to get charged”; (2) therefore [the jurors infer], Louissaint was not a\\n\\ncooperating witness, and in no way beholden to the government; (3) therefore [the\\n\\njurors continue], Louissaint’s testimony is more persuasive than it would have\\n\\nbeen if he weren’t “going to get charged.” The second statement doesn’t\\n\\nnecessarily follow from the first, nor the third from the second. Absent more\\n\\ncompelling evidence that Louissaint actually cut a deal, we decline to indulge the\\n\\nsorts of leaps that Baptiste’s argument requires.\\n\\n b\\n\\n The prosecutor told the jury during his closing argument that “everything out\\n\\nof [Francesse Chery’s] mouth was a complete lie,” that Baptiste “paid [a witness]\\n\\nto lie on the stand,” and that “the only thing that smells like a fish . . . is the\\n\\ntestimony of Francesse Chery.” Baptiste argues that these statements\\n\\nmischaracterized the evidence presented at trial. He also takes issue with the\\n\\nprosecutor’s summary of his defense as being “duped” by a “dead guy.”\\n\\n We’ll start with the fish. “[A] prosecutor is justified in arguing during\\n\\nclosing arguments that a particular witness is lying, if that is an inference\\n\\nsupported by the evidence at trial.” United States v. Schmitz, 634 F.3d 1247, 1270\\n\\n(11th Cir. 2011). Anael Chery’s testimony regarding his sister’s truthfulness (or\\n\\nlack thereof) provided adequate support for the inference that Francesse lied on the\\n\\n\\n 13\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 14 of 28\\n\\n\\nstand. The government’s description of the piscatorial odor emanating from\\n\\nFrancesse’s testimony was also permissible, if a little cheesy. Cf. Tucker v. Kemp,\\n\\n762 F.2d 1496, 1507 (11th Cir. 1985) (en banc) (references to defendant as “less\\n\\nthan human,” “not somebody in our society that we can afford to keep,” and “a\\n\\ndanger like a time bomb” permitted because “supported by the evidence.”). 2\\n\\n What about the government’s duped-by-a-dead-guy description of Baptiste’s\\n\\ndefense theory? While Baptiste undoubtedly dislikes the government’s turn of\\n\\nphrase, a passage from his own brief suffices to indicate its essential accuracy. To\\n\\nrecap, in closing, the government’s lawyer said that Baptiste was trying to “make it\\n\\nlook like some dead guy duped [him].” Baptiste’s brief on appeal summarizes his\\n\\nposition as follows: “Pagnon masterminded the scheme to dupe the unsuspecting\\n\\nBaptiste.” Br. of Appellant at 19. Pretty similar.\\n\\n In sum, although “[l]ittle time and no discussion is necessary to conclude\\n\\nthat it is improper for a prosecutor to use misstatements and falsehoods,” Davis v.\\n\\nZant, 36 F.3d 1538, 1548 (11th Cir. 1994), we see neither here.\\n\\n B\\n\\n\\n\\n2\\n We’ll cut the government a bit of slack given the context. The government’s “fish” comment\\ncame in response to Baptiste’s attorney’s extended analogy at closing comparing Baptiste to a\\nstarfish tossed ashore by a hurricane. Baptiste’s attorney concluded, “My client, Junior Jean\\nBaptiste, is the starfish,” asking the jury “to fulfill your oath, look in your heart, do what the law\\nrequires, apply reasonable doubt and throw [Baptiste] back, throw him back to his family.”\\nOoooooookay.\\n\\n 14\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 15 of 28\\n\\n\\n Now for the feature presentation. Recall Anael Chery testified that his\\n\\nsister, Francesse, told him that Baptiste was going to give her a Mercedes CLA 45\\n\\nin exchange for favorable testimony. Responding to Baptiste’s objection, the\\n\\ndistrict court said that’s “got to be hearsay.” Yep—textbook: “‘Hearsay’ means a\\n\\nstatement that: (1) the declarant does not make while testifying at the current trial\\n\\nor hearing; and (2) a party offers in evidence to prove the truth of the matter\\n\\nasserted in the statement.” Fed. R. Evid. 801(c). Even so, the district court\\n\\nultimately admitted Anael’s testimony under the statement-against-interest\\n\\nexception to the rule against hearsay, see Fed. R. Evid. 804(b)(3), because\\n\\nFrancesse’s underlying statement “could subject [her] to perjury” or other\\n\\n“charges.”\\n\\n 1\\n\\n Perhaps anticipating some rough sledding, the government makes only a\\n\\ntoken effort to defend the district court’s ruling—the statement-against-interest\\n\\nexception typically requires that the declarant be “unavailable,” after all, and\\n\\nFrancesse testified at trial. See Fed. R. Evid. 804(b) (“The following are not\\n\\nexcluded by the rule against hearsay if the declarant is unavailable as a witness . . .\\n\\n.”). The government goes on, though, to float three alternative rationales for why\\n\\nAnael’s testimony was admissible: as the statement of a party opponent under\\n\\nFederal Rule of Evidence 801(d)(2)(D); as a statement by a co-conspirator under\\n\\n\\n 15\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 16 of 28\\n\\n\\nRule 801(d)(2)(E); and, finally, under Rule 807’s residual exception. For good\\n\\nmeasure, the government concludes its tour of the Federal Rules of Evidence by\\n\\nadding that even if the admission of Anael’s testimony was erroneous, this error\\n\\nwas harmless in light of the weight of the other (undoubtedly admissible) evidence\\n\\nagainst Baptiste.\\n\\n We’ll forgo the tour and head straight to the finish line. We conclude that\\n\\nwe needn’t determine whether the district court erred in admitting Anael’s\\n\\ntestimony—as a statement against interest, or via any other exclusion or exception\\n\\nto the hearsay rule—because any error was indeed harmless. “Even where an\\n\\nabuse of discretion is shown, nonconstitutional evidentiary errors are not grounds\\n\\nfor reversal absent a reasonable likelihood that the defendant’s substantial rights\\n\\nwere affected.” United States v. Sellers, 906 F.2d 597, 601 (11th Cir. 1990); Fed.\\n\\nR. Evid. 103(a); see also United States v. Guzman, 167 F.3d 1350, 1353 (11th Cir.\\n\\n1999) (quoting Kotteakos v. United States, 328 U.S. 750, 776 (1946), for the\\n\\nproposition that “a nonconstitutional error requires reversal only if it resulted ‘in\\n\\nactual prejudice because it had substantial and injurious effect or influence in\\n\\ndetermining the jury’s verdict’”) (cleaned up).\\n\\n It’s the government’s burden to demonstrate that an alleged error was\\n\\nharmless, United States v. Sweat, 555 F.3d 1364, 1367 (11th Cir. 2009), and we\\n\\nconclude that it has met that burden here. Strike Anael’s testimony and the jury\\n\\n\\n 16\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 17 of 28\\n\\n\\nwould still have had before it—and we’re only hitting the highlights here—\\n\\nBaptiste’s curiously high income, gobs of documentary evidence showing\\n\\nfraudulent behavior, and the testimony of Moltimer and Louissaint, including that\\n\\nBaptiste planned to (and at trial did) “blame it on Marvin.” Under a long line of\\n\\nprecedent, see United States v. Hornaday, 392 F.3d 1306, 1316 (11th Cir. 2004)\\n\\n(collecting applications of Kotteakos), such overwhelming evidence of guilt\\n\\nsuffices to demonstrate that whatever error the district court might have committed\\n\\ndidn’t have a “substantial and injurious effect.” Kotteakos, 328 U.S. at 776. Nor\\n\\ncould the alleged error have influenced a reasonable juror’s verdict.\\n\\n 2\\n\\n Anael’s testimony relating Francesse’s statement about the Mercedes also\\n\\nunderlay the district court’s imposition of a sentencing enhancement for\\n\\nobstruction of justice. Baptiste objects that because Francesse’s statement was\\n\\ninadmissible hearsay, the district court improperly relied on it in finding\\n\\nobstruction.\\n\\n Section 3C1.1 of the Sentencing Guidelines provides for a two-level\\n\\nenhancement if “the defendant willfully obstructed or impeded, or attempted to\\n\\nobstruct or impede, the administration of justice with respect to the investigation,\\n\\nprosecution, or sentencing of the instant offense of conviction.” U.S.S.G. § 3C1.1.\\n\\n\\n\\n\\n 17\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 18 of 28\\n\\n\\nExamples of covered conduct include “committing, suborning, or attempting to\\n\\nsuborn perjury.” Id. at cmt. 4(B).\\n\\n If Anael’s testimony were admissible under one of the exceptions or\\n\\nexclusions to the rule against hearsay, there wouldn’t be any question that the\\n\\ndistrict court was on firm footing in relying on it for the enhancement. But if the\\n\\ntestimony was inadmissible, then our earlier conclusion that its admission at trial\\n\\nwas harmless wouldn’t (in and of itself) validate the district court’s reliance on it at\\n\\nsentencing. It all boils down to harm. Even if we strike Anael’s testimony, there\\n\\nwas still ample evidence supporting Baptiste’s conviction. With regard to that\\n\\ncomponent of the trial, Anael’s testimony was harmless. But Anael’s testimony\\n\\nprovided the only evidence supporting Baptiste’s enhancement for obstruction.\\n\\nWith regard to that component of the trial, Anael’s testimony was necessarily\\n\\nharmful.\\n\\n So have we reached the end of the road, such that we finally have to make a\\n\\ncall on the admissibility of Anael’s testimony? Not quite. Thanks to the “reliable\\n\\nhearsay” doctrine—which permits sentencing courts to rely on otherwise-\\n\\ninadmissible hearsay evidence in certain circumstances—we can reserve judgment\\n\\non the evidence’s admissibility (as we’ve done) while still addressing the district\\n\\ncourt’s enhancement.\\n\\n\\n\\n\\n 18\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 19 of 28\\n\\n\\n The doctrine is rooted in § 6A1.3 of the Sentencing Guidelines, which states\\n\\nthat a court “may consider relevant information without regard to its admissibility\\n\\nunder the rules of evidence applicable at trial, provided that the information has\\n\\nsufficient indicia of reliability to support its probable accuracy.” U.S.S.G. §\\n\\n6A1.3. Indeed, it’s well established that “[i]n determining the relevant facts,\\n\\nsentencing judges are not restricted to information that would be admissible at\\n\\ntrial.” Id. at cmt. The reason, the Supreme Court has explained, is that “modern\\n\\nconcepts individualizing punishment have made it all the more necessary that a\\n\\nsentencing judge not be denied an opportunity to obtain pertinent information by a\\n\\nrequirement of rigid adherence to restrictive rules of evidence properly applicable\\n\\nto the trial.” Williams v. New York, 337 U.S. 241, 247 (1949). Williams’ “modern\\n\\nconcepts” are old hat today, but no less important for it. Hence the need for a\\n\\nclear-headed understanding of when—and under what circumstances—hearsay\\n\\nevidence, in particular, can be deemed sufficiently reliable to support a sentencing\\n\\ndecision.\\n\\n The Sentencing Guidelines’ text—just quoted—is crystal clear. It articulates\\n\\na single precondition to a district court’s consideration of otherwise inadmissible\\n\\nevidence, including hearsay, as part of its sentencing calculus: The evidence is fair\\n\\ngame “provided that the information has sufficient indicia of reliability to support\\n\\nits probable accuracy.” U.S.S.G. § 6A1.3. The trickier question is whether our\\n\\n\\n 19\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 20 of 28\\n\\n\\ncase law has engrafted another, extratextual criterion. In particular, we’ve sent\\n\\nvaguely mixed signals about whether, as a prerequisite to relying on hearsay\\n\\nevidence at sentencing, a district court must make express findings on the record\\n\\nthat the evidence is reliable—or if, instead, as the Guidelines’ language indicates,\\n\\nit’s enough that the record as a whole provides “sufficient indicia of reliability.”\\n\\nThe distinction matters here because it’s undisputed that the district court did not\\n\\nmake any explicit findings about the reliability of Francesse’s testimony.\\n\\n Some have read our decision in United States v. Lee, 68 F.3d 1267 (11th Cir.\\n\\n1995), as requiring express findings that the hearsay evidence is reliable. For\\n\\nreasons we have explained before—and will reiterate here—that is incorrect. The\\n\\ndefendant in Lee objected that the district court had relied on the hearsay evidence\\n\\nof one of his co-conspirators—Russell—in determining his sentence. See id. at\\n\\n1275. We observed, as an initial matter, that “[a]ccording to the Sentencing\\n\\nGuidelines, and our case law interpreting them, the district court ‘may consider any\\n\\ninformation, including reliable hearsay, regardless of the information’s\\n\\nadmissibility at trial, provided that there are sufficient indicia of reliability to\\n\\nsupport its probable accuracy.’” Id. (quoting United States v. Castellanos, 904\\n\\nF.2d 1490, 1495 (11th Cir. 1990), and citing U.S.S.G. § 6A1.3). We stated that\\n\\n“the focus” of the inquiry “is upon the question of [the hearsay evidence’s]\\n\\nreliability,” and we further explained that reliability “must be determined on a case\\n\\n\\n 20\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 21 of 28\\n\\n\\nby case basis.” Id. The government there contended that Russell’s hearsay\\n\\nstatements were reliable because they were “against his penal interest,” and we\\n\\nacknowledged that as a general matter “a statement made against one’s penal\\n\\ninterest may bolster its credibility.” Id. at 1275–76. Even so, we concluded that\\n\\n“[i]n Russell’s case,” his penal interest “fail[ed] to qualify as a ‘sufficient\\n\\nindic[ium] of reliability’” given his “status as a fugitive from justice.” Id. at 1276.\\n\\nAccordingly, we concluded that “[s]pecific findings on Russell’s credibility [were]\\n\\nnecessary before the district court c[ould] use this evidence as a basis for\\n\\ndetermining” the defendant’s sentence, and we remanded for the district court “to\\n\\nmake further findings as to the reliability of Russell’s hearsay statement.” Id.\\n\\n Our subsequent decision in United States v. Gordon, 231 F.3d 750 (11th Cir.\\n\\n2000), clarified Lee’s holding and reach. The defendant there cited Lee in support\\n\\nof his contention “that the district court was required to make specific findings\\n\\nregarding the reliability” of hearsay testimony on which the government sought to\\n\\nrely at sentencing. Id. at 760. In response, we explained that Lee was never\\n\\nintended to—and didn’t—prescribe a per se rule that findings are always\\n\\nnecessary. Id. at 760–61. Rather, we made clear, Lee’s holding—that findings\\n\\nwere required there—was a product of the “circumstances” that underlay that\\n\\nparticular case. Id. Specifically, we said, it was because the traditional against-\\n\\npenal-interest justification didn’t cut it, and because the hearsay declarant was a\\n\\n\\n 21\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 22 of 28\\n\\n\\nfugitive, that the Lee court found “the absence of findings … particularly glaring\\n\\nand troubling.” Id. By contrast, we emphasized in Gordon, where the record and\\n\\nthe circumstances of the case “demonstrate adequate indicia of reliability,”\\n\\nfindings are not strictly necessary. Id. at 761. In such a case, a “district court’s\\n\\nfailure to make separate findings regarding the reliability of [hearsay] statements\\n\\n[is] not error.” Id. Put differently, “the absence of such findings does not\\n\\nnecessarily require reversal or remand where the reliability of the statements is\\n\\napparent from the record.” Id.3\\n\\n So … is this a “Lee” case—in which the traditional indicia of reliability fall\\n\\nshort and findings are thus required—or a “Gordon” case—in which the traditional\\n\\nindicia provide the necessary comfort? Clearly the latter, we think. For one,\\n\\n\\n\\n\\n3\\n One loose end: After Lee but before Gordon, in United States v. Anderton we addressed a\\nsimilar complaint that a district court had erroneously based its sentencing decision on\\n“unreliable hearsay.” 136 F.3d 747, 751 (11th Cir. 1998). In the course of rejecting that\\ncontention, we recited that “a court may rely on hearsay at sentencing, as long as the evidence\\nhas sufficient indicia of reliability, the court makes explicit findings of fact as to credibility, and\\nthe defendant has an opportunity to rebut the evidence.” Id. Seemingly in support of the\\n“explicit findings” aspect of that statement—albeit without saying so expressly—we cited our\\nearlier decision in Lee. Id. Anderton, though, doesn’t stand for the sort of per se rule that the\\nGordon defendant sought and that the Gordon panel rejected. The parties in Anderton never\\ndisputed whether “explicit findings” were required, and nothing rode on that issue. Id. Our\\ndecision simply assumed (at least in the circumstances presented) that they were, observed that\\nthe district court had made them, and held that they weren’t clearly erroneous. Id. The same is\\ntrue of our post-Gordon decisions in United States v. Zlatogur, 271 F.3d 1025, 1031 (11th Cir.\\n2001), and United States v. Patti, 337 F.3d 1317, 1326 (11th Cir. 2003), which merely recite\\nAnderton’s boilerplate. And to the extent that one could read Anderton as contradicting Lee, and\\nas holding that express findings of credibility are always required, we are bound by our prior-\\nprecedent rule to apply our earlier decision in Lee. See United States v. Steele, 147 F.3d 1316,\\n1318 (11th Cir. 1998) (en banc).\\n 22\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 23 of 28\\n\\n\\nFrancesse’s statement—that Baptiste had promised her a Mercedes in exchange for\\n\\nfavorable testimony—was clearly against both her and Baptiste’s penal interests,4\\n\\nand there’s no “fugitive from justice”-like reason here, as there was in Lee, to\\n\\ndiscount the force of that established gauge of reliability. For another, Francesse’s\\n\\nstatement aligned with Louissaint’s testimony regarding how Baptiste planned to\\n\\n“blame everything” on Pagnon. And for yet another, her statement specifically\\n\\nidentified the make and model of car she was set to receive. All of these are\\n\\n“indicia of reliability,” and by our lights, collectively clear the “sufficien[cy]”\\n\\nthreshold. See Lee, 68 F.3d at 1276.\\n\\n Because this isn’t the sort of case in which explicit findings were required,\\n\\nthe district court didn’t err in not making any. Nor did it err in relying on\\n\\nFrancesse’s sufficiently-reliable testimony to impose the obstruction-of-justice\\n\\nenhancement.\\n\\n C\\n\\n That leaves a few sentencing-related loose ends.\\n\\n 1\\n\\n As an initial matter, Baptiste challenges the district court’s imposition of\\n\\nsentencing enhancements for the amount of money that the government lost and\\n\\n\\n4\\n Even if, as we have said, it might (?) not formally qualify for the statement-against-interest\\nexception to the hearsay rule because Francesse testified at trial and thus might (?) not have been\\n“unavailable,” as Rule 804(b)(3) requires. See supra at 16–17.\\n\\n 23\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 24 of 28\\n\\n\\nthe number and vulnerability of the scheme’s victims. U.S.S.G. § 2B1.1(b)(1)(K)\\n\\n(amount lost exceeding $9,500,000); § 2B1.1(b)(2)(A) (offense involving more\\n\\nthan 10 victims); § 3A1.1(b)(1)–(2) (vulnerable victims).\\n\\n a\\n\\n As for the amount of money involved, we note first that Baptiste expressly\\n\\nwithdrew all objections except those relating to the enhancements for obstruction\\n\\nof justice and the number and vulnerability of his victims. We needn’t determine\\n\\nwhether in doing so Baptiste invited any alleged error, because even under the\\n\\ntypical clear-error standard, see United States v. Castaneda-Pozo, 877 F.3d 1249,\\n\\n1251 (11th Cir. 2017), Baptiste’s argument fails. The government submitted\\n\\nevidence indicating that Surveillance Master and its associated bank accounts were\\n\\nall under Baptiste’s name. The identification documents reviewed by Customs and\\n\\nImmigration Services also demonstrated that it was reasonable for the district court\\n\\nto conclude that virtually all of Surveillance Master’s check-cashing business was\\n\\nfraudulent—629 of the 630 IDs reviewed were fakes. Finally, “[t]he district court\\n\\nis only required to make a reasonable estimate of the loss.” United States v. Cobb,\\n\\n842 F.3d 1213, 1218 (11th Cir. 2016); U.S.S.G. § 2B1.1, cmt. 3(C); see also\\n\\nUnited States v. Medina, 485 F.3d 1291, 1304 (11th Cir. 2007) (indicating that a\\n\\nreasonable estimate is particularly appropriate in fraud cases because the amount\\n\\nlost is often difficult to determine with precision). We thus have evidence\\n\\n\\n 24\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 25 of 28\\n\\n\\nindicating that the sums lost went to Baptiste, that virtually all of the money that\\n\\npassed through the check-cashing business was linked to the scheme, as well as\\n\\nprecedent underscoring that the district court could apply this enhancement even if\\n\\nit couldn’t determine a precise figure. Taken together, we don’t see error here,\\n\\nmuch less of the “clear” variety.\\n\\n b\\n\\n Baptiste’s challenges to the enhancements based on the vulnerability and\\n\\nnumber of the scheme’s victims travel together. The Sentencing Guidelines\\n\\nprovide for a two-level increase on the ground that he “knew or should have\\n\\nknown that a victim of the offense was a vulnerable victim.” U.S.S.G. §\\n\\n3A1.1(b)(1). We have held that “both circumstances and immutable characteristics\\n\\ncan render a victim vulnerable,” United States v. Bradley, 644 F.3d 1213, 1288\\n\\n(11th Cir. 2011), and that “[n]either bodily injury nor financial loss is required” for\\n\\nthe enhancement, United States v. Moran, 778 F.3d 942, 978 (11th Cir. 2015). Our\\n\\ndecision in United States v. Pierre, 825 F.3d 1183 (11th Cir. 2016), settles the\\n\\nmatter with regard to Baptiste’s use of prisoners’ identities. We explained there\\n\\nthat “inmates have unique circumstances and immutable characteristics that make\\n\\nthem . . . vulnerable to . . . fraudulent activity.” Id. at 1196; see also id. (“Inmates\\n\\nusually do not file tax returns during periods of incarceration, and they are less\\n\\nlikely to discover that their identities have been compromised.”). The government\\n\\n\\n 25\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 26 of 28\\n\\n\\npresented evidence showing that Baptiste specifically requested the identities of\\n\\nprisoners, so the vulnerability-of-victims enhancement checks out.5 So too the\\n\\nnumber-of-victims enhancement; the district could reasonably infer from the\\n\\nevidence detailing the number of checks that Baptiste cashed that he had targeted a\\n\\n“large number” of people. See U.S.S.G. § 3A1.1(b)(2) (stipulating that if (i) §\\n\\n3A1.1(b)(1) applies and (ii) “the offense involved a large number of vulnerable\\n\\nvictims,” the defendant’s offense level should be raised an additional two levels.).\\n\\n 2\\n\\n Finally—and simply for us—the parties agree that the district court plainly\\n\\nerred in failing to “address the defendant personally in order to permit the\\n\\ndefendant to speak” at sentencing, as Federal Rule of Criminal Procedure\\n\\n32(i)(4)(A)(ii) expressly requires. 6 In response to the district court’s query, “Does\\n\\nthe defendant wish to address me?”, Baptiste’s attorney responded on his client’s\\n\\nbehalf: “Mr. Baptiste respectfully has nothing to add.” However formalistic it may\\n\\nseem, and even reviewing for plain error, we agree with the parties that this\\n\\n5\\n Because the record supports the district court’s finding that Baptiste targeted incarcerated\\nvictims, we can affirm the enhancement for vulnerable victims on that ground alone. We needn’t\\ndetermine whether the deceased count as “vulnerable” as that term is used in § 3A1.1(b).\\nCompare United States v. Roberson, 872 F.2d 597, 609 (5th Cir. 1989) (sentence of a defendant\\nconvicted of fraud properly enhanced where the defendant used a dead companion’s credit card\\nafter burning his corpse), with United States v. Shumway, 112 F.3d 1413, 1424 (10th Cir. 1997)\\n(refusing to find that skeletal remains are a “vulnerable victim”).\\n6\\n “Before imposing sentence, the court must . . . (ii) address the defendant personally in order to\\npermit the defendant to speak or present any information to mitigate the sentence . . . .” Fed. R.\\nCrim. P. 32(i)(4)(A)(ii).\\n\\n 26\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 27 of 28\\n\\n\\nexchange—between court and counsel rather than court and defendant—amounted\\n\\nto “(1) error, (2) that is plain and (3) that affect[ed] [Baptiste’s] substantial rights,”\\n\\nand that the error “[4] seriously affect[ed] the fairness, integrity, or public\\n\\nreputation of judicial proceedings.” United States v. Turner, 474 F.3d 1265, 1276\\n\\n(11th Cir. 2007); United States v. Doyle, 857 F.3d 1115, 1120 (11th Cir. 2017)\\n\\n(“[T]he presumption of prejudice in denial of allocution cases is still the rule when\\n\\na defendant is not sentenced at the bottom of his guidelines range . . . .”); United\\n\\nStates v. Prouty, 303 F.3d 1249, 1251 n.1 (11th Cir. 2002) (rejecting “the\\n\\nGovernment’s argument that directing comments to the lawyers suffices to comply\\n\\nwith the rule”). On remand, Baptiste should be afforded the opportunity to address\\n\\nthe court directly.\\n\\n III\\n\\n Accordingly, we hold as follows:\\n\\n (A) The district court did not abuse its discretion in refusing to allow\\n\\nBaptiste to introduce evidence that Marvin Pagnon had duped others into\\n\\nparticipating in similar schemes. Nor did the government fail to disclose any\\n\\nmaterial information concerning its witnesses or mischaracterize or exaggerate the\\n\\nevidence during its closing argument.\\n\\n (B) Any error that the district court might have committed in admitting\\n\\nAnael Chery’s testimony relaying his sister Francesse’s statement that Baptiste had\\n\\n\\n 27\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 28 of 28\\n\\n\\npromised her a car in exchange for favorable testimony was harmless. And the\\n\\ndistrict court did not err in relying on the same evidence to support an obstruction-\\n\\nrelated sentencing enhancement because the record reveals that the evidence bore\\n\\n“sufficient indicia of reliability.”\\n\\n (C) The district court did not err in imposing sentencing enhancements\\n\\npertaining to the amount of money that the government lost and the number and\\n\\nvulnerability of Baptiste’s scheme’s victims. The district court did reversibly err,\\n\\nhowever, when it failed to permit Baptiste to allocute personally.\\n\\n AFFIRMED IN PART AND REMANDED.\\n\\n\\n\\n\\n 28\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434220/\", \"author_raw\": \"NEWSOM, Circuit Judge:\"}]}"},"judge1":{"kind":"string","value":"PRYOR"},"judge2":{"kind":"string","value":"NEWSOM"},"judge3":{"kind":"string","value":"BRANCH"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4656967/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: UNITED STATES of America, Plaintiff - Appellee, v. Junior Jean BAPTISTE, Defendant - Appellant.\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-28\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nCase: 16-17175 Date Filed: 08/28/2019 Page: 1 of 28\n\n\n [PUBLISH]\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n ________________________\n\n Nos. 16-17175, 16-17595\n ________________________\n\n\n D.C. Docket No. 1:15-cr-20777-JEM-1\n\n\nUNITED STATES OF AMERICA,\n\n Plaintiff - Appellee,\n versus\n\nJUNIOR JEAN BAPTISTE,\n\n Defendant - Appellant.\n\n\n ________________________\n\n Appeals from the United States District Court\n for the Southern District of Florida\n ________________________\n\n (August 28, 2019)\n\n\n\nBefore WILLIAM PRYOR, NEWSOM, and BRANCH, Circuit Judges.\n\nNEWSOM, Circuit Judge:\n\n Jean Junior Baptiste was a businessman. In addition to a portfolio of media\n\nand entertainment companies, Baptiste owned and operated Surveillance Master\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 2 of 28\n\n\nCCTV. Surveillance Master installed security cameras, repaired computers, and—\n\none of these things is not like the others—cashed checks. In 2016, a jury found\n\nthat Baptiste had taken part in an $11 million scheme that involved cashing tax-\n\nrefund checks that he had fraudulently obtained in the names of inmates, minors,\n\ndead people, and other insentient or otherwise unsuspecting “clients.” Baptiste\n\nnow appeals his conviction and 212-month sentence.\n\n Although Baptiste raises a number of issues on appeal, we focus primarily\n\non two questions related to the hearsay testimony of a government witness. The\n\nabridged version of the story: Francesse Chery was one of Baptiste’s key\n\nwitnesses. The government countered with her brother, Anael Chery, who testified\n\n(among other things) that Francesse had told him that, in exchange for her (false)\n\ntestimony supporting Baptiste’s narrative, Baptiste would give her a Mercedes.\n\nBaptiste argues that Anael’s testimony was inadmissible hearsay and that the\n\ndistrict court’s error in allowing the jury to hear it tainted both his conviction and\n\nhis sentence.\n\n Baptiste’s challenge presents two questions. First, was Anael’s testimony\n\nindeed inadmissible hearsay? The district court admitted the testimony pursuant to\n\nthe statement-against-interest exception to the general prohibition on hearsay\n\nevidence, and on appeal the government has offered a smattering of additional\n\ntheories of admissibility. We conclude that we needn’t decide whether Anael’s\n\n\n 2\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 3 of 28\n\n\ntestimony was inadmissible hearsay because even if the district court did err in\n\nallowing it, the error was harmless. There was more than enough compelling—and\n\nundoubtedly admissible—evidence to support Baptiste’s conviction.\n\n Second, and (sort of) relatedly, did the district court err in relying on Anael’s\n\ntestimony when it imposed a sentencing enhancement for obstructing justice? If\n\nyou’re saying, “Didn’t they just say they weren’t going to decide whether the\n\ntestimony was admissible?”—we hear you. As it turns out, though, thanks to a\n\ndoctrine called (somewhat oxymoronically) “reliable hearsay” we can answer the\n\nsecond question without deciding the first. Under the reliable-hearsay doctrine, so\n\nlong as certain preconditions are met, a sentencing court can rely on evidence that\n\nwould be off-limits in the guilt phase. For Baptiste, this means that even if Anael’s\n\ndescription of his sister’s supposed deal was inadmissible hearsay (and we aren’t\n\nsaying either way) the district court might not have erred in relying on that\n\ntestimony for the obstruction enhancement—again, so long as the preconditions are\n\nmet.\n\n So, what are they? Well, our case law has arguably sent mixed signals about\n\nthat. There is, though, a synthesis. We hold (and clarify) today that the Sentencing\n\nGuidelines permit use of hearsay testimony so long as the overall record provides\n\n“sufficient indicia of reliability”—and we conclude that the indicia of reliability\n\nhere are sufficient.\n\n\n 3\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 4 of 28\n\n\n Accordingly—and because we find that none of Baptiste’s other arguments\n\nhas merit—we affirm Baptiste’s conviction and, with one asterisk, his sentence.\n\nThe asterisk? Pursuant to the parties’ joint request, we remand the case to the\n\ndistrict court for the limited purpose of allowing Baptiste to allocute, a right that he\n\nwas denied at sentencing.\n\n I\n\n During Baptiste’s trial, two cooperating witnesses, Andy Louissaint and Karl\n\nMoltimer, described the tax-refund scheme. Baptiste would first provide tax\n\npreparers such as Louissaint and Moltimer with a list of individuals—many of\n\nthem young, incarcerated, dead, or otherwise unlikely to discover the use of their\n\nidentities—along with accompanying personal information such as birthdates and\n\nsocial security numbers. The tax preparers used that information to (fraudulently)\n\nobtain refund checks from the IRS, which they then brought to Baptiste’s\n\nSurveillance Master for cashing.\n\n Baptiste contracted with a third party to process the checks. He would\n\nfabricate the necessary documents—work permits, passports, green cards, driver’s\n\nlicenses—to avoid raising red flags. Baptiste and the tax preparers would divide\n\nthe proceeds, with Baptiste often taking a cut of between 25% and 50%. Not bad,\n\ngiven that typical check-cashing fees range from 1.5% to 3%.\n\n\n\n\n 4\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 5 of 28\n\n\n When the IRS reviewed Surveillance Master’s records in conjunction with a\n\nroutine inspection, an agent noticed that several refund checks were for identical\n\n(or nearly identical) amounts. During a subsequent investigation, the IRS obtained\n\ncopies of identification documents, Treasury checks, and lists containing personal\n\ninformation for both the quick and the dead. Customs and Immigration Services\n\nconducted an evaluation of approximately 630 permanent-resident and\n\nemployment-authorization cards that Baptiste had used to verify the checks; all but\n\none—belonging to Baptiste’s wife—were fraudulent.\n\n The ensuing indictment charged Baptiste with money laundering, conspiracy\n\nto commit money laundering, possession of false identification documents with\n\nunlawful intent, theft of government money, and aggravated identity theft. At trial,\n\nBaptiste’s story was that he was an unwitting participant in a broader scheme\n\nconcocted by a now-dead business partner, Marvin Pagnon. Pagnon’s ex-\n\ngirlfriend, the aforementioned Francesse Chery, echoed Baptiste’s account in her\n\ntestimony and explained how Pagnon operated. In particular, she explained that\n\nbecause Pagnon had poor credit, he would ask would-be associates such as\n\nBaptiste to partner with him. Pagnon, she said, laid out a collaboration that was\n\nstraightforward and above board: Pagnon would obtain (real) refund checks from\n\ntax preparers; Baptiste would cash whatever Pagnon handed him. Baptiste signed\n\nup, and he seems to have been well compensated for his efforts. During the\n\n\n 5\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 6 of 28\n\n\nscheme, Baptiste purchased a Mercedes Benz CL 63, a Mercedes Benz E Class, a\n\nRange Rover, and—again, one of these things is not like the others—a $345,000\n\nshare in a 200-foot long cargo ship.\n\n The government painted a different picture. Moltimer and Louissaint both\n\ntestified, for instance, that Pagnon played no role in Baptiste’s check-cashing\n\noperation. Louissaint added that when Baptiste first became aware that an\n\nindictment could be coming down the pike, he had told Louissaint that he planned\n\nto “basically blame everything on Marvin [Pagnon].” The government also offered\n\ntestimony from Francesse Chery’s brother, Anael. Anael first opined, as a general\n\nmatter, that his sister just wasn’t a very truthful person. He went on to testify,\n\nmuch more specifically, that Francesse had told him that Baptiste was going to\n\ngive her a Mercedes CLA 45 so long as she “h[e]ld up her end of the bargain” by\n\nbacking Baptiste’s it-was-all-Pagnon’s-idea narrative. Baptiste objected that\n\nAnael’s testimony—relaying what his sister had allegedly told him—was\n\ninadmissible hearsay. The district court ultimately admitted Anael’s statement,\n\nciting the statement-against-interest exception to the general prohibition on hearsay\n\nevidence. See Fed. R. Evid. 804(b)(3).\n\n The jury convicted Baptiste on all counts. At sentencing, the district court\n\napplied enhancements for the amount of money that Baptiste had fraudulently\n\nobtained from the government (exceeding $11 million), for the number of victims\n\n\n 6\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 7 of 28\n\n\n(perhaps hundreds), for the victims’ vulnerable status (many incapacitated in some\n\nform or fashion), and for obstructing justice (for the perjury-for-a-Mercedes deal\n\nwith Francesse). The district court sentenced Baptiste to 212 months’\n\nimprisonment and restitution of more than $11 million. 1 This is Baptiste’s appeal.\n\n II\n\n As already noted, our focus here will be the related contentions that the\n\ndistrict court (1) abused its discretion by admitting Anael’s hearsay testimony as\n\nevidence against Baptiste and (2) violated the Sentencing Guidelines by imposing\n\nan obstruction-based enhancement that relied on the same. Before getting there,\n\nthough, we’ll address a few other guilt-phase matters—namely, whether the district\n\ncourt erroneously refused to admit Rule 404(b) evidence, whether the government\n\nfailed to reveal (alleged) deals with its witnesses, and whether the prosecutor made\n\nimproper statements to the jury during his closing argument. Then, after dealing\n\nwith the Anael-related issues, which span the trial and sentencing phases, we’ll\n\nconclude with a few additional, sentencing-specific arguments—namely, that the\n\ndistrict court erred in increasing Baptiste’s sentence due to the amount of money\n\nthat the government lost as well as the number and vulnerability of the scheme’s\n\nvictims, as well as in declining to permit Baptiste to allocute personally.\n\n A\n\n1\n The court ordered that $70,000 of that total be paid jointly and severally with Moltimer.\n\n 7\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 8 of 28\n\n\n 1\n\n During the trial, the district court refused to allow Baptiste to put on\n\nevidence that Marvin Pagnon had duped others into participating in similar\n\nschemes. On appeal, Baptiste contends that, in so doing, the district court\n\nmisapplied Federal Rule of Evidence 404(b), which states, in relevant part—\n\n (1) Prohibited Uses. Evidence of a crime, wrong, or other act is not\n admissible to prove a person’s character in order to show that on a\n particular occasion the person acted in accordance with the character.\n (2) Permitted Uses; Notice in a Criminal Case. This evidence may\n be admissible for another purpose, such as proving motive,\n opportunity, intent, preparation, plan, knowledge, identity, absence of\n mistake, or lack of accident.\n\nFed. R. Evid. 404(b). In particular, Baptiste argues that his evidence would have\n\ndemonstrated only Pagnon’s capacity to implement his scheme—he had done so\n\nbefore, and so could do so again—rather than Pagnon’s character or propensity\n\nmore generally, and was therefore admissible as “reverse” 404(b) evidence—i.e.,\n\nbad-acts evidence that “tends to negate the defendant’s guilt of the crime charged\n\nagainst him.” United States v. Seals, 419 F.3d 600, 606 (7th Cir. 2005); see also,\n\ne.g., United States v. Cohen, 888 F.2d 770, 776 (11th Cir. 1989) (admitting\n\nevidence showing that the government’s witness “was capable of concocting and\n\nmanaging [a] fraudulent scheme”).\n\n The government doesn’t really dispute Baptiste’s premise—that the\n\nevidence bore on whether Pagnon was competent of pulling off such a scheme—\n\n\n 8\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 9 of 28\n\n\nand, indeed, it quotes Baptiste’s own brief back at him to make its point. In his\n\nbrief, the government says, Baptiste admits that “[t]he purpose of the proffered\n\nevidence was to show that Pagnon was capable of implementing his fraudulent tax-\n\nrefund scheme without the knowledge and involvement of the nominal owners.”\n\nBr. of Appellant at 41. In so stating, the government contends, Baptiste has\n\neffectively conceded that the evidence was of the “pure propensity” variety,\n\nprobative only of Pagnon’s general mendacity.\n\n The parties thus read the same bit of testimony in two very different ways.\n\nAnd not implausibly so. The evidence demonstrating that Pagnon had previously\n\ndeceived others could be construed as indicating either Pagnon’s capacity to do so\n\nagain and/or his propensity to do so again—either “Pagnon is well-equipped to lie\n\nin these circumstances” or “Pagnon is a liar.” The two interpretations are\n\ninterrelated—mutually reinforcing, you might say—but only the former is\n\ngenerally admissible under Rule 404(b).\n\n The trouble for Baptiste is that even scales won’t get the job done.\n\n“Determinations of admissibility of evidence rest largely within the discretion of\n\nthe trial judge and will not be disturbed on appeal absent a clear showing of an\n\nabuse of discretion.” United States v. Russell, 703 F.2d 1243, 1249 (11th Cir.\n\n1983). That this evidence can reasonably be read as (permissible) capacity\n\n\n\n\n 9\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 10 of 28\n\n\nevidence or (impermissible) character evidence demonstrates that Baptiste hasn’t\n\nmade the required “clear showing.”\n\n 2\n\n Baptiste also objects to how the government prosecuted his case. He makes\n\ntwo claims: first, that the government failed to disclose material issues involving\n\nkey witnesses; and second, that the prosecutor’s closing argument mischaracterized\n\nand exaggerated the evidence.\n\n a\n\n Baptiste alleges three disclosure violations. First, he argues that the\n\ngovernment failed to correct false testimony when one of its key witnesses,\n\nMoltimer, stated that he had derivative U.S. citizenship. Baptiste contends that, in\n\nfact, Moltimer was granted citizenship in exchange for testifying against Baptiste.\n\nSecond, Baptiste claims that Moltimer lied—and the government failed to\n\ncorrect—when he testified that he only met Pagnon in 2013. Pagnon died in 2011,\n\nand according to Baptiste, the government’s failure to set the record straight\n\nundermined Baptiste’s defense that “Pagnon masterminded the scheme to dupe the\n\nunsuspecting Baptiste.” Br. of Appellant at 19. Finally, Baptiste asserts that the\n\ngovernment failed to disclose a cooperation agreement with Louissaint, preventing\n\nthe jury from taking the true measure of Louissaint’s incentives.\n\n\n\n\n 10\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 11 of 28\n\n\n All three claims fall short. Baptiste’s first argument—about the\n\ngovernment’s alleged citizenship deal with Moltimer—rests on misleadingly\n\nselective quotations of the record. Baptiste explains that Moltimer’s attorney\n\nindicated that Moltimer had been “granted American citizenship,” encouraging us\n\nto infer that this “grant[]” was in exchange for Moltimer’s testimony against\n\nBaptiste. But just before the snippet that Baptiste highlights, Moltimer’s attorney\n\nstated that Moltimer “has been granted American citizenship because of his being\n\nbrought here when he was very young” and that Moltimer’s “mother was an\n\nAmerican citizen.” (emphasis added). And just after the passage excerpted by\n\nBaptiste, Moltimer’s attorney continued, “ICE came to see him and went forward,\n\nand he has been granted American citizenship, which he should have had from the\n\nstart.” (emphasis added). Come on. Baptiste’s assertion that the government\n\nfailed to disclose a citizenship-for-testimony deal with Moltimer is worse than\n\nconjectural—it appears to be flatly contradicted by the record.\n\n As for Moltimer’s testimony that he first met Pagnon three years after\n\nPagnon’s death, Baptiste is of course correct that—barring either time travel or\n\nnecromancy—Moltimer was wrong. But the focus for our purposes is not the\n\nshakiness of Moltimer’s chronology but rather the government’s failure to correct\n\nit. See Giglio v. United States, 405 U.S. 150, 153 (1972). To prove that the\n\ngovernment’s inaction violated his rights, Baptiste must show not only that the\n\n\n 11\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 12 of 28\n\n\ngovernment “[1] knowingly used perjured testimony, or failed to correct what [it]\n\nsubsequently learned was false testimony, and [2] that the falsehood was material,”\n\nbut also that the testimony was “[3] given with the willful intent to provide false\n\ntestimony and not as a result of a mistake, confusion, or faulty memory.” United\n\nStates v. Horner, 853 F.3d 1201, 1206 (11th Cir. 2017) (quoting another source),\n\ncert. denied, 138 S. Ct. 674 (2018). Even setting aside that the jury was repeatedly\n\nreminded—including by Baptiste’s attorney—that Pagnon was long dead by 2014,\n\nand that the falsehood was thus unlikely to have been material, Moltimer’s\n\ntestimony clearly indicates “mistake, confusion, or faulty memory.” See id. He\n\nsaid, for instance: “I’ve been in prison for almost two years. I think met Marvin\n\n[Pagnon] in 2014. I don’t remember the exact date . . . .” That testimony, we\n\nthink, hardly suggests a “willful intent” to deceive on the government’s part.\n\n Finally, Baptiste alleges that the government failed to disclose a cooperation\n\nagreement with Louissaint. Baptiste seizes on Louissaint’s testimony that he was\n\n“still going to get charged.” Baptiste points out that, in fact, Louissaint was never\n\ncharged, and he asks us to infer that the government cut a(nother) secret deal.\n\nAgain, we aren’t persuaded—in part because the claim is highly speculative, and in\n\npart because, in order to prove materiality, Baptiste must show that the jury made\n\ncredibility inferences from Louissaint’s comment that don’t necessarily follow.\n\nFor Louissaint’s comment to be material, the jurors would have had to follow\n\n\n 12\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 13 of 28\n\n\nsomething like the following chain of logic: (1) Louissaint testified that he was\n\n“still going to get charged”; (2) therefore [the jurors infer], Louissaint was not a\n\ncooperating witness, and in no way beholden to the government; (3) therefore [the\n\njurors continue], Louissaint’s testimony is more persuasive than it would have\n\nbeen if he weren’t “going to get charged.” The second statement doesn’t\n\nnecessarily follow from the first, nor the third from the second. Absent more\n\ncompelling evidence that Louissaint actually cut a deal, we decline to indulge the\n\nsorts of leaps that Baptiste’s argument requires.\n\n b\n\n The prosecutor told the jury during his closing argument that “everything out\n\nof [Francesse Chery’s] mouth was a complete lie,” that Baptiste “paid [a witness]\n\nto lie on the stand,” and that “the only thing that smells like a fish . . . is the\n\ntestimony of Francesse Chery.” Baptiste argues that these statements\n\nmischaracterized the evidence presented at trial. He also takes issue with the\n\nprosecutor’s summary of his defense as being “duped” by a “dead guy.”\n\n We’ll start with the fish. “[A] prosecutor is justified in arguing during\n\nclosing arguments that a particular witness is lying, if that is an inference\n\nsupported by the evidence at trial.” United States v. Schmitz, 634 F.3d 1247, 1270\n\n(11th Cir. 2011). Anael Chery’s testimony regarding his sister’s truthfulness (or\n\nlack thereof) provided adequate support for the inference that Francesse lied on the\n\n\n 13\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 14 of 28\n\n\nstand. The government’s description of the piscatorial odor emanating from\n\nFrancesse’s testimony was also permissible, if a little cheesy. Cf. Tucker v. Kemp,\n\n762 F.2d 1496, 1507 (11th Cir. 1985) (en banc) (references to defendant as “less\n\nthan human,” “not somebody in our society that we can afford to keep,” and “a\n\ndanger like a time bomb” permitted because “supported by the evidence.”). 2\n\n What about the government’s duped-by-a-dead-guy description of Baptiste’s\n\ndefense theory? While Baptiste undoubtedly dislikes the government’s turn of\n\nphrase, a passage from his own brief suffices to indicate its essential accuracy. To\n\nrecap, in closing, the government’s lawyer said that Baptiste was trying to “make it\n\nlook like some dead guy duped [him].” Baptiste’s brief on appeal summarizes his\n\nposition as follows: “Pagnon masterminded the scheme to dupe the unsuspecting\n\nBaptiste.” Br. of Appellant at 19. Pretty similar.\n\n In sum, although “[l]ittle time and no discussion is necessary to conclude\n\nthat it is improper for a prosecutor to use misstatements and falsehoods,” Davis v.\n\nZant, 36 F.3d 1538, 1548 (11th Cir. 1994), we see neither here.\n\n B\n\n\n\n2\n We’ll cut the government a bit of slack given the context. The government’s “fish” comment\ncame in response to Baptiste’s attorney’s extended analogy at closing comparing Baptiste to a\nstarfish tossed ashore by a hurricane. Baptiste’s attorney concluded, “My client, Junior Jean\nBaptiste, is the starfish,” asking the jury “to fulfill your oath, look in your heart, do what the law\nrequires, apply reasonable doubt and throw [Baptiste] back, throw him back to his family.”\nOoooooookay.\n\n 14\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 15 of 28\n\n\n Now for the feature presentation. Recall Anael Chery testified that his\n\nsister, Francesse, told him that Baptiste was going to give her a Mercedes CLA 45\n\nin exchange for favorable testimony. Responding to Baptiste’s objection, the\n\ndistrict court said that’s “got to be hearsay.” Yep—textbook: “‘Hearsay’ means a\n\nstatement that: (1) the declarant does not make while testifying at the current trial\n\nor hearing; and (2) a party offers in evidence to prove the truth of the matter\n\nasserted in the statement.” Fed. R. Evid. 801(c). Even so, the district court\n\nultimately admitted Anael’s testimony under the statement-against-interest\n\nexception to the rule against hearsay, see Fed. R. Evid. 804(b)(3), because\n\nFrancesse’s underlying statement “could subject [her] to perjury” or other\n\n“charges.”\n\n 1\n\n Perhaps anticipating some rough sledding, the government makes only a\n\ntoken effort to defend the district court’s ruling—the statement-against-interest\n\nexception typically requires that the declarant be “unavailable,” after all, and\n\nFrancesse testified at trial. See Fed. R. Evid. 804(b) (“The following are not\n\nexcluded by the rule against hearsay if the declarant is unavailable as a witness . . .\n\n.”). The government goes on, though, to float three alternative rationales for why\n\nAnael’s testimony was admissible: as the statement of a party opponent under\n\nFederal Rule of Evidence 801(d)(2)(D); as a statement by a co-conspirator under\n\n\n 15\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 16 of 28\n\n\nRule 801(d)(2)(E); and, finally, under Rule 807’s residual exception. For good\n\nmeasure, the government concludes its tour of the Federal Rules of Evidence by\n\nadding that even if the admission of Anael’s testimony was erroneous, this error\n\nwas harmless in light of the weight of the other (undoubtedly admissible) evidence\n\nagainst Baptiste.\n\n We’ll forgo the tour and head straight to the finish line. We conclude that\n\nwe needn’t determine whether the district court erred in admitting Anael’s\n\ntestimony—as a statement against interest, or via any other exclusion or exception\n\nto the hearsay rule—because any error was indeed harmless. “Even where an\n\nabuse of discretion is shown, nonconstitutional evidentiary errors are not grounds\n\nfor reversal absent a reasonable likelihood that the defendant’s substantial rights\n\nwere affected.” United States v. Sellers, 906 F.2d 597, 601 (11th Cir. 1990); Fed.\n\nR. Evid. 103(a); see also United States v. Guzman, 167 F.3d 1350, 1353 (11th Cir.\n\n1999) (quoting Kotteakos v. United States, 328 U.S. 750, 776 (1946), for the\n\nproposition that “a nonconstitutional error requires reversal only if it resulted ‘in\n\nactual prejudice because it had substantial and injurious effect or influence in\n\ndetermining the jury’s verdict’”) (cleaned up).\n\n It’s the government’s burden to demonstrate that an alleged error was\n\nharmless, United States v. Sweat, 555 F.3d 1364, 1367 (11th Cir. 2009), and we\n\nconclude that it has met that burden here. Strike Anael’s testimony and the jury\n\n\n 16\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 17 of 28\n\n\nwould still have had before it—and we’re only hitting the highlights here—\n\nBaptiste’s curiously high income, gobs of documentary evidence showing\n\nfraudulent behavior, and the testimony of Moltimer and Louissaint, including that\n\nBaptiste planned to (and at trial did) “blame it on Marvin.” Under a long line of\n\nprecedent, see United States v. Hornaday, 392 F.3d 1306, 1316 (11th Cir. 2004)\n\n(collecting applications of Kotteakos), such overwhelming evidence of guilt\n\nsuffices to demonstrate that whatever error the district court might have committed\n\ndidn’t have a “substantial and injurious effect.” Kotteakos, 328 U.S. at 776. Nor\n\ncould the alleged error have influenced a reasonable juror’s verdict.\n\n 2\n\n Anael’s testimony relating Francesse’s statement about the Mercedes also\n\nunderlay the district court’s imposition of a sentencing enhancement for\n\nobstruction of justice. Baptiste objects that because Francesse’s statement was\n\ninadmissible hearsay, the district court improperly relied on it in finding\n\nobstruction.\n\n Section 3C1.1 of the Sentencing Guidelines provides for a two-level\n\nenhancement if “the defendant willfully obstructed or impeded, or attempted to\n\nobstruct or impede, the administration of justice with respect to the investigation,\n\nprosecution, or sentencing of the instant offense of conviction.” U.S.S.G. § 3C1.1.\n\n\n\n\n 17\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 18 of 28\n\n\nExamples of covered conduct include “committing, suborning, or attempting to\n\nsuborn perjury.” Id. at cmt. 4(B).\n\n If Anael’s testimony were admissible under one of the exceptions or\n\nexclusions to the rule against hearsay, there wouldn’t be any question that the\n\ndistrict court was on firm footing in relying on it for the enhancement. But if the\n\ntestimony was inadmissible, then our earlier conclusion that its admission at trial\n\nwas harmless wouldn’t (in and of itself) validate the district court’s reliance on it at\n\nsentencing. It all boils down to harm. Even if we strike Anael’s testimony, there\n\nwas still ample evidence supporting Baptiste’s conviction. With regard to that\n\ncomponent of the trial, Anael’s testimony was harmless. But Anael’s testimony\n\nprovided the only evidence supporting Baptiste’s enhancement for obstruction.\n\nWith regard to that component of the trial, Anael’s testimony was necessarily\n\nharmful.\n\n So have we reached the end of the road, such that we finally have to make a\n\ncall on the admissibility of Anael’s testimony? Not quite. Thanks to the “reliable\n\nhearsay” doctrine—which permits sentencing courts to rely on otherwise-\n\ninadmissible hearsay evidence in certain circumstances—we can reserve judgment\n\non the evidence’s admissibility (as we’ve done) while still addressing the district\n\ncourt’s enhancement.\n\n\n\n\n 18\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 19 of 28\n\n\n The doctrine is rooted in § 6A1.3 of the Sentencing Guidelines, which states\n\nthat a court “may consider relevant information without regard to its admissibility\n\nunder the rules of evidence applicable at trial, provided that the information has\n\nsufficient indicia of reliability to support its probable accuracy.” U.S.S.G. §\n\n6A1.3. Indeed, it’s well established that “[i]n determining the relevant facts,\n\nsentencing judges are not restricted to information that would be admissible at\n\ntrial.” Id. at cmt. The reason, the Supreme Court has explained, is that “modern\n\nconcepts individualizing punishment have made it all the more necessary that a\n\nsentencing judge not be denied an opportunity to obtain pertinent information by a\n\nrequirement of rigid adherence to restrictive rules of evidence properly applicable\n\nto the trial.” Williams v. New York, 337 U.S. 241, 247 (1949). Williams’ “modern\n\nconcepts” are old hat today, but no less important for it. Hence the need for a\n\nclear-headed understanding of when—and under what circumstances—hearsay\n\nevidence, in particular, can be deemed sufficiently reliable to support a sentencing\n\ndecision.\n\n The Sentencing Guidelines’ text—just quoted—is crystal clear. It articulates\n\na single precondition to a district court’s consideration of otherwise inadmissible\n\nevidence, including hearsay, as part of its sentencing calculus: The evidence is fair\n\ngame “provided that the information has sufficient indicia of reliability to support\n\nits probable accuracy.” U.S.S.G. § 6A1.3. The trickier question is whether our\n\n\n 19\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 20 of 28\n\n\ncase law has engrafted another, extratextual criterion. In particular, we’ve sent\n\nvaguely mixed signals about whether, as a prerequisite to relying on hearsay\n\nevidence at sentencing, a district court must make express findings on the record\n\nthat the evidence is reliable—or if, instead, as the Guidelines’ language indicates,\n\nit’s enough that the record as a whole provides “sufficient indicia of reliability.”\n\nThe distinction matters here because it’s undisputed that the district court did not\n\nmake any explicit findings about the reliability of Francesse’s testimony.\n\n Some have read our decision in United States v. Lee, 68 F.3d 1267 (11th Cir.\n\n1995), as requiring express findings that the hearsay evidence is reliable. For\n\nreasons we have explained before—and will reiterate here—that is incorrect. The\n\ndefendant in Lee objected that the district court had relied on the hearsay evidence\n\nof one of his co-conspirators—Russell—in determining his sentence. See id. at\n\n1275. We observed, as an initial matter, that “[a]ccording to the Sentencing\n\nGuidelines, and our case law interpreting them, the district court ‘may consider any\n\ninformation, including reliable hearsay, regardless of the information’s\n\nadmissibility at trial, provided that there are sufficient indicia of reliability to\n\nsupport its probable accuracy.’” Id. (quoting United States v. Castellanos, 904\n\nF.2d 1490, 1495 (11th Cir. 1990), and citing U.S.S.G. § 6A1.3). We stated that\n\n“the focus” of the inquiry “is upon the question of [the hearsay evidence’s]\n\nreliability,” and we further explained that reliability “must be determined on a case\n\n\n 20\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 21 of 28\n\n\nby case basis.” Id. The government there contended that Russell’s hearsay\n\nstatements were reliable because they were “against his penal interest,” and we\n\nacknowledged that as a general matter “a statement made against one’s penal\n\ninterest may bolster its credibility.” Id. at 1275–76. Even so, we concluded that\n\n“[i]n Russell’s case,” his penal interest “fail[ed] to qualify as a ‘sufficient\n\nindic[ium] of reliability’” given his “status as a fugitive from justice.” Id. at 1276.\n\nAccordingly, we concluded that “[s]pecific findings on Russell’s credibility [were]\n\nnecessary before the district court c[ould] use this evidence as a basis for\n\ndetermining” the defendant’s sentence, and we remanded for the district court “to\n\nmake further findings as to the reliability of Russell’s hearsay statement.” Id.\n\n Our subsequent decision in United States v. Gordon, 231 F.3d 750 (11th Cir.\n\n2000), clarified Lee’s holding and reach. The defendant there cited Lee in support\n\nof his contention “that the district court was required to make specific findings\n\nregarding the reliability” of hearsay testimony on which the government sought to\n\nrely at sentencing. Id. at 760. In response, we explained that Lee was never\n\nintended to—and didn’t—prescribe a per se rule that findings are always\n\nnecessary. Id. at 760–61. Rather, we made clear, Lee’s holding—that findings\n\nwere required there—was a product of the “circumstances” that underlay that\n\nparticular case. Id. Specifically, we said, it was because the traditional against-\n\npenal-interest justification didn’t cut it, and because the hearsay declarant was a\n\n\n 21\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 22 of 28\n\n\nfugitive, that the Lee court found “the absence of findings … particularly glaring\n\nand troubling.” Id. By contrast, we emphasized in Gordon, where the record and\n\nthe circumstances of the case “demonstrate adequate indicia of reliability,”\n\nfindings are not strictly necessary. Id. at 761. In such a case, a “district court’s\n\nfailure to make separate findings regarding the reliability of [hearsay] statements\n\n[is] not error.” Id. Put differently, “the absence of such findings does not\n\nnecessarily require reversal or remand where the reliability of the statements is\n\napparent from the record.” Id.3\n\n So … is this a “Lee” case—in which the traditional indicia of reliability fall\n\nshort and findings are thus required—or a “Gordon” case—in which the traditional\n\nindicia provide the necessary comfort? Clearly the latter, we think. For one,\n\n\n\n\n3\n One loose end: After Lee but before Gordon, in United States v. Anderton we addressed a\nsimilar complaint that a district court had erroneously based its sentencing decision on\n“unreliable hearsay.” 136 F.3d 747, 751 (11th Cir. 1998). In the course of rejecting that\ncontention, we recited that “a court may rely on hearsay at sentencing, as long as the evidence\nhas sufficient indicia of reliability, the court makes explicit findings of fact as to credibility, and\nthe defendant has an opportunity to rebut the evidence.” Id. Seemingly in support of the\n“explicit findings” aspect of that statement—albeit without saying so expressly—we cited our\nearlier decision in Lee. Id. Anderton, though, doesn’t stand for the sort of per se rule that the\nGordon defendant sought and that the Gordon panel rejected. The parties in Anderton never\ndisputed whether “explicit findings” were required, and nothing rode on that issue. Id. Our\ndecision simply assumed (at least in the circumstances presented) that they were, observed that\nthe district court had made them, and held that they weren’t clearly erroneous. Id. The same is\ntrue of our post-Gordon decisions in United States v. Zlatogur, 271 F.3d 1025, 1031 (11th Cir.\n2001), and United States v. Patti, 337 F.3d 1317, 1326 (11th Cir. 2003), which merely recite\nAnderton’s boilerplate. And to the extent that one could read Anderton as contradicting Lee, and\nas holding that express findings of credibility are always required, we are bound by our prior-\nprecedent rule to apply our earlier decision in Lee. See United States v. Steele, 147 F.3d 1316,\n1318 (11th Cir. 1998) (en banc).\n 22\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 23 of 28\n\n\nFrancesse’s statement—that Baptiste had promised her a Mercedes in exchange for\n\nfavorable testimony—was clearly against both her and Baptiste’s penal interests,4\n\nand there’s no “fugitive from justice”-like reason here, as there was in Lee, to\n\ndiscount the force of that established gauge of reliability. For another, Francesse’s\n\nstatement aligned with Louissaint’s testimony regarding how Baptiste planned to\n\n“blame everything” on Pagnon. And for yet another, her statement specifically\n\nidentified the make and model of car she was set to receive. All of these are\n\n“indicia of reliability,” and by our lights, collectively clear the “sufficien[cy]”\n\nthreshold. See Lee, 68 F.3d at 1276.\n\n Because this isn’t the sort of case in which explicit findings were required,\n\nthe district court didn’t err in not making any. Nor did it err in relying on\n\nFrancesse’s sufficiently-reliable testimony to impose the obstruction-of-justice\n\nenhancement.\n\n C\n\n That leaves a few sentencing-related loose ends.\n\n 1\n\n As an initial matter, Baptiste challenges the district court’s imposition of\n\nsentencing enhancements for the amount of money that the government lost and\n\n\n4\n Even if, as we have said, it might (?) not formally qualify for the statement-against-interest\nexception to the hearsay rule because Francesse testified at trial and thus might (?) not have been\n“unavailable,” as Rule 804(b)(3) requires. See supra at 16–17.\n\n 23\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 24 of 28\n\n\nthe number and vulnerability of the scheme’s victims. U.S.S.G. § 2B1.1(b)(1)(K)\n\n(amount lost exceeding $9,500,000); § 2B1.1(b)(2)(A) (offense involving more\n\nthan 10 victims); § 3A1.1(b)(1)–(2) (vulnerable victims).\n\n a\n\n As for the amount of money involved, we note first that Baptiste expressly\n\nwithdrew all objections except those relating to the enhancements for obstruction\n\nof justice and the number and vulnerability of his victims. We needn’t determine\n\nwhether in doing so Baptiste invited any alleged error, because even under the\n\ntypical clear-error standard, see United States v. Castaneda-Pozo, 877 F.3d 1249,\n\n1251 (11th Cir. 2017), Baptiste’s argument fails. The government submitted\n\nevidence indicating that Surveillance Master and its associated bank accounts were\n\nall under Baptiste’s name. The identification documents reviewed by Customs and\n\nImmigration Services also demonstrated that it was reasonable for the district court\n\nto conclude that virtually all of Surveillance Master’s check-cashing business was\n\nfraudulent—629 of the 630 IDs reviewed were fakes. Finally, “[t]he district court\n\nis only required to make a reasonable estimate of the loss.” United States v. Cobb,\n\n842 F.3d 1213, 1218 (11th Cir. 2016); U.S.S.G. § 2B1.1, cmt. 3(C); see also\n\nUnited States v. Medina, 485 F.3d 1291, 1304 (11th Cir. 2007) (indicating that a\n\nreasonable estimate is particularly appropriate in fraud cases because the amount\n\nlost is often difficult to determine with precision). We thus have evidence\n\n\n 24\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 25 of 28\n\n\nindicating that the sums lost went to Baptiste, that virtually all of the money that\n\npassed through the check-cashing business was linked to the scheme, as well as\n\nprecedent underscoring that the district court could apply this enhancement even if\n\nit couldn’t determine a precise figure. Taken together, we don’t see error here,\n\nmuch less of the “clear” variety.\n\n b\n\n Baptiste’s challenges to the enhancements based on the vulnerability and\n\nnumber of the scheme’s victims travel together. The Sentencing Guidelines\n\nprovide for a two-level increase on the ground that he “knew or should have\n\nknown that a victim of the offense was a vulnerable victim.” U.S.S.G. §\n\n3A1.1(b)(1). We have held that “both circumstances and immutable characteristics\n\ncan render a victim vulnerable,” United States v. Bradley, 644 F.3d 1213, 1288\n\n(11th Cir. 2011), and that “[n]either bodily injury nor financial loss is required” for\n\nthe enhancement, United States v. Moran, 778 F.3d 942, 978 (11th Cir. 2015). Our\n\ndecision in United States v. Pierre, 825 F.3d 1183 (11th Cir. 2016), settles the\n\nmatter with regard to Baptiste’s use of prisoners’ identities. We explained there\n\nthat “inmates have unique circumstances and immutable characteristics that make\n\nthem . . . vulnerable to . . . fraudulent activity.” Id. at 1196; see also id. (“Inmates\n\nusually do not file tax returns during periods of incarceration, and they are less\n\nlikely to discover that their identities have been compromised.”). The government\n\n\n 25\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 26 of 28\n\n\npresented evidence showing that Baptiste specifically requested the identities of\n\nprisoners, so the vulnerability-of-victims enhancement checks out.5 So too the\n\nnumber-of-victims enhancement; the district could reasonably infer from the\n\nevidence detailing the number of checks that Baptiste cashed that he had targeted a\n\n“large number” of people. See U.S.S.G. § 3A1.1(b)(2) (stipulating that if (i) §\n\n3A1.1(b)(1) applies and (ii) “the offense involved a large number of vulnerable\n\nvictims,” the defendant’s offense level should be raised an additional two levels.).\n\n 2\n\n Finally—and simply for us—the parties agree that the district court plainly\n\nerred in failing to “address the defendant personally in order to permit the\n\ndefendant to speak” at sentencing, as Federal Rule of Criminal Procedure\n\n32(i)(4)(A)(ii) expressly requires. 6 In response to the district court’s query, “Does\n\nthe defendant wish to address me?”, Baptiste’s attorney responded on his client’s\n\nbehalf: “Mr. Baptiste respectfully has nothing to add.” However formalistic it may\n\nseem, and even reviewing for plain error, we agree with the parties that this\n\n5\n Because the record supports the district court’s finding that Baptiste targeted incarcerated\nvictims, we can affirm the enhancement for vulnerable victims on that ground alone. We needn’t\ndetermine whether the deceased count as “vulnerable” as that term is used in § 3A1.1(b).\nCompare United States v. Roberson, 872 F.2d 597, 609 (5th Cir. 1989) (sentence of a defendant\nconvicted of fraud properly enhanced where the defendant used a dead companion’s credit card\nafter burning his corpse), with United States v. Shumway, 112 F.3d 1413, 1424 (10th Cir. 1997)\n(refusing to find that skeletal remains are a “vulnerable victim”).\n6\n “Before imposing sentence, the court must . . . (ii) address the defendant personally in order to\npermit the defendant to speak or present any information to mitigate the sentence . . . .” Fed. R.\nCrim. P. 32(i)(4)(A)(ii).\n\n 26\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 27 of 28\n\n\nexchange—between court and counsel rather than court and defendant—amounted\n\nto “(1) error, (2) that is plain and (3) that affect[ed] [Baptiste’s] substantial rights,”\n\nand that the error “[4] seriously affect[ed] the fairness, integrity, or public\n\nreputation of judicial proceedings.” United States v. Turner, 474 F.3d 1265, 1276\n\n(11th Cir. 2007); United States v. Doyle, 857 F.3d 1115, 1120 (11th Cir. 2017)\n\n(“[T]he presumption of prejudice in denial of allocution cases is still the rule when\n\na defendant is not sentenced at the bottom of his guidelines range . . . .”); United\n\nStates v. Prouty, 303 F.3d 1249, 1251 n.1 (11th Cir. 2002) (rejecting “the\n\nGovernment’s argument that directing comments to the lawyers suffices to comply\n\nwith the rule”). On remand, Baptiste should be afforded the opportunity to address\n\nthe court directly.\n\n III\n\n Accordingly, we hold as follows:\n\n (A) The district court did not abuse its discretion in refusing to allow\n\nBaptiste to introduce evidence that Marvin Pagnon had duped others into\n\nparticipating in similar schemes. Nor did the government fail to disclose any\n\nmaterial information concerning its witnesses or mischaracterize or exaggerate the\n\nevidence during its closing argument.\n\n (B) Any error that the district court might have committed in admitting\n\nAnael Chery’s testimony relaying his sister Francesse’s statement that Baptiste had\n\n\n 27\n\f Case: 16-17175 Date Filed: 08/28/2019 Page: 28 of 28\n\n\npromised her a car in exchange for favorable testimony was harmless. And the\n\ndistrict court did not err in relying on the same evidence to support an obstruction-\n\nrelated sentencing enhancement because the record reveals that the evidence bore\n\n“sufficient indicia of reliability.”\n\n (C) The district court did not err in imposing sentencing enhancements\n\npertaining to the amount of money that the government lost and the number and\n\nvulnerability of Baptiste’s scheme’s victims. The district court did reversibly err,\n\nhowever, when it failed to permit Baptiste to allocute personally.\n\n AFFIRMED IN PART AND REMANDED.\n\n\n\n\n 28\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: UNITED STATES of America, Plaintiff - Appellee, v. Junior Jean BAPTISTE, Defendant - Appellant.\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-28\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 16-17175 Date Filed: 08/28/2019 Page: 1 of 28\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n Nos. 16-17175, 16-17595\\n ________________________\\n\\n\\n D.C. Docket No. 1:15-cr-20777-JEM-1\\n\\n\\nUNITED STATES OF AMERICA,\\n\\n Plaintiff - Appellee,\\n versus\\n\\nJUNIOR JEAN BAPTISTE,\\n\\n Defendant - Appellant.\\n\\n\\n ________________________\\n\\n Appeals from the United States District Court\\n for the Southern District of Florida\\n ________________________\\n\\n (August 28, 2019)\\n\\n\\n\\nBefore WILLIAM PRYOR, NEWSOM, and BRANCH, Circuit Judges.\\n\\nNEWSOM, Circuit Judge:\\n\\n Jean Junior Baptiste was a businessman. In addition to a portfolio of media\\n\\nand entertainment companies, Baptiste owned and operated Surveillance Master\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 2 of 28\\n\\n\\nCCTV. Surveillance Master installed security cameras, repaired computers, and—\\n\\none of these things is not like the others—cashed checks. In 2016, a jury found\\n\\nthat Baptiste had taken part in an $11 million scheme that involved cashing tax-\\n\\nrefund checks that he had fraudulently obtained in the names of inmates, minors,\\n\\ndead people, and other insentient or otherwise unsuspecting “clients.” Baptiste\\n\\nnow appeals his conviction and 212-month sentence.\\n\\n Although Baptiste raises a number of issues on appeal, we focus primarily\\n\\non two questions related to the hearsay testimony of a government witness. The\\n\\nabridged version of the story: Francesse Chery was one of Baptiste’s key\\n\\nwitnesses. The government countered with her brother, Anael Chery, who testified\\n\\n(among other things) that Francesse had told him that, in exchange for her (false)\\n\\ntestimony supporting Baptiste’s narrative, Baptiste would give her a Mercedes.\\n\\nBaptiste argues that Anael’s testimony was inadmissible hearsay and that the\\n\\ndistrict court’s error in allowing the jury to hear it tainted both his conviction and\\n\\nhis sentence.\\n\\n Baptiste’s challenge presents two questions. First, was Anael’s testimony\\n\\nindeed inadmissible hearsay? The district court admitted the testimony pursuant to\\n\\nthe statement-against-interest exception to the general prohibition on hearsay\\n\\nevidence, and on appeal the government has offered a smattering of additional\\n\\ntheories of admissibility. We conclude that we needn’t decide whether Anael’s\\n\\n\\n 2\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 3 of 28\\n\\n\\ntestimony was inadmissible hearsay because even if the district court did err in\\n\\nallowing it, the error was harmless. There was more than enough compelling—and\\n\\nundoubtedly admissible—evidence to support Baptiste’s conviction.\\n\\n Second, and (sort of) relatedly, did the district court err in relying on Anael’s\\n\\ntestimony when it imposed a sentencing enhancement for obstructing justice? If\\n\\nyou’re saying, “Didn’t they just say they weren’t going to decide whether the\\n\\ntestimony was admissible?”—we hear you. As it turns out, though, thanks to a\\n\\ndoctrine called (somewhat oxymoronically) “reliable hearsay” we can answer the\\n\\nsecond question without deciding the first. Under the reliable-hearsay doctrine, so\\n\\nlong as certain preconditions are met, a sentencing court can rely on evidence that\\n\\nwould be off-limits in the guilt phase. For Baptiste, this means that even if Anael’s\\n\\ndescription of his sister’s supposed deal was inadmissible hearsay (and we aren’t\\n\\nsaying either way) the district court might not have erred in relying on that\\n\\ntestimony for the obstruction enhancement—again, so long as the preconditions are\\n\\nmet.\\n\\n So, what are they? Well, our case law has arguably sent mixed signals about\\n\\nthat. There is, though, a synthesis. We hold (and clarify) today that the Sentencing\\n\\nGuidelines permit use of hearsay testimony so long as the overall record provides\\n\\n“sufficient indicia of reliability”—and we conclude that the indicia of reliability\\n\\nhere are sufficient.\\n\\n\\n 3\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 4 of 28\\n\\n\\n Accordingly—and because we find that none of Baptiste’s other arguments\\n\\nhas merit—we affirm Baptiste’s conviction and, with one asterisk, his sentence.\\n\\nThe asterisk? Pursuant to the parties’ joint request, we remand the case to the\\n\\ndistrict court for the limited purpose of allowing Baptiste to allocute, a right that he\\n\\nwas denied at sentencing.\\n\\n I\\n\\n During Baptiste’s trial, two cooperating witnesses, Andy Louissaint and Karl\\n\\nMoltimer, described the tax-refund scheme. Baptiste would first provide tax\\n\\npreparers such as Louissaint and Moltimer with a list of individuals—many of\\n\\nthem young, incarcerated, dead, or otherwise unlikely to discover the use of their\\n\\nidentities—along with accompanying personal information such as birthdates and\\n\\nsocial security numbers. The tax preparers used that information to (fraudulently)\\n\\nobtain refund checks from the IRS, which they then brought to Baptiste’s\\n\\nSurveillance Master for cashing.\\n\\n Baptiste contracted with a third party to process the checks. He would\\n\\nfabricate the necessary documents—work permits, passports, green cards, driver’s\\n\\nlicenses—to avoid raising red flags. Baptiste and the tax preparers would divide\\n\\nthe proceeds, with Baptiste often taking a cut of between 25% and 50%. Not bad,\\n\\ngiven that typical check-cashing fees range from 1.5% to 3%.\\n\\n\\n\\n\\n 4\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 5 of 28\\n\\n\\n When the IRS reviewed Surveillance Master’s records in conjunction with a\\n\\nroutine inspection, an agent noticed that several refund checks were for identical\\n\\n(or nearly identical) amounts. During a subsequent investigation, the IRS obtained\\n\\ncopies of identification documents, Treasury checks, and lists containing personal\\n\\ninformation for both the quick and the dead. Customs and Immigration Services\\n\\nconducted an evaluation of approximately 630 permanent-resident and\\n\\nemployment-authorization cards that Baptiste had used to verify the checks; all but\\n\\none—belonging to Baptiste’s wife—were fraudulent.\\n\\n The ensuing indictment charged Baptiste with money laundering, conspiracy\\n\\nto commit money laundering, possession of false identification documents with\\n\\nunlawful intent, theft of government money, and aggravated identity theft. At trial,\\n\\nBaptiste’s story was that he was an unwitting participant in a broader scheme\\n\\nconcocted by a now-dead business partner, Marvin Pagnon. Pagnon’s ex-\\n\\ngirlfriend, the aforementioned Francesse Chery, echoed Baptiste’s account in her\\n\\ntestimony and explained how Pagnon operated. In particular, she explained that\\n\\nbecause Pagnon had poor credit, he would ask would-be associates such as\\n\\nBaptiste to partner with him. Pagnon, she said, laid out a collaboration that was\\n\\nstraightforward and above board: Pagnon would obtain (real) refund checks from\\n\\ntax preparers; Baptiste would cash whatever Pagnon handed him. Baptiste signed\\n\\nup, and he seems to have been well compensated for his efforts. During the\\n\\n\\n 5\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 6 of 28\\n\\n\\nscheme, Baptiste purchased a Mercedes Benz CL 63, a Mercedes Benz E Class, a\\n\\nRange Rover, and—again, one of these things is not like the others—a $345,000\\n\\nshare in a 200-foot long cargo ship.\\n\\n The government painted a different picture. Moltimer and Louissaint both\\n\\ntestified, for instance, that Pagnon played no role in Baptiste’s check-cashing\\n\\noperation. Louissaint added that when Baptiste first became aware that an\\n\\nindictment could be coming down the pike, he had told Louissaint that he planned\\n\\nto “basically blame everything on Marvin [Pagnon].” The government also offered\\n\\ntestimony from Francesse Chery’s brother, Anael. Anael first opined, as a general\\n\\nmatter, that his sister just wasn’t a very truthful person. He went on to testify,\\n\\nmuch more specifically, that Francesse had told him that Baptiste was going to\\n\\ngive her a Mercedes CLA 45 so long as she “h[e]ld up her end of the bargain” by\\n\\nbacking Baptiste’s it-was-all-Pagnon’s-idea narrative. Baptiste objected that\\n\\nAnael’s testimony—relaying what his sister had allegedly told him—was\\n\\ninadmissible hearsay. The district court ultimately admitted Anael’s statement,\\n\\nciting the statement-against-interest exception to the general prohibition on hearsay\\n\\nevidence. See Fed. R. Evid. 804(b)(3).\\n\\n The jury convicted Baptiste on all counts. At sentencing, the district court\\n\\napplied enhancements for the amount of money that Baptiste had fraudulently\\n\\nobtained from the government (exceeding $11 million), for the number of victims\\n\\n\\n 6\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 7 of 28\\n\\n\\n(perhaps hundreds), for the victims’ vulnerable status (many incapacitated in some\\n\\nform or fashion), and for obstructing justice (for the perjury-for-a-Mercedes deal\\n\\nwith Francesse). The district court sentenced Baptiste to 212 months’\\n\\nimprisonment and restitution of more than $11 million. 1 This is Baptiste’s appeal.\\n\\n II\\n\\n As already noted, our focus here will be the related contentions that the\\n\\ndistrict court (1) abused its discretion by admitting Anael’s hearsay testimony as\\n\\nevidence against Baptiste and (2) violated the Sentencing Guidelines by imposing\\n\\nan obstruction-based enhancement that relied on the same. Before getting there,\\n\\nthough, we’ll address a few other guilt-phase matters—namely, whether the district\\n\\ncourt erroneously refused to admit Rule 404(b) evidence, whether the government\\n\\nfailed to reveal (alleged) deals with its witnesses, and whether the prosecutor made\\n\\nimproper statements to the jury during his closing argument. Then, after dealing\\n\\nwith the Anael-related issues, which span the trial and sentencing phases, we’ll\\n\\nconclude with a few additional, sentencing-specific arguments—namely, that the\\n\\ndistrict court erred in increasing Baptiste’s sentence due to the amount of money\\n\\nthat the government lost as well as the number and vulnerability of the scheme’s\\n\\nvictims, as well as in declining to permit Baptiste to allocute personally.\\n\\n A\\n\\n1\\n The court ordered that $70,000 of that total be paid jointly and severally with Moltimer.\\n\\n 7\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 8 of 28\\n\\n\\n 1\\n\\n During the trial, the district court refused to allow Baptiste to put on\\n\\nevidence that Marvin Pagnon had duped others into participating in similar\\n\\nschemes. On appeal, Baptiste contends that, in so doing, the district court\\n\\nmisapplied Federal Rule of Evidence 404(b), which states, in relevant part—\\n\\n (1) Prohibited Uses. Evidence of a crime, wrong, or other act is not\\n admissible to prove a person’s character in order to show that on a\\n particular occasion the person acted in accordance with the character.\\n (2) Permitted Uses; Notice in a Criminal Case. This evidence may\\n be admissible for another purpose, such as proving motive,\\n opportunity, intent, preparation, plan, knowledge, identity, absence of\\n mistake, or lack of accident.\\n\\nFed. R. Evid. 404(b). In particular, Baptiste argues that his evidence would have\\n\\ndemonstrated only Pagnon’s capacity to implement his scheme—he had done so\\n\\nbefore, and so could do so again—rather than Pagnon’s character or propensity\\n\\nmore generally, and was therefore admissible as “reverse” 404(b) evidence—i.e.,\\n\\nbad-acts evidence that “tends to negate the defendant’s guilt of the crime charged\\n\\nagainst him.” United States v. Seals, 419 F.3d 600, 606 (7th Cir. 2005); see also,\\n\\ne.g., United States v. Cohen, 888 F.2d 770, 776 (11th Cir. 1989) (admitting\\n\\nevidence showing that the government’s witness “was capable of concocting and\\n\\nmanaging [a] fraudulent scheme”).\\n\\n The government doesn’t really dispute Baptiste’s premise—that the\\n\\nevidence bore on whether Pagnon was competent of pulling off such a scheme—\\n\\n\\n 8\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 9 of 28\\n\\n\\nand, indeed, it quotes Baptiste’s own brief back at him to make its point. In his\\n\\nbrief, the government says, Baptiste admits that “[t]he purpose of the proffered\\n\\nevidence was to show that Pagnon was capable of implementing his fraudulent tax-\\n\\nrefund scheme without the knowledge and involvement of the nominal owners.”\\n\\nBr. of Appellant at 41. In so stating, the government contends, Baptiste has\\n\\neffectively conceded that the evidence was of the “pure propensity” variety,\\n\\nprobative only of Pagnon’s general mendacity.\\n\\n The parties thus read the same bit of testimony in two very different ways.\\n\\nAnd not implausibly so. The evidence demonstrating that Pagnon had previously\\n\\ndeceived others could be construed as indicating either Pagnon’s capacity to do so\\n\\nagain and/or his propensity to do so again—either “Pagnon is well-equipped to lie\\n\\nin these circumstances” or “Pagnon is a liar.” The two interpretations are\\n\\ninterrelated—mutually reinforcing, you might say—but only the former is\\n\\ngenerally admissible under Rule 404(b).\\n\\n The trouble for Baptiste is that even scales won’t get the job done.\\n\\n“Determinations of admissibility of evidence rest largely within the discretion of\\n\\nthe trial judge and will not be disturbed on appeal absent a clear showing of an\\n\\nabuse of discretion.” United States v. Russell, 703 F.2d 1243, 1249 (11th Cir.\\n\\n1983). That this evidence can reasonably be read as (permissible) capacity\\n\\n\\n\\n\\n 9\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 10 of 28\\n\\n\\nevidence or (impermissible) character evidence demonstrates that Baptiste hasn’t\\n\\nmade the required “clear showing.”\\n\\n 2\\n\\n Baptiste also objects to how the government prosecuted his case. He makes\\n\\ntwo claims: first, that the government failed to disclose material issues involving\\n\\nkey witnesses; and second, that the prosecutor’s closing argument mischaracterized\\n\\nand exaggerated the evidence.\\n\\n a\\n\\n Baptiste alleges three disclosure violations. First, he argues that the\\n\\ngovernment failed to correct false testimony when one of its key witnesses,\\n\\nMoltimer, stated that he had derivative U.S. citizenship. Baptiste contends that, in\\n\\nfact, Moltimer was granted citizenship in exchange for testifying against Baptiste.\\n\\nSecond, Baptiste claims that Moltimer lied—and the government failed to\\n\\ncorrect—when he testified that he only met Pagnon in 2013. Pagnon died in 2011,\\n\\nand according to Baptiste, the government’s failure to set the record straight\\n\\nundermined Baptiste’s defense that “Pagnon masterminded the scheme to dupe the\\n\\nunsuspecting Baptiste.” Br. of Appellant at 19. Finally, Baptiste asserts that the\\n\\ngovernment failed to disclose a cooperation agreement with Louissaint, preventing\\n\\nthe jury from taking the true measure of Louissaint’s incentives.\\n\\n\\n\\n\\n 10\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 11 of 28\\n\\n\\n All three claims fall short. Baptiste’s first argument—about the\\n\\ngovernment’s alleged citizenship deal with Moltimer—rests on misleadingly\\n\\nselective quotations of the record. Baptiste explains that Moltimer’s attorney\\n\\nindicated that Moltimer had been “granted American citizenship,” encouraging us\\n\\nto infer that this “grant[]” was in exchange for Moltimer’s testimony against\\n\\nBaptiste. But just before the snippet that Baptiste highlights, Moltimer’s attorney\\n\\nstated that Moltimer “has been granted American citizenship because of his being\\n\\nbrought here when he was very young” and that Moltimer’s “mother was an\\n\\nAmerican citizen.” (emphasis added). And just after the passage excerpted by\\n\\nBaptiste, Moltimer’s attorney continued, “ICE came to see him and went forward,\\n\\nand he has been granted American citizenship, which he should have had from the\\n\\nstart.” (emphasis added). Come on. Baptiste’s assertion that the government\\n\\nfailed to disclose a citizenship-for-testimony deal with Moltimer is worse than\\n\\nconjectural—it appears to be flatly contradicted by the record.\\n\\n As for Moltimer’s testimony that he first met Pagnon three years after\\n\\nPagnon’s death, Baptiste is of course correct that—barring either time travel or\\n\\nnecromancy—Moltimer was wrong. But the focus for our purposes is not the\\n\\nshakiness of Moltimer’s chronology but rather the government’s failure to correct\\n\\nit. See Giglio v. United States, 405 U.S. 150, 153 (1972). To prove that the\\n\\ngovernment’s inaction violated his rights, Baptiste must show not only that the\\n\\n\\n 11\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 12 of 28\\n\\n\\ngovernment “[1] knowingly used perjured testimony, or failed to correct what [it]\\n\\nsubsequently learned was false testimony, and [2] that the falsehood was material,”\\n\\nbut also that the testimony was “[3] given with the willful intent to provide false\\n\\ntestimony and not as a result of a mistake, confusion, or faulty memory.” United\\n\\nStates v. Horner, 853 F.3d 1201, 1206 (11th Cir. 2017) (quoting another source),\\n\\ncert. denied, 138 S. Ct. 674 (2018). Even setting aside that the jury was repeatedly\\n\\nreminded—including by Baptiste’s attorney—that Pagnon was long dead by 2014,\\n\\nand that the falsehood was thus unlikely to have been material, Moltimer’s\\n\\ntestimony clearly indicates “mistake, confusion, or faulty memory.” See id. He\\n\\nsaid, for instance: “I’ve been in prison for almost two years. I think met Marvin\\n\\n[Pagnon] in 2014. I don’t remember the exact date . . . .” That testimony, we\\n\\nthink, hardly suggests a “willful intent” to deceive on the government’s part.\\n\\n Finally, Baptiste alleges that the government failed to disclose a cooperation\\n\\nagreement with Louissaint. Baptiste seizes on Louissaint’s testimony that he was\\n\\n“still going to get charged.” Baptiste points out that, in fact, Louissaint was never\\n\\ncharged, and he asks us to infer that the government cut a(nother) secret deal.\\n\\nAgain, we aren’t persuaded—in part because the claim is highly speculative, and in\\n\\npart because, in order to prove materiality, Baptiste must show that the jury made\\n\\ncredibility inferences from Louissaint’s comment that don’t necessarily follow.\\n\\nFor Louissaint’s comment to be material, the jurors would have had to follow\\n\\n\\n 12\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 13 of 28\\n\\n\\nsomething like the following chain of logic: (1) Louissaint testified that he was\\n\\n“still going to get charged”; (2) therefore [the jurors infer], Louissaint was not a\\n\\ncooperating witness, and in no way beholden to the government; (3) therefore [the\\n\\njurors continue], Louissaint’s testimony is more persuasive than it would have\\n\\nbeen if he weren’t “going to get charged.” The second statement doesn’t\\n\\nnecessarily follow from the first, nor the third from the second. Absent more\\n\\ncompelling evidence that Louissaint actually cut a deal, we decline to indulge the\\n\\nsorts of leaps that Baptiste’s argument requires.\\n\\n b\\n\\n The prosecutor told the jury during his closing argument that “everything out\\n\\nof [Francesse Chery’s] mouth was a complete lie,” that Baptiste “paid [a witness]\\n\\nto lie on the stand,” and that “the only thing that smells like a fish . . . is the\\n\\ntestimony of Francesse Chery.” Baptiste argues that these statements\\n\\nmischaracterized the evidence presented at trial. He also takes issue with the\\n\\nprosecutor’s summary of his defense as being “duped” by a “dead guy.”\\n\\n We’ll start with the fish. “[A] prosecutor is justified in arguing during\\n\\nclosing arguments that a particular witness is lying, if that is an inference\\n\\nsupported by the evidence at trial.” United States v. Schmitz, 634 F.3d 1247, 1270\\n\\n(11th Cir. 2011). Anael Chery’s testimony regarding his sister’s truthfulness (or\\n\\nlack thereof) provided adequate support for the inference that Francesse lied on the\\n\\n\\n 13\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 14 of 28\\n\\n\\nstand. The government’s description of the piscatorial odor emanating from\\n\\nFrancesse’s testimony was also permissible, if a little cheesy. Cf. Tucker v. Kemp,\\n\\n762 F.2d 1496, 1507 (11th Cir. 1985) (en banc) (references to defendant as “less\\n\\nthan human,” “not somebody in our society that we can afford to keep,” and “a\\n\\ndanger like a time bomb” permitted because “supported by the evidence.”). 2\\n\\n What about the government’s duped-by-a-dead-guy description of Baptiste’s\\n\\ndefense theory? While Baptiste undoubtedly dislikes the government’s turn of\\n\\nphrase, a passage from his own brief suffices to indicate its essential accuracy. To\\n\\nrecap, in closing, the government’s lawyer said that Baptiste was trying to “make it\\n\\nlook like some dead guy duped [him].” Baptiste’s brief on appeal summarizes his\\n\\nposition as follows: “Pagnon masterminded the scheme to dupe the unsuspecting\\n\\nBaptiste.” Br. of Appellant at 19. Pretty similar.\\n\\n In sum, although “[l]ittle time and no discussion is necessary to conclude\\n\\nthat it is improper for a prosecutor to use misstatements and falsehoods,” Davis v.\\n\\nZant, 36 F.3d 1538, 1548 (11th Cir. 1994), we see neither here.\\n\\n B\\n\\n\\n\\n2\\n We’ll cut the government a bit of slack given the context. The government’s “fish” comment\\ncame in response to Baptiste’s attorney’s extended analogy at closing comparing Baptiste to a\\nstarfish tossed ashore by a hurricane. Baptiste’s attorney concluded, “My client, Junior Jean\\nBaptiste, is the starfish,” asking the jury “to fulfill your oath, look in your heart, do what the law\\nrequires, apply reasonable doubt and throw [Baptiste] back, throw him back to his family.”\\nOoooooookay.\\n\\n 14\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 15 of 28\\n\\n\\n Now for the feature presentation. Recall Anael Chery testified that his\\n\\nsister, Francesse, told him that Baptiste was going to give her a Mercedes CLA 45\\n\\nin exchange for favorable testimony. Responding to Baptiste’s objection, the\\n\\ndistrict court said that’s “got to be hearsay.” Yep—textbook: “‘Hearsay’ means a\\n\\nstatement that: (1) the declarant does not make while testifying at the current trial\\n\\nor hearing; and (2) a party offers in evidence to prove the truth of the matter\\n\\nasserted in the statement.” Fed. R. Evid. 801(c). Even so, the district court\\n\\nultimately admitted Anael’s testimony under the statement-against-interest\\n\\nexception to the rule against hearsay, see Fed. R. Evid. 804(b)(3), because\\n\\nFrancesse’s underlying statement “could subject [her] to perjury” or other\\n\\n“charges.”\\n\\n 1\\n\\n Perhaps anticipating some rough sledding, the government makes only a\\n\\ntoken effort to defend the district court’s ruling—the statement-against-interest\\n\\nexception typically requires that the declarant be “unavailable,” after all, and\\n\\nFrancesse testified at trial. See Fed. R. Evid. 804(b) (“The following are not\\n\\nexcluded by the rule against hearsay if the declarant is unavailable as a witness . . .\\n\\n.”). The government goes on, though, to float three alternative rationales for why\\n\\nAnael’s testimony was admissible: as the statement of a party opponent under\\n\\nFederal Rule of Evidence 801(d)(2)(D); as a statement by a co-conspirator under\\n\\n\\n 15\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 16 of 28\\n\\n\\nRule 801(d)(2)(E); and, finally, under Rule 807’s residual exception. For good\\n\\nmeasure, the government concludes its tour of the Federal Rules of Evidence by\\n\\nadding that even if the admission of Anael’s testimony was erroneous, this error\\n\\nwas harmless in light of the weight of the other (undoubtedly admissible) evidence\\n\\nagainst Baptiste.\\n\\n We’ll forgo the tour and head straight to the finish line. We conclude that\\n\\nwe needn’t determine whether the district court erred in admitting Anael’s\\n\\ntestimony—as a statement against interest, or via any other exclusion or exception\\n\\nto the hearsay rule—because any error was indeed harmless. “Even where an\\n\\nabuse of discretion is shown, nonconstitutional evidentiary errors are not grounds\\n\\nfor reversal absent a reasonable likelihood that the defendant’s substantial rights\\n\\nwere affected.” United States v. Sellers, 906 F.2d 597, 601 (11th Cir. 1990); Fed.\\n\\nR. Evid. 103(a); see also United States v. Guzman, 167 F.3d 1350, 1353 (11th Cir.\\n\\n1999) (quoting Kotteakos v. United States, 328 U.S. 750, 776 (1946), for the\\n\\nproposition that “a nonconstitutional error requires reversal only if it resulted ‘in\\n\\nactual prejudice because it had substantial and injurious effect or influence in\\n\\ndetermining the jury’s verdict’”) (cleaned up).\\n\\n It’s the government’s burden to demonstrate that an alleged error was\\n\\nharmless, United States v. Sweat, 555 F.3d 1364, 1367 (11th Cir. 2009), and we\\n\\nconclude that it has met that burden here. Strike Anael’s testimony and the jury\\n\\n\\n 16\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 17 of 28\\n\\n\\nwould still have had before it—and we’re only hitting the highlights here—\\n\\nBaptiste’s curiously high income, gobs of documentary evidence showing\\n\\nfraudulent behavior, and the testimony of Moltimer and Louissaint, including that\\n\\nBaptiste planned to (and at trial did) “blame it on Marvin.” Under a long line of\\n\\nprecedent, see United States v. Hornaday, 392 F.3d 1306, 1316 (11th Cir. 2004)\\n\\n(collecting applications of Kotteakos), such overwhelming evidence of guilt\\n\\nsuffices to demonstrate that whatever error the district court might have committed\\n\\ndidn’t have a “substantial and injurious effect.” Kotteakos, 328 U.S. at 776. Nor\\n\\ncould the alleged error have influenced a reasonable juror’s verdict.\\n\\n 2\\n\\n Anael’s testimony relating Francesse’s statement about the Mercedes also\\n\\nunderlay the district court’s imposition of a sentencing enhancement for\\n\\nobstruction of justice. Baptiste objects that because Francesse’s statement was\\n\\ninadmissible hearsay, the district court improperly relied on it in finding\\n\\nobstruction.\\n\\n Section 3C1.1 of the Sentencing Guidelines provides for a two-level\\n\\nenhancement if “the defendant willfully obstructed or impeded, or attempted to\\n\\nobstruct or impede, the administration of justice with respect to the investigation,\\n\\nprosecution, or sentencing of the instant offense of conviction.” U.S.S.G. § 3C1.1.\\n\\n\\n\\n\\n 17\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 18 of 28\\n\\n\\nExamples of covered conduct include “committing, suborning, or attempting to\\n\\nsuborn perjury.” Id. at cmt. 4(B).\\n\\n If Anael’s testimony were admissible under one of the exceptions or\\n\\nexclusions to the rule against hearsay, there wouldn’t be any question that the\\n\\ndistrict court was on firm footing in relying on it for the enhancement. But if the\\n\\ntestimony was inadmissible, then our earlier conclusion that its admission at trial\\n\\nwas harmless wouldn’t (in and of itself) validate the district court’s reliance on it at\\n\\nsentencing. It all boils down to harm. Even if we strike Anael’s testimony, there\\n\\nwas still ample evidence supporting Baptiste’s conviction. With regard to that\\n\\ncomponent of the trial, Anael’s testimony was harmless. But Anael’s testimony\\n\\nprovided the only evidence supporting Baptiste’s enhancement for obstruction.\\n\\nWith regard to that component of the trial, Anael’s testimony was necessarily\\n\\nharmful.\\n\\n So have we reached the end of the road, such that we finally have to make a\\n\\ncall on the admissibility of Anael’s testimony? Not quite. Thanks to the “reliable\\n\\nhearsay” doctrine—which permits sentencing courts to rely on otherwise-\\n\\ninadmissible hearsay evidence in certain circumstances—we can reserve judgment\\n\\non the evidence’s admissibility (as we’ve done) while still addressing the district\\n\\ncourt’s enhancement.\\n\\n\\n\\n\\n 18\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 19 of 28\\n\\n\\n The doctrine is rooted in § 6A1.3 of the Sentencing Guidelines, which states\\n\\nthat a court “may consider relevant information without regard to its admissibility\\n\\nunder the rules of evidence applicable at trial, provided that the information has\\n\\nsufficient indicia of reliability to support its probable accuracy.” U.S.S.G. §\\n\\n6A1.3. Indeed, it’s well established that “[i]n determining the relevant facts,\\n\\nsentencing judges are not restricted to information that would be admissible at\\n\\ntrial.” Id. at cmt. The reason, the Supreme Court has explained, is that “modern\\n\\nconcepts individualizing punishment have made it all the more necessary that a\\n\\nsentencing judge not be denied an opportunity to obtain pertinent information by a\\n\\nrequirement of rigid adherence to restrictive rules of evidence properly applicable\\n\\nto the trial.” Williams v. New York, 337 U.S. 241, 247 (1949). Williams’ “modern\\n\\nconcepts” are old hat today, but no less important for it. Hence the need for a\\n\\nclear-headed understanding of when—and under what circumstances—hearsay\\n\\nevidence, in particular, can be deemed sufficiently reliable to support a sentencing\\n\\ndecision.\\n\\n The Sentencing Guidelines’ text—just quoted—is crystal clear. It articulates\\n\\na single precondition to a district court’s consideration of otherwise inadmissible\\n\\nevidence, including hearsay, as part of its sentencing calculus: The evidence is fair\\n\\ngame “provided that the information has sufficient indicia of reliability to support\\n\\nits probable accuracy.” U.S.S.G. § 6A1.3. The trickier question is whether our\\n\\n\\n 19\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 20 of 28\\n\\n\\ncase law has engrafted another, extratextual criterion. In particular, we’ve sent\\n\\nvaguely mixed signals about whether, as a prerequisite to relying on hearsay\\n\\nevidence at sentencing, a district court must make express findings on the record\\n\\nthat the evidence is reliable—or if, instead, as the Guidelines’ language indicates,\\n\\nit’s enough that the record as a whole provides “sufficient indicia of reliability.”\\n\\nThe distinction matters here because it’s undisputed that the district court did not\\n\\nmake any explicit findings about the reliability of Francesse’s testimony.\\n\\n Some have read our decision in United States v. Lee, 68 F.3d 1267 (11th Cir.\\n\\n1995), as requiring express findings that the hearsay evidence is reliable. For\\n\\nreasons we have explained before—and will reiterate here—that is incorrect. The\\n\\ndefendant in Lee objected that the district court had relied on the hearsay evidence\\n\\nof one of his co-conspirators—Russell—in determining his sentence. See id. at\\n\\n1275. We observed, as an initial matter, that “[a]ccording to the Sentencing\\n\\nGuidelines, and our case law interpreting them, the district court ‘may consider any\\n\\ninformation, including reliable hearsay, regardless of the information’s\\n\\nadmissibility at trial, provided that there are sufficient indicia of reliability to\\n\\nsupport its probable accuracy.’” Id. (quoting United States v. Castellanos, 904\\n\\nF.2d 1490, 1495 (11th Cir. 1990), and citing U.S.S.G. § 6A1.3). We stated that\\n\\n“the focus” of the inquiry “is upon the question of [the hearsay evidence’s]\\n\\nreliability,” and we further explained that reliability “must be determined on a case\\n\\n\\n 20\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 21 of 28\\n\\n\\nby case basis.” Id. The government there contended that Russell’s hearsay\\n\\nstatements were reliable because they were “against his penal interest,” and we\\n\\nacknowledged that as a general matter “a statement made against one’s penal\\n\\ninterest may bolster its credibility.” Id. at 1275–76. Even so, we concluded that\\n\\n“[i]n Russell’s case,” his penal interest “fail[ed] to qualify as a ‘sufficient\\n\\nindic[ium] of reliability’” given his “status as a fugitive from justice.” Id. at 1276.\\n\\nAccordingly, we concluded that “[s]pecific findings on Russell’s credibility [were]\\n\\nnecessary before the district court c[ould] use this evidence as a basis for\\n\\ndetermining” the defendant’s sentence, and we remanded for the district court “to\\n\\nmake further findings as to the reliability of Russell’s hearsay statement.” Id.\\n\\n Our subsequent decision in United States v. Gordon, 231 F.3d 750 (11th Cir.\\n\\n2000), clarified Lee’s holding and reach. The defendant there cited Lee in support\\n\\nof his contention “that the district court was required to make specific findings\\n\\nregarding the reliability” of hearsay testimony on which the government sought to\\n\\nrely at sentencing. Id. at 760. In response, we explained that Lee was never\\n\\nintended to—and didn’t—prescribe a per se rule that findings are always\\n\\nnecessary. Id. at 760–61. Rather, we made clear, Lee’s holding—that findings\\n\\nwere required there—was a product of the “circumstances” that underlay that\\n\\nparticular case. Id. Specifically, we said, it was because the traditional against-\\n\\npenal-interest justification didn’t cut it, and because the hearsay declarant was a\\n\\n\\n 21\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 22 of 28\\n\\n\\nfugitive, that the Lee court found “the absence of findings … particularly glaring\\n\\nand troubling.” Id. By contrast, we emphasized in Gordon, where the record and\\n\\nthe circumstances of the case “demonstrate adequate indicia of reliability,”\\n\\nfindings are not strictly necessary. Id. at 761. In such a case, a “district court’s\\n\\nfailure to make separate findings regarding the reliability of [hearsay] statements\\n\\n[is] not error.” Id. Put differently, “the absence of such findings does not\\n\\nnecessarily require reversal or remand where the reliability of the statements is\\n\\napparent from the record.” Id.3\\n\\n So … is this a “Lee” case—in which the traditional indicia of reliability fall\\n\\nshort and findings are thus required—or a “Gordon” case—in which the traditional\\n\\nindicia provide the necessary comfort? Clearly the latter, we think. For one,\\n\\n\\n\\n\\n3\\n One loose end: After Lee but before Gordon, in United States v. Anderton we addressed a\\nsimilar complaint that a district court had erroneously based its sentencing decision on\\n“unreliable hearsay.” 136 F.3d 747, 751 (11th Cir. 1998). In the course of rejecting that\\ncontention, we recited that “a court may rely on hearsay at sentencing, as long as the evidence\\nhas sufficient indicia of reliability, the court makes explicit findings of fact as to credibility, and\\nthe defendant has an opportunity to rebut the evidence.” Id. Seemingly in support of the\\n“explicit findings” aspect of that statement—albeit without saying so expressly—we cited our\\nearlier decision in Lee. Id. Anderton, though, doesn’t stand for the sort of per se rule that the\\nGordon defendant sought and that the Gordon panel rejected. The parties in Anderton never\\ndisputed whether “explicit findings” were required, and nothing rode on that issue. Id. Our\\ndecision simply assumed (at least in the circumstances presented) that they were, observed that\\nthe district court had made them, and held that they weren’t clearly erroneous. Id. The same is\\ntrue of our post-Gordon decisions in United States v. Zlatogur, 271 F.3d 1025, 1031 (11th Cir.\\n2001), and United States v. Patti, 337 F.3d 1317, 1326 (11th Cir. 2003), which merely recite\\nAnderton’s boilerplate. And to the extent that one could read Anderton as contradicting Lee, and\\nas holding that express findings of credibility are always required, we are bound by our prior-\\nprecedent rule to apply our earlier decision in Lee. See United States v. Steele, 147 F.3d 1316,\\n1318 (11th Cir. 1998) (en banc).\\n 22\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 23 of 28\\n\\n\\nFrancesse’s statement—that Baptiste had promised her a Mercedes in exchange for\\n\\nfavorable testimony—was clearly against both her and Baptiste’s penal interests,4\\n\\nand there’s no “fugitive from justice”-like reason here, as there was in Lee, to\\n\\ndiscount the force of that established gauge of reliability. For another, Francesse’s\\n\\nstatement aligned with Louissaint’s testimony regarding how Baptiste planned to\\n\\n“blame everything” on Pagnon. And for yet another, her statement specifically\\n\\nidentified the make and model of car she was set to receive. All of these are\\n\\n“indicia of reliability,” and by our lights, collectively clear the “sufficien[cy]”\\n\\nthreshold. See Lee, 68 F.3d at 1276.\\n\\n Because this isn’t the sort of case in which explicit findings were required,\\n\\nthe district court didn’t err in not making any. Nor did it err in relying on\\n\\nFrancesse’s sufficiently-reliable testimony to impose the obstruction-of-justice\\n\\nenhancement.\\n\\n C\\n\\n That leaves a few sentencing-related loose ends.\\n\\n 1\\n\\n As an initial matter, Baptiste challenges the district court’s imposition of\\n\\nsentencing enhancements for the amount of money that the government lost and\\n\\n\\n4\\n Even if, as we have said, it might (?) not formally qualify for the statement-against-interest\\nexception to the hearsay rule because Francesse testified at trial and thus might (?) not have been\\n“unavailable,” as Rule 804(b)(3) requires. See supra at 16–17.\\n\\n 23\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 24 of 28\\n\\n\\nthe number and vulnerability of the scheme’s victims. U.S.S.G. § 2B1.1(b)(1)(K)\\n\\n(amount lost exceeding $9,500,000); § 2B1.1(b)(2)(A) (offense involving more\\n\\nthan 10 victims); § 3A1.1(b)(1)–(2) (vulnerable victims).\\n\\n a\\n\\n As for the amount of money involved, we note first that Baptiste expressly\\n\\nwithdrew all objections except those relating to the enhancements for obstruction\\n\\nof justice and the number and vulnerability of his victims. We needn’t determine\\n\\nwhether in doing so Baptiste invited any alleged error, because even under the\\n\\ntypical clear-error standard, see United States v. Castaneda-Pozo, 877 F.3d 1249,\\n\\n1251 (11th Cir. 2017), Baptiste’s argument fails. The government submitted\\n\\nevidence indicating that Surveillance Master and its associated bank accounts were\\n\\nall under Baptiste’s name. The identification documents reviewed by Customs and\\n\\nImmigration Services also demonstrated that it was reasonable for the district court\\n\\nto conclude that virtually all of Surveillance Master’s check-cashing business was\\n\\nfraudulent—629 of the 630 IDs reviewed were fakes. Finally, “[t]he district court\\n\\nis only required to make a reasonable estimate of the loss.” United States v. Cobb,\\n\\n842 F.3d 1213, 1218 (11th Cir. 2016); U.S.S.G. § 2B1.1, cmt. 3(C); see also\\n\\nUnited States v. Medina, 485 F.3d 1291, 1304 (11th Cir. 2007) (indicating that a\\n\\nreasonable estimate is particularly appropriate in fraud cases because the amount\\n\\nlost is often difficult to determine with precision). We thus have evidence\\n\\n\\n 24\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 25 of 28\\n\\n\\nindicating that the sums lost went to Baptiste, that virtually all of the money that\\n\\npassed through the check-cashing business was linked to the scheme, as well as\\n\\nprecedent underscoring that the district court could apply this enhancement even if\\n\\nit couldn’t determine a precise figure. Taken together, we don’t see error here,\\n\\nmuch less of the “clear” variety.\\n\\n b\\n\\n Baptiste’s challenges to the enhancements based on the vulnerability and\\n\\nnumber of the scheme’s victims travel together. The Sentencing Guidelines\\n\\nprovide for a two-level increase on the ground that he “knew or should have\\n\\nknown that a victim of the offense was a vulnerable victim.” U.S.S.G. §\\n\\n3A1.1(b)(1). We have held that “both circumstances and immutable characteristics\\n\\ncan render a victim vulnerable,” United States v. Bradley, 644 F.3d 1213, 1288\\n\\n(11th Cir. 2011), and that “[n]either bodily injury nor financial loss is required” for\\n\\nthe enhancement, United States v. Moran, 778 F.3d 942, 978 (11th Cir. 2015). Our\\n\\ndecision in United States v. Pierre, 825 F.3d 1183 (11th Cir. 2016), settles the\\n\\nmatter with regard to Baptiste’s use of prisoners’ identities. We explained there\\n\\nthat “inmates have unique circumstances and immutable characteristics that make\\n\\nthem . . . vulnerable to . . . fraudulent activity.” Id. at 1196; see also id. (“Inmates\\n\\nusually do not file tax returns during periods of incarceration, and they are less\\n\\nlikely to discover that their identities have been compromised.”). The government\\n\\n\\n 25\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 26 of 28\\n\\n\\npresented evidence showing that Baptiste specifically requested the identities of\\n\\nprisoners, so the vulnerability-of-victims enhancement checks out.5 So too the\\n\\nnumber-of-victims enhancement; the district could reasonably infer from the\\n\\nevidence detailing the number of checks that Baptiste cashed that he had targeted a\\n\\n“large number” of people. See U.S.S.G. § 3A1.1(b)(2) (stipulating that if (i) §\\n\\n3A1.1(b)(1) applies and (ii) “the offense involved a large number of vulnerable\\n\\nvictims,” the defendant’s offense level should be raised an additional two levels.).\\n\\n 2\\n\\n Finally—and simply for us—the parties agree that the district court plainly\\n\\nerred in failing to “address the defendant personally in order to permit the\\n\\ndefendant to speak” at sentencing, as Federal Rule of Criminal Procedure\\n\\n32(i)(4)(A)(ii) expressly requires. 6 In response to the district court’s query, “Does\\n\\nthe defendant wish to address me?”, Baptiste’s attorney responded on his client’s\\n\\nbehalf: “Mr. Baptiste respectfully has nothing to add.” However formalistic it may\\n\\nseem, and even reviewing for plain error, we agree with the parties that this\\n\\n5\\n Because the record supports the district court’s finding that Baptiste targeted incarcerated\\nvictims, we can affirm the enhancement for vulnerable victims on that ground alone. We needn’t\\ndetermine whether the deceased count as “vulnerable” as that term is used in § 3A1.1(b).\\nCompare United States v. Roberson, 872 F.2d 597, 609 (5th Cir. 1989) (sentence of a defendant\\nconvicted of fraud properly enhanced where the defendant used a dead companion’s credit card\\nafter burning his corpse), with United States v. Shumway, 112 F.3d 1413, 1424 (10th Cir. 1997)\\n(refusing to find that skeletal remains are a “vulnerable victim”).\\n6\\n “Before imposing sentence, the court must . . . (ii) address the defendant personally in order to\\npermit the defendant to speak or present any information to mitigate the sentence . . . .” Fed. R.\\nCrim. P. 32(i)(4)(A)(ii).\\n\\n 26\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 27 of 28\\n\\n\\nexchange—between court and counsel rather than court and defendant—amounted\\n\\nto “(1) error, (2) that is plain and (3) that affect[ed] [Baptiste’s] substantial rights,”\\n\\nand that the error “[4] seriously affect[ed] the fairness, integrity, or public\\n\\nreputation of judicial proceedings.” United States v. Turner, 474 F.3d 1265, 1276\\n\\n(11th Cir. 2007); United States v. Doyle, 857 F.3d 1115, 1120 (11th Cir. 2017)\\n\\n(“[T]he presumption of prejudice in denial of allocution cases is still the rule when\\n\\na defendant is not sentenced at the bottom of his guidelines range . . . .”); United\\n\\nStates v. Prouty, 303 F.3d 1249, 1251 n.1 (11th Cir. 2002) (rejecting “the\\n\\nGovernment’s argument that directing comments to the lawyers suffices to comply\\n\\nwith the rule”). On remand, Baptiste should be afforded the opportunity to address\\n\\nthe court directly.\\n\\n III\\n\\n Accordingly, we hold as follows:\\n\\n (A) The district court did not abuse its discretion in refusing to allow\\n\\nBaptiste to introduce evidence that Marvin Pagnon had duped others into\\n\\nparticipating in similar schemes. Nor did the government fail to disclose any\\n\\nmaterial information concerning its witnesses or mischaracterize or exaggerate the\\n\\nevidence during its closing argument.\\n\\n (B) Any error that the district court might have committed in admitting\\n\\nAnael Chery’s testimony relaying his sister Francesse’s statement that Baptiste had\\n\\n\\n 27\\n\\f Case: 16-17175 Date Filed: 08/28/2019 Page: 28 of 28\\n\\n\\npromised her a car in exchange for favorable testimony was harmless. And the\\n\\ndistrict court did not err in relying on the same evidence to support an obstruction-\\n\\nrelated sentencing enhancement because the record reveals that the evidence bore\\n\\n“sufficient indicia of reliability.”\\n\\n (C) The district court did not err in imposing sentencing enhancements\\n\\npertaining to the amount of money that the government lost and the number and\\n\\nvulnerability of Baptiste’s scheme’s victims. The district court did reversibly err,\\n\\nhowever, when it failed to permit Baptiste to allocute personally.\\n\\n AFFIRMED IN PART AND REMANDED.\\n\\n\\n\\n\\n 28\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4159,"cells":{"id":{"kind":"number","value":4656968,"string":"4,656,968"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"Lizzie Davis v. Oasis Legal Finance Operating Company, LLC"},"decision_date":{"kind":"string","value":"2019-08-28"},"docket_number":{"kind":"string","value":"18-10526"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before TJOFLAT and JORDAN, Circuit Judges, and SCHLESINGER, District Judge.\", \"parties\": \"\", \"opinions\": [{\"author\": \"JORDAN, Circuit Judge:\", \"type\": \"010combined\", \"text\": \"Case: 18-10526 Date Filed: 08/28/2019 Page: 1 of 19\\n\\n\\n [PUBLISH]\\n\\n\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 18-10526\\n ________________________\\n\\n D.C. Docket No. 3:17-cv-00022-DHB-BKE\\n\\n\\n\\nLIZZIE DAVIS,\\nindividually and on behalf of all others similarly situated,\\nDENNIS GREEN,\\nindividually and on behalf of all others similarly situated,\\nJOHNNY MOODY,\\nindividually and on behalf of all others similarly situated,\\nJOHN SUBER,\\nindividually and on behalf of all others similarly situated,\\nSHIRLEY WILLIAMS,\\nindividually and on behalf of all others similarly situated,\\nPAMELA DAVIS,\\nindividually and on behalf of all others similarly situated,\\n\\n Plaintiffs-Appellees,\\n\\nversus\\n\\nOASIS LEGAL FINANCE OPERATING COMPANY, LLC,\\nOASIS LEGAL FINANCE, LLC,\\nOASIS LEGAL FINANCE HOLDING COMPANY, LLC,\\n\\n Defendants-Appellants.\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 2 of 19\\n\\n\\n _______________________\\n\\n Appeals from the United States District Court\\n for the Southern District of Georgia\\n ________________________\\n\\n (August 28, 2019)\\n\\nBefore TJOFLAT and JORDAN, Circuit Judges, and SCHLESINGER, District\\nJudge.∗\\n\\nJORDAN, Circuit Judge:\\n\\n American courts have long refused to enforce contractual provisions that\\n\\ncontravene public policy. See, e.g., Marshall v. Baltimore and Ohio R.R., 57 U.S.\\n\\n314, 334 (1853) (“It is an undoubted principle of the common law that it will not\\n\\nlend its aid to enforce a contract to do an act that is illegal, or which is inconsistent\\n\\nwith sound morals or public policy. . . .”). In Georgia, “[n]o principle of\\n\\njurisprudence is better settled than this.” Glass v. Childs, 71 S.E. 920, 921 (Ga. Ct.\\n\\nApp. 1911). 1\\n\\n Courts have said that “[i]t is the duty of all courts of justice to keep their eye\\n\\nsteadily up on the interests of the public, . . . and when they find an action is founded\\n\\n\\n\\n∗The Honorable Harvey E. Schlesinger, United States District Judge for the Middle District of\\nFlorida, sitting by designation.\\n1\\n For an interesting early Georgia case holding a contract void as against public policy, see\\nChancely v. Bailey, 37 Ga. 532, 533, 541–42 (Ga. 1868) (holding that a contract purporting to pay\\nMr. Chancely $2,500 if he served in Mr. Bailey’s place in the Army of the Confederate States\\nduring the Civil War was void because Georgia did not have the right to forcibly secede from the\\nUnion and Mr. Chancely could not contract to illegally engage in war against the United States).\\n 2\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 3 of 19\\n\\n\\nup on a claim injurious to the public . . . to give no countenance or assistance in foro\\n\\ncivili.” Elisha Greenhood, The Doctrine of Public Policy: Reduced to Rules 2 (1886)\\n\\n(quoting C.J. Wilmot’s Opinion in Low v. Peers, (1770) 97 Eng. Rep. 138 (Ex. Ch.)).\\n\\nOthers, however, have characterized the public policy defense as “a very unruly\\n\\nhorse, and when once you get astride it you never know where it will carry you.”\\n\\nRichardson v. Mellish, (1824) 130 Eng. Rep. 294, 303 (H.L.) (Burrough, J.).\\n\\n In this case, a class of borrowers filed suit in Georgia against their lenders,\\n\\nalleging that their loan agreements violated state usury laws. The lenders moved to\\n\\ndismiss the action based on the forum selection clause and class action waiver in the\\n\\nagreements. The district court concluded that both provisions were unenforceable\\n\\nas against Georgia public policy, and the lenders appealed.\\n\\n Following oral argument and a review of the relevant authorities, we agree\\n\\nwith the district court. Georgia’s Payday Lending Act and Industrial Loan Act\\n\\narticulate a clear public policy against enforcing forum selection clauses in payday\\n\\nloan agreements and in favor of preserving class actions as a remedy for those\\n\\naggrieved by predatory lenders. If Georgia’s public policy regarding payday lenders\\n\\nis a horse, as Justice Burrough suggested, it carries these borrowers safely to a\\n\\nGeorgia courthouse.\\n\\n\\n\\n\\n 3\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 4 of 19\\n\\n\\n I\\n\\n The plaintiffs entered into identical loan agreements with Oasis Legal\\n\\nFinance, LLC; Oasis Legal Finance Operating Company, LLC; and Oasis Legal\\n\\nFinance Holding Company, LLC ( the “Oasis lenders”). The loans generally\\n\\namounted to less than $3,000 and were to be repaid from any recoveries that the\\n\\nplaintiffs received in their separate personal injury lawsuits. The plaintiffs’\\n\\nobligations to repay the loans were therefore contingent on success in the underlying\\n\\nlawsuits.\\n\\n In February of 2017, the plaintiffs filed a class action complaint against the\\n\\nOasis lenders in Georgia state court, alleging that the loan agreements violated\\n\\nGeorgia’s Payday Lending Act, O.C.G.A. § 16-17-1 et seq., Industrial Loan Act,\\n\\nO.C.G.A. § 7-3-1 et seq., and usury laws, O.C.G.A. § 7-4-18. The Oasis lenders\\n\\nremoved the suit to federal court and moved to dismiss the complaint under Federal\\n\\nRule of Civil Procedure 12(b)(6) and to strike the plaintiffs’ class allegations under\\n\\nFederal Rule of Civil Procedure 12(f). The Oasis lenders argued, among other\\n\\nthings, that the loan agreements’ forum selection clause required the plaintiffs to\\n\\nbring suit in Illinois, and that the class action waiver barred their ability to file a class\\n\\n\\n\\n\\n 4\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 5 of 19\\n\\n\\naction. The plaintiffs responded that these provisions violated Georgia public policy\\n\\nand, therefore, were unenforceable. 2\\n\\n Applying Georgia law, the district court rejected both of the arguments made\\n\\nby the Oasis lenders and held that the forum section clause and class action waiver\\n\\nwere unenforceable. See Davis v. Oasis Legal Fin. Operating Co., No. CV 317-022,\\n\\n2017 WL 5490919, at *4–5 (S.D. Ga. Nov. 15, 2017) The district court concluded\\n\\nthat “the enforcement of forum selections clauses in payday lending contracts would\\n\\ncontravene [Georgia’s] public policy” as established by the Payday Lending Act. Id.\\n\\nat *4. It explained that “[c]ertain payday lenders have attempted to use forum\\n\\nselection clauses contained in payday loan documents in order to avoid the courts of\\n\\nthe State of Georgia, and the General Assembly has determined that such practices\\n\\nare unconscionable and should be prohibited.” Id. at *3 (quoting § 16-17-1(d)). The\\n\\ndistrict court similarly ruled that the class action waiver contravened public policy\\n\\nbecause, when the Georgia Legislature enacted the PLA and the GILA, it expressly\\n\\nincluded class actions as a remedy for those aggrieved by payday lenders. The\\n\\ndistrict court reasoned that the Georgia Legislature would not create such a remedy\\n\\n\\n2\\n The loan agreements’ forum selection clause states: “THE PARTIES HEREBY\\nIRREVOCABLY AND UNCONDITIONALLY CONSENT TO SUBMIT TO THE EXCLUSIVE\\nJURISDICTION OF THE CIRCUIT COURT OF COOK COUNTY, ILLINOIS FOR ANY\\nDISPUTES . . . .” D.E. 10-8 at 25 ¶6.5 (capitalization in original). The loan agreements’ class\\naction waiver states: “THE PARTIES HEREBY . . . WAIVE ANY RIGHT TO CONSOLIDATE\\nOR TO HAVE HANDLED AS A CLASS ACTION ANY PROCEEDING ON ANY LAWSUIT\\n. . . .” D.E. 10-8 at 25 ¶6.6 (capitalization in original).\\n\\n 5\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 6 of 19\\n\\n\\nand then allow lenders to “effectively wipe away this consumer protection with a\\n\\nwaiver in a single paragraph of a six-page, single-spaced agreement.” Id. at *5.\\n\\n The district court certified its decision for interlocutory review, and the Oasis\\n\\nlenders petitioned us to consider whether the district court erred in concluding that\\n\\nthe loan agreements’ forum selection clause and class action waiver are\\n\\nunenforceable. We agreed to consider both issues, and now affirm. 3\\n\\n II\\n\\n The enforceability of a forum selection clause is a question of law that we\\n\\nreview de novo. See Rucker v. Oasis Legal Fin., L.L.C., 632 F.3d 1231, 1235 (11th\\n\\nCir. 2011). The same plenary standard governs the enforceability of a class action\\n\\nwaiver. See Cruz v. Cingular Wireless, LLC, 648 F.3d 1205, 1210 (11th Cir. 2011).\\n\\n A\\n\\n A forum selection clause, when properly bargained for, “protects [the parties’]\\n\\nlegitimate expectations and furthers the vital interest of the justice system.” Atl.\\n\\nMarine Constr. Co. v. U.S. Dist. Ct. for the W. Dist. of Tex., 571 U.S. 49, 63 (2013)\\n\\n\\n\\n\\n3\\n We recognize that in Ruth v. Cherokee Funding, LLC, 820 S.E.2d 704, 710 (Ga. 2018), the\\nGeorgia Supreme Court held that loans in which the borrowers’ obligations to repay are contingent\\nupon success in an underlying personal-injury lawsuit are not “loans” under the Georgia Payday\\nLending Act and the Industrial Loan Act. The Oasis lenders, however, concede that they did not\\nargue in the district court that the loans at issue here are not loans under the PLA and GILA. [See\\nO.A. Recording at 5:16] That argument, therefore, is not properly before us. See Access Now,\\nInc. v. Sw. Airlines Co., 385 F.3d 1324, 1331 (11th Cir. 2004).\\n\\n\\n 6\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 7 of 19\\n\\n\\n(quotation marks omitted). Such clauses “should be given controlling weight in all\\n\\nbut the most exceptional cases.” Id. See also M/S Bremen v. Zapata Off-Shore Co.,\\n\\n407 U.S. 1, 15 (1972) (stating that a forum selection clause “should control absent a\\n\\nstrong showing that it should be set aside”).\\n\\n In Bremen, 407 U.S. at 15–18, the Supreme Court identified four grounds on\\n\\nwhich a court can refuse to enforce an otherwise-valid forum selection clause. These\\n\\ninclude if “(1) [its] formation was induced by fraud or overreaching; (2) the plaintiff\\n\\neffectively would be deprived of its day in court because of the inconvenience or\\n\\nunfairness of the chosen forum; (3) the fundamental unfairness of the chosen law\\n\\nwould deprive the plaintiff of a remedy; or (4) enforcement of the [forum selection\\n\\nclause] would contravene a strong public policy.” Lipcon v. Underwriters at\\n\\nLloyd’s, London, 148 F.3d 1285, 1292 (11th Cir. 1998) (citing Bremen, 407 U.S. at\\n\\n15–18). Here, we consider the fourth ground—when enforcing the forum selection\\n\\nclause “would contravene a strong public policy of the forum in which suit is\\n\\nbrought, whether declared by statute or judicial decision.” Bremen, 407 U.S. at 15.\\n\\n “Public policy is an amorphous concept . . . . Accordingly, it has been held\\n\\nthat, the delicate and undefined power of courts to declare a contract void as\\n\\ncontravening public policy should be exercised with great caution, and only in cases\\n\\nfree from substantial doubt.” Dep’t of Transp. v. Brooks, 328 S.E.2d 705, 713 (Ga.\\n\\n1985) (alterations adopted and quotations omitted). In Georgia, a contractual\\n\\n\\n 7\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 8 of 19\\n\\n\\nprovision generally does not violate public policy unless the Legislature has declared\\n\\nit so or enforcement of the provision would flout the very purpose of the law. See\\n\\nid.; Robinson v. Reynolds, 21 S.E.2d 214, 215 (Ga. 1942) (“Contracts that obviously\\n\\nand directly tend in a marked degree to bring about results that the law seeks to\\n\\nprevent cannot be made the ground of a successful suit . . . [and are] void as against\\n\\npublic policy.”).\\n\\n Although amorphous in concept, Georgia’s public policy bar is built on a solid\\n\\nfoundation—the Georgia Constitution and state statutes. “[T]hese are the sources\\n\\nthat are first to be considered and that often may be conclusive.” Strickland v. Gulf\\n\\nLife Ins. Co., 242 S.E.2d 148, 151 (Ga. 1978) (quotation omitted). “Statutes, of\\n\\ncourse, are perhaps the clearest expressions of the public policy of [Georgia].” Dove\\n\\nv. Dove, 680 S.E.2d 839, 842 (Ga. 2009). See also Sonja Laesen, 7 Georgia\\n\\nJurisprudence § 3:3 (updated 2019) (“The only authentic and admissible evidence of\\n\\npublic policy of a state is its constitution, laws, and judicial decisions.”). For\\n\\nexample, O.C.G.A. § 13-8-2 provides that “[a] contract which is against the policy\\n\\nof the law cannot be enforced,” and lists particular agreements that are unenforceable\\n\\nas against public policy. Although § 13-8-2 does not address the specific provisions\\n\\nat issue in this case, the statute makes clear that its list is “nonexclusive” and Georgia\\n\\ncourts have held that they “are not restricted by this list in voiding contracts that are\\n\\nagainst public policy.” Edwards v. Grapefields, Inc., 599 S.E.2d 489, 493 (Ga. Ct.\\n\\n\\n 8\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 9 of 19\\n\\n\\nApp. 2004). We may therefore look to other Georgia statutes to determine whether\\n\\nthe state has a strong public policy against enforcing forum selection clauses in favor\\n\\nof out-of-state payday lenders.\\n\\n The Georgia Payday Lending Act, O.C.G.A. § 16-17-1 et seq., states in\\n\\nrelevant part as follows:\\n\\n A payday lender shall not . . . nor shall the loan contract\\n designate a court for the resolution of disputes concerning\\n the contract other than a court of competent jurisdiction in\\n and for the county in which the borrower resides or the\\n loan office is located.\\n\\n§ 16-17-2(c)(1). The Georgia Legislature explained the PLA’s prohibition on out-\\n\\nof-state forum selection clauses in this way:\\n\\n Certain payday lenders have attempted to use forum\\n selection clauses contained in payday loan documents in\\n order to avoid the courts of the State of Georgia, and the\\n General Assembly has determined that such practices are\\n unconscionable and should be prohibited.\\n\\n§ 16-17-1(d) (emphasis added).\\n\\n Based on this language, the Georgia Supreme Court ruled that the PLA\\n\\nestablished a clear public policy against “payday lenders [that] have attempted to\\n\\nskirt the laws of Georgia by use of forum selection clauses.” W. Sky Fin., LLC v.\\n\\nGeorgia, 793 S.E.2d 357, 363 (Ga. 2016). In ruling that the loan agreements’ forum\\n\\nselection clause is unenforceable in this case, the district court found that § 16-17-\\n\\n1(d) and § 16-17-2(c)(1) were “conclusive” as to Georgia’s public policy. See\\n\\n\\n 9\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 10 of 19\\n\\n\\nStrickland, 242 S.E.2d at 151. On appeal, the Oasis lenders argue that two other\\n\\nsubsections of the PLA support the opposite conclusion.4\\n\\n First, the Oasis lenders assert that the word “county” in § 16-17-2(c)(1) is\\n\\nunqualified, meaning that by permitting parties to select a forum “in and for the\\n\\ncounty in which . . . the loan office is located” the PLA allows parties to select a\\n\\ncounty outside Georgia for litigation if it is the county where the lenders’ loan office\\n\\nis located. The parties could therefore select as a forum the county where the Oasis\\n\\nlenders operate their loan office—Cook County, Illinois. This argument has some\\n\\nsuperficial appeal, but the district court rejected it, concluding that the word\\n\\n“county” in § 16-17-2(c)(1) refers only to Georgia counties because that subsection\\n\\nis a venue provision through which the Legislature intended to allow parties to select\\n\\nfora within Georgia while prohibiting outbound forum-selection clauses. The Oasis\\n\\nlenders contend that the district court erred by reading in the word “Georgia” in front\\n\\nof “county” in § 16-17-2(c)(1). We think the district court got it right.\\n\\n Georgia venue provisions commonly use the term “county” or “counties” in\\n\\nrefence to Georgia counties, without explicitly saying so. See Ga. Const., Art. VI, §\\n\\n2, ¶¶ III, IV, VI; O.C.G.A. § 9-10-93. For example, Georgia’s Constitution provides\\n\\n\\n4\\n The Oasis lenders also claim (in a footnote of their reply brief) that interpreting the PLA to require\\nout-of-state lenders to defend these kinds of lawsuits in Georgia may violate the dormant\\nCommerce Clause. The Oasis lenders, however, did not present this argument in the district court\\nor in their initial brief, and they also failed to support the argument with legal authorities. We\\ntherefore do not address it. See Sapuppo v. Allstate Floridian Ins. Co., 739 F.3d 678, 681 (11th\\nCir. 2014); Lovett v. Ray, 327 F.3d 1181, 1183 (11th Cir. 2003).\\n 10\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 11 of 19\\n\\n\\nthat “joint trespassers residing in different counties may be tried in either county,”\\n\\nGa. Const., Art. VI, § 2, ¶ IV, and Georgia courts have interpreted that provision to\\n\\nallow “joint tortfeasor residents of different Georgia counties [to] be sued in either\\n\\ncounty.” Robinson v. Star Gas of Hawkinsville, Inc., 533 S.E.2d 97, 98 (Ga. Ct. App.\\n\\n2000) (emphasis added). See also Cartwright v. Fuji Photo Film U.S.A., Inc., 720\\n\\nS.E.2d 200, 204 n.4, 206 (Ga. Ct. App. 2011) (“[T]he proper venue for a suit is in\\n\\nwhich Georgia county the defendant resides[.]”); Cunningham v. Estate of\\n\\nCunningham, 697 S.E.2d 280, 282 (Ga. Ct. App. 2010) (“Georgia’s constitutional\\n\\nvenue provision . . . has no application to a suit against a nonresident . . . because in\\n\\nsuch case there is no Georgia county in which the nonresident resides.”). The same\\n\\ninterpretation applies to Georgia statutes with similar language. See Lloyd Adams,\\n\\nInc. v. Liberty Mut. Ins. Co., 10 S.E.2d 46, 48 (Ga. 1940) (holding that a similar\\n\\nvenue provision in Georgia’s Nonresident Motorist Act, now O.C.G.A. § 40-12-3,\\n\\nafforded venue in “any county in the State”). This makes sense given that the\\n\\nGeorgia Legislature generally has authority to establish venue rules within Georgia.\\n\\n Given the PLA’s condemnation of out-of-state lenders using forum selection\\n\\nclauses to avoid Georgia courts in § 16-17-1(d), it would make little sense to\\n\\nconclude that the Georgia Legislature intended the word “county” in the next\\n\\nsubsection to include places like Cook County, Illinois. See West v. City of Albany,\\n\\n797 S.E.2d 809, 811 (2017) (“We do not limit our consideration to the words of one\\n\\n\\n 11\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 12 of 19\\n\\n\\nsubsection of a statute alone, but consider a particular provision in the context of the\\n\\nstatute as a whole . . . .”). The Oasis lenders’ argument—that by using the term\\n\\n“county” instead of “Georgia county” in § 16-17-2(c)(1), the Legislature opened a\\n\\nback door for out-of-state forum selection clauses—would effectively “hide\\n\\nelephants in mouseholes.” Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 468\\n\\n(2001). Moreover, it would also render the PLA’s prohibition of certain forum\\n\\nselection clauses in § 16-17-1(d) “meaningless, redundant, or mere surplusage.”\\n\\nUnited States v. Canals-Jimenez, 943 F.2d 1284, 1287 (11th Cir. 1991). We disfavor\\n\\nsuch interpretations. See Cadwell v. Kaufman, Englett & Lynd, PPLC, 886 F.3d\\n\\n1153, 1159 (11th Cir. 2018).\\n\\n Next, the Oasis lenders contend that the PLA does not apply to loan\\n\\nagreements between a Georgia borrower and an out-of-state lender because § 16-17-\\n\\n1(d) states that “[p]ayday lending involves relatively small loans and does not\\n\\nencompass loans that involve interstate commerce.” We disagree for a few reasons.\\n\\nFirst, the Oasis lenders’ arguments are mutually exclusive. Arguing that the PLA\\n\\ndoes not apply to loans by an out-of-state lender contradicts the lenders’ argument\\n\\nthat the Georgia Legislature meant for the term “county” in § 16-17-2(c)(1) to\\n\\ninclude the county where an out-of-state lender maintains its home office. See\\n\\nFootstar, Inc. v. Liberty Mut. Ins. Co., 637 S.E.2d 692, 695 (2006) (“[C]ourts should\\n\\nconstrue a statute to give sensible and intelligent effect to all of its provisions . . .”).\\n\\n\\n 12\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 13 of 19\\n\\n\\nSecond, other provisions of the PLA make clear that the Act governs “any business”\\n\\nthat “consists in whole or in part of making . . . loans of $3,000.00 or less” unless\\n\\nthose entities are specifically exempted. See W. Sky Fin., 793 S.E.2d at 363 (quoting\\n\\n§ 16-17-2(a)). Out-of-state lenders are not exempt. See § 16-17-2(a)(1)–(4). Third,\\n\\nexcluding loans involving out-of-state lenders from coverage—as the Oasis lenders\\n\\nargue—would render the PLA’s prohibition of out-of-state forum selection clauses\\n\\nmeaningless. See Cadwell, 886 F.3d at 1159; Canals-Jimenez, 943 F.2d at 1287.\\n\\n In Western Sky, 793 S.E.2d at 363, the defendant argued that the PLA’s\\n\\n“interstate commerce” language rendered the statute inapplicable to loans involving\\n\\nout-of-state lenders. The Georgia Supreme Court “reject[ed] the notion that this\\n\\nphrase was meant to exclude loans that involve interstate commerce from the scope\\n\\nof the [PLA].” Id. “If that were so, the [PLA] would be virtually meaningless\\n\\nbecause it would prohibit nothing.” Id. See also id. at 364 (“If, as Defendants argue,\\n\\nthe [PLA] is limited only to transactions that do not involve interstate commerce, the\\n\\n[PLA] would effectively prohibit nothing at all.”). The Georgia Supreme Court went\\n\\non to articulate how excluding loans involving an out-of-state lender would\\n\\nundermine the entire purpose of the PLA.\\n\\n Given the clear and unambiguous scope of the [PLA] as a\\n whole, to interpret that phrase as a definitional limitation\\n upon payday lending and thereby exempt loans that\\n involve interstate commerce from the prohibitions of the\\n [PLA] would create such a contradiction and absurdity as\\n\\n\\n 13\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 14 of 19\\n\\n\\n to demonstrate that the legislature did not mean it to\\n create such a limitation.\\n\\nId. at 365 (emphasis added).\\n\\n Based on our review, Georgia statutes establish a clear public policy against\\n\\nout-of-state lenders using forum selection clauses to avoid litigation in Georgia\\n\\ncourts. See O.C.G.A. §§ 16-17-1(d), 2(c)(1). Enforcing a forum selection clause\\n\\nlike the ones here “would [therefore] contravene a strong public policy of the forum\\n\\nin which suit is brought,” and the district court correctly denied the Oasis lenders’\\n\\nmotion to dismiss on that ground. See Bremen, 407 U.S. at 15.5\\n\\n B\\n\\n The plaintiffs argue, and the district court concluded, that the class action\\n\\nwaivers contained in the loan agreements also contravene the purpose of the PLA\\n\\nand the Georgia Industrial Loan Act, O.C.G.A. § 7-3-29. We start with the text of\\n\\nboth acts. See Dove, 680 S.E.2d at 842.\\n\\n The PLA contains a provision stating that “a civil action may be brought on\\n\\nbehalf of an individual borrower or on behalf of an ascertainable class of borrowers.”\\n\\n\\n\\n\\n5\\n Because we conclude that Georgia statutes establish a public policy against the forum selection\\nclause at issue, we are not bound by our opinion in Rucker, 632 F.3d at 1238. In that case, we held\\nthat an identical forum selection clause was valid and enforceable under Alabama law. See id. Our\\ndecision to invalidate the forum selection clause in this case depends entirely on Georgia law.\\n 14\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 15 of 19\\n\\n\\n§ 16-7-3. The GILA provides that “a claim for violation of this chapter against an\\n\\nunlicensed lender may be asserted in a class action.” § 7-3-29(e). 6\\n\\n The district court concluded that, in passing these laws, the Georgia\\n\\nLegislature “expressly contemplated a specific remedy—[a] class action—for\\n\\npersons aggrieved by predatory lending . . . [and] did not expressly create the class\\n\\naction remedy so that predatory lenders could effectively wipe away this consumer\\n\\nprotection with a waiver . . . .” Davis, 2017 WL 5490919, at *5. On appeal, the\\n\\nOasis lenders contend that the district court erred by not considering whether the\\n\\nprovision was procedurally or substantively unconscionable. They also argue that\\n\\nneither the PLA nor the GILA prohibit class action waivers or create a statutory right\\n\\nto pursue a class action.\\n\\n These arguments miss the point. The district court’s ruling flowed from its\\n\\nconclusion that enforcing class action waivers in this context would allow payday\\n\\nlenders to eliminate a remedy that was expressly contemplated by the Georgia\\n\\nLegislature, and thereby undermine the purpose of the statutory scheme. That\\n\\nconclusion, if correct, renders the class action waiver unenforceable under Georgia\\n\\nlaw regardless of whether the provision is also procedurally or substantively\\n\\nunconscionable. See Brookfield Country Club, Inc. v. St. James-Brookfield, LLC,\\n\\n\\n6\\n We note that this subsection of the GILA applies to “unlicensed lenders,” but the district court\\nquoted this provision in ruling that the class action waiver is unenforceable, and the Oasis lenders\\ndo not argue that it does not apply. See Davis, 2017 WL 5490919, at *5.\\n 15\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 16 of 19\\n\\n\\n696 S.E.2d 663, 667 (Ga. 2010) (refusing to “enforce a contractual provision which\\n\\ncontravenes the statutory law of this state”); Terry v. St. Farm Fire & Cas. Ins. Co.,\\n\\n504 S.E.2d 194, 195 (Ga. 1998) (considering whether “[t]he underlying purpose” of\\n\\nthe statute would be “contravened by enforcement of the specific terms of the\\n\\nagreement”). Accord Atl. Flooring Design Ctrs., Inc. v. R.G. Williams Const., Inc.,\\n\\n773 S.E.2d 868, 870 (Ga. Ct. App. 2015) (voiding a contractual provision because it\\n\\nwould frustrate the public policy expressed in the Georgia Arbitration Code);\\n\\nLangford v. Royal Indem. Co., 430 S.E.2d 98, 102 (Ga. Ct. App. 1993) (“A contract\\n\\nprovision normally should not be enforced where it conflicts with the general policy\\n\\nand spirit of the statute which governs it, although there may be no literal conflict;\\n\\nthat is, it makes no difference whether the statutory prohibition or command is\\n\\nexpressed or implied.”).\\n\\n The Oasis lenders may be correct in arguing that, when Georgia courts address\\n\\nwhether a contractual provision is substantively unconscionable, they also “consider\\n\\n‘the commercial reasonableness of the contract terms, the purpose and effect of the\\n\\nterms, the allocation of the risks between the parties, and similar public policy\\n\\nconcerns.’” Jenkins v. First Am. Cash Advance of Ga., LLC, 400 F.3d 868, 876 (11th\\n\\nCir. 2005) (quoting NEC Techs., Inc. v. Nelson, 478 S.E.2d 769, 772 (Ga. 1996)).\\n\\nThe public policy bar, however, remains an independent basis to hold a contractual\\n\\nprovision unenforceable. See Glosser v. Powers, 71 S.E.2d 230, 231 (Ga. 1952)\\n\\n\\n 16\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 17 of 19\\n\\n\\n(“Even in the absence of fraud in the procurement of a contract, a contract which is\\n\\nagainst the policy of the law is void and unenforceable.”) A hornbook example of\\n\\nthe public policy defense is that a court will not enforce a contractual provision that\\n\\nis illegal regardless of whether its obligations are mutual, its terms are conspicuous,\\n\\nand the parties are well represented. See 17A Am. Jur. 2d Contracts § 247 (updated\\n\\n2019). See also Stoudemire v. HSBC Bank USA, N.A., 776 S.E.2d 483, 484 (Ga. Ct.\\n\\nApp. 2015) (highlighting the difference between “[c]ontracts to do illegal or immoral\\n\\nthings, contracts against public policy, and gambling contracts,” which are legally\\n\\nvoid, and “fraudulent contracts and contracts entered under duress,” which are\\n\\nvoidable by the injured party) (citations omitted).\\n\\n The district court did not, and needed not, conclude that the PLA or the GILA\\n\\nexpressly prohibited class action waivers or created a statutory right to pursue a class\\n\\naction. A contractual provision need not “literal[ly] conflict” with Georgia law to\\n\\ncontravene public policy. See Langford, 430 S.E.2d at 102; Robinson, 21 S.E.2d at\\n\\n215. Instead, the district court concluded that enforcement of the class action\\n\\nwaivers in this context would eliminate a remedy contemplated by the Georgia\\n\\nLegislature and undermine the purpose of the PLA and the GILA. Davis, 2017 WL\\n\\n5490919, at *5.\\n\\n To that point, the Oasis lenders make one additional argument warranting\\n\\nfurther discussion—that the PLA’s fee-shifting provision eliminates the risk that\\n\\n\\n 17\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 18 of 19\\n\\n\\nenforcing the class action waiver would effectively prevent the plaintiffs from\\n\\nlitigating their claims. In past cases, we have held that certain class action waivers\\n\\nwere not unconscionable, in part because the agreements at issue contained a fee-\\n\\nshifting provision that, in practice, permitted the plaintiffs to pursue their claims\\n\\nindividually. See Jenkins, 400 F.3d at 877–80; Randolph v. Green Tree Fin. Corp.,\\n\\n244 F.3d 814, 816–18 (11th Cir. 2001). Those cases, however, concerned whether\\n\\nthe class action waivers were unconscionable because enforcement would prevent\\n\\nthe plaintiffs from having their day in court. See Jenkins, 400 F.3d at 877–78. As\\n\\ndiscussed above, the district court here did not conclude that the class action waiver\\n\\nwas procedurally or substantively unconscionable—or that the plaintiffs in this case\\n\\ncould never litigate their claims individually—but relied on Georgia’s public policy\\n\\nas expressed in the text and purpose of the PLA and the GILA.\\n\\n The Oasis lenders point out that in Jenkins, 400 F.3d at 877–80, and Bowen v.\\n\\nFirst Family Financial Services, Inc., 233 F.3d 1331, 1336 (11th Cir. 2000), we held\\n\\nthat class action waivers in arbitration clauses were not void as against public policy.\\n\\nBut those cases do not call for a different result here. Neither Jenkins nor Bowen\\n\\nanalyzed the PLA, the GILA, or Georgia’s public policy against class action waivers\\n\\nin payday loan agreements. Instead, those cases concerned class action waivers in\\n\\narbitration agreements—where the Federal Arbitration Act “create[s] a strong\\n\\nfederal policy in favor of arbitration.” Picard v. Credit Solutions, Inc., 564 F.3d\\n\\n\\n 18\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 19 of 19\\n\\n\\n1249, 1253 (11th Cir. 2009). The Supreme Court, in multiple cases, has ruled that\\n\\n§ 2 of the FAA overrides a state statute or common-law doctrine that attempts to\\n\\nundercut the enforceability of an arbitration agreement. See Kindred Nursing Ctrs.\\n\\nLtd. P’ship v. Clark, 137 S. Ct. 1421, 1425, 1426 (2017); DIRECTV, Inc. v.\\n\\nImburgia, 136 S. Ct. 463, 468, 471 (2015). The class action waiver here is not\\n\\ncontained in an arbitration agreement, so § 2 of the FAA does not stand in the way\\n\\nof enforcing Georgia’s public policy.\\n\\n The PLA and the GILA establish the Georgia Legislature’s intent to preserve\\n\\nclass actions as a remedy for those aggrieved by payday lenders. Enforcing the class\\n\\naction waiver here would undermine the purpose and spirit of Georgia’s statutory\\n\\nscheme. The class action waiver is therefore unenforceable, and the district court\\n\\ndid not err in denying the Oasis lenders’ motion to strike the plaintiffs’ class\\n\\nallegations.\\n\\n III\\n\\n We affirm the district court’s denial of the motion to dismiss and motion to\\n\\nstrike.\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 19\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434221/\", \"author_raw\": \"JORDAN, Circuit Judge:\"}]}"},"judge1":{"kind":"string","value":"TJOFLAT"},"judge2":{"kind":"string","value":"JORDAN"},"judge3":{"kind":"string","value":"SCHLESINGER"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"TJOFLAT\": \", Circuit\", \"JORDAN\": \", Circuit\", \"SCHLESINGER\": \", District\"}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4656968/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Lizzie Davis v. Oasis Legal Finance Operating Company, LLC\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-28\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nCase: 18-10526 Date Filed: 08/28/2019 Page: 1 of 19\n\n\n [PUBLISH]\n\n\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n ________________________\n\n No. 18-10526\n ________________________\n\n D.C. Docket No. 3:17-cv-00022-DHB-BKE\n\n\n\nLIZZIE DAVIS,\nindividually and on behalf of all others similarly situated,\nDENNIS GREEN,\nindividually and on behalf of all others similarly situated,\nJOHNNY MOODY,\nindividually and on behalf of all others similarly situated,\nJOHN SUBER,\nindividually and on behalf of all others similarly situated,\nSHIRLEY WILLIAMS,\nindividually and on behalf of all others similarly situated,\nPAMELA DAVIS,\nindividually and on behalf of all others similarly situated,\n\n Plaintiffs-Appellees,\n\nversus\n\nOASIS LEGAL FINANCE OPERATING COMPANY, LLC,\nOASIS LEGAL FINANCE, LLC,\nOASIS LEGAL FINANCE HOLDING COMPANY, LLC,\n\n Defendants-Appellants.\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 2 of 19\n\n\n _______________________\n\n Appeals from the United States District Court\n for the Southern District of Georgia\n ________________________\n\n (August 28, 2019)\n\nBefore TJOFLAT and JORDAN, Circuit Judges, and SCHLESINGER, District\nJudge.∗\n\nJORDAN, Circuit Judge:\n\n American courts have long refused to enforce contractual provisions that\n\ncontravene public policy. See, e.g., Marshall v. Baltimore and Ohio R.R., 57 U.S.\n\n314, 334 (1853) (“It is an undoubted principle of the common law that it will not\n\nlend its aid to enforce a contract to do an act that is illegal, or which is inconsistent\n\nwith sound morals or public policy. . . .”). In Georgia, “[n]o principle of\n\njurisprudence is better settled than this.” Glass v. Childs, 71 S.E. 920, 921 (Ga. Ct.\n\nApp. 1911). 1\n\n Courts have said that “[i]t is the duty of all courts of justice to keep their eye\n\nsteadily up on the interests of the public, . . . and when they find an action is founded\n\n\n\n∗The Honorable Harvey E. Schlesinger, United States District Judge for the Middle District of\nFlorida, sitting by designation.\n1\n For an interesting early Georgia case holding a contract void as against public policy, see\nChancely v. Bailey, 37 Ga. 532, 533, 541–42 (Ga. 1868) (holding that a contract purporting to pay\nMr. Chancely $2,500 if he served in Mr. Bailey’s place in the Army of the Confederate States\nduring the Civil War was void because Georgia did not have the right to forcibly secede from the\nUnion and Mr. Chancely could not contract to illegally engage in war against the United States).\n 2\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 3 of 19\n\n\nup on a claim injurious to the public . . . to give no countenance or assistance in foro\n\ncivili.” Elisha Greenhood, The Doctrine of Public Policy: Reduced to Rules 2 (1886)\n\n(quoting C.J. Wilmot’s Opinion in Low v. Peers, (1770) 97 Eng. Rep. 138 (Ex. Ch.)).\n\nOthers, however, have characterized the public policy defense as “a very unruly\n\nhorse, and when once you get astride it you never know where it will carry you.”\n\nRichardson v. Mellish, (1824) 130 Eng. Rep. 294, 303 (H.L.) (Burrough, J.).\n\n In this case, a class of borrowers filed suit in Georgia against their lenders,\n\nalleging that their loan agreements violated state usury laws. The lenders moved to\n\ndismiss the action based on the forum selection clause and class action waiver in the\n\nagreements. The district court concluded that both provisions were unenforceable\n\nas against Georgia public policy, and the lenders appealed.\n\n Following oral argument and a review of the relevant authorities, we agree\n\nwith the district court. Georgia’s Payday Lending Act and Industrial Loan Act\n\narticulate a clear public policy against enforcing forum selection clauses in payday\n\nloan agreements and in favor of preserving class actions as a remedy for those\n\naggrieved by predatory lenders. If Georgia’s public policy regarding payday lenders\n\nis a horse, as Justice Burrough suggested, it carries these borrowers safely to a\n\nGeorgia courthouse.\n\n\n\n\n 3\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 4 of 19\n\n\n I\n\n The plaintiffs entered into identical loan agreements with Oasis Legal\n\nFinance, LLC; Oasis Legal Finance Operating Company, LLC; and Oasis Legal\n\nFinance Holding Company, LLC ( the “Oasis lenders”). The loans generally\n\namounted to less than $3,000 and were to be repaid from any recoveries that the\n\nplaintiffs received in their separate personal injury lawsuits. The plaintiffs’\n\nobligations to repay the loans were therefore contingent on success in the underlying\n\nlawsuits.\n\n In February of 2017, the plaintiffs filed a class action complaint against the\n\nOasis lenders in Georgia state court, alleging that the loan agreements violated\n\nGeorgia’s Payday Lending Act, O.C.G.A. § 16-17-1 et seq., Industrial Loan Act,\n\nO.C.G.A. § 7-3-1 et seq., and usury laws, O.C.G.A. § 7-4-18. The Oasis lenders\n\nremoved the suit to federal court and moved to dismiss the complaint under Federal\n\nRule of Civil Procedure 12(b)(6) and to strike the plaintiffs’ class allegations under\n\nFederal Rule of Civil Procedure 12(f). The Oasis lenders argued, among other\n\nthings, that the loan agreements’ forum selection clause required the plaintiffs to\n\nbring suit in Illinois, and that the class action waiver barred their ability to file a class\n\n\n\n\n 4\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 5 of 19\n\n\naction. The plaintiffs responded that these provisions violated Georgia public policy\n\nand, therefore, were unenforceable. 2\n\n Applying Georgia law, the district court rejected both of the arguments made\n\nby the Oasis lenders and held that the forum section clause and class action waiver\n\nwere unenforceable. See Davis v. Oasis Legal Fin. Operating Co., No. CV 317-022,\n\n2017 WL 5490919, at *4–5 (S.D. Ga. Nov. 15, 2017) The district court concluded\n\nthat “the enforcement of forum selections clauses in payday lending contracts would\n\ncontravene [Georgia’s] public policy” as established by the Payday Lending Act. Id.\n\nat *4. It explained that “[c]ertain payday lenders have attempted to use forum\n\nselection clauses contained in payday loan documents in order to avoid the courts of\n\nthe State of Georgia, and the General Assembly has determined that such practices\n\nare unconscionable and should be prohibited.” Id. at *3 (quoting § 16-17-1(d)). The\n\ndistrict court similarly ruled that the class action waiver contravened public policy\n\nbecause, when the Georgia Legislature enacted the PLA and the GILA, it expressly\n\nincluded class actions as a remedy for those aggrieved by payday lenders. The\n\ndistrict court reasoned that the Georgia Legislature would not create such a remedy\n\n\n2\n The loan agreements’ forum selection clause states: “THE PARTIES HEREBY\nIRREVOCABLY AND UNCONDITIONALLY CONSENT TO SUBMIT TO THE EXCLUSIVE\nJURISDICTION OF THE CIRCUIT COURT OF COOK COUNTY, ILLINOIS FOR ANY\nDISPUTES . . . .” D.E. 10-8 at 25 ¶6.5 (capitalization in original). The loan agreements’ class\naction waiver states: “THE PARTIES HEREBY . . . WAIVE ANY RIGHT TO CONSOLIDATE\nOR TO HAVE HANDLED AS A CLASS ACTION ANY PROCEEDING ON ANY LAWSUIT\n. . . .” D.E. 10-8 at 25 ¶6.6 (capitalization in original).\n\n 5\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 6 of 19\n\n\nand then allow lenders to “effectively wipe away this consumer protection with a\n\nwaiver in a single paragraph of a six-page, single-spaced agreement.” Id. at *5.\n\n The district court certified its decision for interlocutory review, and the Oasis\n\nlenders petitioned us to consider whether the district court erred in concluding that\n\nthe loan agreements’ forum selection clause and class action waiver are\n\nunenforceable. We agreed to consider both issues, and now affirm. 3\n\n II\n\n The enforceability of a forum selection clause is a question of law that we\n\nreview de novo. See Rucker v. Oasis Legal Fin., L.L.C., 632 F.3d 1231, 1235 (11th\n\nCir. 2011). The same plenary standard governs the enforceability of a class action\n\nwaiver. See Cruz v. Cingular Wireless, LLC, 648 F.3d 1205, 1210 (11th Cir. 2011).\n\n A\n\n A forum selection clause, when properly bargained for, “protects [the parties’]\n\nlegitimate expectations and furthers the vital interest of the justice system.” Atl.\n\nMarine Constr. Co. v. U.S. Dist. Ct. for the W. Dist. of Tex., 571 U.S. 49, 63 (2013)\n\n\n\n\n3\n We recognize that in Ruth v. Cherokee Funding, LLC, 820 S.E.2d 704, 710 (Ga. 2018), the\nGeorgia Supreme Court held that loans in which the borrowers’ obligations to repay are contingent\nupon success in an underlying personal-injury lawsuit are not “loans” under the Georgia Payday\nLending Act and the Industrial Loan Act. The Oasis lenders, however, concede that they did not\nargue in the district court that the loans at issue here are not loans under the PLA and GILA. [See\nO.A. Recording at 5:16] That argument, therefore, is not properly before us. See Access Now,\nInc. v. Sw. Airlines Co., 385 F.3d 1324, 1331 (11th Cir. 2004).\n\n\n 6\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 7 of 19\n\n\n(quotation marks omitted). Such clauses “should be given controlling weight in all\n\nbut the most exceptional cases.” Id. See also M/S Bremen v. Zapata Off-Shore Co.,\n\n407 U.S. 1, 15 (1972) (stating that a forum selection clause “should control absent a\n\nstrong showing that it should be set aside”).\n\n In Bremen, 407 U.S. at 15–18, the Supreme Court identified four grounds on\n\nwhich a court can refuse to enforce an otherwise-valid forum selection clause. These\n\ninclude if “(1) [its] formation was induced by fraud or overreaching; (2) the plaintiff\n\neffectively would be deprived of its day in court because of the inconvenience or\n\nunfairness of the chosen forum; (3) the fundamental unfairness of the chosen law\n\nwould deprive the plaintiff of a remedy; or (4) enforcement of the [forum selection\n\nclause] would contravene a strong public policy.” Lipcon v. Underwriters at\n\nLloyd’s, London, 148 F.3d 1285, 1292 (11th Cir. 1998) (citing Bremen, 407 U.S. at\n\n15–18). Here, we consider the fourth ground—when enforcing the forum selection\n\nclause “would contravene a strong public policy of the forum in which suit is\n\nbrought, whether declared by statute or judicial decision.” Bremen, 407 U.S. at 15.\n\n “Public policy is an amorphous concept . . . . Accordingly, it has been held\n\nthat, the delicate and undefined power of courts to declare a contract void as\n\ncontravening public policy should be exercised with great caution, and only in cases\n\nfree from substantial doubt.” Dep’t of Transp. v. Brooks, 328 S.E.2d 705, 713 (Ga.\n\n1985) (alterations adopted and quotations omitted). In Georgia, a contractual\n\n\n 7\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 8 of 19\n\n\nprovision generally does not violate public policy unless the Legislature has declared\n\nit so or enforcement of the provision would flout the very purpose of the law. See\n\nid.; Robinson v. Reynolds, 21 S.E.2d 214, 215 (Ga. 1942) (“Contracts that obviously\n\nand directly tend in a marked degree to bring about results that the law seeks to\n\nprevent cannot be made the ground of a successful suit . . . [and are] void as against\n\npublic policy.”).\n\n Although amorphous in concept, Georgia’s public policy bar is built on a solid\n\nfoundation—the Georgia Constitution and state statutes. “[T]hese are the sources\n\nthat are first to be considered and that often may be conclusive.” Strickland v. Gulf\n\nLife Ins. Co., 242 S.E.2d 148, 151 (Ga. 1978) (quotation omitted). “Statutes, of\n\ncourse, are perhaps the clearest expressions of the public policy of [Georgia].” Dove\n\nv. Dove, 680 S.E.2d 839, 842 (Ga. 2009). See also Sonja Laesen, 7 Georgia\n\nJurisprudence § 3:3 (updated 2019) (“The only authentic and admissible evidence of\n\npublic policy of a state is its constitution, laws, and judicial decisions.”). For\n\nexample, O.C.G.A. § 13-8-2 provides that “[a] contract which is against the policy\n\nof the law cannot be enforced,” and lists particular agreements that are unenforceable\n\nas against public policy. Although § 13-8-2 does not address the specific provisions\n\nat issue in this case, the statute makes clear that its list is “nonexclusive” and Georgia\n\ncourts have held that they “are not restricted by this list in voiding contracts that are\n\nagainst public policy.” Edwards v. Grapefields, Inc., 599 S.E.2d 489, 493 (Ga. Ct.\n\n\n 8\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 9 of 19\n\n\nApp. 2004). We may therefore look to other Georgia statutes to determine whether\n\nthe state has a strong public policy against enforcing forum selection clauses in favor\n\nof out-of-state payday lenders.\n\n The Georgia Payday Lending Act, O.C.G.A. § 16-17-1 et seq., states in\n\nrelevant part as follows:\n\n A payday lender shall not . . . nor shall the loan contract\n designate a court for the resolution of disputes concerning\n the contract other than a court of competent jurisdiction in\n and for the county in which the borrower resides or the\n loan office is located.\n\n§ 16-17-2(c)(1). The Georgia Legislature explained the PLA’s prohibition on out-\n\nof-state forum selection clauses in this way:\n\n Certain payday lenders have attempted to use forum\n selection clauses contained in payday loan documents in\n order to avoid the courts of the State of Georgia, and the\n General Assembly has determined that such practices are\n unconscionable and should be prohibited.\n\n§ 16-17-1(d) (emphasis added).\n\n Based on this language, the Georgia Supreme Court ruled that the PLA\n\nestablished a clear public policy against “payday lenders [that] have attempted to\n\nskirt the laws of Georgia by use of forum selection clauses.” W. Sky Fin., LLC v.\n\nGeorgia, 793 S.E.2d 357, 363 (Ga. 2016). In ruling that the loan agreements’ forum\n\nselection clause is unenforceable in this case, the district court found that § 16-17-\n\n1(d) and § 16-17-2(c)(1) were “conclusive” as to Georgia’s public policy. See\n\n\n 9\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 10 of 19\n\n\nStrickland, 242 S.E.2d at 151. On appeal, the Oasis lenders argue that two other\n\nsubsections of the PLA support the opposite conclusion.4\n\n First, the Oasis lenders assert that the word “county” in § 16-17-2(c)(1) is\n\nunqualified, meaning that by permitting parties to select a forum “in and for the\n\ncounty in which . . . the loan office is located” the PLA allows parties to select a\n\ncounty outside Georgia for litigation if it is the county where the lenders’ loan office\n\nis located. The parties could therefore select as a forum the county where the Oasis\n\nlenders operate their loan office—Cook County, Illinois. This argument has some\n\nsuperficial appeal, but the district court rejected it, concluding that the word\n\n“county” in § 16-17-2(c)(1) refers only to Georgia counties because that subsection\n\nis a venue provision through which the Legislature intended to allow parties to select\n\nfora within Georgia while prohibiting outbound forum-selection clauses. The Oasis\n\nlenders contend that the district court erred by reading in the word “Georgia” in front\n\nof “county” in § 16-17-2(c)(1). We think the district court got it right.\n\n Georgia venue provisions commonly use the term “county” or “counties” in\n\nrefence to Georgia counties, without explicitly saying so. See Ga. Const., Art. VI, §\n\n2, ¶¶ III, IV, VI; O.C.G.A. § 9-10-93. For example, Georgia’s Constitution provides\n\n\n4\n The Oasis lenders also claim (in a footnote of their reply brief) that interpreting the PLA to require\nout-of-state lenders to defend these kinds of lawsuits in Georgia may violate the dormant\nCommerce Clause. The Oasis lenders, however, did not present this argument in the district court\nor in their initial brief, and they also failed to support the argument with legal authorities. We\ntherefore do not address it. See Sapuppo v. Allstate Floridian Ins. Co., 739 F.3d 678, 681 (11th\nCir. 2014); Lovett v. Ray, 327 F.3d 1181, 1183 (11th Cir. 2003).\n 10\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 11 of 19\n\n\nthat “joint trespassers residing in different counties may be tried in either county,”\n\nGa. Const., Art. VI, § 2, ¶ IV, and Georgia courts have interpreted that provision to\n\nallow “joint tortfeasor residents of different Georgia counties [to] be sued in either\n\ncounty.” Robinson v. Star Gas of Hawkinsville, Inc., 533 S.E.2d 97, 98 (Ga. Ct. App.\n\n2000) (emphasis added). See also Cartwright v. Fuji Photo Film U.S.A., Inc., 720\n\nS.E.2d 200, 204 n.4, 206 (Ga. Ct. App. 2011) (“[T]he proper venue for a suit is in\n\nwhich Georgia county the defendant resides[.]”); Cunningham v. Estate of\n\nCunningham, 697 S.E.2d 280, 282 (Ga. Ct. App. 2010) (“Georgia’s constitutional\n\nvenue provision . . . has no application to a suit against a nonresident . . . because in\n\nsuch case there is no Georgia county in which the nonresident resides.”). The same\n\ninterpretation applies to Georgia statutes with similar language. See Lloyd Adams,\n\nInc. v. Liberty Mut. Ins. Co., 10 S.E.2d 46, 48 (Ga. 1940) (holding that a similar\n\nvenue provision in Georgia’s Nonresident Motorist Act, now O.C.G.A. § 40-12-3,\n\nafforded venue in “any county in the State”). This makes sense given that the\n\nGeorgia Legislature generally has authority to establish venue rules within Georgia.\n\n Given the PLA’s condemnation of out-of-state lenders using forum selection\n\nclauses to avoid Georgia courts in § 16-17-1(d), it would make little sense to\n\nconclude that the Georgia Legislature intended the word “county” in the next\n\nsubsection to include places like Cook County, Illinois. See West v. City of Albany,\n\n797 S.E.2d 809, 811 (2017) (“We do not limit our consideration to the words of one\n\n\n 11\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 12 of 19\n\n\nsubsection of a statute alone, but consider a particular provision in the context of the\n\nstatute as a whole . . . .”). The Oasis lenders’ argument—that by using the term\n\n“county” instead of “Georgia county” in § 16-17-2(c)(1), the Legislature opened a\n\nback door for out-of-state forum selection clauses—would effectively “hide\n\nelephants in mouseholes.” Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 468\n\n(2001). Moreover, it would also render the PLA’s prohibition of certain forum\n\nselection clauses in § 16-17-1(d) “meaningless, redundant, or mere surplusage.”\n\nUnited States v. Canals-Jimenez, 943 F.2d 1284, 1287 (11th Cir. 1991). We disfavor\n\nsuch interpretations. See Cadwell v. Kaufman, Englett & Lynd, PPLC, 886 F.3d\n\n1153, 1159 (11th Cir. 2018).\n\n Next, the Oasis lenders contend that the PLA does not apply to loan\n\nagreements between a Georgia borrower and an out-of-state lender because § 16-17-\n\n1(d) states that “[p]ayday lending involves relatively small loans and does not\n\nencompass loans that involve interstate commerce.” We disagree for a few reasons.\n\nFirst, the Oasis lenders’ arguments are mutually exclusive. Arguing that the PLA\n\ndoes not apply to loans by an out-of-state lender contradicts the lenders’ argument\n\nthat the Georgia Legislature meant for the term “county” in § 16-17-2(c)(1) to\n\ninclude the county where an out-of-state lender maintains its home office. See\n\nFootstar, Inc. v. Liberty Mut. Ins. Co., 637 S.E.2d 692, 695 (2006) (“[C]ourts should\n\nconstrue a statute to give sensible and intelligent effect to all of its provisions . . .”).\n\n\n 12\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 13 of 19\n\n\nSecond, other provisions of the PLA make clear that the Act governs “any business”\n\nthat “consists in whole or in part of making . . . loans of $3,000.00 or less” unless\n\nthose entities are specifically exempted. See W. Sky Fin., 793 S.E.2d at 363 (quoting\n\n§ 16-17-2(a)). Out-of-state lenders are not exempt. See § 16-17-2(a)(1)–(4). Third,\n\nexcluding loans involving out-of-state lenders from coverage—as the Oasis lenders\n\nargue—would render the PLA’s prohibition of out-of-state forum selection clauses\n\nmeaningless. See Cadwell, 886 F.3d at 1159; Canals-Jimenez, 943 F.2d at 1287.\n\n In Western Sky, 793 S.E.2d at 363, the defendant argued that the PLA’s\n\n“interstate commerce” language rendered the statute inapplicable to loans involving\n\nout-of-state lenders. The Georgia Supreme Court “reject[ed] the notion that this\n\nphrase was meant to exclude loans that involve interstate commerce from the scope\n\nof the [PLA].” Id. “If that were so, the [PLA] would be virtually meaningless\n\nbecause it would prohibit nothing.” Id. See also id. at 364 (“If, as Defendants argue,\n\nthe [PLA] is limited only to transactions that do not involve interstate commerce, the\n\n[PLA] would effectively prohibit nothing at all.”). The Georgia Supreme Court went\n\non to articulate how excluding loans involving an out-of-state lender would\n\nundermine the entire purpose of the PLA.\n\n Given the clear and unambiguous scope of the [PLA] as a\n whole, to interpret that phrase as a definitional limitation\n upon payday lending and thereby exempt loans that\n involve interstate commerce from the prohibitions of the\n [PLA] would create such a contradiction and absurdity as\n\n\n 13\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 14 of 19\n\n\n to demonstrate that the legislature did not mean it to\n create such a limitation.\n\nId. at 365 (emphasis added).\n\n Based on our review, Georgia statutes establish a clear public policy against\n\nout-of-state lenders using forum selection clauses to avoid litigation in Georgia\n\ncourts. See O.C.G.A. §§ 16-17-1(d), 2(c)(1). Enforcing a forum selection clause\n\nlike the ones here “would [therefore] contravene a strong public policy of the forum\n\nin which suit is brought,” and the district court correctly denied the Oasis lenders’\n\nmotion to dismiss on that ground. See Bremen, 407 U.S. at 15.5\n\n B\n\n The plaintiffs argue, and the district court concluded, that the class action\n\nwaivers contained in the loan agreements also contravene the purpose of the PLA\n\nand the Georgia Industrial Loan Act, O.C.G.A. § 7-3-29. We start with the text of\n\nboth acts. See Dove, 680 S.E.2d at 842.\n\n The PLA contains a provision stating that “a civil action may be brought on\n\nbehalf of an individual borrower or on behalf of an ascertainable class of borrowers.”\n\n\n\n\n5\n Because we conclude that Georgia statutes establish a public policy against the forum selection\nclause at issue, we are not bound by our opinion in Rucker, 632 F.3d at 1238. In that case, we held\nthat an identical forum selection clause was valid and enforceable under Alabama law. See id. Our\ndecision to invalidate the forum selection clause in this case depends entirely on Georgia law.\n 14\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 15 of 19\n\n\n§ 16-7-3. The GILA provides that “a claim for violation of this chapter against an\n\nunlicensed lender may be asserted in a class action.” § 7-3-29(e). 6\n\n The district court concluded that, in passing these laws, the Georgia\n\nLegislature “expressly contemplated a specific remedy—[a] class action—for\n\npersons aggrieved by predatory lending . . . [and] did not expressly create the class\n\naction remedy so that predatory lenders could effectively wipe away this consumer\n\nprotection with a waiver . . . .” Davis, 2017 WL 5490919, at *5. On appeal, the\n\nOasis lenders contend that the district court erred by not considering whether the\n\nprovision was procedurally or substantively unconscionable. They also argue that\n\nneither the PLA nor the GILA prohibit class action waivers or create a statutory right\n\nto pursue a class action.\n\n These arguments miss the point. The district court’s ruling flowed from its\n\nconclusion that enforcing class action waivers in this context would allow payday\n\nlenders to eliminate a remedy that was expressly contemplated by the Georgia\n\nLegislature, and thereby undermine the purpose of the statutory scheme. That\n\nconclusion, if correct, renders the class action waiver unenforceable under Georgia\n\nlaw regardless of whether the provision is also procedurally or substantively\n\nunconscionable. See Brookfield Country Club, Inc. v. St. James-Brookfield, LLC,\n\n\n6\n We note that this subsection of the GILA applies to “unlicensed lenders,” but the district court\nquoted this provision in ruling that the class action waiver is unenforceable, and the Oasis lenders\ndo not argue that it does not apply. See Davis, 2017 WL 5490919, at *5.\n 15\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 16 of 19\n\n\n696 S.E.2d 663, 667 (Ga. 2010) (refusing to “enforce a contractual provision which\n\ncontravenes the statutory law of this state”); Terry v. St. Farm Fire & Cas. Ins. Co.,\n\n504 S.E.2d 194, 195 (Ga. 1998) (considering whether “[t]he underlying purpose” of\n\nthe statute would be “contravened by enforcement of the specific terms of the\n\nagreement”). Accord Atl. Flooring Design Ctrs., Inc. v. R.G. Williams Const., Inc.,\n\n773 S.E.2d 868, 870 (Ga. Ct. App. 2015) (voiding a contractual provision because it\n\nwould frustrate the public policy expressed in the Georgia Arbitration Code);\n\nLangford v. Royal Indem. Co., 430 S.E.2d 98, 102 (Ga. Ct. App. 1993) (“A contract\n\nprovision normally should not be enforced where it conflicts with the general policy\n\nand spirit of the statute which governs it, although there may be no literal conflict;\n\nthat is, it makes no difference whether the statutory prohibition or command is\n\nexpressed or implied.”).\n\n The Oasis lenders may be correct in arguing that, when Georgia courts address\n\nwhether a contractual provision is substantively unconscionable, they also “consider\n\n‘the commercial reasonableness of the contract terms, the purpose and effect of the\n\nterms, the allocation of the risks between the parties, and similar public policy\n\nconcerns.’” Jenkins v. First Am. Cash Advance of Ga., LLC, 400 F.3d 868, 876 (11th\n\nCir. 2005) (quoting NEC Techs., Inc. v. Nelson, 478 S.E.2d 769, 772 (Ga. 1996)).\n\nThe public policy bar, however, remains an independent basis to hold a contractual\n\nprovision unenforceable. See Glosser v. Powers, 71 S.E.2d 230, 231 (Ga. 1952)\n\n\n 16\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 17 of 19\n\n\n(“Even in the absence of fraud in the procurement of a contract, a contract which is\n\nagainst the policy of the law is void and unenforceable.”) A hornbook example of\n\nthe public policy defense is that a court will not enforce a contractual provision that\n\nis illegal regardless of whether its obligations are mutual, its terms are conspicuous,\n\nand the parties are well represented. See 17A Am. Jur. 2d Contracts § 247 (updated\n\n2019). See also Stoudemire v. HSBC Bank USA, N.A., 776 S.E.2d 483, 484 (Ga. Ct.\n\nApp. 2015) (highlighting the difference between “[c]ontracts to do illegal or immoral\n\nthings, contracts against public policy, and gambling contracts,” which are legally\n\nvoid, and “fraudulent contracts and contracts entered under duress,” which are\n\nvoidable by the injured party) (citations omitted).\n\n The district court did not, and needed not, conclude that the PLA or the GILA\n\nexpressly prohibited class action waivers or created a statutory right to pursue a class\n\naction. A contractual provision need not “literal[ly] conflict” with Georgia law to\n\ncontravene public policy. See Langford, 430 S.E.2d at 102; Robinson, 21 S.E.2d at\n\n215. Instead, the district court concluded that enforcement of the class action\n\nwaivers in this context would eliminate a remedy contemplated by the Georgia\n\nLegislature and undermine the purpose of the PLA and the GILA. Davis, 2017 WL\n\n5490919, at *5.\n\n To that point, the Oasis lenders make one additional argument warranting\n\nfurther discussion—that the PLA’s fee-shifting provision eliminates the risk that\n\n\n 17\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 18 of 19\n\n\nenforcing the class action waiver would effectively prevent the plaintiffs from\n\nlitigating their claims. In past cases, we have held that certain class action waivers\n\nwere not unconscionable, in part because the agreements at issue contained a fee-\n\nshifting provision that, in practice, permitted the plaintiffs to pursue their claims\n\nindividually. See Jenkins, 400 F.3d at 877–80; Randolph v. Green Tree Fin. Corp.,\n\n244 F.3d 814, 816–18 (11th Cir. 2001). Those cases, however, concerned whether\n\nthe class action waivers were unconscionable because enforcement would prevent\n\nthe plaintiffs from having their day in court. See Jenkins, 400 F.3d at 877–78. As\n\ndiscussed above, the district court here did not conclude that the class action waiver\n\nwas procedurally or substantively unconscionable—or that the plaintiffs in this case\n\ncould never litigate their claims individually—but relied on Georgia’s public policy\n\nas expressed in the text and purpose of the PLA and the GILA.\n\n The Oasis lenders point out that in Jenkins, 400 F.3d at 877–80, and Bowen v.\n\nFirst Family Financial Services, Inc., 233 F.3d 1331, 1336 (11th Cir. 2000), we held\n\nthat class action waivers in arbitration clauses were not void as against public policy.\n\nBut those cases do not call for a different result here. Neither Jenkins nor Bowen\n\nanalyzed the PLA, the GILA, or Georgia’s public policy against class action waivers\n\nin payday loan agreements. Instead, those cases concerned class action waivers in\n\narbitration agreements—where the Federal Arbitration Act “create[s] a strong\n\nfederal policy in favor of arbitration.” Picard v. Credit Solutions, Inc., 564 F.3d\n\n\n 18\n\f Case: 18-10526 Date Filed: 08/28/2019 Page: 19 of 19\n\n\n1249, 1253 (11th Cir. 2009). The Supreme Court, in multiple cases, has ruled that\n\n§ 2 of the FAA overrides a state statute or common-law doctrine that attempts to\n\nundercut the enforceability of an arbitration agreement. See Kindred Nursing Ctrs.\n\nLtd. P’ship v. Clark, 137 S. Ct. 1421, 1425, 1426 (2017); DIRECTV, Inc. v.\n\nImburgia, 136 S. Ct. 463, 468, 471 (2015). The class action waiver here is not\n\ncontained in an arbitration agreement, so § 2 of the FAA does not stand in the way\n\nof enforcing Georgia’s public policy.\n\n The PLA and the GILA establish the Georgia Legislature’s intent to preserve\n\nclass actions as a remedy for those aggrieved by payday lenders. Enforcing the class\n\naction waiver here would undermine the purpose and spirit of Georgia’s statutory\n\nscheme. The class action waiver is therefore unenforceable, and the district court\n\ndid not err in denying the Oasis lenders’ motion to strike the plaintiffs’ class\n\nallegations.\n\n III\n\n We affirm the district court’s denial of the motion to dismiss and motion to\n\nstrike.\n\n AFFIRMED.\n\n\n\n\n 19\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Lizzie Davis v. Oasis Legal Finance Operating Company, LLC\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-28\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 18-10526 Date Filed: 08/28/2019 Page: 1 of 19\\n\\n\\n [PUBLISH]\\n\\n\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 18-10526\\n ________________________\\n\\n D.C. Docket No. 3:17-cv-00022-DHB-BKE\\n\\n\\n\\nLIZZIE DAVIS,\\nindividually and on behalf of all others similarly situated,\\nDENNIS GREEN,\\nindividually and on behalf of all others similarly situated,\\nJOHNNY MOODY,\\nindividually and on behalf of all others similarly situated,\\nJOHN SUBER,\\nindividually and on behalf of all others similarly situated,\\nSHIRLEY WILLIAMS,\\nindividually and on behalf of all others similarly situated,\\nPAMELA DAVIS,\\nindividually and on behalf of all others similarly situated,\\n\\n Plaintiffs-Appellees,\\n\\nversus\\n\\nOASIS LEGAL FINANCE OPERATING COMPANY, LLC,\\nOASIS LEGAL FINANCE, LLC,\\nOASIS LEGAL FINANCE HOLDING COMPANY, LLC,\\n\\n Defendants-Appellants.\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 2 of 19\\n\\n\\n _______________________\\n\\n Appeals from the United States District Court\\n for the Southern District of Georgia\\n ________________________\\n\\n (August 28, 2019)\\n\\nBefore TJOFLAT and JORDAN, Circuit Judges, and SCHLESINGER, District\\nJudge.∗\\n\\nJORDAN, Circuit Judge:\\n\\n American courts have long refused to enforce contractual provisions that\\n\\ncontravene public policy. See, e.g., Marshall v. Baltimore and Ohio R.R., 57 U.S.\\n\\n314, 334 (1853) (“It is an undoubted principle of the common law that it will not\\n\\nlend its aid to enforce a contract to do an act that is illegal, or which is inconsistent\\n\\nwith sound morals or public policy. . . .”). In Georgia, “[n]o principle of\\n\\njurisprudence is better settled than this.” Glass v. Childs, 71 S.E. 920, 921 (Ga. Ct.\\n\\nApp. 1911). 1\\n\\n Courts have said that “[i]t is the duty of all courts of justice to keep their eye\\n\\nsteadily up on the interests of the public, . . . and when they find an action is founded\\n\\n\\n\\n∗The Honorable Harvey E. Schlesinger, United States District Judge for the Middle District of\\nFlorida, sitting by designation.\\n1\\n For an interesting early Georgia case holding a contract void as against public policy, see\\nChancely v. Bailey, 37 Ga. 532, 533, 541–42 (Ga. 1868) (holding that a contract purporting to pay\\nMr. Chancely $2,500 if he served in Mr. Bailey’s place in the Army of the Confederate States\\nduring the Civil War was void because Georgia did not have the right to forcibly secede from the\\nUnion and Mr. Chancely could not contract to illegally engage in war against the United States).\\n 2\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 3 of 19\\n\\n\\nup on a claim injurious to the public . . . to give no countenance or assistance in foro\\n\\ncivili.” Elisha Greenhood, The Doctrine of Public Policy: Reduced to Rules 2 (1886)\\n\\n(quoting C.J. Wilmot’s Opinion in Low v. Peers, (1770) 97 Eng. Rep. 138 (Ex. Ch.)).\\n\\nOthers, however, have characterized the public policy defense as “a very unruly\\n\\nhorse, and when once you get astride it you never know where it will carry you.”\\n\\nRichardson v. Mellish, (1824) 130 Eng. Rep. 294, 303 (H.L.) (Burrough, J.).\\n\\n In this case, a class of borrowers filed suit in Georgia against their lenders,\\n\\nalleging that their loan agreements violated state usury laws. The lenders moved to\\n\\ndismiss the action based on the forum selection clause and class action waiver in the\\n\\nagreements. The district court concluded that both provisions were unenforceable\\n\\nas against Georgia public policy, and the lenders appealed.\\n\\n Following oral argument and a review of the relevant authorities, we agree\\n\\nwith the district court. Georgia’s Payday Lending Act and Industrial Loan Act\\n\\narticulate a clear public policy against enforcing forum selection clauses in payday\\n\\nloan agreements and in favor of preserving class actions as a remedy for those\\n\\naggrieved by predatory lenders. If Georgia’s public policy regarding payday lenders\\n\\nis a horse, as Justice Burrough suggested, it carries these borrowers safely to a\\n\\nGeorgia courthouse.\\n\\n\\n\\n\\n 3\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 4 of 19\\n\\n\\n I\\n\\n The plaintiffs entered into identical loan agreements with Oasis Legal\\n\\nFinance, LLC; Oasis Legal Finance Operating Company, LLC; and Oasis Legal\\n\\nFinance Holding Company, LLC ( the “Oasis lenders”). The loans generally\\n\\namounted to less than $3,000 and were to be repaid from any recoveries that the\\n\\nplaintiffs received in their separate personal injury lawsuits. The plaintiffs’\\n\\nobligations to repay the loans were therefore contingent on success in the underlying\\n\\nlawsuits.\\n\\n In February of 2017, the plaintiffs filed a class action complaint against the\\n\\nOasis lenders in Georgia state court, alleging that the loan agreements violated\\n\\nGeorgia’s Payday Lending Act, O.C.G.A. § 16-17-1 et seq., Industrial Loan Act,\\n\\nO.C.G.A. § 7-3-1 et seq., and usury laws, O.C.G.A. § 7-4-18. The Oasis lenders\\n\\nremoved the suit to federal court and moved to dismiss the complaint under Federal\\n\\nRule of Civil Procedure 12(b)(6) and to strike the plaintiffs’ class allegations under\\n\\nFederal Rule of Civil Procedure 12(f). The Oasis lenders argued, among other\\n\\nthings, that the loan agreements’ forum selection clause required the plaintiffs to\\n\\nbring suit in Illinois, and that the class action waiver barred their ability to file a class\\n\\n\\n\\n\\n 4\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 5 of 19\\n\\n\\naction. The plaintiffs responded that these provisions violated Georgia public policy\\n\\nand, therefore, were unenforceable. 2\\n\\n Applying Georgia law, the district court rejected both of the arguments made\\n\\nby the Oasis lenders and held that the forum section clause and class action waiver\\n\\nwere unenforceable. See Davis v. Oasis Legal Fin. Operating Co., No. CV 317-022,\\n\\n2017 WL 5490919, at *4–5 (S.D. Ga. Nov. 15, 2017) The district court concluded\\n\\nthat “the enforcement of forum selections clauses in payday lending contracts would\\n\\ncontravene [Georgia’s] public policy” as established by the Payday Lending Act. Id.\\n\\nat *4. It explained that “[c]ertain payday lenders have attempted to use forum\\n\\nselection clauses contained in payday loan documents in order to avoid the courts of\\n\\nthe State of Georgia, and the General Assembly has determined that such practices\\n\\nare unconscionable and should be prohibited.” Id. at *3 (quoting § 16-17-1(d)). The\\n\\ndistrict court similarly ruled that the class action waiver contravened public policy\\n\\nbecause, when the Georgia Legislature enacted the PLA and the GILA, it expressly\\n\\nincluded class actions as a remedy for those aggrieved by payday lenders. The\\n\\ndistrict court reasoned that the Georgia Legislature would not create such a remedy\\n\\n\\n2\\n The loan agreements’ forum selection clause states: “THE PARTIES HEREBY\\nIRREVOCABLY AND UNCONDITIONALLY CONSENT TO SUBMIT TO THE EXCLUSIVE\\nJURISDICTION OF THE CIRCUIT COURT OF COOK COUNTY, ILLINOIS FOR ANY\\nDISPUTES . . . .” D.E. 10-8 at 25 ¶6.5 (capitalization in original). The loan agreements’ class\\naction waiver states: “THE PARTIES HEREBY . . . WAIVE ANY RIGHT TO CONSOLIDATE\\nOR TO HAVE HANDLED AS A CLASS ACTION ANY PROCEEDING ON ANY LAWSUIT\\n. . . .” D.E. 10-8 at 25 ¶6.6 (capitalization in original).\\n\\n 5\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 6 of 19\\n\\n\\nand then allow lenders to “effectively wipe away this consumer protection with a\\n\\nwaiver in a single paragraph of a six-page, single-spaced agreement.” Id. at *5.\\n\\n The district court certified its decision for interlocutory review, and the Oasis\\n\\nlenders petitioned us to consider whether the district court erred in concluding that\\n\\nthe loan agreements’ forum selection clause and class action waiver are\\n\\nunenforceable. We agreed to consider both issues, and now affirm. 3\\n\\n II\\n\\n The enforceability of a forum selection clause is a question of law that we\\n\\nreview de novo. See Rucker v. Oasis Legal Fin., L.L.C., 632 F.3d 1231, 1235 (11th\\n\\nCir. 2011). The same plenary standard governs the enforceability of a class action\\n\\nwaiver. See Cruz v. Cingular Wireless, LLC, 648 F.3d 1205, 1210 (11th Cir. 2011).\\n\\n A\\n\\n A forum selection clause, when properly bargained for, “protects [the parties’]\\n\\nlegitimate expectations and furthers the vital interest of the justice system.” Atl.\\n\\nMarine Constr. Co. v. U.S. Dist. Ct. for the W. Dist. of Tex., 571 U.S. 49, 63 (2013)\\n\\n\\n\\n\\n3\\n We recognize that in Ruth v. Cherokee Funding, LLC, 820 S.E.2d 704, 710 (Ga. 2018), the\\nGeorgia Supreme Court held that loans in which the borrowers’ obligations to repay are contingent\\nupon success in an underlying personal-injury lawsuit are not “loans” under the Georgia Payday\\nLending Act and the Industrial Loan Act. The Oasis lenders, however, concede that they did not\\nargue in the district court that the loans at issue here are not loans under the PLA and GILA. [See\\nO.A. Recording at 5:16] That argument, therefore, is not properly before us. See Access Now,\\nInc. v. Sw. Airlines Co., 385 F.3d 1324, 1331 (11th Cir. 2004).\\n\\n\\n 6\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 7 of 19\\n\\n\\n(quotation marks omitted). Such clauses “should be given controlling weight in all\\n\\nbut the most exceptional cases.” Id. See also M/S Bremen v. Zapata Off-Shore Co.,\\n\\n407 U.S. 1, 15 (1972) (stating that a forum selection clause “should control absent a\\n\\nstrong showing that it should be set aside”).\\n\\n In Bremen, 407 U.S. at 15–18, the Supreme Court identified four grounds on\\n\\nwhich a court can refuse to enforce an otherwise-valid forum selection clause. These\\n\\ninclude if “(1) [its] formation was induced by fraud or overreaching; (2) the plaintiff\\n\\neffectively would be deprived of its day in court because of the inconvenience or\\n\\nunfairness of the chosen forum; (3) the fundamental unfairness of the chosen law\\n\\nwould deprive the plaintiff of a remedy; or (4) enforcement of the [forum selection\\n\\nclause] would contravene a strong public policy.” Lipcon v. Underwriters at\\n\\nLloyd’s, London, 148 F.3d 1285, 1292 (11th Cir. 1998) (citing Bremen, 407 U.S. at\\n\\n15–18). Here, we consider the fourth ground—when enforcing the forum selection\\n\\nclause “would contravene a strong public policy of the forum in which suit is\\n\\nbrought, whether declared by statute or judicial decision.” Bremen, 407 U.S. at 15.\\n\\n “Public policy is an amorphous concept . . . . Accordingly, it has been held\\n\\nthat, the delicate and undefined power of courts to declare a contract void as\\n\\ncontravening public policy should be exercised with great caution, and only in cases\\n\\nfree from substantial doubt.” Dep’t of Transp. v. Brooks, 328 S.E.2d 705, 713 (Ga.\\n\\n1985) (alterations adopted and quotations omitted). In Georgia, a contractual\\n\\n\\n 7\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 8 of 19\\n\\n\\nprovision generally does not violate public policy unless the Legislature has declared\\n\\nit so or enforcement of the provision would flout the very purpose of the law. See\\n\\nid.; Robinson v. Reynolds, 21 S.E.2d 214, 215 (Ga. 1942) (“Contracts that obviously\\n\\nand directly tend in a marked degree to bring about results that the law seeks to\\n\\nprevent cannot be made the ground of a successful suit . . . [and are] void as against\\n\\npublic policy.”).\\n\\n Although amorphous in concept, Georgia’s public policy bar is built on a solid\\n\\nfoundation—the Georgia Constitution and state statutes. “[T]hese are the sources\\n\\nthat are first to be considered and that often may be conclusive.” Strickland v. Gulf\\n\\nLife Ins. Co., 242 S.E.2d 148, 151 (Ga. 1978) (quotation omitted). “Statutes, of\\n\\ncourse, are perhaps the clearest expressions of the public policy of [Georgia].” Dove\\n\\nv. Dove, 680 S.E.2d 839, 842 (Ga. 2009). See also Sonja Laesen, 7 Georgia\\n\\nJurisprudence § 3:3 (updated 2019) (“The only authentic and admissible evidence of\\n\\npublic policy of a state is its constitution, laws, and judicial decisions.”). For\\n\\nexample, O.C.G.A. § 13-8-2 provides that “[a] contract which is against the policy\\n\\nof the law cannot be enforced,” and lists particular agreements that are unenforceable\\n\\nas against public policy. Although § 13-8-2 does not address the specific provisions\\n\\nat issue in this case, the statute makes clear that its list is “nonexclusive” and Georgia\\n\\ncourts have held that they “are not restricted by this list in voiding contracts that are\\n\\nagainst public policy.” Edwards v. Grapefields, Inc., 599 S.E.2d 489, 493 (Ga. Ct.\\n\\n\\n 8\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 9 of 19\\n\\n\\nApp. 2004). We may therefore look to other Georgia statutes to determine whether\\n\\nthe state has a strong public policy against enforcing forum selection clauses in favor\\n\\nof out-of-state payday lenders.\\n\\n The Georgia Payday Lending Act, O.C.G.A. § 16-17-1 et seq., states in\\n\\nrelevant part as follows:\\n\\n A payday lender shall not . . . nor shall the loan contract\\n designate a court for the resolution of disputes concerning\\n the contract other than a court of competent jurisdiction in\\n and for the county in which the borrower resides or the\\n loan office is located.\\n\\n§ 16-17-2(c)(1). The Georgia Legislature explained the PLA’s prohibition on out-\\n\\nof-state forum selection clauses in this way:\\n\\n Certain payday lenders have attempted to use forum\\n selection clauses contained in payday loan documents in\\n order to avoid the courts of the State of Georgia, and the\\n General Assembly has determined that such practices are\\n unconscionable and should be prohibited.\\n\\n§ 16-17-1(d) (emphasis added).\\n\\n Based on this language, the Georgia Supreme Court ruled that the PLA\\n\\nestablished a clear public policy against “payday lenders [that] have attempted to\\n\\nskirt the laws of Georgia by use of forum selection clauses.” W. Sky Fin., LLC v.\\n\\nGeorgia, 793 S.E.2d 357, 363 (Ga. 2016). In ruling that the loan agreements’ forum\\n\\nselection clause is unenforceable in this case, the district court found that § 16-17-\\n\\n1(d) and § 16-17-2(c)(1) were “conclusive” as to Georgia’s public policy. See\\n\\n\\n 9\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 10 of 19\\n\\n\\nStrickland, 242 S.E.2d at 151. On appeal, the Oasis lenders argue that two other\\n\\nsubsections of the PLA support the opposite conclusion.4\\n\\n First, the Oasis lenders assert that the word “county” in § 16-17-2(c)(1) is\\n\\nunqualified, meaning that by permitting parties to select a forum “in and for the\\n\\ncounty in which . . . the loan office is located” the PLA allows parties to select a\\n\\ncounty outside Georgia for litigation if it is the county where the lenders’ loan office\\n\\nis located. The parties could therefore select as a forum the county where the Oasis\\n\\nlenders operate their loan office—Cook County, Illinois. This argument has some\\n\\nsuperficial appeal, but the district court rejected it, concluding that the word\\n\\n“county” in § 16-17-2(c)(1) refers only to Georgia counties because that subsection\\n\\nis a venue provision through which the Legislature intended to allow parties to select\\n\\nfora within Georgia while prohibiting outbound forum-selection clauses. The Oasis\\n\\nlenders contend that the district court erred by reading in the word “Georgia” in front\\n\\nof “county” in § 16-17-2(c)(1). We think the district court got it right.\\n\\n Georgia venue provisions commonly use the term “county” or “counties” in\\n\\nrefence to Georgia counties, without explicitly saying so. See Ga. Const., Art. VI, §\\n\\n2, ¶¶ III, IV, VI; O.C.G.A. § 9-10-93. For example, Georgia’s Constitution provides\\n\\n\\n4\\n The Oasis lenders also claim (in a footnote of their reply brief) that interpreting the PLA to require\\nout-of-state lenders to defend these kinds of lawsuits in Georgia may violate the dormant\\nCommerce Clause. The Oasis lenders, however, did not present this argument in the district court\\nor in their initial brief, and they also failed to support the argument with legal authorities. We\\ntherefore do not address it. See Sapuppo v. Allstate Floridian Ins. Co., 739 F.3d 678, 681 (11th\\nCir. 2014); Lovett v. Ray, 327 F.3d 1181, 1183 (11th Cir. 2003).\\n 10\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 11 of 19\\n\\n\\nthat “joint trespassers residing in different counties may be tried in either county,”\\n\\nGa. Const., Art. VI, § 2, ¶ IV, and Georgia courts have interpreted that provision to\\n\\nallow “joint tortfeasor residents of different Georgia counties [to] be sued in either\\n\\ncounty.” Robinson v. Star Gas of Hawkinsville, Inc., 533 S.E.2d 97, 98 (Ga. Ct. App.\\n\\n2000) (emphasis added). See also Cartwright v. Fuji Photo Film U.S.A., Inc., 720\\n\\nS.E.2d 200, 204 n.4, 206 (Ga. Ct. App. 2011) (“[T]he proper venue for a suit is in\\n\\nwhich Georgia county the defendant resides[.]”); Cunningham v. Estate of\\n\\nCunningham, 697 S.E.2d 280, 282 (Ga. Ct. App. 2010) (“Georgia’s constitutional\\n\\nvenue provision . . . has no application to a suit against a nonresident . . . because in\\n\\nsuch case there is no Georgia county in which the nonresident resides.”). The same\\n\\ninterpretation applies to Georgia statutes with similar language. See Lloyd Adams,\\n\\nInc. v. Liberty Mut. Ins. Co., 10 S.E.2d 46, 48 (Ga. 1940) (holding that a similar\\n\\nvenue provision in Georgia’s Nonresident Motorist Act, now O.C.G.A. § 40-12-3,\\n\\nafforded venue in “any county in the State”). This makes sense given that the\\n\\nGeorgia Legislature generally has authority to establish venue rules within Georgia.\\n\\n Given the PLA’s condemnation of out-of-state lenders using forum selection\\n\\nclauses to avoid Georgia courts in § 16-17-1(d), it would make little sense to\\n\\nconclude that the Georgia Legislature intended the word “county” in the next\\n\\nsubsection to include places like Cook County, Illinois. See West v. City of Albany,\\n\\n797 S.E.2d 809, 811 (2017) (“We do not limit our consideration to the words of one\\n\\n\\n 11\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 12 of 19\\n\\n\\nsubsection of a statute alone, but consider a particular provision in the context of the\\n\\nstatute as a whole . . . .”). The Oasis lenders’ argument—that by using the term\\n\\n“county” instead of “Georgia county” in § 16-17-2(c)(1), the Legislature opened a\\n\\nback door for out-of-state forum selection clauses—would effectively “hide\\n\\nelephants in mouseholes.” Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 468\\n\\n(2001). Moreover, it would also render the PLA’s prohibition of certain forum\\n\\nselection clauses in § 16-17-1(d) “meaningless, redundant, or mere surplusage.”\\n\\nUnited States v. Canals-Jimenez, 943 F.2d 1284, 1287 (11th Cir. 1991). We disfavor\\n\\nsuch interpretations. See Cadwell v. Kaufman, Englett & Lynd, PPLC, 886 F.3d\\n\\n1153, 1159 (11th Cir. 2018).\\n\\n Next, the Oasis lenders contend that the PLA does not apply to loan\\n\\nagreements between a Georgia borrower and an out-of-state lender because § 16-17-\\n\\n1(d) states that “[p]ayday lending involves relatively small loans and does not\\n\\nencompass loans that involve interstate commerce.” We disagree for a few reasons.\\n\\nFirst, the Oasis lenders’ arguments are mutually exclusive. Arguing that the PLA\\n\\ndoes not apply to loans by an out-of-state lender contradicts the lenders’ argument\\n\\nthat the Georgia Legislature meant for the term “county” in § 16-17-2(c)(1) to\\n\\ninclude the county where an out-of-state lender maintains its home office. See\\n\\nFootstar, Inc. v. Liberty Mut. Ins. Co., 637 S.E.2d 692, 695 (2006) (“[C]ourts should\\n\\nconstrue a statute to give sensible and intelligent effect to all of its provisions . . .”).\\n\\n\\n 12\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 13 of 19\\n\\n\\nSecond, other provisions of the PLA make clear that the Act governs “any business”\\n\\nthat “consists in whole or in part of making . . . loans of $3,000.00 or less” unless\\n\\nthose entities are specifically exempted. See W. Sky Fin., 793 S.E.2d at 363 (quoting\\n\\n§ 16-17-2(a)). Out-of-state lenders are not exempt. See § 16-17-2(a)(1)–(4). Third,\\n\\nexcluding loans involving out-of-state lenders from coverage—as the Oasis lenders\\n\\nargue—would render the PLA’s prohibition of out-of-state forum selection clauses\\n\\nmeaningless. See Cadwell, 886 F.3d at 1159; Canals-Jimenez, 943 F.2d at 1287.\\n\\n In Western Sky, 793 S.E.2d at 363, the defendant argued that the PLA’s\\n\\n“interstate commerce” language rendered the statute inapplicable to loans involving\\n\\nout-of-state lenders. The Georgia Supreme Court “reject[ed] the notion that this\\n\\nphrase was meant to exclude loans that involve interstate commerce from the scope\\n\\nof the [PLA].” Id. “If that were so, the [PLA] would be virtually meaningless\\n\\nbecause it would prohibit nothing.” Id. See also id. at 364 (“If, as Defendants argue,\\n\\nthe [PLA] is limited only to transactions that do not involve interstate commerce, the\\n\\n[PLA] would effectively prohibit nothing at all.”). The Georgia Supreme Court went\\n\\non to articulate how excluding loans involving an out-of-state lender would\\n\\nundermine the entire purpose of the PLA.\\n\\n Given the clear and unambiguous scope of the [PLA] as a\\n whole, to interpret that phrase as a definitional limitation\\n upon payday lending and thereby exempt loans that\\n involve interstate commerce from the prohibitions of the\\n [PLA] would create such a contradiction and absurdity as\\n\\n\\n 13\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 14 of 19\\n\\n\\n to demonstrate that the legislature did not mean it to\\n create such a limitation.\\n\\nId. at 365 (emphasis added).\\n\\n Based on our review, Georgia statutes establish a clear public policy against\\n\\nout-of-state lenders using forum selection clauses to avoid litigation in Georgia\\n\\ncourts. See O.C.G.A. §§ 16-17-1(d), 2(c)(1). Enforcing a forum selection clause\\n\\nlike the ones here “would [therefore] contravene a strong public policy of the forum\\n\\nin which suit is brought,” and the district court correctly denied the Oasis lenders’\\n\\nmotion to dismiss on that ground. See Bremen, 407 U.S. at 15.5\\n\\n B\\n\\n The plaintiffs argue, and the district court concluded, that the class action\\n\\nwaivers contained in the loan agreements also contravene the purpose of the PLA\\n\\nand the Georgia Industrial Loan Act, O.C.G.A. § 7-3-29. We start with the text of\\n\\nboth acts. See Dove, 680 S.E.2d at 842.\\n\\n The PLA contains a provision stating that “a civil action may be brought on\\n\\nbehalf of an individual borrower or on behalf of an ascertainable class of borrowers.”\\n\\n\\n\\n\\n5\\n Because we conclude that Georgia statutes establish a public policy against the forum selection\\nclause at issue, we are not bound by our opinion in Rucker, 632 F.3d at 1238. In that case, we held\\nthat an identical forum selection clause was valid and enforceable under Alabama law. See id. Our\\ndecision to invalidate the forum selection clause in this case depends entirely on Georgia law.\\n 14\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 15 of 19\\n\\n\\n§ 16-7-3. The GILA provides that “a claim for violation of this chapter against an\\n\\nunlicensed lender may be asserted in a class action.” § 7-3-29(e). 6\\n\\n The district court concluded that, in passing these laws, the Georgia\\n\\nLegislature “expressly contemplated a specific remedy—[a] class action—for\\n\\npersons aggrieved by predatory lending . . . [and] did not expressly create the class\\n\\naction remedy so that predatory lenders could effectively wipe away this consumer\\n\\nprotection with a waiver . . . .” Davis, 2017 WL 5490919, at *5. On appeal, the\\n\\nOasis lenders contend that the district court erred by not considering whether the\\n\\nprovision was procedurally or substantively unconscionable. They also argue that\\n\\nneither the PLA nor the GILA prohibit class action waivers or create a statutory right\\n\\nto pursue a class action.\\n\\n These arguments miss the point. The district court’s ruling flowed from its\\n\\nconclusion that enforcing class action waivers in this context would allow payday\\n\\nlenders to eliminate a remedy that was expressly contemplated by the Georgia\\n\\nLegislature, and thereby undermine the purpose of the statutory scheme. That\\n\\nconclusion, if correct, renders the class action waiver unenforceable under Georgia\\n\\nlaw regardless of whether the provision is also procedurally or substantively\\n\\nunconscionable. See Brookfield Country Club, Inc. v. St. James-Brookfield, LLC,\\n\\n\\n6\\n We note that this subsection of the GILA applies to “unlicensed lenders,” but the district court\\nquoted this provision in ruling that the class action waiver is unenforceable, and the Oasis lenders\\ndo not argue that it does not apply. See Davis, 2017 WL 5490919, at *5.\\n 15\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 16 of 19\\n\\n\\n696 S.E.2d 663, 667 (Ga. 2010) (refusing to “enforce a contractual provision which\\n\\ncontravenes the statutory law of this state”); Terry v. St. Farm Fire & Cas. Ins. Co.,\\n\\n504 S.E.2d 194, 195 (Ga. 1998) (considering whether “[t]he underlying purpose” of\\n\\nthe statute would be “contravened by enforcement of the specific terms of the\\n\\nagreement”). Accord Atl. Flooring Design Ctrs., Inc. v. R.G. Williams Const., Inc.,\\n\\n773 S.E.2d 868, 870 (Ga. Ct. App. 2015) (voiding a contractual provision because it\\n\\nwould frustrate the public policy expressed in the Georgia Arbitration Code);\\n\\nLangford v. Royal Indem. Co., 430 S.E.2d 98, 102 (Ga. Ct. App. 1993) (“A contract\\n\\nprovision normally should not be enforced where it conflicts with the general policy\\n\\nand spirit of the statute which governs it, although there may be no literal conflict;\\n\\nthat is, it makes no difference whether the statutory prohibition or command is\\n\\nexpressed or implied.”).\\n\\n The Oasis lenders may be correct in arguing that, when Georgia courts address\\n\\nwhether a contractual provision is substantively unconscionable, they also “consider\\n\\n‘the commercial reasonableness of the contract terms, the purpose and effect of the\\n\\nterms, the allocation of the risks between the parties, and similar public policy\\n\\nconcerns.’” Jenkins v. First Am. Cash Advance of Ga., LLC, 400 F.3d 868, 876 (11th\\n\\nCir. 2005) (quoting NEC Techs., Inc. v. Nelson, 478 S.E.2d 769, 772 (Ga. 1996)).\\n\\nThe public policy bar, however, remains an independent basis to hold a contractual\\n\\nprovision unenforceable. See Glosser v. Powers, 71 S.E.2d 230, 231 (Ga. 1952)\\n\\n\\n 16\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 17 of 19\\n\\n\\n(“Even in the absence of fraud in the procurement of a contract, a contract which is\\n\\nagainst the policy of the law is void and unenforceable.”) A hornbook example of\\n\\nthe public policy defense is that a court will not enforce a contractual provision that\\n\\nis illegal regardless of whether its obligations are mutual, its terms are conspicuous,\\n\\nand the parties are well represented. See 17A Am. Jur. 2d Contracts § 247 (updated\\n\\n2019). See also Stoudemire v. HSBC Bank USA, N.A., 776 S.E.2d 483, 484 (Ga. Ct.\\n\\nApp. 2015) (highlighting the difference between “[c]ontracts to do illegal or immoral\\n\\nthings, contracts against public policy, and gambling contracts,” which are legally\\n\\nvoid, and “fraudulent contracts and contracts entered under duress,” which are\\n\\nvoidable by the injured party) (citations omitted).\\n\\n The district court did not, and needed not, conclude that the PLA or the GILA\\n\\nexpressly prohibited class action waivers or created a statutory right to pursue a class\\n\\naction. A contractual provision need not “literal[ly] conflict” with Georgia law to\\n\\ncontravene public policy. See Langford, 430 S.E.2d at 102; Robinson, 21 S.E.2d at\\n\\n215. Instead, the district court concluded that enforcement of the class action\\n\\nwaivers in this context would eliminate a remedy contemplated by the Georgia\\n\\nLegislature and undermine the purpose of the PLA and the GILA. Davis, 2017 WL\\n\\n5490919, at *5.\\n\\n To that point, the Oasis lenders make one additional argument warranting\\n\\nfurther discussion—that the PLA’s fee-shifting provision eliminates the risk that\\n\\n\\n 17\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 18 of 19\\n\\n\\nenforcing the class action waiver would effectively prevent the plaintiffs from\\n\\nlitigating their claims. In past cases, we have held that certain class action waivers\\n\\nwere not unconscionable, in part because the agreements at issue contained a fee-\\n\\nshifting provision that, in practice, permitted the plaintiffs to pursue their claims\\n\\nindividually. See Jenkins, 400 F.3d at 877–80; Randolph v. Green Tree Fin. Corp.,\\n\\n244 F.3d 814, 816–18 (11th Cir. 2001). Those cases, however, concerned whether\\n\\nthe class action waivers were unconscionable because enforcement would prevent\\n\\nthe plaintiffs from having their day in court. See Jenkins, 400 F.3d at 877–78. As\\n\\ndiscussed above, the district court here did not conclude that the class action waiver\\n\\nwas procedurally or substantively unconscionable—or that the plaintiffs in this case\\n\\ncould never litigate their claims individually—but relied on Georgia’s public policy\\n\\nas expressed in the text and purpose of the PLA and the GILA.\\n\\n The Oasis lenders point out that in Jenkins, 400 F.3d at 877–80, and Bowen v.\\n\\nFirst Family Financial Services, Inc., 233 F.3d 1331, 1336 (11th Cir. 2000), we held\\n\\nthat class action waivers in arbitration clauses were not void as against public policy.\\n\\nBut those cases do not call for a different result here. Neither Jenkins nor Bowen\\n\\nanalyzed the PLA, the GILA, or Georgia’s public policy against class action waivers\\n\\nin payday loan agreements. Instead, those cases concerned class action waivers in\\n\\narbitration agreements—where the Federal Arbitration Act “create[s] a strong\\n\\nfederal policy in favor of arbitration.” Picard v. Credit Solutions, Inc., 564 F.3d\\n\\n\\n 18\\n\\f Case: 18-10526 Date Filed: 08/28/2019 Page: 19 of 19\\n\\n\\n1249, 1253 (11th Cir. 2009). The Supreme Court, in multiple cases, has ruled that\\n\\n§ 2 of the FAA overrides a state statute or common-law doctrine that attempts to\\n\\nundercut the enforceability of an arbitration agreement. See Kindred Nursing Ctrs.\\n\\nLtd. P’ship v. Clark, 137 S. Ct. 1421, 1425, 1426 (2017); DIRECTV, Inc. v.\\n\\nImburgia, 136 S. Ct. 463, 468, 471 (2015). The class action waiver here is not\\n\\ncontained in an arbitration agreement, so § 2 of the FAA does not stand in the way\\n\\nof enforcing Georgia’s public policy.\\n\\n The PLA and the GILA establish the Georgia Legislature’s intent to preserve\\n\\nclass actions as a remedy for those aggrieved by payday lenders. Enforcing the class\\n\\naction waiver here would undermine the purpose and spirit of Georgia’s statutory\\n\\nscheme. The class action waiver is therefore unenforceable, and the district court\\n\\ndid not err in denying the Oasis lenders’ motion to strike the plaintiffs’ class\\n\\nallegations.\\n\\n III\\n\\n We affirm the district court’s denial of the motion to dismiss and motion to\\n\\nstrike.\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 19\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4160,"cells":{"id":{"kind":"number","value":4657195,"string":"4,657,195"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"Mitchell Marbury v. Warden"},"decision_date":{"kind":"string","value":"2019-08-29"},"docket_number":{"kind":"string","value":"17-12589"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before ROSENBAUM, BRANCH, and HIGGINBOTHAM, * Circuit Judges.\", \"parties\": \"\", \"opinions\": [{\"author\": \"Honorable Patrick E. Higginbotham, United States Circuit Judge for the Fifth Circuit\", \"type\": \"010combined\", \"text\": \"Case: 17-12589 Date Filed: 08/29/2019 Page: 1 of 50\\n\\n\\n [PUBLISH]\\n\\n\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 17-12589\\n ________________________\\n\\n D.C. Docket No. 4:16-cv-01152-AKK-JHE\\n\\n\\n\\nMITCHELL MARBURY,\\n\\n Plaintiff-Appellant,\\n\\n versus\\n\\n\\n\\nWARDEN, et al.,\\n\\n Defendants-Appellees.\\n\\n ________________________\\n\\n Appeal from the United States District Court\\n for the Northern District of Alabama\\n ________________________\\n\\n (August 29, 2019)\\n\\nBefore ROSENBAUM, BRANCH, and HIGGINBOTHAM, * Circuit Judges.\\n\\n\\n *\\n Honorable Patrick E. Higginbotham, United States Circuit Judge for the Fifth Circuit,\\nsitting by designation.\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 2 of 50\\n\\n\\nPER CURIAM:\\n Mitchell Marbury, a prisoner at Alabama’s St. Clair Correctional Facility,\\n\\nwas attacked by a fellow prisoner after making multiple requests to be transferred\\n\\nto a different dormitory or put in protective lock-up. He sued an officer in his cell\\n\\nblock and warden of the prison under 42 U.S.C. § 1983, arguing that both were\\n\\ndeliberately indifferent to a substantial risk to his safety. Marbury appeals the\\n\\ndistrict court’s grant of summary judgment to both defendants. We affirm.\\n\\n I\\n\\n The following reflects the facts as contained in the limited summary\\n\\njudgment record in this case. Between February and April 2016, Marbury\\n\\nrepeatedly attempted to be transferred to another dormitory. He sent a written\\n\\nrequest to Warden Dewayne Estes on February 12, 2016, stating that he had\\n\\nwitnessed over fifteen inmate-on-inmate stabbing incidents that appeared to be\\n\\ngang related and asking to be assigned to a “more sociable” living area with\\n\\ninmates closer to his own age. Estes did not act on this request. 1 Marbury’s sworn\\n\\ncomplaint avers that around the same time, he made several in-person transfer\\n\\nrequests to Officer Beverly Warren and that Warren told him that if he continued\\n\\nto make transfer requests, she “would personally see to it, that it be fixed, where\\n\\nshe’ll have a legal reason to deny [the] requests.” The complaint also alleges that\\n\\n\\n 1\\n Estes stated in his affidavit that he was unaware of the request.\\n 2\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 3 of 50\\n\\n\\nWarren made other comments like “[Y]ou don’t enjoy hanging out with the\\n\\n‘thugs,’ afraid you might get shanked!” and “I got the keys to the city baby, you[]\\n\\nlocked in.”\\n\\n On April 5, Marbury sent another letter to Estes asking why he had not heard\\n\\nback regarding his February 12 letter. 2 He said that he had seen prisoners\\n\\ndisrespect and attack security staff and felt “nothing is being done to correct the\\n\\nproblem,” and requested to be moved “from a[n] over-rated gang affiliated block to\\n\\na program block or one where [he could] feel safe and secure[ ].”\\n\\n On April 18, Marbury asked Warren to have the captain put him in lock-up\\n\\nuntil he could be transferred because he had heard from a friend that another\\n\\ninmate wanted to hurt him. He avers that Warren responded, “[D]o you really think\\n\\nI’ma act upon your requests, after you’ve filed complaints and requests against\\n\\nme,” started laughing, then said, “You don’t have a shank, . . . you need to get one,\\n\\n[because you aren’t] going to lock-up, there’s no cells available, so seem[s] like to\\n\\nme you’ve got a problem.” One day later, Marbury again wrote to Estes saying, “I\\n\\nwas told by a friend to watch my back, because he got word someone was out to do\\n\\nharm to me.” He requested that Estes place him in lock-up as soon as he received\\n\\nthe complaint because he was “in fear of [being] hurt or possibl[y] killed.” He also\\n\\n\\n 2\\n The day before, Marbury had sent a letter to another prison official—not a defendant in\\nthis lawsuit—accusing Warren of retaliating against him for requesting to be transferred. While\\nMarbury also brought a retaliation claim in the district court, he does not pursue it on appeal.\\n 3\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 4 of 50\\n\\n\\nreported to Estes that Warren had laughed at him when he told her about his\\n\\nconcerns and told him to get a knife. The captain in charge of placing inmates in\\n\\nlock-up says that she did not receive any information about Marbury’s request to\\n\\nbe locked up.\\n\\n On April 23, 2016, Marbury was stabbed and hit in the face in the prison’s\\n\\ndayroom. He was treated for multiple stab wounds and a broken nose. Prison staff\\n\\nstated that they were unable to identify who attacked him, though they learned\\n\\nfrom other inmates that he was stabbed because he had called another inmate’s\\n\\ngirlfriend.\\n\\n Marbury filed a pro se § 1983 complaint against Estes and Warren in the\\n\\nNorthern District of Alabama, alleging that they failed to protect him from unsafe\\n\\nconditions, were deliberately indifferent to those conditions, and retaliated against\\n\\nhim for exercising his constitutionally protected rights. The magistrate judge\\n\\nconstrued the defendants’ special report as a motion for summary judgment and\\n\\nissued a report and recommendation to grant them summary judgment on all\\n\\nclaims. Marbury objected on the deliberate-indifference issue. The district court\\n\\noverruled Marbury’s objections, adopted the report, and accepted its\\n\\nrecommendation. Marbury now appeals the grant of summary judgment to the\\n\\ndefendants.\\n\\n II\\n\\n\\n 4\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 5 of 50\\n\\n\\n We review the district court’s grant of summary judgment de novo,\\n\\n“view[ing] all the evidence and draw[ing] all reasonable inferences in the light\\n\\nmost favorable to the non-moving party.” 3 Summary judgment is warranted where\\n\\nthe evidence in the record “presents no genuine issue of material fact and compels\\n\\njudgment as a matter of law in favor of the moving party.” 4 Where, as here, an\\n\\ninmate proceeded pro se in the district court, his summary judgment pleadings are\\n\\nconstrued liberally and “specific facts” alleged in his sworn complaint can suffice\\n\\nto generate a genuine dispute of fact.5\\n\\n III\\n\\n “The defense of qualified immunity completely protects government\\n\\nofficials performing discretionary functions from suit in their individual capacities\\n\\nunless their conduct violates ‘clearly established statutory or constitutional rights\\n\\nof which a reasonable person would have known.’’ Gonzalez v. Reno, 325 F.3d\\n\\n1228, 1233 (11th Cir. 2003) (quoting Hope v. Pelzer, 536 U.S. 730, 739 (2002)).\\n\\nMarbury concedes that the defendants were performing their discretionary duties at\\n\\nall times relevant to this appeal. Once it has been determined that the official was\\n\\nacting within his discretionary duties, the burden shifts to the plaintiff to show (1)\\n\\n\\n\\n\\n 3\\n Caldwell v. Warden, 748 F.3d 1090, 1098 (11th Cir. 2014).\\n 4\\n Id. (quoting Owusu-Ansah v. Coca-Cola Co., 715 F.3d 1306, 1307 (11th Cir. 2013)).\\n 5\\n Id. (citing Perry v. Thompson, 786 F.2d 1093, 1095 (11th Cir. 1986); and Sammons v.\\nTaylor, 967 F.2d 1533, 1545 n.5 (11th Cir. 1992)).\\n 5\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 6 of 50\\n\\n\\nthat the official violated a constitutional right and (2) that the right was clearly\\n\\nestablished at the time of the alleged violation. Caldwell, 748 F.3d at 1099. Our\\n\\ninquiry “can begin with either prong.” Morris v. Town of Lexington, 748 F.3d\\n\\n1316, 1322 (11th Cir. 2014).\\n\\n It is well settled that prison officials must “take reasonable measures to\\n\\nguarantee the safety of the inmates,” 6 and “[a] prison official violates the Eighth\\n\\nAmendment’s prohibition against cruel and unusual punishment if [the official] is\\n\\ndeliberately indifferent to a substantial risk of serious harm to an inmate who\\n\\nsuffers injury.” 7 Not “every injury suffered by one inmate at the hands of another,”\\n\\nhowever, “translates into a constitutional liability for prison officials responsible\\n\\nfor the victim’s safety.” 8 To establish a § 1983 claim for deliberate indifference, a\\n\\nplaintiff must show “(1) a substantial risk of serious harm; (2) the defendants’\\n\\ndeliberate indifference to that risk; and (3) causation.”9\\n\\n The first element of deliberate indifference—whether there was a substantial\\n\\nrisk of serious harm—is assessed objectively and requires the plaintiff to show\\n\\n“conditions that were extreme and posed an unreasonable risk of serious injury to\\n\\n\\n\\n 6\\n Caldwell, 748 F.3d at 1099 (quoting Farmer v. Brennan, 511 U.S. 825, 832 (1994)).\\n 7\\n Lane v. Philbin, 835 F.3d 1302, 1307 (11th Cir. 2016).\\n 8\\n Carter v. Galloway, 352 F.3d 1346, 1349 (11th Cir. 2003) (per curiam) (quoting Farmer,\\n511 U.S. at 834).\\n 9\\n Lane, 835 F.3d at 1307 (quoting Hale v. Tallapoosa Cty., 50 F.3d 1579, 1582 (11th Cir.\\n1995)).\\n\\n\\n 6\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 7 of 50\\n\\n\\nhis future health or safety.” 10 The second element—whether the defendant was\\n\\ndeliberately indifferent to that risk—has both a subjective and an objective\\n\\ncomponent. Subjectively, the “official must both be aware of facts from which the\\n\\ninference could be drawn that a substantial risk of serious harm exists, and . . . also\\n\\ndraw the inference.”11 Objectively, the official must have responded to the known\\n\\nrisk in an unreasonable manner, in that he or she “knew of ways to reduce the\\n\\nharm” but knowingly or recklessly declined to act. 12 Finally, the plaintiff must\\n\\nshow a “necessary causal link” between the officer’s failure to act reasonably and\\n\\nthe plaintiff’s injury. 13\\n\\n Marbury argues that the defendants were subjectively aware of two distinct\\n\\ntypes of risk. The first is the general threat posed by inmate-on-inmate violence in\\n\\nMarbury’s cell block based on his statement that he had witnessed fifteen stabbings\\n\\nat the prison. The second is the threat Marbury identified in April 2016, when he\\n\\ntold the defendants that he had heard from a friend that someone intended to harm\\n\\nhim. He also argues that the defendants were deliberately indifferent in failing to\\n\\nfurther investigate his claims before the attack.\\n\\n\\n\\n\\n 10\\n Id.\\n 11\\n Rodriguez v. Sec’y for Dep’t of Corrs., 508 F.3d 611, 617 (11th Cir. 2007) (quoting\\nFarmer, 511 U.S. at 837).\\n 12\\n Id. at 620 (quoting Hale, 50 F.3d at 1583).\\n 13\\n Id. at 622–23.\\n 7\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 8 of 50\\n\\n\\n As we explain, Marbury’s deliberate-indifference claim fails because he has\\n\\nnot demonstrated a genuine factual issue as to whether the defendants were\\n\\ndeliberately indifferent to a substantial risk of serious harm to Marbury. Since\\n\\nMarbury has not met his burden to show the violation of a constitutional right, we\\n\\nneed not proceed past step one of the qualified-immunity analysis.\\n\\n A\\n\\n Marbury repeatedly asked to be transferred because he was concerned about\\n\\na general lack of safety in his cell block. In his letters to Estes and the verbal\\n\\nrequests he says he made to Warren, he told them that he had witnessed fifteen\\n\\ninmate-on-inmate stabbings that he attributed to gang affiliations, expressing fear\\n\\nfor his own safety. Granting Marbury the resolution of all disputed facts in his\\n\\nfavor, our caselaw nevertheless establishes that this evidence was insufficient to\\n\\nestablish deliberate indifference to a substantial risk of serious harm.\\n\\n In general, a plaintiff must show “more than a generalized awareness of\\n\\nrisk” to make out a deliberate-indifference claim. 14 While “occasional, isolated\\n\\nattacks by one prisoner on another may not constitute cruel and unusual\\n\\npunishment, . . . . confinement in a prison where violence and terror reign is\\n\\n\\n\\n\\n 14\\n Caldwell, 748 F.3d at 1101 (internal quotation marks omitted).\\n\\n\\n 8\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 9 of 50\\n\\n\\nactionable.” 15 To establish deliberate indifference based on a generalized risk, the\\n\\nplaintiff must show “that serious inmate-on-inmate violence was the norm or\\n\\nsomething close to it.” 16\\n\\n Turning to this case, we look to only those facts contained in the summary\\n\\njudgment record. While Marbury’s sworn allegation may support the inference that\\n\\nMarbury faced some generalized risk of attack, such evidence does not support the\\n\\nconclusion that serious inmate-on-inmate violence was so pervasive that it\\n\\nconstitutes a substantial risk of serious harm to which defendants were deliberately\\n\\nindifferent. The only allegation Marbury makes about inmate-on-inmate violence\\n\\nis his statement that he personally witnessed fifteen inmate-on-inmate stabbings\\n\\nduring his time at St. Clair. 17 There is no evidence in the record of the total prison\\n\\npopulation or the sections of the prison in which the attacks occurred that would\\n\\nplace Marbury’s statement in context. Also, it is not precisely clear from the record\\n\\nover what period of time these incidents occurred. But we can tell from the record\\n\\nthat Marbury was at St. Clair between 2002 and 2007, went to a different\\n\\n\\n\\n\\n 15\\n Purcell ex rel. Estate of Morgan v. Toombs Cty., 400 F.3d 1313, 1320 (11th Cir. 2005)\\n(quoting Woodhous v. Virginia, 487 F.2d 889, 890 (4th Cir. 1973)); accord Harrison v. Culliver,\\n746 F.3d 1288, 1299 (11th Cir. 2014).\\n 16\\n Purcell, 400 F.3d at 1322.\\n 17\\n In Marbury’s objection to the report and recommendation, he argued that additional\\ndiscovery would substantiate his claim of widespread violence “a year prior to[ ] and a year\\nafter” Marbury’s departure from St. Clair. His affidavit also stated he had witnessed over fifteen\\nstabbings over the “short time” he had been at St. Clair.\\n 9\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 10 of 50\\n\\n\\ncorrectional facility, and then returned in 2015. So, the fifteen stabbings may have\\n\\noccurred over the course of 6 years, for a rate of 2.5 per year.\\n\\n Marbury’s allegation stands in sharp contrast to those at issue in Harrison v.\\n\\nCulliver, for example, where we held that even though a prison warden was on\\n\\nnotice that inmate-on-inmate assaults occurred throughout the prison and in a\\n\\nparticular location where a prisoner was attacked, “the evidence of inmate-on-\\n\\ninmate assault involving weapons [did] not . . . indicate that inmates were ‘exposed\\n\\nto something even approaching the constant threat of violence.’” 18 We emphasized\\n\\nthat the institution in Harrison was large, housing between 830 and 990 inmates\\n\\nbetween 2005 and 2008, and the fact that there had been thirty-three incidents\\n\\ninvolving weapons during the same time period was “hardly sufficient to\\n\\ndemonstrate that [the institution] was a prison ‘where violence and terror reign.’” 19\\n\\nAnd Marbury has presented far less evidence about the level of violence than the\\n\\nevidence presented in Harrison that we found “hardly sufficient” to establish\\n\\ndeliberate indifference to a substantial risk of serious harm.\\n\\n Moreover, when we have held that a generalized risk of violence from a\\n\\nprison population could support a claim of deliberate indifference to a substantial\\n\\nrisk of serious harm, the plaintiff has pointed to specific features of a facility or its\\n\\n\\n\\n 18\\n 746 F.3d at 1299–1300 (quoting Purcell, 400 F.3d at 1321).\\n 19\\n Id. at 1300 (quoting Purcell, 400 F.3d at 1320).\\n 10\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 11 of 50\\n\\n\\npopulation rendering it particularly violent. This evidence has included pervasive\\n\\nstaffing and logistical issues rendering prison officials unable to address near-\\n\\nconstant violence, 20 tensions between different subsets of a prison population,21\\n\\nand unique risks posed by individual prisoners or groups of prisoners due to\\n\\ncharacteristics like mental illness. 22 Even if Marbury had shown a risk of\\n\\ngeneralized prison violence, he has made no allegations regarding the specific\\n\\nfeatures of the prison that would make it particularly violent.\\n\\n We emphasize that the record before the district court was limited. 23 While\\n\\nwe are sensitive to Marbury’s pro se status before the district court, the evidence\\n\\nMarbury has presented regarding a general risk of inmate-on-inmate violence does\\n\\nnot rise to the level necessary to show deliberate indifference to a substantial risk\\n\\n\\n\\n\\n 20\\n See, e.g., Lane, 835 F.3d at 1307–08 (finding allegations that only one officer\\nsupervised two separate dorms, that inmates regularly brought back weapons from their work\\ndetail and fashioned them from prison materials, and that officials did not confiscate weapons\\nsufficient); Hale, 50 F.3d at 1582–83 (finding potential awareness of a substantial risk of serious\\nharm where a defendant was aware that a prison had severe overcrowding problems and the\\nplaintiff presented evidence that “inmate-on-inmate violence occurred regularly when the jail\\nwas overcrowded”).\\n 21\\n See Lane, 835 F.3d at 1307–08 (finding potential awareness of a substantial risk of\\nserious harm where a plaintiff alleged that a particular prison building was composed of 90%\\ngang members, it was common for the non-gang-affiliate inmates or non-Muslim inmates to be\\nrobbed or stabbed, and the prison had inadequate supervision to prevent inmates from making\\nhomemade weapons).\\n 22\\n See Cottone v. Jenne, 326 F.3d 1352, 1355–56, 1358–59 (11th Cir. 2003) (finding\\npotential awareness of a substantial risk of serious harm where mentally ill inmates were\\nseparated from the general population but kept in unlocked cells where they could interact with\\neach other, and guards were aware of a particular prisoner’s history of violent schizophrenic\\noutbursts).\\n 23\\n Marbury does not argue that the district court improperly limited discovery.\\n 11\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 12 of 50\\n\\n\\nof serious harm required by our caselaw. This sparse record at most shows that\\n\\ninmates at St. Clair faced some risk of assaults by fellow prisoners, but we have\\n\\nsaid that some risk of harm is insufficient. Marbury has thus failed to produce\\n\\nevidence that he was in an environment so beset by violence that confinement, by\\n\\nits nature, threatened him with the substantial risk of serious harm.\\n\\n B\\n\\n Marbury also argues that even if the defendants were not deliberately\\n\\nindifferent to a general lack of safety in his cell block, they were at least\\n\\ndeliberately indifferent to the more specific threat he warned them about in April\\n\\n2016—that he had heard from a friend that someone intended to harm him. As we\\n\\nhave explained, Marbury’s sworn complaint alleges that when he told Warren\\n\\nabout this threat she laughed at him, told him she was not going to act on his\\n\\nrequest because he had filed internal complaints against her, and told him to get a\\n\\nknife or a shank if he did not already have one. Marbury then wrote another letter\\n\\nto Estes telling Estes about the threat and Warren’s response, requesting to be\\n\\nplaced in lock-up, and stating that he was afraid of being hurt or possibly killed.\\n\\nOn the summary judgment record, Marbury appears to have provided no further\\n\\ninformation to either defendant about the nature of the threat.\\n\\n We must therefore decide whether a reasonable jury could find Marbury’s\\n\\nstatement that he had heard from a friend that an unnamed prisoner intended to\\n\\n\\n 12\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 13 of 50\\n\\n\\nhurt him, and that he was afraid of being hurt or killed, without any further details,\\n\\nsufficient to make the defendants aware of a substantial risk of serious harm. While\\n\\nthis question is a close one, we conclude that our precedent does not allow\\n\\nMarbury’s deliberate-indifference claim to proceed.\\n\\n On the one hand, it is settled that “a prison official [cannot] escape liability\\n\\nfor deliberate indifference by showing that . . . he did not know that the\\n\\ncomplainant was especially likely to be assaulted by the specific prisoner who\\n\\neventually committed the assault,” as long as the official was otherwise aware that\\n\\nthe victim faced a substantial risk of serious harm. 24 Our caselaw also establishes,\\n\\nhowever, that officials must possess enough details about a threat to enable them to\\n\\nconclude that it presents a “strong likelihood” of injury, not a “mere possibility.” 25\\n\\nThe unfortunate reality is that “threats between inmates are common and do not,\\n\\nunder all circumstances, serve to impute actual knowledge of a substantial risk of\\n\\nharm.” 26\\n\\n We have therefore upheld dismissal of or summary judgment against\\n\\ndeliberate-indifference claims where, although a plaintiff told prison officials about\\n\\na threat by another inmate or inmates, the prison officials were not deliberately\\n\\n\\n\\n\\n 24\\n Rodriguez, 508 F.3d at 619 (quoting Farmer, 511 U.S. at 843) (alterations in original).\\n 25\\n E.g., Brooks v. Warden, 800 F.3d 1295, 1301 (11th Cir. 2015); Brown v. Hughes, 894\\nF.2d 1533, 1537 (11th Cir. 1990).\\n 26\\n Prater v. Dahm, 89 F.3d 538, 542 (8th Cir. 1996).\\n 13\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 14 of 50\\n\\n\\nindifferent to a substantial risk of harm. For example, we held that a plaintiff had\\n\\nfailed to allege plausibly that prison officials were deliberately indifferent to a\\n\\nstrong likelihood of serious harm where he was attacked by an inmate in the next\\n\\ncell when all of the cell doors opened and a riot began.27 We noted that even\\n\\nthough the plaintiff alleged that the attacker had threatened him previously and that\\n\\nan attack was possible, he had not alleged any information showing a strong\\n\\nlikelihood that he and his attacker would be released at the same time in a chaotic\\n\\nenvironment. 28 Successful deliberate-indifference claims will generally require\\n\\nsome further reason—beyond the plaintiff having informed the defendant officers\\n\\nof the threat—that a prison official could have concluded that a particular threat\\n\\nevidenced a substantial threat, rather than the mere possibility, of serious harm.\\n\\n For example, a plaintiff can establish deliberate indifference to a substantial\\n\\nrisk of serious harm where he has given prison officials further information\\n\\nenabling them to conclude that the risk was substantial and not merely possible. In\\n\\nRodriguez v. Secretary for Department of Corrections, a prisoner informed prison\\n\\nstaff that members of his former gang had threatened to kill him upon release into\\n\\nthe general prison population. 29 We concluded that if true, this allegation was\\n\\n\\n\\n\\n 27\\n Brooks v. Warden, 800 F.3d 1295, 1298–99, 1301 (11th Cir. 2015).\\n 28\\n Id. at 1301.\\n 29\\n See Rodriguez, 508 F.3d at 612–15.\\n\\n\\n 14\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 15 of 50\\n\\n\\nenough to place the defendants on notice of a substantial risk of serious harm—so\\n\\nwe allowed the case to proceed.30 The key distinction between Rodriguez and\\n\\nMarbury’s case is that Rodriguez involved a series of threats that contained enough\\n\\nspecific information—the threats came from members of the plaintiff’s former\\n\\ngang—that prison officials could have reasonably inferred that there was a\\n\\nsubstantial, not merely possible, risk of harm. In fact, Rodriguez observed that a\\n\\nvague statement like “I have a problem with another inmate in this compound,”\\n\\nabsent some information “about the nature of the anticipated risk,” would not have\\n\\ncreated a genuine issue of fact regarding deliberate indifference to a substantial\\n\\nrisk of serious harm. 31 Marbury’s statement that an another inmate told him\\n\\nanother inmate intended to harm him is precisely this type of vague statement that\\n\\nconveys nothing about the nature of the anticipated risk that we cautioned in\\n\\nRodriguez would not rise to the level of deliberate indifference to a substantial risk.\\n\\n Marbury’s argument is essentially that every prisoner who tells prison\\n\\nofficials about an unspecified threat from an unspecified inmate without more is\\n\\nentitled to protective custody or a transfer. But, as already explained, our caselaw\\n\\nestablishes a higher standard for deliberate indifference. To be clear, Marbury was\\n\\n\\n\\n\\n 30\\n Id. at 621–22.\\n 31\\n Rodriguez, 508 F.3d at 619 n.15.\\n\\n\\n 15\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 16 of 50\\n\\n\\nnot required to identify the person who was threatening him by name, 32 or even\\n\\nnecessarily to give the defendants advance notice of a potential attack,33 so long as\\n\\nother facts put the defendants on notice that he faced a substantial risk of serious\\n\\nharm. It may be possible for a general threat of inmate-on-inmate violence in a\\n\\nprison to bolster an otherwise insufficient unspecified threat of harm. But, as\\n\\nalready discussed, Marbury has not shown anything close to such a substantial\\n\\nthreat from the generally violent nature of the prison environment. And because\\n\\nMarbury has not presented anything else that would bolster the unspecified threat,\\n\\nhe has not met the requirement of showing deliberate indifference to a substantial\\n\\nrisk of serious harm.\\n\\n Marbury further emphasizes that when he told Warren about the threat from\\n\\nan unidentified inmate, she laughed at him, told him she was not going to act on\\n\\nhis request because he had filed complaints against her, and advised him to get a\\n\\nknife or a shank because it seemed as though he had a “problem.” He suggests that\\n\\nthis reaction demonstrated her awareness that he faced a substantial risk. The\\n\\ndistrict court did not explicitly address Marbury’s argument, noting that\\n\\n“[a]ccording to defendant Warren, she never said this and asserts that she did not\\n\\nknow the plaintiff was in danger of being attacked by another inmate on April 23,\\n\\n\\n\\n 32\\n See, e.g., Rodriguez, 508 F.3d at 619 (discussing Farmer, 511 U.S. at 843).\\n 33\\n See Farmer, 511 U.S. at 848.\\n 16\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 17 of 50\\n\\n\\n2016.” Again, “specific facts” alleged in a pro se plaintiff’s sworn complaint can\\n\\nsuffice to generate a genuine factual issue. 34 We must therefore accept as true, for\\n\\nthe purposes of summary judgment, that Warren made these statements.\\n\\n But the statements ultimately do not affect Warren’s entitlement to summary\\n\\njudgment. The Supreme Court has made clear that “[w]hether a prison official had\\n\\nthe requisite knowledge of a substantial risk is a question of fact subject to\\n\\ndemonstration in the usual ways, including inference from circumstantial\\n\\nevidence.”35 While Marbury can establish Warren’s subjective knowledge through\\n\\ncircumstantial proof, he must nevertheless present evidence sufficient to “support a\\n\\nreasonable jury’s finding that [Warren] harbored a subjective awareness that\\n\\n[Marbury] was in serious danger.”36 Viewed in the light most favorable to\\n\\nMarbury, the summary judgment record allows one to conclude that Warren was\\n\\naware of the potential for inmate-on-inmate violence at the prison, Marbury told\\n\\nher that another inmate was looking to harm him, and she responded by laughing at\\n\\nMarbury and telling him to get a knife because he had a “problem.” A reasonable\\n\\njury could not extrapolate from such statements that Warren was subjectively\\n\\n\\n\\n\\n 34\\n See Caldwell, 748 F.3d at 1098.\\n 35\\n Rodriguez, 508 F.3d at 617 (quoting Farmer, 511 U.S. at 842).\\n 36\\n Goodman v. Kimbrough, 718 F.3d 1325, 1332 (11th Cir. 2013).\\n\\n\\n 17\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 18 of 50\\n\\n\\naware of a substantial risk of serious harm to Marbury at the time. Indeed,\\n\\nMarbury concedes in his briefing that Warren’s statements were callous jokes.\\n\\n Again, it bears mention that subjective awareness of only some risk of harm\\n\\nto a prisoner is insufficient for a deliberate-indifference claim. 37 At most, the\\n\\nevidence Marbury has put forward would allow a jury to conclude that Warren was\\n\\nput on notice that Marbury faced some unspecified risk of harm to his well-\\n\\nbeing—not that she was aware he faced the type of substantial risk of serious harm\\n\\nnecessary to establish deliberate indifference. Marbury has not marshaled enough\\n\\nevidence to establish a genuine issue of fact on this necessary element of his case.\\n\\n C\\n\\n Finally, we turn to Marbury’s suggestion that the defendants were\\n\\ndeliberately indifferent in failing to investigate his report that someone was out to\\n\\nharm him or in otherwise failing to abide by prison policy. We have explained that\\n\\n“merely negligent failure to protect an inmate from attack does not justify liability\\n\\nunder section 1983.”38 To allow Marbury’s deliberate-indifference claim to\\n\\nproceed absent sufficient evidence that the defendants were subjectively aware that\\n\\nhe faced a substantial risk of serious harm would elide the “subtle distinction”\\n\\nbetween deliberate indifference and mere negligence.39 We cannot condone the\\n\\n\\n 37\\n See, e.g., Brooks, 800 F.3d at 1301; Brown, 894 F.2d at 1537.\\n 38\\n Carter, 352 F.3d at 1350.\\n 39\\n See Goodman, 718 F.3d at 1333–34.\\n 18\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 19 of 50\\n\\n\\nfailure to investigate inmates’ allegations of threats or to follow policy in reporting\\n\\npotential threats up the chain of command. But our caselaw does not allow these\\n\\nfailures, without corresponding subjective awareness of a serious risk of harm, to\\n\\nestablish deliberate indifference.\\n\\n IV\\n\\n We affirm the judgment of the district court.\\n\\n\\n\\n\\n 19\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 20 of 50\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434448/\", \"author_raw\": \"Honorable Patrick E. Higginbotham, United States Circuit Judge for the Fifth Circuit\"}, {\"author\": \"ROSENBAUM, Circuit Judge, dissenting\", \"type\": \"dissent\", \"text\": \"ROSENBAUM, Circuit Judge, dissenting:\\n\\n We do not sentence people to be stabbed and beaten. But we might as well,\\n\\nif the Majority Opinion is correct.\\n\\n Time and again, Mitchell Marbury pleaded with Warden Dewayne Estes and\\n\\nOfficer Beverly Warren to be transferred to safety; he had witnessed abject\\n\\nlawlessness and unmitigated violence—even against the guards—in his less-than-\\n\\nseven months at St. Clair prison, and he had been warned by a friend that he was the\\n\\ntarget of an imminent attack. Estes ignored Marbury’s pleas. Warren was even\\n\\nworse: she gloated about Marbury’s predicament and told him to find a “shank”—\\n\\na makeshift knife—to fend for himself. Just a few days later, an inmate blocked\\n\\nMarbury’s path. Then Marbury was repeatedly stabbed from behind. The attack left\\n\\nMarbury in the infirmary with a puncture wound to the base of his skull, multiple\\n\\nstab wounds to his shoulder area, a broken nose, and a gash two centimeters deep in\\n\\nhis back.\\n\\n Yet somehow, the Majority Opinion concludes that no reasonable juror could\\n\\never find that Estes and Warren were deliberately indifferent to a substantial risk of\\n\\nserious harm to Marbury. This misguided decision allows corrections officers, with\\n\\nimpunity, to refuse to take any action whatsoever to protect a prisoner in the face of\\n\\na known threat. Just as bad, today’s decision also gives corrections officers license\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 21 of 50\\n\\n\\nto perpetuate prison violence by advising a prisoner who reports a threat that\\n\\n“get[ting]” a “shank” is his only protection option.\\n\\n Lord of the Flies is supposed to be a work of fiction; it should not describe\\n\\nthe environment in our prisons. Indeed, the Eighth Amendment strictly prohibits\\n\\nprison officials from allowing such treacherous environments to exist. As the\\n\\nSupreme Court has explained, “[h]aving incarcerated persons with demonstrated\\n\\nproclivities for antisocial criminal, and often violent, conduct, having stripped them\\n\\nof virtually every means of self-protection and foreclosed their access to outside aid,\\n\\nthe government and its officials are not free to let the state of nature take its course”\\n\\nand allow inmates to hurt one another. Farmer v. Brennan, 51 U.S. 825, 833 (1994)\\n\\n(alterations adopted) (citation and quotation marks omitted).\\n\\n “[G]ratuitously allowing the beating [and stabbing] of one prisoner by another\\n\\nserves no legitimate penological objective.” Id. (alteration adopted) (quotation\\n\\nmarks omitted). It also does not square with society’s “evolving standards of\\n\\ndecency.” Id. (quotation marks omitted). Rather, under the Eighth Amendment,\\n\\nprison officials must “provide humane conditions of confinement.” Id. at 832. And\\n\\nthey must “take reasonable measures to guarantee the safety of the inmates.” Id.\\n\\n(quoting Hudson v. Palmer, 468 U.S. 517, 526-27 (1984)). This means prison\\n\\nofficials cannot be deliberately indifferent to a known risk to inmate safety. Id. at\\n\\n833-34.\\n\\n\\n 21\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 22 of 50\\n\\n\\n When we look at the record in the light most favorable to Marbury, as we must\\n\\non Estes and Warren’s motion for summary judgment, Estes and Warren were, at\\n\\nbest, deliberately indifferent to the known threat posed to Marbury. 1 The Majority\\n\\nOpinion mistakenly reaches the opposite conclusion because of three errors it makes.\\n\\nFirst, the Majority Opinion fails to view the facts in the light most favorable to\\n\\nMarbury and to draw all reasonable inferences in his favor. Second, the Majority\\n\\nOpinion does not account for important facts in its analysis. And third, the Majority\\n\\nOpinion evaluates the evidentiary components of Marbury’s claim separately, rather\\n\\nthan considering them as a whole. As a result, the Majority Opinion misses the forest\\n\\nfor the trees. When these errors are corrected, the record here yields only one\\n\\npossible answer under the Eighth Amendment: Estes and Warren must be denied\\n\\nsummary judgment. I therefore respectfully dissent from the Majority Opinion.\\n\\n I divide my discussion into two substantive parts. Section I demonstrates that\\n\\nMarbury provided sufficient evidence to establish Warren and Estes violated his\\n\\nEighth Amendment right to be free from deliberate indifference. And Section II\\n\\nexplains why Warren and Estes are not entitled to qualified immunity for their\\n\\nalleged violations.\\n\\n\\n\\n\\n 1\\n Warren and Estes deny Marbury’s allegations. Summary-judgment review of the record,\\nhowever, requires us to take the facts in the light most favorable to the non-moving party—here,\\nMarbury. Goodman v. Kimbrough, 718 F.3d 1325, 1331 (11th Cir. 2013). Of course, if a party\\nsurvives summary judgment, he must convince a jury that the facts are what he says they are.\\n 22\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 23 of 50\\n\\n\\n I.\\n\\n As I have noted, “[a] prison official’s ‘deliberate indifference’ to a substantial\\n\\nrisk of serious harm to an inmate violates the Eighth Amendment.” Farmer, 511\\n\\nU.S. at 828. To establish a deliberate-indifference Eighth Amendment claim, a\\n\\nprisoner must show (1) he was subjected to a substantial risk of serious harm; (2) the\\n\\ndefendants were deliberately indifferent to that risk; and (3) a causal connection\\n\\nexists between the prison official’s conduct and the Eighth Amendment violation.\\n\\nHale v. Tallapoosa Cty., 50 F.3d 1579, 1582 (11th Cir. 1995). Here, when we view\\n\\nthe record in the light most favorable to Marbury and draw all reasonable inferences\\n\\nin his favor, Marbury satisfies each of these elements.\\n\\n A. Marbury presented enough evidence for the jury to conclude that he\\n faced a substantial risk of serious harm.\\n First, Marbury was exposed to a substantial risk of serious harm. We use an\\n\\nobjective standard to assess whether complained-of circumstances constitute a\\n\\nsubstantial risk of serious harm. Caldwell v. Warden, FCI Talladega, 748 F.3d 1090,\\n\\n1099 (11th Cir. 2014).\\n\\n Under that standard, we have held that while occasional, isolated attacks do\\n\\nnot create a substantial risk of serious harm, “an excessive risk of inmate-on-inmate\\n\\nviolence at a jail” does, and “confinement in a prison where violence and terror reign\\n\\nis actionable.” Purcell ex rel. Estate of Morgan v. Toombs Cty., Ga, 400 F.3d 1313,\\n\\n1320 (11th Cir. 2005). For example, in Purcell, we held that the level of violence\\n\\n 23\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 24 of 50\\n\\n\\nthere was insufficient to establish a substantial risk of harm where the plaintiff could\\n\\nproduce evidence of only “two to three pretty serious inmate fights over a period of\\n\\nnine months and of not very many other fights over a four-year span.” Id. at 1323\\n\\nn.21. Given those facts, we simply could not say that “inmate-on-inmate violence\\n\\nwas the norm or something close to it.” Id. at 1322.\\n\\n On the other hand, we held in Marsh v. Butler County that “conditions in a\\n\\njail facility that allow prisoners ready access to weapons, fail to provide an ability to\\n\\nlock down inmates, and fail to allow for surveillance of inmates pose a substantial\\n\\nrisk of serious harm to inmates.” 268 F.3d 1014, 1028 (11th Cir. 2001) (en banc),\\n\\nabrogated on other grounds by Bell Atl. Corp. v. Twombly, 550 U.S. 544, 561-63\\n\\n(2007). This is so because such conditions create “an excessive risk of inmate-on-\\n\\ninmate violence.” See Purcell, 400 F.3d at 1320-21. Notably, we reached the\\n\\nconclusion that the conditions in Marsh established a significant risk of serious harm\\n\\neven though the record there did not show a history of inmate assaults resulting in\\n\\nserious injuries. Marsh, 268 F.3d at 1034.\\n\\n But when conditions like those in Marsh are actually accompanied by a\\n\\nhistory of inmate-on-inmate violence that is “severe enough to require medical\\n\\nattention and even hospitalization on occasion,” Hale, 50 F.3d at 1583, that\\n\\ncircumstance only reinforces the conclusion that a substantial risk of serious harm\\n\\n\\n\\n\\n 24\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 25 of 50\\n\\n\\nexists. This is true even when the complaining prisoner was not expressly\\n\\nthreatened. See Lane v. Philbin, 835 F.3d 1302, 1307-08 (11th Cir. 2016).\\n\\n Here, Marbury has described a prison environment where the worst of all\\n\\nworlds collided. Like in Marsh, prisoners at St. Clair enjoyed ready access to\\n\\nweapons, and prison officials regularly failed to surveil and control inmates. As\\n\\nMarbury describes it, St. Clair was a place of open lawlessness where, because of a\\n\\n“lack of security staff,” 2 inmates could assault each other and even brazenly pull\\n\\nknives on and stab security staff.3 In addition, Marbury also alleged a remarkable\\n\\nhistory of violence during his time at St. Clair, as he recounted witnessing “over 15\\n\\ninmates-on-inmates stabbing incidents” in less than seven months there.\\n\\n\\n\\n\\n 2\\n The U.S. Department of Justice investigated and reported on conditions in Alabama’s\\nprisons from 2017-18. See U.S. Dep’t of Justice, Investigation of Alabama’s State Prisons for\\nMen (April 2, 2019), https://www.justice.gov/crt/case-document/file/1149971/download (“DOJ\\nReport”). According to the Department of Justice’s report, St. Clair employs only 28% of the\\ncorrectional officers authorized for that prison. DOJ Report at 9-10. To be clear, I mention\\nfindings from the DOJ Report as an aside only, for those who may be interested in the continuing\\nconditions at St. Clair, and I do not rely on facts from the DOJ Report in my analysis. See infra\\nnotes 3-7. Nevertheless, I note that Rule 201 of the Federal Rules of Evidence appears to allow\\nme to rely on the report in adjudicating this motion, since the DOJ Report is a government-agency\\nreport, and counsel was put on notice at oral argument that we might consider it. See K.T. v. Royal\\nCaribbean Cruises, Ltd., --- F.3d ----, 2019 WL 3312530, at *5-7 (11th Cir. July 24, 2019) (Carnes,\\nC.J., concurring).\\n 3\\n These problems persisted following the events in Marbury’s case. For example, the DOJ\\nReport catalogued a March 2018 incident at St. Clair where seven prisoners, with homemade\\nknives drawn, surrounded a correctional officer. DOJ Report at 28. One prisoner used a knife to\\ncut the officer in his stomach before help arrived. Id. Similarly, at St. Clair in December 2017, a\\ncorrectional officer told several inmates to leave a dormitory. Id. at 29. In response, one prisoner\\npunched the officer repeatedly in the face and then stabbed him in the face with a homemade ice\\npick. Id. And in October 2017 at St. Clair, an officer directed an inmate to put a shirt on. Id. That\\nprisoner returned with a 26-inch-long shank and attempted to strike four officers. Id.\\n 25\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 26 of 50\\n\\n\\n The Majority Opinion incorrectly asserts that we must assume Marbury meant\\n\\nhe witnessed more than fifteen stabbings in six years. See Maj. Op. at 9-10. So I\\n\\npause to explain why the summary-judgment standard requires us to accept that\\n\\nMarbury saw more than 15 inmate-on-inmate stabbings in less than seven months.\\n\\nMarbury attested that he had witnessed “over [fifteen] stabbing[s] during the short\\n\\ntime” he was housed there (emphasis added). According to the record, as of\\n\\nFebruary 12, 2016, when Marbury first claimed in a letter to Estes that he had\\n\\nwitnessed all these stabbings in his short time while at P-Block of St. Clair, he had\\n\\nbeen housed in St. Clair since July 29, 2015. That is a six-and-a-half-month period.\\n\\nThe Majority arrives at its six-year figure only by reaching back fourteen years\\n\\nearlier, when Marbury was previously housed at St. Clair during a five-year period\\n\\nthat ended nine years before Marbury’s complaint here. Six years, of course, is not\\n\\na “short time.” But six-and-a-half months is.\\n\\n And even if six years could be considered a “short time,” the summary-\\n\\njudgment standard requires us to make all reasonable inferences in Marbury’s favor.\\n\\nThat means the relevant question is not whether it is reasonable to assume that\\n\\nMarbury was referring to his total six years at St. Clair during the last decade-and-\\n\\na-half. Rather, the question is whether it is reasonable to infer that Marbury was\\n\\nreferring to his most recent six-and-a-half months at St. Clair. Here, it is certainly a\\n\\nreasonable inference that when Marbury spoke of the “short time” in which he saw\\n\\n\\n 26\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 27 of 50\\n\\n\\nall the stabbings, he was referring to his then-current six-and-a-half-month period of\\n\\nincarceration at St. Clair. Indeed, the Majority Opinion offers no reason why that\\n\\ninference is not reasonable. So we must accept for purposes of evaluating the motion\\n\\nfor summary judgment that Marbury witnessed the more-than-fifteen stabbings in\\n\\nsix-and-a-half months.\\n\\n That means a stabbing occurred roughly once every 1.75 weeks, on average,\\n\\nwhich is three times as often as did the non-specified “weapons” incidents found to\\n\\nbe insufficient in Harrison v. Culliver, 746 F.3d 1288 (11th Cir. 2014), on which the\\n\\nMajority Opinion relies. Maj. Op. at 10; Harrison, 746 F.3d at 1299-1300 & n.17.\\n\\nPlus, Marbury’s numbers include only stabbings, while Harrison’s total of 33\\n\\nincidents over three-and-a-half years was composed of stabbings and all other types\\n\\nof violence involving weapons.4 Harrison, 746 F.3d at 1299-1300 & n.17. Simply\\n\\nput, the frequency of stabbings at St. Clair far, far exceeded those at the prison in\\n\\nHarrison, and we should not pretend that Harrison supports the Majority Opinion’s\\n\\nposition. See Maj. Op. at 10.\\n\\n\\n\\n\\n 4\\n St. Clair likewise experienced violence involving other weapons. For example, in\\nSeptember 2017, two prisoners beat a third with a sock filled with metal locks—and this occurred\\nin St. Clair’s honor dormitory, which was reserved for prisoners with good behavior. DOJ Report\\nat 2. And in July 2017, a St. Clair prisoner was found tied up and strangled to death. Id. at 13.\\nSimilarly, in May 2016, another prisoner was strangled to death. Id. at 15. In October 2017, a\\nprisoner was assaulted and severely beaten, and he was bound with tape around his hands, ankles,\\nmouth, and head. Id. at 19. He also had a fresh burn mark on his face. Id.\\n 27\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 28 of 50\\n\\n\\n The danger at St. Clair becomes even more apparent when we account for\\n\\nreasonable inferences in favor of Marbury. For example, Marbury surely was not\\n\\nalways in exactly the wrong place at precisely the wrong time, so he could not have\\n\\nwitnessed every single stabbing that occurred in his 900-person prison during his\\n\\nbrief time there. When we factor this common-sense reasoning into the equation,\\n\\nwe must conclude that stabbing incidents occurred even more frequently than once\\n\\nevery 1.75 weeks, on average.5\\n\\n And contrary to the Majority Opinion’s suggestion in its analysis, Marbury\\n\\ndid not just allege that he saw over fifteen stabbings in under seven months. See\\n\\nMaj. Op. at 9 (“The only allegation Marbury makes about inmate-on-inmate\\n\\nviolence is his statement that he personally witnessed fifteen inmate-on-inmate\\n\\n\\n\\n 5\\n Indeed, as the Majority Opinion notes, “in Marbury’s objection to the report and\\nrecommendation, he argued that additional discovery would substantiate his claim of widespread\\nviolence ‘a year prior to[ ] and a year after’ Marbury’s departure from St. Clair.” Maj. Op. at 9\\nn.17. As it turns out, the DOJ Report confirms he was right. See id.; supra notes 3-4; infra notes\\n6-7; see also DOJ Report at 48-49 (“Our investigation into the violence, contraband, corruption,\\nand harm occurring in Alabama’s prisons evidences issues previously known to ADOC [Alabama\\nDepartment of Corrections]. For instance, several years before we initiated our investigation,\\nADOC was acutely aware of extensive problems at St. Clair. In 2014 alone, there were at least\\nthree publicly reported prisoner-on-prisoner homicides. In April 2014, the Equal Justice Initiative\\n(“EJI”) urged ADOC to investigate, among other violence, the fatal and non-fatal stabbings that\\nwere escalating at St. Clair. Following another homicide in June 2014, EJI renewed its formal\\nrequest that ADOC address the violence at St. Clair, including six homicides in the preceding three\\nyears. . . . Three years later, in November 2017, the plaintiffs and ADOC reached a settlement.\\nADOC promised many reforms in the settlement. For instance, ADOC promised to ask the\\nAlabama Legislature for funding to install video cameras for monitoring at the prison. ADOC did\\nnot make good on that promise. By June 2018, ADOC had not satisfied several of the settlement\\nrequirements. The parties went back into mediation in June 2018—only eight months after ADOC\\nmade all of its promises to reform St. Clair.” (emphasis added)).\\n\\n\\n 28\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 29 of 50\\n\\n\\nstabbings during his time at St. Clair.”). Rather, Marbury also alleged that the prison\\n\\nwas so understaffed and that conditions were so out of control that prisoners also\\n\\nstabbed security staff and even the warden, and nothing was done to correct the\\n\\nproblem. 6\\n\\n Against this toxic background, on April 18, 2016—five days before he was\\n\\nbrutally attacked—Marbury also directly reported to Warren that a friend had\\n\\nwarned him that another inmate wanted to hurt him. For this reason, Marbury asked\\n\\nWarren to send him to lock-up until he could be transferred to a different prison or\\n\\nthe threat passed. He also asked Warren to have Captain Carla Graham speak with\\n\\nhim, so he could tell her about the threat.\\n\\n But according to Marbury, Warren refused to pass along his request to\\n\\nGraham and took no action at all: she did not report the threat to anyone; she did\\n\\nnot seek to investigate the threat; and she did not do anything to provide Marbury\\n\\nwith any type of protection. Instead, Warren allegedly used the opportunity to\\n\\n\\n\\n 6\\n Again, for those interested, the DOJ Report provides some detail concerning the severity\\nof some of the stabbings that regularly occurred at St. Clair during the period on which it reported.\\nFor example, in addition to the stabbings described in note 3, supra, the Report stated that in\\nSeptember 2018, a prisoner was stabbed to death at St. Clair and noted that he had previously been\\nstabbed in July 2017 while imprisoned at the same facility. DOJ Report at 13. Similarly, in\\nFebruary 2018, one inmate killed another in a knife fight at St. Clair. Id. at 14. In March 2017 at\\nSt. Clair, two prisoners stabbed another inmate. Id. at 18. When an officer yelled for them to stop,\\none of the prisoners just kept stabbing the victim. Id. After the assault was finally stopped, the\\nvictim had to be taken to an emergency room to treat stab wounds to the back, a perforated lung,\\nand a stab wound to the head. Id. Then, in February 2018, an officer found a prisoner roaming\\nthe prison with a knife stuck in his head. Id. at 21. He also had to be taken to an emergency room,\\nwhere an eight-inch, metal homemade knife was removed from the back of his head. Id.\\n 29\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 30 of 50\\n\\n\\npunish Marbury for previously having filed complaints against her. Warren derided\\n\\nMarbury: “[D]o you really think I’m[] [going to] act upon your request, after you’ve\\n\\nfiled complaints and requests against me[?]” Then she laughed at him and said,\\n\\n“[Y]ou don’t have a shank[?] [I]f not[,] you need to get one, cause you [aren’t]\\n\\ngoing to lock-up[.] [T]here[’]s no cells available, so seem[s] like to me, you’ve got\\n\\na problem.”\\n\\n Warren’s statement that Marbury better get himself a “shank” further supports\\n\\nthe notion that stabbings were regular occurrences at St. Clair and that “shanks” were\\n\\nreadily available. 7 And her assertion that no cells were open to protect Marbury\\n\\nfrom the threat likewise suggests that St. Clair was overflowing with prisoners who\\n\\nwere either misbehaving or under threat of serious harm, further supporting\\n\\nMarbury’s contention that violence at St. Clair was rampant.\\n\\n So at bottom, Marbury alleged that stabbings were the norm at St. Clair, that\\n\\nthe prison was understaffed, that inmates were out of control and could stab guards\\n\\nand even the warden, that he had been expressly threatened and had reported that\\n\\n\\n\\n 7\\n Makeshift knives remained readily available at St. Clair even after Marbury’s stabbing.\\nFor example, the DOJ Report recounted one incident that occurred at St. Clair in September 2017\\nand involved four different makeshift knives. In that incident, a prison official saw two prisoners\\nfighting with box cutters and homemade knives. DOJ Report at 17. He radioed for assistance. Id.\\nWhen one officer began taking one of the prisoners involved to the healthcare unit, a third prisoner\\nsprang into action and stabbed the escorted prisoner in the back. Id. While the officers were\\npepper-spraying the third prisoner, a fourth prisoner tried to stab the third prisoner with a knife.\\nId. One of the prisoners had to be taken to an emergency room (as opposed to the prison’s\\nhealthcare facility) for treatment of his stab wounds. Id.\\n 30\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 31 of 50\\n\\n\\nthreat to prison officials, and that the one prison official directly responsible for\\n\\nhelping him specifically refused to do so and instead goaded him to get a “shank.”\\n\\n Yet in its analysis of the overall prison environment, the Majority Opinion\\n\\nseverely understates the situation. It isolates the more-than-fifteen stabbings\\n\\nMarbury witnessed and evaluates only whether, in and of themselves, those created\\n\\na pervasively violent and dangerous prison environment. It then compounds its error\\n\\nby improperly stretching back a decade-and-a-half to make the inference in favor of\\n\\nthe defendants that the stabbings occurred over six years instead of over six-and-a-\\n\\nhalf months. And those facts the Majority Opinion doesn’t understate, it simply\\n\\nignores: nowhere in its analysis does the Majority Opinion discuss Marbury’s\\n\\nallegations about the prison’s staffing issues or the utter lack of order that allowed\\n\\nprisoners to stab even the guards and the warden. The Majority Opinion also ignores\\n\\nthe dire implications of the words and actions of Warren—who was supposed to be\\n\\nMarbury’s protector.\\n\\n Our precedent requires us to consider the entirety of the prison environment\\n\\nin Marbury’s favor. When we do that, we can see that, without a doubt, the\\n\\nconditions Marbury faced created a substantial risk of serious harm—both under our\\n\\nbinding precedent and as a matter of common sense. If the circumstances here do\\n\\nnot describe an environment “where violence and terror reign,” Purcell, 400 F.3d at\\n\\n1320, it is difficult to imagine what would.\\n\\n\\n 31\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 32 of 50\\n\\n\\n B. A reasonable juror could conclude that both Warden Estes and Officer\\n Warren acted with deliberate indifference to Marbury’s pleas.\\n\\n To establish deliberate indifference, a plaintiff must satisfy both a subjective\\n\\nand objective component of the standard. Caldwell, 748 F.3d at 1099. The\\n\\nsubjective aspect requires the plaintiff to show that the defendant “actually\\n\\n(subjectively) knew that an inmate faced a substantial risk of serious harm.” Id.\\n\\n(cleaned up). This means the plaintiff must produce evidence that the defendant was\\n\\n“both . . . aware of facts from which the inference could be drawn that a substantial\\n\\nrisk of serious harm exist[ed], and he . . . dr[ew] the inference.” Id. at 1099-1100\\n\\n(quoting Farmer, 511 U.S. at 837). We have explained that this inquiry presents a\\n\\nquestion of fact. Id. at 1100. As for the objective component of the standard, the\\n\\nplaintiff must demonstrate that the defendant “disregarded [the] known risk by\\n\\nfailing to respond to it in an (objectively) reasonable manner.” Id. at 1099 (cleaned\\n\\nup).\\n\\n Here, read in the light most favorable to Marbury, the record establishes\\n\\nEstes’s and Warren’s deliberate indifference to the substantial risk of serious harm\\n\\nMarbury faced.\\n\\n i. The record contains enough evidence for a jury to conclude that Estes\\n and Warren knew that Marbury faced a substantial risk of harm.\\n\\n As I have noted, whether Estes and Warren knew of facts that would have\\n\\nallowed them to conclude that Marbury faced a substantial risk of serious harm, and\\n\\n\\n 32\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 33 of 50\\n\\n\\nwhether they actually drew that inference from those facts, present questions of fact.\\n\\nAnd “a factfinder may conclude that a prison official knew of a substantial risk from\\n\\nthe very fact that the risk was obvious.” Farmer, 511 U.S. at 842. For example, a\\n\\ntrier of fact can find that the defendant had actual knowledge of the risk if the\\n\\nplaintiff “presents evidence showing that a substantial risk of inmate attacks was\\n\\nlongstanding, pervasive, well-documented, or expressly noted by prison officials in\\n\\nthe past, and the circumstances suggest that the defendant-official being sued had\\n\\nbeen exposed to information concerning the risk and thus must have known about\\n\\nit.” Id. at 842-43 (quotation marks omitted).\\n\\n That’s what happened here. First, stabbing incidents occurred regularly at St.\\n\\nClair, and even guards and a warden had been victims of the violence. The prison\\n\\nofficials could not maintain order, and prisoners regularly disrespected them. On\\n\\ntop of that, Marbury had reported an express threat to both Warren (in person) and\\n\\nEstes (in writing). So the risk the jail conditions and the express threat posed was\\n\\nobvious.\\n\\n And there can be no doubt that Warren clearly understood Marbury’s\\n\\ndangerous dilemma because she literally said as much. Indeed, when Marbury\\n\\nreported the threat against him to Warren, Warren expressly confirmed that Marbury\\n\\nhad “a problem.” Then she went so far as to suggest that Marbury get a “shank.”\\n\\nShe may as well have left a smoking gun.\\n\\n\\n 33\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 34 of 50\\n\\n\\n Yet the Majority Opinion dismisses Marbury’s admissions as insufficient\\n\\nbecause Warren laughed and “callous[ly] joke[d]” when she made these comments.\\n\\nSee Maj. Op. at 18. This reasoning, however, again betrays that the Majority\\n\\nOpinion has impermissibly construed the facts in the light most favorable to Warren,\\n\\nnot Marbury. Here, it is certainly reasonable to infer from the record that Warren’s\\n\\nlaughter implicates, not absolves, her: Warren told Marbury she would not help him\\n\\nbecause he had previously filed complaints about her. It was in this context that\\n\\nWarren laughed at Marbury while acknowledging Marbury’s “problem” and telling\\n\\nhim to get a “shank.” These facts require us to make the more-than-reasonable\\n\\ninference here that Warren fully understood the danger Marbury faced when she\\n\\n“callous[ly] joke[d]” about it. Cf. Lane, 835 F.3d at 1309-10 (prison official’s threat\\n\\nto punish prisoners by putting them into a building known for its violence supported\\n\\nthe reasonable inference that the official was aware of the violent conditions at that\\n\\nbuilding). Notably, the Majority Opinion does not explain why this inference in\\n\\nfavor of Marbury is not reasonable.\\n\\n Next, the Majority Opinion tries to minimize the evidence against Warren as\\n\\nshowing only “that Warren was put on notice that Marbury faced some unspecified\\n\\nrisk of harm to his well-being—not that she was aware he faced the type of\\n\\nsubstantial risk of serious harm necessary to establish deliberate indifference.” Maj.\\n\\nOp. at 18. But this, too, is error, as the Majority Opinion both ignores inconvenient\\n\\n\\n 34\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 35 of 50\\n\\n\\nfacts and incorrectly suggests that a prisoner must be able to identify with specificity\\n\\nthe precise threat he faces before a prison official has any responsibility to protect\\n\\nhim. Maj. Op. at 12-16.\\n\\n As the Supreme Court has explained, when prison conditions allow for\\n\\nviolence and terror to reign, “it [is] obviously . . . irrelevant to liability that the\\n\\nofficials could not guess beforehand precisely who would attack whom.” Farmer,\\n\\n511 U.S. at 843-44. Based on this principle, we have rejected the notion that a\\n\\nprisoner must identify his attacker to receive protection. In Rodriguez v. Secretary\\n\\nfor Department of Corrections, for example, the prison official argued that the\\n\\nprisoner’s declaration did not put the prison official on notice of the risk of harm\\n\\nbecause it did not “furnish any specifics as to who was posing the alleged threats.”\\n\\n508 F.3d 611, 619 (11th Cir. 2007) (quotation marks omitted). The Rodriguez panel\\n\\nrejected that argument, because even though vague statements like “I have a problem\\n\\nwith another inmate in this compound” would be insufficient, the inmate in that case\\n\\nhad provided additional information—that he had been threatened by members of\\n\\nhis former gang—that corroborated otherwise vague statements. Id. & n.15.\\n\\n As with the case in Rodriguez, and contrary to the Majority Opinion’s\\n\\nassertion, Marbury did not just generally allege that he had been threatened. Maj.\\n\\nOp. at 15-16 (“Marbury’s argument is essentially that every prisoner who tells prison\\n\\nofficials about an unspecified threat from an unspecified inmate without more is\\n\\n\\n 35\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 36 of 50\\n\\n\\nentitled to protective custody or a transfer.”). Instead, Marbury also averred many\\n\\nfacts that showed a prison culture of unchecked violence: consistent inmate-on-\\n\\ninmate stabbing attacks requiring medical attention, understaffing of guards,\\n\\nineffective guarding where even guards were attacked, and no corrective response\\n\\nto these conditions from prison officials. These allegations are enough under our\\n\\nprecedent to establish a serious risk of harm, so “it [was] obviously . . . irrelevant to\\n\\nliability that the officials could not guess beforehand precisely who would attack\\n\\nwhom.” Farmer, 511 U.S. at 844. Indeed, in LaMarca v. Turner, we have\\n\\nrecognized that when the prison environment demonstrates the prison’s inability to\\n\\nprotect prisoners from violence, that environment alone can support an inference that\\n\\nthe prison officials knew and understood that the inmates were not protected. 995\\n\\nF.2d 1526, 1536-37 (11th Cir. 1993). Because a norm of unchecked violence at a\\n\\nprison more than sufficiently corroborates a threat, Rodriguez supports, not hurts,\\n\\nMarbury’s case.\\n\\n The Majority Opinion also suggests that Brooks v. Warden, 800 F.3d 1295\\n\\n(11th Cir. 2015), shows that the defendants were not on notice of a substantial risk\\n\\nof harm to Marbury. See Maj. Op. at 14. But that case is wholly inapposite. In\\n\\nBrooks, Brooks alleged that he had reported threats of physical and sexual assault he\\n\\nhad received from the inmate in the adjacent cell, but nothing was done. Brooks,\\n\\n800 F.3d at 1298. Then, one day, a freak accident occurred—all 32 doors in\\n\\n\\n 36\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 37 of 50\\n\\n\\nBrooks’s dormitory opened simultaneously—leading to a riot, and Brooks was\\n\\nattacked by the inmate who had threatened him. Id. We held that Brooks had failed\\n\\nto allege a substantial risk of serious harm existed before the prison riot, since harm\\n\\nto Brooks depended on both Brooks’s and the threatening inmate’s release from their\\n\\ncells at the same time in an unsupervised or chaotic environment where an attack\\n\\ncould occur—an unlikely scenario that no one could have been expected to\\n\\nanticipate. Id. at 1301. In contrast, Marbury warned that he had been threatened in\\n\\na prison where deficient supervision was the norm, stabbings were par for the course,\\n\\nand even the guards were constantly at risk. And the conditions described in\\n\\nMarbury’s allegations—not an unforeseen accident—were what enabled Marbury to\\n\\nbe attacked in the regular course of the prison’s business.\\n\\n In short, that Marbury did not know precisely how or why he was going to be\\n\\nattacked, or who was going to attack him, could not relieve Warren of all\\n\\nresponsibility under the circumstances of this case. Based on the prison environment\\n\\nalone, Warren was on notice that Marbury faced a substantial risk of serious harm.\\n\\nIn addition, Marbury also reported to Warren that he had received an express threat,\\n\\nand Warren’s response effectively admitted that she knew Marbury faced a major\\n\\nthreat. A jury would be well within its rights to conclude Warren knew good and\\n\\nwell that Marbury was in serious danger. Under our precedent and these facts, and\\n\\n\\n\\n\\n 37\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 38 of 50\\n\\n\\nat this summary-judgment posture, Warren may not escape liability on the adverse\\n\\ninference that she was not aware of the risk to Marbury.\\n\\n Estes fares no better on this record. Over and over, Marbury wrote to him,\\n\\nworried about prison conditions. On February 12, 2016, Marbury told Estes that he\\n\\nhad seen over fifteen stabbings and implored the warden to “look into [the] matter\\n\\nas soon as possible” for the sake of his safety. On April 5, Marbury again wrote to\\n\\nthe warden. He told him that he had seen blatant examples of how unsafe his cell\\n\\nblock was, as inmates had even pulled knives on and stabbed security staff. He\\n\\ntalked about how he wanted to move from his block, which was teeming with gang\\n\\nmembers, to one where he could be safe. And significantly, he mentioned that he\\n\\nhad antagonized a prison officer—Warren—by filing complaints about her, and she\\n\\nhad, in turn, falsely accused him of infractions and told him that she “[had] it in for\\n\\n[him].” Then on April 19, four days before he was stabbed multiple times, Marbury\\n\\nwrote again. He described how a friend had told him to watch himself, since the\\n\\nword was that someone wanted to hurt him. He also noted he had immediately told\\n\\nWarren of the threat, and, instead of helping him, she had told him to find a knife\\n\\nbecause he had a “problem.” Marbury begged Estes to put him in lockup because\\n\\nhe was afraid of being hurt or killed.\\n\\n Taking the record in Marbury’s favor, we can see that Estes had more than\\n\\nenough information to allow him to realize that Marbury was in danger. Estes was\\n\\n\\n 38\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 39 of 50\\n\\n\\ntold that violence was rampant and that things were out of control. Plus, serving as\\n\\nthe warden at St. Clair, Estes necessarily knew firsthand of all the stabbings, inmate-\\n\\non-inmate violence, and inmate attacks on guards. He also knew of the lack of prison\\n\\nguards. In addition, Estes knew about the direct threat against Marbury, and he knew\\n\\nthat Marbury had a hostile relationship with his would-be protector, Warren.\\n\\nFurther, Estes knew that Warren, after acknowledging that Marbury had a problem,\\n\\nhad refused to help him and instead told him to find a “shank.”\\n\\n Of course, for Marbury’s claim against Estes to be actionable, Estes must have\\n\\nactually drawn the inference from all this information that Marbury was in danger.\\n\\nFarmer, 511 U.S. at 837. So he may certainly argue to the jury that he wouldn’t\\n\\nknow substantial risk of serious harm if it hit him in the face. But on this record, a\\n\\nreasonable jury could conclude that Estes knew Marbury was in serious danger\\n\\n“from the very fact that the risk was obvious.” Id. at 842. For our purposes on\\n\\nsummary judgment, then, we must conclude that Estes knew Marbury faced a\\n\\nsubstantial risk of harm.\\n\\n ii. Estes and Warren responded unreasonably given their knowledge that\\n Marbury faced a substantial risk of harm.\\n\\n A reasonable jury could also find that Estes and Warren responded in an\\n\\nobjectively unreasonable manner to the danger Marbury faced. “An official\\n\\nresponds to a known risk in an objectively unreasonable manner if he knew of ways\\n\\nto reduce the harm but knowingly declined to act or if he knew of ways to reduce\\n\\n 39\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 40 of 50\\n\\n\\nthe harm but recklessly declined to act.” Rodriguez, 508 F.3d at 620 (quotation\\n\\nmarks omitted). Both Estes and Warren knew of ways to reduce the risk of harm to\\n\\nMarbury but failed to take advantage of those opportunities.\\n\\n I begin with Warren. Warren responded unreasonably to the known risk of\\n\\nharm to Marbury when she refused to take steps known to her that could have\\n\\nreduced the risk. Estes declared in his affidavit that “[w]hen inmates are\\n\\nexperiencing issues[,] all they need to do is report the situation to any Correctional\\n\\nOfficer[,] and the Correctional Officer will report it up their chain of command”\\n\\n(emphasis added). But when Marbury went to Warren on April 18 with information\\n\\nthat he was being targeted for attack, Warren rebuffed him. She refused to report\\n\\nthe threat up the chain of command or to do anything else at all to help him. And\\n\\nshe did nothing even though she knew she was supposed to report Marbury’s\\n\\nconcerns to her superior. 8 Under these circumstances, Warren’s total inaction to\\n\\nreduce the risk of harm to Marbury was unreasonable.\\n\\n Estes also failed to take known and reasonable action to mitigate the risk to\\n\\nMarbury. Estes noted in his affidavit that when an inmate requests to be moved, he\\n\\n“always refer[s] those requests to the appropriate Correctional Captain.” He\\n\\n\\n\\n\\n 8\\n Warren claims that after one of Marbury’s complaints, she spoke with a captain about\\nMarbury’s safety concerns. But she cannot recall the captain with whom she spoke, and no captain\\nhas corroborated her claim. At best, this presents a question of fact for the jury.\\n 40\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 41 of 50\\n\\n\\nidentified the appropriate captain during the relevant period as Captain Carla\\n\\nGraham.\\n\\n But according to Graham, Estes did not mention Marbury to her until after he\\n\\nhad been sued for deliberate indifference (long after Marbury had been attacked).\\n\\nAnd Estes provided no evidence that he did anything in response to Marbury’s\\n\\nrepeated pleas for help. Instead, he based his defense of his actions on the existence\\n\\nof the prison’s metal detectors that inmates leaving the Trade School and Alabama\\n\\nCorrectional Industries must go through as they exit that area, and the metal detectors\\n\\nthat scan inmates entering the prison infirmary. But neither of those generalized\\n\\nsecurity measures was geared to protect against the threat to Marbury: the metal\\n\\ndetectors at the Trade School and Correctional Industries obviously do nothing to\\n\\nsniff out shanks made on the prison grounds, and the metal detector at the infirmary\\n\\nwas useless to Marbury (until, ironically, after the stabbing), since he was not in the\\n\\ninfirmary when he complained about the prison conditions and the express threat.\\n\\nAs for Estes’s assertion that the jail has a portable “cell sense” it “deploy[s] daily at\\n\\ndifferent locations” to detect cell phones or weapons, Estes’s affidavit provides no\\n\\ninformation concerning the success of the technology or how many inmates it\\n\\nassesses on a daily basis. And since this is a motion for summary judgment, we must\\n\\ndraw from Estes’s failure to include any information on the efficacy of such\\n\\ntechnology the inference in favor of Marbury that the “cell sense” is, at best,\\n\\n\\n 41\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 42 of 50\\n\\n\\nminimally effective, if effective at all, in making a dent in the prison’s widespread\\n\\n“shank” problem. 9\\n\\n More fundamentally, these measures do not even attempt to address the\\n\\nexpress threat made to Marbury. It cannot be a defense to deliberate indifference to\\n\\nparade protocols designed for general safety to cover up what occurred here—failure\\n\\nto take any steps at all to alleviate the particularized substantial risk of serious harm\\n\\nto an inmate despite knowledge of the threat the inmate faced.\\n\\n In sum, both Estes and Warren knew that Marbury was in danger and yet did\\n\\nnothing to help. That was objectively not reasonable. As a result, a jury could find\\n\\nthat they both engaged in deliberate indifference.\\n\\n C. A jury could reasonably find that Estes’s and Warren’s deliberate\\n indifference caused Marbury’s injuries.\\n\\n The critical question for purposes of determining whether a defendant’s\\n\\ndeliberate indifference caused the subsequent injury is whether the defendant was\\n\\n“in a position to take steps that could have averted the stabbing incident . . . but,\\n\\nthrough deliberate indifference, failed to do so.” Rodriguez, 508 F.3d at 622\\n\\n(alterations adopted) (quoting Williams v. Bennett, 689 F.2d 1370, 1384 (11th Cir.\\n\\n1982)). In answering this question, we look to whether the prison official (1) “‘had\\n\\nthe means substantially to improve’ the inmate’s safety,” (2) “knew that the actions\\n\\n\\n 9\\n In fact, as the DOJ Report shows, neither measure was all that effective at preventing\\nattacks at St. Clair. See supra notes 3-7.\\n 42\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 43 of 50\\n\\n\\nhe undertook would be insufficient to provide the inmate with reasonable protection\\n\\nfrom violence,” and (3) “had other means available to him which he nevertheless\\n\\ndisregarded.” Id. (alterations adopted) (quoting LaMarca, 995 F.2d at 1539).\\n\\n When we applied this standard in Rodriguez, we found that a prison official\\n\\ncaused the plaintiff’s injuries when the official was able—but failed—to put the\\n\\nplaintiff in administrative segregation and initiate a protective review. Id. at 622-\\n\\n23. On the first factor, Rodriguez explained that a jury could have found that the\\n\\nofficial had the means to improve the plaintiff’s safety, since the official was\\n\\nauthorized to put the plaintiff in lockup and start a protective review. Id. at 623.\\n\\nAnd if the protective review had concluded that the plaintiff was in danger, the\\n\\nplaintiff would have been transferred, which would have averted the stabbing he\\n\\nsuffered. Id. As to the second factor, Rodriguez held that a jury could have found\\n\\nthat the official knew his inaction would be insufficient to protect the plaintiff. Id.\\n\\nOn the final factor, Rodriguez noted that the prison had an established protocol for\\n\\nhandling an inmate’s security concerns—immediate segregation and protective\\n\\nreview—that showed the official disregarded available means that could have helped\\n\\nthe plaintiff. Id.\\n\\n Rodriguez maps perfectly onto Marbury’s case. On the first factor, a jury\\n\\ncould conclude that both Estes and Warren had the means to substantially improve\\n\\nthe inmate’s safety. Estes, as the warden, had the authority to put Marbury in\\n\\n\\n 43\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 44 of 50\\n\\n\\nprotective lockup. And as I have noted, both Estes and Warren could have taken\\n\\nadvantage of the prison’s established protocol for dealing with threats to prisoners\\n\\nby triggering a protective review for Marbury by referring Marbury’s concerns to\\n\\nGraham for investigation. See supra at 40-41. Like the case in Rodriguez, had\\n\\nGraham’s investigation discovered the threat Marbury faced, the stabbing Marbury\\n\\nsuffered just a few days later could have been averted.\\n\\n On the second factor, a reasonable jury could find that both Estes and Warren\\n\\nknew their inaction would be insufficient to protect Marbury. Just as the jury in\\n\\nRodriguez could have determined that the defendant there knew his total inaction\\n\\nwould be insufficient to protect the inmate, a jury in this case could find that Estes\\n\\nand Warren knew their total inaction would be insufficient to protect Marbury.\\n\\n And on the final factor, a reasonable jury could conclude that Estes and\\n\\nWarren had other means available to them to help Marbury. As Estes admitted, they\\n\\ncould have at least referred Marbury’s concerns to the appropriate captain—here,\\n\\nGraham—for investigation.\\n\\n Estes and Warren thus failed to take steps that could have averted the attack\\n\\non Marbury, despite their being in a position to take those steps. Under our caselaw,\\n\\na jury could reasonably find from this evidence that Estes and Warren caused\\n\\nMarbury’s injuries.\\n\\n\\n\\n\\n 44\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 45 of 50\\n\\n\\n II.\\n\\n Because Marbury satisfied all the elements for proving a successful\\n\\ndeliberate-indifference claim, I next consider whether qualified immunity\\n\\nnonetheless shields Estes and Warren from suit. Qualified immunity “completely\\n\\nprotects government officials performing discretionary functions from suit in their\\n\\nindividual capacities unless their conduct violates clearly established statutory or\\n\\nconstitutional rights of which a reasonable person would have known.” Cottone v.\\n\\nJenne, 326 F.3d 1352, 1357 (11th Cir. 2003) (quoting Gonzalez v. Reno, 325 F.3d\\n\\n1228, 1233 (11th Cir. 2003)). To take advantage of qualified immunity, the\\n\\ngovernment official must first demonstrate that she acted within her discretionary\\n\\nauthority. Id. If she makes that showing, the burden shifts to the plaintiff to establish\\n\\nthat his allegations, if true, demonstrate a violation of clearly established law. Id. at\\n\\n1358.\\n\\n Estes and Warren were acting within their duties as prison officers when they\\n\\nengaged in the acts and omissions at issue here, so Marbury must allege a violation\\n\\nof clearly established law. As I have explained, see supra at Section I, Marbury has\\n\\nsuccessfully alleged all the elements of a deliberate-indifference violation. The only\\n\\nquestion is whether the law of deliberate indifference that Estes and Warren violated\\n\\nwas clearly established at the time of the events.\\n\\n\\n\\n\\n 45\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 46 of 50\\n\\n\\n The touchstone inquiry under the “clearly established” prong of qualified\\n\\nimmunity is whether the officials charged had “fair warning” that their conduct\\n\\nconstituted a constitutional violation. Gaines v. Wardynski, 871 F.3d 1203, 1208\\n\\n(11th Cir. 2017). As relevant here, a plaintiff can establish fair warning in two ways:\\n\\nhe can show that the unlawfulness of the defendants’ actions was apparent in light\\n\\nof pre-existing caselaw, Hope v. Pelzer, 536 U.S. 730, 739 (2002), or he may\\n\\ndemonstrate that the defendants’ actions were so egregious that they violated a\\n\\nbroader constitutional rule with obvious clarity, Gaines, 871 F.3d at 1208-09.\\n\\n A. Caselaw at the time of the events gave Estes and Warren fair notice\\n that their actions violated the Eighth Amendment.\\n\\n Prison officials have fair notice of their obligations if controlling caselaw at\\n\\nthe time of the events at issue previously established the governing principles\\n\\ninvolved, even if the existing cases were not factually identical. Hope, 536 U.S. at\\n\\n741-42. For example, in Hope, the Supreme Court denied qualified immunity to\\n\\nprison officials who had handcuffed the plaintiff to a hitching post for an extended\\n\\nperiod. Id. at 744-46. It did so because it found that two prior Supreme Court cases\\n\\nand a Department of Justice report put the officials on notice that the practice was\\n\\nillegal. In particular, the Court pointed to Gates v. Collier, 501 F.2d 1291, 1306 (5th\\n\\nCir. 1974), which held that handcuffing inmates to a fence for long periods is\\n\\nunconstitutional, and Ort v. White, 813 F.2d 318, 324 (11th Cir. 1987), which stood\\n\\nfor the proposition that “physical abuse directed at a prisoner after he terminates his\\n\\n 46\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 47 of 50\\n\\n\\nresistance to authority” is unconstitutional. Hope, 536 U.S. at 741-43 (alterations\\n\\nadopted) (quotation marks omitted). As for the Department of Justice report, the\\n\\nCourt noted it condemned the practice of using a hitching post to punish prisoners.\\n\\nId. at 744-45. In light of these authorities, the Supreme Court concluded that a\\n\\nreasonable officer was on notice that handcuffing the plaintiff to a hitching post was\\n\\nunconstitutional. Id. at 745-46.\\n\\n Here, our long-standing precedent gave Estes and Warren fair notice that their\\n\\ninaction was unconstitutional. In Purcell, we explained that an inmate faces a\\n\\nsubstantial risk of harm if “serious inmate-on-inmate violence was the norm or\\n\\nsomething close to it.” Purcell, 400 F.3d at 1322. In Lane, we found that violent\\n\\nprison conditions could pose a substantial risk of serious harm to an inmate, even\\n\\nwhen the complaining prisoner was not expressly threatened. See Lane, 835 F.3d at\\n\\n1307-08. And when a prisoner is threatened, we have held in Rodriguez that even a\\n\\nvague threat should be taken seriously if prison conditions corroborate the weight of\\n\\nthat threat. Rodriguez, 508 F.3d at 618-19 & n.15. We have also noted that regular\\n\\ninmate-on-inmate violence requiring medical attention creates a substantial risk of\\n\\nserious harm. Hale, 50 F.3d at 1583. And for over a decade, we have maintained\\n\\nthe commonsense notion that “it is an unreasonable response for an official to do\\n\\nnothing when confronted with prison conditions . . . that pose a risk of serious\\n\\nphysical harm to inmates.” Marsh, 268 F.3d at 1034.\\n\\n\\n 47\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 48 of 50\\n\\n\\n Given the preexisting caselaw, it is certainly fair to say that Estes and Warren\\n\\nhad notice that in a violent environment where stabbings were the norm, it was\\n\\nunreasonable to do absolutely nothing in the face of a prisoner’s report that he was\\n\\nbeing targeted for attack. For this reason, qualified immunity does not protect Estes\\n\\nand Warren.\\n\\n B. Warren’s conduct was so egregious that it violated the Eighth\\n Amendment with obvious clarity, even without the guidance of prior\\n caselaw.\\n\\n The foundational principle behind much of the previously discussed caselaw\\n\\nis the one I began this dissent with, the one the Supreme Court articulated over two\\n\\ndecades ago in Farmer: being attacked in prison is not part of the constitutional\\n\\nscheme of punishment that we mete out to individuals, and the Eighth Amendment\\n\\nprohibits prison officials from being deliberately indifferent when they know that an\\n\\ninmate is in danger. Farmer, 51 U.S. at 833-34. Warren’s alleged conduct here\\n\\nstands in such stark contrast to the range of behavior acceptable under the Eighth\\n\\nAmendment that if a jury believes Marbury’s account, Warren engaged in deliberate\\n\\nindifference with obvious clarity.\\n\\n Under Marbury’s account, Marbury was housed in a brutal environment\\n\\nwhere stabbings occurred regularly. He personally witnessed more than fifteen\\n\\nstabbings in his brief time at St. Clair. And Warren, who was supposed to protect\\n\\nhim, expressed outright hostility towards Marbury when he reported a threat against\\n\\n\\n 48\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 49 of 50\\n\\n\\nhim. She refused to help in any way, laughed and expressed glee after\\n\\nacknowledging the seriousness of his dilemma, and told him to get a “shank.” Five\\n\\ndays later, Marbury was repeatedly stabbed and beaten. Short of stabbing Marbury\\n\\nherself, it is hard to imagine how Warren could have been less helpful. A prison\\n\\nofficial does not need a court to tell her that this is not okay to know that it is not\\n\\nokay.\\n\\n Clearly, this is not a case where the good-faith efforts of an official charged\\n\\nwith making tough decisions turned out, in hindsight, to be insufficient. When I\\n\\nmake all reasonable inferences for Marbury, I must conclude Warren’s actions were\\n\\nnot only reckless, they were malicious. And no reasonable prison official could have\\n\\nbelieved that maliciously refusing to protect a prisoner from a known threat\\n\\ncomports with her Eighth Amendment duty to protect prisoners from harm.\\n\\nQualified immunity was never designed to protect actions like these. So Warren is\\n\\nnot entitled to qualified immunity on this ground as well. See Lee v. Ferraro, 284\\n\\nF.3d 1188, 1199 (11th Cir. 2002) (denying qualified immunity to an officer who\\n\\nused such grossly disproportionate force that no reasonable officer could have\\n\\nthought the action was legal).\\n\\n III.\\n\\n The Eighth Amendment does not allow prisons to be modern-day settings for\\n\\nLord of the Flies. When a prison official knows of a substantial threat of serious\\n\\n\\n 49\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 50 of 50\\n\\n\\nharm to an inmate, she must undertake reasonable action to protect that inmate. It\\n\\nshould go without saying that refusing to help in any way—and worse yet, laughing\\n\\nat the prisoner’s predicament and telling him to get a “shank”—is simply not an\\n\\noption. Yet by declining to allow Marbury to hold Warren and Estes responsible\\n\\nhere, the Majority Opinion condones this behavior and ensures it will occur again. I\\n\\ntherefore dissent.\\n\\n\\n\\n\\n 50\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434448/\", \"author_raw\": \"ROSENBAUM, Circuit Judge, dissenting\"}]}"},"judge1":{"kind":"string","value":"ROSENBAUM"},"judge2":{"kind":"string","value":"BRANCH"},"judge3":{"kind":"string","value":"HIGGINBOTHAM"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"ROSENBAUM\": \", Circuit\", \"BRANCH\": \", Circuit\", \"HIGGINBOTHAM\": \", Circuit\"}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":1,"string":"1"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4657195/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Mitchell Marbury v. Warden\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-29\nOpinion type: majority\nHas dissenting opinion: Yes\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nCase: 17-12589 Date Filed: 08/29/2019 Page: 1 of 50\n\n\n [PUBLISH]\n\n\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n ________________________\n\n No. 17-12589\n ________________________\n\n D.C. Docket No. 4:16-cv-01152-AKK-JHE\n\n\n\nMITCHELL MARBURY,\n\n Plaintiff-Appellant,\n\n versus\n\n\n\nWARDEN, et al.,\n\n Defendants-Appellees.\n\n ________________________\n\n Appeal from the United States District Court\n for the Northern District of Alabama\n ________________________\n\n (August 29, 2019)\n\nBefore ROSENBAUM, BRANCH, and HIGGINBOTHAM, * Circuit Judges.\n\n\n *\n Honorable Patrick E. Higginbotham, United States Circuit Judge for the Fifth Circuit,\nsitting by designation.\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 2 of 50\n\n\nPER CURIAM:\n Mitchell Marbury, a prisoner at Alabama’s St. Clair Correctional Facility,\n\nwas attacked by a fellow prisoner after making multiple requests to be transferred\n\nto a different dormitory or put in protective lock-up. He sued an officer in his cell\n\nblock and warden of the prison under 42 U.S.C. § 1983, arguing that both were\n\ndeliberately indifferent to a substantial risk to his safety. Marbury appeals the\n\ndistrict court’s grant of summary judgment to both defendants. We affirm.\n\n I\n\n The following reflects the facts as contained in the limited summary\n\njudgment record in this case. Between February and April 2016, Marbury\n\nrepeatedly attempted to be transferred to another dormitory. He sent a written\n\nrequest to Warden Dewayne Estes on February 12, 2016, stating that he had\n\nwitnessed over fifteen inmate-on-inmate stabbing incidents that appeared to be\n\ngang related and asking to be assigned to a “more sociable” living area with\n\ninmates closer to his own age. Estes did not act on this request. 1 Marbury’s sworn\n\ncomplaint avers that around the same time, he made several in-person transfer\n\nrequests to Officer Beverly Warren and that Warren told him that if he continued\n\nto make transfer requests, she “would personally see to it, that it be fixed, where\n\nshe’ll have a legal reason to deny [the] requests.” The complaint also alleges that\n\n\n 1\n Estes stated in his affidavit that he was unaware of the request.\n 2\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 3 of 50\n\n\nWarren made other comments like “[Y]ou don’t enjoy hanging out with the\n\n‘thugs,’ afraid you might get shanked!” and “I got the keys to the city baby, you[]\n\nlocked in.”\n\n On April 5, Marbury sent another letter to Estes asking why he had not heard\n\nback regarding his February 12 letter. 2 He said that he had seen prisoners\n\ndisrespect and attack security staff and felt “nothing is being done to correct the\n\nproblem,” and requested to be moved “from a[n] over-rated gang affiliated block to\n\na program block or one where [he could] feel safe and secure[ ].”\n\n On April 18, Marbury asked Warren to have the captain put him in lock-up\n\nuntil he could be transferred because he had heard from a friend that another\n\ninmate wanted to hurt him. He avers that Warren responded, “[D]o you really think\n\nI’ma act upon your requests, after you’ve filed complaints and requests against\n\nme,” started laughing, then said, “You don’t have a shank, . . . you need to get one,\n\n[because you aren’t] going to lock-up, there’s no cells available, so seem[s] like to\n\nme you’ve got a problem.” One day later, Marbury again wrote to Estes saying, “I\n\nwas told by a friend to watch my back, because he got word someone was out to do\n\nharm to me.” He requested that Estes place him in lock-up as soon as he received\n\nthe complaint because he was “in fear of [being] hurt or possibl[y] killed.” He also\n\n\n 2\n The day before, Marbury had sent a letter to another prison official—not a defendant in\nthis lawsuit—accusing Warren of retaliating against him for requesting to be transferred. While\nMarbury also brought a retaliation claim in the district court, he does not pursue it on appeal.\n 3\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 4 of 50\n\n\nreported to Estes that Warren had laughed at him when he told her about his\n\nconcerns and told him to get a knife. The captain in charge of placing inmates in\n\nlock-up says that she did not receive any information about Marbury’s request to\n\nbe locked up.\n\n On April 23, 2016, Marbury was stabbed and hit in the face in the prison’s\n\ndayroom. He was treated for multiple stab wounds and a broken nose. Prison staff\n\nstated that they were unable to identify who attacked him, though they learned\n\nfrom other inmates that he was stabbed because he had called another inmate’s\n\ngirlfriend.\n\n Marbury filed a pro se § 1983 complaint against Estes and Warren in the\n\nNorthern District of Alabama, alleging that they failed to protect him from unsafe\n\nconditions, were deliberately indifferent to those conditions, and retaliated against\n\nhim for exercising his constitutionally protected rights. The magistrate judge\n\nconstrued the defendants’ special report as a motion for summary judgment and\n\nissued a report and recommendation to grant them summary judgment on all\n\nclaims. Marbury objected on the deliberate-indifference issue. The district court\n\noverruled Marbury’s objections, adopted the report, and accepted its\n\nrecommendation. Marbury now appeals the grant of summary judgment to the\n\ndefendants.\n\n II\n\n\n 4\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 5 of 50\n\n\n We review the district court’s grant of summary judgment de novo,\n\n“view[ing] all the evidence and draw[ing] all reasonable inferences in the light\n\nmost favorable to the non-moving party.” 3 Summary judgment is warranted where\n\nthe evidence in the record “presents no genuine issue of material fact and compels\n\njudgment as a matter of law in favor of the moving party.” 4 Where, as here, an\n\ninmate proceeded pro se in the district court, his summary judgment pleadings are\n\nconstrued liberally and “specific facts” alleged in his sworn complaint can suffice\n\nto generate a genuine dispute of fact.5\n\n III\n\n “The defense of qualified immunity completely protects government\n\nofficials performing discretionary functions from suit in their individual capacities\n\nunless their conduct violates ‘clearly established statutory or constitutional rights\n\nof which a reasonable person would have known.’’ Gonzalez v. Reno, 325 F.3d\n\n1228, 1233 (11th Cir. 2003) (quoting Hope v. Pelzer, 536 U.S. 730, 739 (2002)).\n\nMarbury concedes that the defendants were performing their discretionary duties at\n\nall times relevant to this appeal. Once it has been determined that the official was\n\nacting within his discretionary duties, the burden shifts to the plaintiff to show (1)\n\n\n\n\n 3\n Caldwell v. Warden, 748 F.3d 1090, 1098 (11th Cir. 2014).\n 4\n Id. (quoting Owusu-Ansah v. Coca-Cola Co., 715 F.3d 1306, 1307 (11th Cir. 2013)).\n 5\n Id. (citing Perry v. Thompson, 786 F.2d 1093, 1095 (11th Cir. 1986); and Sammons v.\nTaylor, 967 F.2d 1533, 1545 n.5 (11th Cir. 1992)).\n 5\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 6 of 50\n\n\nthat the official violated a constitutional right and (2) that the right was clearly\n\nestablished at the time of the alleged violation. Caldwell, 748 F.3d at 1099. Our\n\ninquiry “can begin with either prong.” Morris v. Town of Lexington, 748 F.3d\n\n1316, 1322 (11th Cir. 2014).\n\n It is well settled that prison officials must “take reasonable measures to\n\nguarantee the safety of the inmates,” 6 and “[a] prison official violates the Eighth\n\nAmendment’s prohibition against cruel and unusual punishment if [the official] is\n\ndeliberately indifferent to a substantial risk of serious harm to an inmate who\n\nsuffers injury.” 7 Not “every injury suffered by one inmate at the hands of another,”\n\nhowever, “translates into a constitutional liability for prison officials responsible\n\nfor the victim’s safety.” 8 To establish a § 1983 claim for deliberate indifference, a\n\nplaintiff must show “(1) a substantial risk of serious harm; (2) the defendants’\n\ndeliberate indifference to that risk; and (3) causation.”9\n\n The first element of deliberate indifference—whether there was a substantial\n\nrisk of serious harm—is assessed objectively and requires the plaintiff to show\n\n“conditions that were extreme and posed an unreasonable risk of serious injury to\n\n\n\n 6\n Caldwell, 748 F.3d at 1099 (quoting Farmer v. Brennan, 511 U.S. 825, 832 (1994)).\n 7\n Lane v. Philbin, 835 F.3d 1302, 1307 (11th Cir. 2016).\n 8\n Carter v. Galloway, 352 F.3d 1346, 1349 (11th Cir. 2003) (per curiam) (quoting Farmer,\n511 U.S. at 834).\n 9\n Lane, 835 F.3d at 1307 (quoting Hale v. Tallapoosa Cty., 50 F.3d 1579, 1582 (11th Cir.\n1995)).\n\n\n 6\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 7 of 50\n\n\nhis future health or safety.” 10 The second element—whether the defendant was\n\ndeliberately indifferent to that risk—has both a subjective and an objective\n\ncomponent. Subjectively, the “official must both be aware of facts from which the\n\ninference could be drawn that a substantial risk of serious harm exists, and . . . also\n\ndraw the inference.”11 Objectively, the official must have responded to the known\n\nrisk in an unreasonable manner, in that he or she “knew of ways to reduce the\n\nharm” but knowingly or recklessly declined to act. 12 Finally, the plaintiff must\n\nshow a “necessary causal link” between the officer’s failure to act reasonably and\n\nthe plaintiff’s injury. 13\n\n Marbury argues that the defendants were subjectively aware of two distinct\n\ntypes of risk. The first is the general threat posed by inmate-on-inmate violence in\n\nMarbury’s cell block based on his statement that he had witnessed fifteen stabbings\n\nat the prison. The second is the threat Marbury identified in April 2016, when he\n\ntold the defendants that he had heard from a friend that someone intended to harm\n\nhim. He also argues that the defendants were deliberately indifferent in failing to\n\nfurther investigate his claims before the attack.\n\n\n\n\n 10\n Id.\n 11\n Rodriguez v. Sec’y for Dep’t of Corrs., 508 F.3d 611, 617 (11th Cir. 2007) (quoting\nFarmer, 511 U.S. at 837).\n 12\n Id. at 620 (quoting Hale, 50 F.3d at 1583).\n 13\n Id. at 622–23.\n 7\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 8 of 50\n\n\n As we explain, Marbury’s deliberate-indifference claim fails because he has\n\nnot demonstrated a genuine factual issue as to whether the defendants were\n\ndeliberately indifferent to a substantial risk of serious harm to Marbury. Since\n\nMarbury has not met his burden to show the violation of a constitutional right, we\n\nneed not proceed past step one of the qualified-immunity analysis.\n\n A\n\n Marbury repeatedly asked to be transferred because he was concerned about\n\na general lack of safety in his cell block. In his letters to Estes and the verbal\n\nrequests he says he made to Warren, he told them that he had witnessed fifteen\n\ninmate-on-inmate stabbings that he attributed to gang affiliations, expressing fear\n\nfor his own safety. Granting Marbury the resolution of all disputed facts in his\n\nfavor, our caselaw nevertheless establishes that this evidence was insufficient to\n\nestablish deliberate indifference to a substantial risk of serious harm.\n\n In general, a plaintiff must show “more than a generalized awareness of\n\nrisk” to make out a deliberate-indifference claim. 14 While “occasional, isolated\n\nattacks by one prisoner on another may not constitute cruel and unusual\n\npunishment, . . . . confinement in a prison where violence and terror reign is\n\n\n\n\n 14\n Caldwell, 748 F.3d at 1101 (internal quotation marks omitted).\n\n\n 8\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 9 of 50\n\n\nactionable.” 15 To establish deliberate indifference based on a generalized risk, the\n\nplaintiff must show “that serious inmate-on-inmate violence was the norm or\n\nsomething close to it.” 16\n\n Turning to this case, we look to only those facts contained in the summary\n\njudgment record. While Marbury’s sworn allegation may support the inference that\n\nMarbury faced some generalized risk of attack, such evidence does not support the\n\nconclusion that serious inmate-on-inmate violence was so pervasive that it\n\nconstitutes a substantial risk of serious harm to which defendants were deliberately\n\nindifferent. The only allegation Marbury makes about inmate-on-inmate violence\n\nis his statement that he personally witnessed fifteen inmate-on-inmate stabbings\n\nduring his time at St. Clair. 17 There is no evidence in the record of the total prison\n\npopulation or the sections of the prison in which the attacks occurred that would\n\nplace Marbury’s statement in context. Also, it is not precisely clear from the record\n\nover what period of time these incidents occurred. But we can tell from the record\n\nthat Marbury was at St. Clair between 2002 and 2007, went to a different\n\n\n\n\n 15\n Purcell ex rel. Estate of Morgan v. Toombs Cty., 400 F.3d 1313, 1320 (11th Cir. 2005)\n(quoting Woodhous v. Virginia, 487 F.2d 889, 890 (4th Cir. 1973)); accord Harrison v. Culliver,\n746 F.3d 1288, 1299 (11th Cir. 2014).\n 16\n Purcell, 400 F.3d at 1322.\n 17\n In Marbury’s objection to the report and recommendation, he argued that additional\ndiscovery would substantiate his claim of widespread violence “a year prior to[ ] and a year\nafter” Marbury’s departure from St. Clair. His affidavit also stated he had witnessed over fifteen\nstabbings over the “short time” he had been at St. Clair.\n 9\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 10 of 50\n\n\ncorrectional facility, and then returned in 2015. So, the fifteen stabbings may have\n\noccurred over the course of 6 years, for a rate of 2.5 per year.\n\n Marbury’s allegation stands in sharp contrast to those at issue in Harrison v.\n\nCulliver, for example, where we held that even though a prison warden was on\n\nnotice that inmate-on-inmate assaults occurred throughout the prison and in a\n\nparticular location where a prisoner was attacked, “the evidence of inmate-on-\n\ninmate assault involving weapons [did] not . . . indicate that inmates were ‘exposed\n\nto something even approaching the constant threat of violence.’” 18 We emphasized\n\nthat the institution in Harrison was large, housing between 830 and 990 inmates\n\nbetween 2005 and 2008, and the fact that there had been thirty-three incidents\n\ninvolving weapons during the same time period was “hardly sufficient to\n\ndemonstrate that [the institution] was a prison ‘where violence and terror reign.’” 19\n\nAnd Marbury has presented far less evidence about the level of violence than the\n\nevidence presented in Harrison that we found “hardly sufficient” to establish\n\ndeliberate indifference to a substantial risk of serious harm.\n\n Moreover, when we have held that a generalized risk of violence from a\n\nprison population could support a claim of deliberate indifference to a substantial\n\nrisk of serious harm, the plaintiff has pointed to specific features of a facility or its\n\n\n\n 18\n 746 F.3d at 1299–1300 (quoting Purcell, 400 F.3d at 1321).\n 19\n Id. at 1300 (quoting Purcell, 400 F.3d at 1320).\n 10\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 11 of 50\n\n\npopulation rendering it particularly violent. This evidence has included pervasive\n\nstaffing and logistical issues rendering prison officials unable to address near-\n\nconstant violence, 20 tensions between different subsets of a prison population,21\n\nand unique risks posed by individual prisoners or groups of prisoners due to\n\ncharacteristics like mental illness. 22 Even if Marbury had shown a risk of\n\ngeneralized prison violence, he has made no allegations regarding the specific\n\nfeatures of the prison that would make it particularly violent.\n\n We emphasize that the record before the district court was limited. 23 While\n\nwe are sensitive to Marbury’s pro se status before the district court, the evidence\n\nMarbury has presented regarding a general risk of inmate-on-inmate violence does\n\nnot rise to the level necessary to show deliberate indifference to a substantial risk\n\n\n\n\n 20\n See, e.g., Lane, 835 F.3d at 1307–08 (finding allegations that only one officer\nsupervised two separate dorms, that inmates regularly brought back weapons from their work\ndetail and fashioned them from prison materials, and that officials did not confiscate weapons\nsufficient); Hale, 50 F.3d at 1582–83 (finding potential awareness of a substantial risk of serious\nharm where a defendant was aware that a prison had severe overcrowding problems and the\nplaintiff presented evidence that “inmate-on-inmate violence occurred regularly when the jail\nwas overcrowded”).\n 21\n See Lane, 835 F.3d at 1307–08 (finding potential awareness of a substantial risk of\nserious harm where a plaintiff alleged that a particular prison building was composed of 90%\ngang members, it was common for the non-gang-affiliate inmates or non-Muslim inmates to be\nrobbed or stabbed, and the prison had inadequate supervision to prevent inmates from making\nhomemade weapons).\n 22\n See Cottone v. Jenne, 326 F.3d 1352, 1355–56, 1358–59 (11th Cir. 2003) (finding\npotential awareness of a substantial risk of serious harm where mentally ill inmates were\nseparated from the general population but kept in unlocked cells where they could interact with\neach other, and guards were aware of a particular prisoner’s history of violent schizophrenic\noutbursts).\n 23\n Marbury does not argue that the district court improperly limited discovery.\n 11\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 12 of 50\n\n\nof serious harm required by our caselaw. This sparse record at most shows that\n\ninmates at St. Clair faced some risk of assaults by fellow prisoners, but we have\n\nsaid that some risk of harm is insufficient. Marbury has thus failed to produce\n\nevidence that he was in an environment so beset by violence that confinement, by\n\nits nature, threatened him with the substantial risk of serious harm.\n\n B\n\n Marbury also argues that even if the defendants were not deliberately\n\nindifferent to a general lack of safety in his cell block, they were at least\n\ndeliberately indifferent to the more specific threat he warned them about in April\n\n2016—that he had heard from a friend that someone intended to harm him. As we\n\nhave explained, Marbury’s sworn complaint alleges that when he told Warren\n\nabout this threat she laughed at him, told him she was not going to act on his\n\nrequest because he had filed internal complaints against her, and told him to get a\n\nknife or a shank if he did not already have one. Marbury then wrote another letter\n\nto Estes telling Estes about the threat and Warren’s response, requesting to be\n\nplaced in lock-up, and stating that he was afraid of being hurt or possibly killed.\n\nOn the summary judgment record, Marbury appears to have provided no further\n\ninformation to either defendant about the nature of the threat.\n\n We must therefore decide whether a reasonable jury could find Marbury’s\n\nstatement that he had heard from a friend that an unnamed prisoner intended to\n\n\n 12\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 13 of 50\n\n\nhurt him, and that he was afraid of being hurt or killed, without any further details,\n\nsufficient to make the defendants aware of a substantial risk of serious harm. While\n\nthis question is a close one, we conclude that our precedent does not allow\n\nMarbury’s deliberate-indifference claim to proceed.\n\n On the one hand, it is settled that “a prison official [cannot] escape liability\n\nfor deliberate indifference by showing that . . . he did not know that the\n\ncomplainant was especially likely to be assaulted by the specific prisoner who\n\neventually committed the assault,” as long as the official was otherwise aware that\n\nthe victim faced a substantial risk of serious harm. 24 Our caselaw also establishes,\n\nhowever, that officials must possess enough details about a threat to enable them to\n\nconclude that it presents a “strong likelihood” of injury, not a “mere possibility.” 25\n\nThe unfortunate reality is that “threats between inmates are common and do not,\n\nunder all circumstances, serve to impute actual knowledge of a substantial risk of\n\nharm.” 26\n\n We have therefore upheld dismissal of or summary judgment against\n\ndeliberate-indifference claims where, although a plaintiff told prison officials about\n\na threat by another inmate or inmates, the prison officials were not deliberately\n\n\n\n\n 24\n Rodriguez, 508 F.3d at 619 (quoting Farmer, 511 U.S. at 843) (alterations in original).\n 25\n E.g., Brooks v. Warden, 800 F.3d 1295, 1301 (11th Cir. 2015); Brown v. Hughes, 894\nF.2d 1533, 1537 (11th Cir. 1990).\n 26\n Prater v. Dahm, 89 F.3d 538, 542 (8th Cir. 1996).\n 13\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 14 of 50\n\n\nindifferent to a substantial risk of harm. For example, we held that a plaintiff had\n\nfailed to allege plausibly that prison officials were deliberately indifferent to a\n\nstrong likelihood of serious harm where he was attacked by an inmate in the next\n\ncell when all of the cell doors opened and a riot began.27 We noted that even\n\nthough the plaintiff alleged that the attacker had threatened him previously and that\n\nan attack was possible, he had not alleged any information showing a strong\n\nlikelihood that he and his attacker would be released at the same time in a chaotic\n\nenvironment. 28 Successful deliberate-indifference claims will generally require\n\nsome further reason—beyond the plaintiff having informed the defendant officers\n\nof the threat—that a prison official could have concluded that a particular threat\n\nevidenced a substantial threat, rather than the mere possibility, of serious harm.\n\n For example, a plaintiff can establish deliberate indifference to a substantial\n\nrisk of serious harm where he has given prison officials further information\n\nenabling them to conclude that the risk was substantial and not merely possible. In\n\nRodriguez v. Secretary for Department of Corrections, a prisoner informed prison\n\nstaff that members of his former gang had threatened to kill him upon release into\n\nthe general prison population. 29 We concluded that if true, this allegation was\n\n\n\n\n 27\n Brooks v. Warden, 800 F.3d 1295, 1298–99, 1301 (11th Cir. 2015).\n 28\n Id. at 1301.\n 29\n See Rodriguez, 508 F.3d at 612–15.\n\n\n 14\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 15 of 50\n\n\nenough to place the defendants on notice of a substantial risk of serious harm—so\n\nwe allowed the case to proceed.30 The key distinction between Rodriguez and\n\nMarbury’s case is that Rodriguez involved a series of threats that contained enough\n\nspecific information—the threats came from members of the plaintiff’s former\n\ngang—that prison officials could have reasonably inferred that there was a\n\nsubstantial, not merely possible, risk of harm. In fact, Rodriguez observed that a\n\nvague statement like “I have a problem with another inmate in this compound,”\n\nabsent some information “about the nature of the anticipated risk,” would not have\n\ncreated a genuine issue of fact regarding deliberate indifference to a substantial\n\nrisk of serious harm. 31 Marbury’s statement that an another inmate told him\n\nanother inmate intended to harm him is precisely this type of vague statement that\n\nconveys nothing about the nature of the anticipated risk that we cautioned in\n\nRodriguez would not rise to the level of deliberate indifference to a substantial risk.\n\n Marbury’s argument is essentially that every prisoner who tells prison\n\nofficials about an unspecified threat from an unspecified inmate without more is\n\nentitled to protective custody or a transfer. But, as already explained, our caselaw\n\nestablishes a higher standard for deliberate indifference. To be clear, Marbury was\n\n\n\n\n 30\n Id. at 621–22.\n 31\n Rodriguez, 508 F.3d at 619 n.15.\n\n\n 15\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 16 of 50\n\n\nnot required to identify the person who was threatening him by name, 32 or even\n\nnecessarily to give the defendants advance notice of a potential attack,33 so long as\n\nother facts put the defendants on notice that he faced a substantial risk of serious\n\nharm. It may be possible for a general threat of inmate-on-inmate violence in a\n\nprison to bolster an otherwise insufficient unspecified threat of harm. But, as\n\nalready discussed, Marbury has not shown anything close to such a substantial\n\nthreat from the generally violent nature of the prison environment. And because\n\nMarbury has not presented anything else that would bolster the unspecified threat,\n\nhe has not met the requirement of showing deliberate indifference to a substantial\n\nrisk of serious harm.\n\n Marbury further emphasizes that when he told Warren about the threat from\n\nan unidentified inmate, she laughed at him, told him she was not going to act on\n\nhis request because he had filed complaints against her, and advised him to get a\n\nknife or a shank because it seemed as though he had a “problem.” He suggests that\n\nthis reaction demonstrated her awareness that he faced a substantial risk. The\n\ndistrict court did not explicitly address Marbury’s argument, noting that\n\n“[a]ccording to defendant Warren, she never said this and asserts that she did not\n\nknow the plaintiff was in danger of being attacked by another inmate on April 23,\n\n\n\n 32\n See, e.g., Rodriguez, 508 F.3d at 619 (discussing Farmer, 511 U.S. at 843).\n 33\n See Farmer, 511 U.S. at 848.\n 16\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 17 of 50\n\n\n2016.” Again, “specific facts” alleged in a pro se plaintiff’s sworn complaint can\n\nsuffice to generate a genuine factual issue. 34 We must therefore accept as true, for\n\nthe purposes of summary judgment, that Warren made these statements.\n\n But the statements ultimately do not affect Warren’s entitlement to summary\n\njudgment. The Supreme Court has made clear that “[w]hether a prison official had\n\nthe requisite knowledge of a substantial risk is a question of fact subject to\n\ndemonstration in the usual ways, including inference from circumstantial\n\nevidence.”35 While Marbury can establish Warren’s subjective knowledge through\n\ncircumstantial proof, he must nevertheless present evidence sufficient to “support a\n\nreasonable jury’s finding that [Warren] harbored a subjective awareness that\n\n[Marbury] was in serious danger.”36 Viewed in the light most favorable to\n\nMarbury, the summary judgment record allows one to conclude that Warren was\n\naware of the potential for inmate-on-inmate violence at the prison, Marbury told\n\nher that another inmate was looking to harm him, and she responded by laughing at\n\nMarbury and telling him to get a knife because he had a “problem.” A reasonable\n\njury could not extrapolate from such statements that Warren was subjectively\n\n\n\n\n 34\n See Caldwell, 748 F.3d at 1098.\n 35\n Rodriguez, 508 F.3d at 617 (quoting Farmer, 511 U.S. at 842).\n 36\n Goodman v. Kimbrough, 718 F.3d 1325, 1332 (11th Cir. 2013).\n\n\n 17\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 18 of 50\n\n\naware of a substantial risk of serious harm to Marbury at the time. Indeed,\n\nMarbury concedes in his briefing that Warren’s statements were callous jokes.\n\n Again, it bears mention that subjective awareness of only some risk of harm\n\nto a prisoner is insufficient for a deliberate-indifference claim. 37 At most, the\n\nevidence Marbury has put forward would allow a jury to conclude that Warren was\n\nput on notice that Marbury faced some unspecified risk of harm to his well-\n\nbeing—not that she was aware he faced the type of substantial risk of serious harm\n\nnecessary to establish deliberate indifference. Marbury has not marshaled enough\n\nevidence to establish a genuine issue of fact on this necessary element of his case.\n\n C\n\n Finally, we turn to Marbury’s suggestion that the defendants were\n\ndeliberately indifferent in failing to investigate his report that someone was out to\n\nharm him or in otherwise failing to abide by prison policy. We have explained that\n\n“merely negligent failure to protect an inmate from attack does not justify liability\n\nunder section 1983.”38 To allow Marbury’s deliberate-indifference claim to\n\nproceed absent sufficient evidence that the defendants were subjectively aware that\n\nhe faced a substantial risk of serious harm would elide the “subtle distinction”\n\nbetween deliberate indifference and mere negligence.39 We cannot condone the\n\n\n 37\n See, e.g., Brooks, 800 F.3d at 1301; Brown, 894 F.2d at 1537.\n 38\n Carter, 352 F.3d at 1350.\n 39\n See Goodman, 718 F.3d at 1333–34.\n 18\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 19 of 50\n\n\nfailure to investigate inmates’ allegations of threats or to follow policy in reporting\n\npotential threats up the chain of command. But our caselaw does not allow these\n\nfailures, without corresponding subjective awareness of a serious risk of harm, to\n\nestablish deliberate indifference.\n\n IV\n\n We affirm the judgment of the district court.\n\n\n\n\n 19\n\n Case: 17-12589 Date Filed: 08/29/2019 Page: 20 of 50\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Mitchell Marbury v. Warden\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-29\\nOpinion type: majority\\nHas dissenting opinion: Yes\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 17-12589 Date Filed: 08/29/2019 Page: 1 of 50\\n\\n\\n [PUBLISH]\\n\\n\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 17-12589\\n ________________________\\n\\n D.C. Docket No. 4:16-cv-01152-AKK-JHE\\n\\n\\n\\nMITCHELL MARBURY,\\n\\n Plaintiff-Appellant,\\n\\n versus\\n\\n\\n\\nWARDEN, et al.,\\n\\n Defendants-Appellees.\\n\\n ________________________\\n\\n Appeal from the United States District Court\\n for the Northern District of Alabama\\n ________________________\\n\\n (August 29, 2019)\\n\\nBefore ROSENBAUM, BRANCH, and HIGGINBOTHAM, * Circuit Judges.\\n\\n\\n *\\n Honorable Patrick E. Higginbotham, United States Circuit Judge for the Fifth Circuit,\\nsitting by designation.\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 2 of 50\\n\\n\\nPER CURIAM:\\n Mitchell Marbury, a prisoner at Alabama’s St. Clair Correctional Facility,\\n\\nwas attacked by a fellow prisoner after making multiple requests to be transferred\\n\\nto a different dormitory or put in protective lock-up. He sued an officer in his cell\\n\\nblock and warden of the prison under 42 U.S.C. § 1983, arguing that both were\\n\\ndeliberately indifferent to a substantial risk to his safety. Marbury appeals the\\n\\ndistrict court’s grant of summary judgment to both defendants. We affirm.\\n\\n I\\n\\n The following reflects the facts as contained in the limited summary\\n\\njudgment record in this case. Between February and April 2016, Marbury\\n\\nrepeatedly attempted to be transferred to another dormitory. He sent a written\\n\\nrequest to Warden Dewayne Estes on February 12, 2016, stating that he had\\n\\nwitnessed over fifteen inmate-on-inmate stabbing incidents that appeared to be\\n\\ngang related and asking to be assigned to a “more sociable” living area with\\n\\ninmates closer to his own age. Estes did not act on this request. 1 Marbury’s sworn\\n\\ncomplaint avers that around the same time, he made several in-person transfer\\n\\nrequests to Officer Beverly Warren and that Warren told him that if he continued\\n\\nto make transfer requests, she “would personally see to it, that it be fixed, where\\n\\nshe’ll have a legal reason to deny [the] requests.” The complaint also alleges that\\n\\n\\n 1\\n Estes stated in his affidavit that he was unaware of the request.\\n 2\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 3 of 50\\n\\n\\nWarren made other comments like “[Y]ou don’t enjoy hanging out with the\\n\\n‘thugs,’ afraid you might get shanked!” and “I got the keys to the city baby, you[]\\n\\nlocked in.”\\n\\n On April 5, Marbury sent another letter to Estes asking why he had not heard\\n\\nback regarding his February 12 letter. 2 He said that he had seen prisoners\\n\\ndisrespect and attack security staff and felt “nothing is being done to correct the\\n\\nproblem,” and requested to be moved “from a[n] over-rated gang affiliated block to\\n\\na program block or one where [he could] feel safe and secure[ ].”\\n\\n On April 18, Marbury asked Warren to have the captain put him in lock-up\\n\\nuntil he could be transferred because he had heard from a friend that another\\n\\ninmate wanted to hurt him. He avers that Warren responded, “[D]o you really think\\n\\nI’ma act upon your requests, after you’ve filed complaints and requests against\\n\\nme,” started laughing, then said, “You don’t have a shank, . . . you need to get one,\\n\\n[because you aren’t] going to lock-up, there’s no cells available, so seem[s] like to\\n\\nme you’ve got a problem.” One day later, Marbury again wrote to Estes saying, “I\\n\\nwas told by a friend to watch my back, because he got word someone was out to do\\n\\nharm to me.” He requested that Estes place him in lock-up as soon as he received\\n\\nthe complaint because he was “in fear of [being] hurt or possibl[y] killed.” He also\\n\\n\\n 2\\n The day before, Marbury had sent a letter to another prison official—not a defendant in\\nthis lawsuit—accusing Warren of retaliating against him for requesting to be transferred. While\\nMarbury also brought a retaliation claim in the district court, he does not pursue it on appeal.\\n 3\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 4 of 50\\n\\n\\nreported to Estes that Warren had laughed at him when he told her about his\\n\\nconcerns and told him to get a knife. The captain in charge of placing inmates in\\n\\nlock-up says that she did not receive any information about Marbury’s request to\\n\\nbe locked up.\\n\\n On April 23, 2016, Marbury was stabbed and hit in the face in the prison’s\\n\\ndayroom. He was treated for multiple stab wounds and a broken nose. Prison staff\\n\\nstated that they were unable to identify who attacked him, though they learned\\n\\nfrom other inmates that he was stabbed because he had called another inmate’s\\n\\ngirlfriend.\\n\\n Marbury filed a pro se § 1983 complaint against Estes and Warren in the\\n\\nNorthern District of Alabama, alleging that they failed to protect him from unsafe\\n\\nconditions, were deliberately indifferent to those conditions, and retaliated against\\n\\nhim for exercising his constitutionally protected rights. The magistrate judge\\n\\nconstrued the defendants’ special report as a motion for summary judgment and\\n\\nissued a report and recommendation to grant them summary judgment on all\\n\\nclaims. Marbury objected on the deliberate-indifference issue. The district court\\n\\noverruled Marbury’s objections, adopted the report, and accepted its\\n\\nrecommendation. Marbury now appeals the grant of summary judgment to the\\n\\ndefendants.\\n\\n II\\n\\n\\n 4\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 5 of 50\\n\\n\\n We review the district court’s grant of summary judgment de novo,\\n\\n“view[ing] all the evidence and draw[ing] all reasonable inferences in the light\\n\\nmost favorable to the non-moving party.” 3 Summary judgment is warranted where\\n\\nthe evidence in the record “presents no genuine issue of material fact and compels\\n\\njudgment as a matter of law in favor of the moving party.” 4 Where, as here, an\\n\\ninmate proceeded pro se in the district court, his summary judgment pleadings are\\n\\nconstrued liberally and “specific facts” alleged in his sworn complaint can suffice\\n\\nto generate a genuine dispute of fact.5\\n\\n III\\n\\n “The defense of qualified immunity completely protects government\\n\\nofficials performing discretionary functions from suit in their individual capacities\\n\\nunless their conduct violates ‘clearly established statutory or constitutional rights\\n\\nof which a reasonable person would have known.’’ Gonzalez v. Reno, 325 F.3d\\n\\n1228, 1233 (11th Cir. 2003) (quoting Hope v. Pelzer, 536 U.S. 730, 739 (2002)).\\n\\nMarbury concedes that the defendants were performing their discretionary duties at\\n\\nall times relevant to this appeal. Once it has been determined that the official was\\n\\nacting within his discretionary duties, the burden shifts to the plaintiff to show (1)\\n\\n\\n\\n\\n 3\\n Caldwell v. Warden, 748 F.3d 1090, 1098 (11th Cir. 2014).\\n 4\\n Id. (quoting Owusu-Ansah v. Coca-Cola Co., 715 F.3d 1306, 1307 (11th Cir. 2013)).\\n 5\\n Id. (citing Perry v. Thompson, 786 F.2d 1093, 1095 (11th Cir. 1986); and Sammons v.\\nTaylor, 967 F.2d 1533, 1545 n.5 (11th Cir. 1992)).\\n 5\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 6 of 50\\n\\n\\nthat the official violated a constitutional right and (2) that the right was clearly\\n\\nestablished at the time of the alleged violation. Caldwell, 748 F.3d at 1099. Our\\n\\ninquiry “can begin with either prong.” Morris v. Town of Lexington, 748 F.3d\\n\\n1316, 1322 (11th Cir. 2014).\\n\\n It is well settled that prison officials must “take reasonable measures to\\n\\nguarantee the safety of the inmates,” 6 and “[a] prison official violates the Eighth\\n\\nAmendment’s prohibition against cruel and unusual punishment if [the official] is\\n\\ndeliberately indifferent to a substantial risk of serious harm to an inmate who\\n\\nsuffers injury.” 7 Not “every injury suffered by one inmate at the hands of another,”\\n\\nhowever, “translates into a constitutional liability for prison officials responsible\\n\\nfor the victim’s safety.” 8 To establish a § 1983 claim for deliberate indifference, a\\n\\nplaintiff must show “(1) a substantial risk of serious harm; (2) the defendants’\\n\\ndeliberate indifference to that risk; and (3) causation.”9\\n\\n The first element of deliberate indifference—whether there was a substantial\\n\\nrisk of serious harm—is assessed objectively and requires the plaintiff to show\\n\\n“conditions that were extreme and posed an unreasonable risk of serious injury to\\n\\n\\n\\n 6\\n Caldwell, 748 F.3d at 1099 (quoting Farmer v. Brennan, 511 U.S. 825, 832 (1994)).\\n 7\\n Lane v. Philbin, 835 F.3d 1302, 1307 (11th Cir. 2016).\\n 8\\n Carter v. Galloway, 352 F.3d 1346, 1349 (11th Cir. 2003) (per curiam) (quoting Farmer,\\n511 U.S. at 834).\\n 9\\n Lane, 835 F.3d at 1307 (quoting Hale v. Tallapoosa Cty., 50 F.3d 1579, 1582 (11th Cir.\\n1995)).\\n\\n\\n 6\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 7 of 50\\n\\n\\nhis future health or safety.” 10 The second element—whether the defendant was\\n\\ndeliberately indifferent to that risk—has both a subjective and an objective\\n\\ncomponent. Subjectively, the “official must both be aware of facts from which the\\n\\ninference could be drawn that a substantial risk of serious harm exists, and . . . also\\n\\ndraw the inference.”11 Objectively, the official must have responded to the known\\n\\nrisk in an unreasonable manner, in that he or she “knew of ways to reduce the\\n\\nharm” but knowingly or recklessly declined to act. 12 Finally, the plaintiff must\\n\\nshow a “necessary causal link” between the officer’s failure to act reasonably and\\n\\nthe plaintiff’s injury. 13\\n\\n Marbury argues that the defendants were subjectively aware of two distinct\\n\\ntypes of risk. The first is the general threat posed by inmate-on-inmate violence in\\n\\nMarbury’s cell block based on his statement that he had witnessed fifteen stabbings\\n\\nat the prison. The second is the threat Marbury identified in April 2016, when he\\n\\ntold the defendants that he had heard from a friend that someone intended to harm\\n\\nhim. He also argues that the defendants were deliberately indifferent in failing to\\n\\nfurther investigate his claims before the attack.\\n\\n\\n\\n\\n 10\\n Id.\\n 11\\n Rodriguez v. Sec’y for Dep’t of Corrs., 508 F.3d 611, 617 (11th Cir. 2007) (quoting\\nFarmer, 511 U.S. at 837).\\n 12\\n Id. at 620 (quoting Hale, 50 F.3d at 1583).\\n 13\\n Id. at 622–23.\\n 7\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 8 of 50\\n\\n\\n As we explain, Marbury’s deliberate-indifference claim fails because he has\\n\\nnot demonstrated a genuine factual issue as to whether the defendants were\\n\\ndeliberately indifferent to a substantial risk of serious harm to Marbury. Since\\n\\nMarbury has not met his burden to show the violation of a constitutional right, we\\n\\nneed not proceed past step one of the qualified-immunity analysis.\\n\\n A\\n\\n Marbury repeatedly asked to be transferred because he was concerned about\\n\\na general lack of safety in his cell block. In his letters to Estes and the verbal\\n\\nrequests he says he made to Warren, he told them that he had witnessed fifteen\\n\\ninmate-on-inmate stabbings that he attributed to gang affiliations, expressing fear\\n\\nfor his own safety. Granting Marbury the resolution of all disputed facts in his\\n\\nfavor, our caselaw nevertheless establishes that this evidence was insufficient to\\n\\nestablish deliberate indifference to a substantial risk of serious harm.\\n\\n In general, a plaintiff must show “more than a generalized awareness of\\n\\nrisk” to make out a deliberate-indifference claim. 14 While “occasional, isolated\\n\\nattacks by one prisoner on another may not constitute cruel and unusual\\n\\npunishment, . . . . confinement in a prison where violence and terror reign is\\n\\n\\n\\n\\n 14\\n Caldwell, 748 F.3d at 1101 (internal quotation marks omitted).\\n\\n\\n 8\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 9 of 50\\n\\n\\nactionable.” 15 To establish deliberate indifference based on a generalized risk, the\\n\\nplaintiff must show “that serious inmate-on-inmate violence was the norm or\\n\\nsomething close to it.” 16\\n\\n Turning to this case, we look to only those facts contained in the summary\\n\\njudgment record. While Marbury’s sworn allegation may support the inference that\\n\\nMarbury faced some generalized risk of attack, such evidence does not support the\\n\\nconclusion that serious inmate-on-inmate violence was so pervasive that it\\n\\nconstitutes a substantial risk of serious harm to which defendants were deliberately\\n\\nindifferent. The only allegation Marbury makes about inmate-on-inmate violence\\n\\nis his statement that he personally witnessed fifteen inmate-on-inmate stabbings\\n\\nduring his time at St. Clair. 17 There is no evidence in the record of the total prison\\n\\npopulation or the sections of the prison in which the attacks occurred that would\\n\\nplace Marbury’s statement in context. Also, it is not precisely clear from the record\\n\\nover what period of time these incidents occurred. But we can tell from the record\\n\\nthat Marbury was at St. Clair between 2002 and 2007, went to a different\\n\\n\\n\\n\\n 15\\n Purcell ex rel. Estate of Morgan v. Toombs Cty., 400 F.3d 1313, 1320 (11th Cir. 2005)\\n(quoting Woodhous v. Virginia, 487 F.2d 889, 890 (4th Cir. 1973)); accord Harrison v. Culliver,\\n746 F.3d 1288, 1299 (11th Cir. 2014).\\n 16\\n Purcell, 400 F.3d at 1322.\\n 17\\n In Marbury’s objection to the report and recommendation, he argued that additional\\ndiscovery would substantiate his claim of widespread violence “a year prior to[ ] and a year\\nafter” Marbury’s departure from St. Clair. His affidavit also stated he had witnessed over fifteen\\nstabbings over the “short time” he had been at St. Clair.\\n 9\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 10 of 50\\n\\n\\ncorrectional facility, and then returned in 2015. So, the fifteen stabbings may have\\n\\noccurred over the course of 6 years, for a rate of 2.5 per year.\\n\\n Marbury’s allegation stands in sharp contrast to those at issue in Harrison v.\\n\\nCulliver, for example, where we held that even though a prison warden was on\\n\\nnotice that inmate-on-inmate assaults occurred throughout the prison and in a\\n\\nparticular location where a prisoner was attacked, “the evidence of inmate-on-\\n\\ninmate assault involving weapons [did] not . . . indicate that inmates were ‘exposed\\n\\nto something even approaching the constant threat of violence.’” 18 We emphasized\\n\\nthat the institution in Harrison was large, housing between 830 and 990 inmates\\n\\nbetween 2005 and 2008, and the fact that there had been thirty-three incidents\\n\\ninvolving weapons during the same time period was “hardly sufficient to\\n\\ndemonstrate that [the institution] was a prison ‘where violence and terror reign.’” 19\\n\\nAnd Marbury has presented far less evidence about the level of violence than the\\n\\nevidence presented in Harrison that we found “hardly sufficient” to establish\\n\\ndeliberate indifference to a substantial risk of serious harm.\\n\\n Moreover, when we have held that a generalized risk of violence from a\\n\\nprison population could support a claim of deliberate indifference to a substantial\\n\\nrisk of serious harm, the plaintiff has pointed to specific features of a facility or its\\n\\n\\n\\n 18\\n 746 F.3d at 1299–1300 (quoting Purcell, 400 F.3d at 1321).\\n 19\\n Id. at 1300 (quoting Purcell, 400 F.3d at 1320).\\n 10\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 11 of 50\\n\\n\\npopulation rendering it particularly violent. This evidence has included pervasive\\n\\nstaffing and logistical issues rendering prison officials unable to address near-\\n\\nconstant violence, 20 tensions between different subsets of a prison population,21\\n\\nand unique risks posed by individual prisoners or groups of prisoners due to\\n\\ncharacteristics like mental illness. 22 Even if Marbury had shown a risk of\\n\\ngeneralized prison violence, he has made no allegations regarding the specific\\n\\nfeatures of the prison that would make it particularly violent.\\n\\n We emphasize that the record before the district court was limited. 23 While\\n\\nwe are sensitive to Marbury’s pro se status before the district court, the evidence\\n\\nMarbury has presented regarding a general risk of inmate-on-inmate violence does\\n\\nnot rise to the level necessary to show deliberate indifference to a substantial risk\\n\\n\\n\\n\\n 20\\n See, e.g., Lane, 835 F.3d at 1307–08 (finding allegations that only one officer\\nsupervised two separate dorms, that inmates regularly brought back weapons from their work\\ndetail and fashioned them from prison materials, and that officials did not confiscate weapons\\nsufficient); Hale, 50 F.3d at 1582–83 (finding potential awareness of a substantial risk of serious\\nharm where a defendant was aware that a prison had severe overcrowding problems and the\\nplaintiff presented evidence that “inmate-on-inmate violence occurred regularly when the jail\\nwas overcrowded”).\\n 21\\n See Lane, 835 F.3d at 1307–08 (finding potential awareness of a substantial risk of\\nserious harm where a plaintiff alleged that a particular prison building was composed of 90%\\ngang members, it was common for the non-gang-affiliate inmates or non-Muslim inmates to be\\nrobbed or stabbed, and the prison had inadequate supervision to prevent inmates from making\\nhomemade weapons).\\n 22\\n See Cottone v. Jenne, 326 F.3d 1352, 1355–56, 1358–59 (11th Cir. 2003) (finding\\npotential awareness of a substantial risk of serious harm where mentally ill inmates were\\nseparated from the general population but kept in unlocked cells where they could interact with\\neach other, and guards were aware of a particular prisoner’s history of violent schizophrenic\\noutbursts).\\n 23\\n Marbury does not argue that the district court improperly limited discovery.\\n 11\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 12 of 50\\n\\n\\nof serious harm required by our caselaw. This sparse record at most shows that\\n\\ninmates at St. Clair faced some risk of assaults by fellow prisoners, but we have\\n\\nsaid that some risk of harm is insufficient. Marbury has thus failed to produce\\n\\nevidence that he was in an environment so beset by violence that confinement, by\\n\\nits nature, threatened him with the substantial risk of serious harm.\\n\\n B\\n\\n Marbury also argues that even if the defendants were not deliberately\\n\\nindifferent to a general lack of safety in his cell block, they were at least\\n\\ndeliberately indifferent to the more specific threat he warned them about in April\\n\\n2016—that he had heard from a friend that someone intended to harm him. As we\\n\\nhave explained, Marbury’s sworn complaint alleges that when he told Warren\\n\\nabout this threat she laughed at him, told him she was not going to act on his\\n\\nrequest because he had filed internal complaints against her, and told him to get a\\n\\nknife or a shank if he did not already have one. Marbury then wrote another letter\\n\\nto Estes telling Estes about the threat and Warren’s response, requesting to be\\n\\nplaced in lock-up, and stating that he was afraid of being hurt or possibly killed.\\n\\nOn the summary judgment record, Marbury appears to have provided no further\\n\\ninformation to either defendant about the nature of the threat.\\n\\n We must therefore decide whether a reasonable jury could find Marbury’s\\n\\nstatement that he had heard from a friend that an unnamed prisoner intended to\\n\\n\\n 12\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 13 of 50\\n\\n\\nhurt him, and that he was afraid of being hurt or killed, without any further details,\\n\\nsufficient to make the defendants aware of a substantial risk of serious harm. While\\n\\nthis question is a close one, we conclude that our precedent does not allow\\n\\nMarbury’s deliberate-indifference claim to proceed.\\n\\n On the one hand, it is settled that “a prison official [cannot] escape liability\\n\\nfor deliberate indifference by showing that . . . he did not know that the\\n\\ncomplainant was especially likely to be assaulted by the specific prisoner who\\n\\neventually committed the assault,” as long as the official was otherwise aware that\\n\\nthe victim faced a substantial risk of serious harm. 24 Our caselaw also establishes,\\n\\nhowever, that officials must possess enough details about a threat to enable them to\\n\\nconclude that it presents a “strong likelihood” of injury, not a “mere possibility.” 25\\n\\nThe unfortunate reality is that “threats between inmates are common and do not,\\n\\nunder all circumstances, serve to impute actual knowledge of a substantial risk of\\n\\nharm.” 26\\n\\n We have therefore upheld dismissal of or summary judgment against\\n\\ndeliberate-indifference claims where, although a plaintiff told prison officials about\\n\\na threat by another inmate or inmates, the prison officials were not deliberately\\n\\n\\n\\n\\n 24\\n Rodriguez, 508 F.3d at 619 (quoting Farmer, 511 U.S. at 843) (alterations in original).\\n 25\\n E.g., Brooks v. Warden, 800 F.3d 1295, 1301 (11th Cir. 2015); Brown v. Hughes, 894\\nF.2d 1533, 1537 (11th Cir. 1990).\\n 26\\n Prater v. Dahm, 89 F.3d 538, 542 (8th Cir. 1996).\\n 13\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 14 of 50\\n\\n\\nindifferent to a substantial risk of harm. For example, we held that a plaintiff had\\n\\nfailed to allege plausibly that prison officials were deliberately indifferent to a\\n\\nstrong likelihood of serious harm where he was attacked by an inmate in the next\\n\\ncell when all of the cell doors opened and a riot began.27 We noted that even\\n\\nthough the plaintiff alleged that the attacker had threatened him previously and that\\n\\nan attack was possible, he had not alleged any information showing a strong\\n\\nlikelihood that he and his attacker would be released at the same time in a chaotic\\n\\nenvironment. 28 Successful deliberate-indifference claims will generally require\\n\\nsome further reason—beyond the plaintiff having informed the defendant officers\\n\\nof the threat—that a prison official could have concluded that a particular threat\\n\\nevidenced a substantial threat, rather than the mere possibility, of serious harm.\\n\\n For example, a plaintiff can establish deliberate indifference to a substantial\\n\\nrisk of serious harm where he has given prison officials further information\\n\\nenabling them to conclude that the risk was substantial and not merely possible. In\\n\\nRodriguez v. Secretary for Department of Corrections, a prisoner informed prison\\n\\nstaff that members of his former gang had threatened to kill him upon release into\\n\\nthe general prison population. 29 We concluded that if true, this allegation was\\n\\n\\n\\n\\n 27\\n Brooks v. Warden, 800 F.3d 1295, 1298–99, 1301 (11th Cir. 2015).\\n 28\\n Id. at 1301.\\n 29\\n See Rodriguez, 508 F.3d at 612–15.\\n\\n\\n 14\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 15 of 50\\n\\n\\nenough to place the defendants on notice of a substantial risk of serious harm—so\\n\\nwe allowed the case to proceed.30 The key distinction between Rodriguez and\\n\\nMarbury’s case is that Rodriguez involved a series of threats that contained enough\\n\\nspecific information—the threats came from members of the plaintiff’s former\\n\\ngang—that prison officials could have reasonably inferred that there was a\\n\\nsubstantial, not merely possible, risk of harm. In fact, Rodriguez observed that a\\n\\nvague statement like “I have a problem with another inmate in this compound,”\\n\\nabsent some information “about the nature of the anticipated risk,” would not have\\n\\ncreated a genuine issue of fact regarding deliberate indifference to a substantial\\n\\nrisk of serious harm. 31 Marbury’s statement that an another inmate told him\\n\\nanother inmate intended to harm him is precisely this type of vague statement that\\n\\nconveys nothing about the nature of the anticipated risk that we cautioned in\\n\\nRodriguez would not rise to the level of deliberate indifference to a substantial risk.\\n\\n Marbury’s argument is essentially that every prisoner who tells prison\\n\\nofficials about an unspecified threat from an unspecified inmate without more is\\n\\nentitled to protective custody or a transfer. But, as already explained, our caselaw\\n\\nestablishes a higher standard for deliberate indifference. To be clear, Marbury was\\n\\n\\n\\n\\n 30\\n Id. at 621–22.\\n 31\\n Rodriguez, 508 F.3d at 619 n.15.\\n\\n\\n 15\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 16 of 50\\n\\n\\nnot required to identify the person who was threatening him by name, 32 or even\\n\\nnecessarily to give the defendants advance notice of a potential attack,33 so long as\\n\\nother facts put the defendants on notice that he faced a substantial risk of serious\\n\\nharm. It may be possible for a general threat of inmate-on-inmate violence in a\\n\\nprison to bolster an otherwise insufficient unspecified threat of harm. But, as\\n\\nalready discussed, Marbury has not shown anything close to such a substantial\\n\\nthreat from the generally violent nature of the prison environment. And because\\n\\nMarbury has not presented anything else that would bolster the unspecified threat,\\n\\nhe has not met the requirement of showing deliberate indifference to a substantial\\n\\nrisk of serious harm.\\n\\n Marbury further emphasizes that when he told Warren about the threat from\\n\\nan unidentified inmate, she laughed at him, told him she was not going to act on\\n\\nhis request because he had filed complaints against her, and advised him to get a\\n\\nknife or a shank because it seemed as though he had a “problem.” He suggests that\\n\\nthis reaction demonstrated her awareness that he faced a substantial risk. The\\n\\ndistrict court did not explicitly address Marbury’s argument, noting that\\n\\n“[a]ccording to defendant Warren, she never said this and asserts that she did not\\n\\nknow the plaintiff was in danger of being attacked by another inmate on April 23,\\n\\n\\n\\n 32\\n See, e.g., Rodriguez, 508 F.3d at 619 (discussing Farmer, 511 U.S. at 843).\\n 33\\n See Farmer, 511 U.S. at 848.\\n 16\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 17 of 50\\n\\n\\n2016.” Again, “specific facts” alleged in a pro se plaintiff’s sworn complaint can\\n\\nsuffice to generate a genuine factual issue. 34 We must therefore accept as true, for\\n\\nthe purposes of summary judgment, that Warren made these statements.\\n\\n But the statements ultimately do not affect Warren’s entitlement to summary\\n\\njudgment. The Supreme Court has made clear that “[w]hether a prison official had\\n\\nthe requisite knowledge of a substantial risk is a question of fact subject to\\n\\ndemonstration in the usual ways, including inference from circumstantial\\n\\nevidence.”35 While Marbury can establish Warren’s subjective knowledge through\\n\\ncircumstantial proof, he must nevertheless present evidence sufficient to “support a\\n\\nreasonable jury’s finding that [Warren] harbored a subjective awareness that\\n\\n[Marbury] was in serious danger.”36 Viewed in the light most favorable to\\n\\nMarbury, the summary judgment record allows one to conclude that Warren was\\n\\naware of the potential for inmate-on-inmate violence at the prison, Marbury told\\n\\nher that another inmate was looking to harm him, and she responded by laughing at\\n\\nMarbury and telling him to get a knife because he had a “problem.” A reasonable\\n\\njury could not extrapolate from such statements that Warren was subjectively\\n\\n\\n\\n\\n 34\\n See Caldwell, 748 F.3d at 1098.\\n 35\\n Rodriguez, 508 F.3d at 617 (quoting Farmer, 511 U.S. at 842).\\n 36\\n Goodman v. Kimbrough, 718 F.3d 1325, 1332 (11th Cir. 2013).\\n\\n\\n 17\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 18 of 50\\n\\n\\naware of a substantial risk of serious harm to Marbury at the time. Indeed,\\n\\nMarbury concedes in his briefing that Warren’s statements were callous jokes.\\n\\n Again, it bears mention that subjective awareness of only some risk of harm\\n\\nto a prisoner is insufficient for a deliberate-indifference claim. 37 At most, the\\n\\nevidence Marbury has put forward would allow a jury to conclude that Warren was\\n\\nput on notice that Marbury faced some unspecified risk of harm to his well-\\n\\nbeing—not that she was aware he faced the type of substantial risk of serious harm\\n\\nnecessary to establish deliberate indifference. Marbury has not marshaled enough\\n\\nevidence to establish a genuine issue of fact on this necessary element of his case.\\n\\n C\\n\\n Finally, we turn to Marbury’s suggestion that the defendants were\\n\\ndeliberately indifferent in failing to investigate his report that someone was out to\\n\\nharm him or in otherwise failing to abide by prison policy. We have explained that\\n\\n“merely negligent failure to protect an inmate from attack does not justify liability\\n\\nunder section 1983.”38 To allow Marbury’s deliberate-indifference claim to\\n\\nproceed absent sufficient evidence that the defendants were subjectively aware that\\n\\nhe faced a substantial risk of serious harm would elide the “subtle distinction”\\n\\nbetween deliberate indifference and mere negligence.39 We cannot condone the\\n\\n\\n 37\\n See, e.g., Brooks, 800 F.3d at 1301; Brown, 894 F.2d at 1537.\\n 38\\n Carter, 352 F.3d at 1350.\\n 39\\n See Goodman, 718 F.3d at 1333–34.\\n 18\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 19 of 50\\n\\n\\nfailure to investigate inmates’ allegations of threats or to follow policy in reporting\\n\\npotential threats up the chain of command. But our caselaw does not allow these\\n\\nfailures, without corresponding subjective awareness of a serious risk of harm, to\\n\\nestablish deliberate indifference.\\n\\n IV\\n\\n We affirm the judgment of the district court.\\n\\n\\n\\n\\n 19\\n\\n Case: 17-12589 Date Filed: 08/29/2019 Page: 20 of 50\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4161,"cells":{"id":{"kind":"number","value":4657692,"string":"4,657,692"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"Mark Jenkins v. Commissioner, Alabama Department of Corrections"},"decision_date":{"kind":"string","value":"2019-08-30"},"docket_number":{"kind":"string","value":"17-12524"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before TJOFLAT, WILSON, and BRANCH, Circuit Judges.\", \"parties\": \"\", \"opinions\": [{\"author\": \"BRANCH, Circuit Judge:\", \"type\": \"010combined\", \"text\": \"[PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n\\n\\n No. 17-12524\\n\\n\\n\\n D.C. Docket No. 4:08-cv-00869-VEH\\n\\nMARK ALLEN JENKINS,\\n\\n Petitioner - Appellant,\\n\\n versus\\n\\n\\n\\nCOMMISSIONER, ALABAMA DEPARTMENT OF CORRECTIONS,\\n\\n Respondent - Appellee.\\n\\n\\n\\n Appeal from the United States District Court\\n for the Northern District of Alabama\\n\\n\\n (August 30, 2019)\\n\\nBefore TJOFLAT, WILSON, and BRANCH, Circuit Judges.\\n\\nBRANCH, Circuit Judge:\\n\\n Mark Allen Jenkins, an Alabama prisoner sentenced to death for the 1989\\n\\nmurder of Tammy Ruth Hogeland, appeals the district court’s denial of his petition\\n\\nfor a writ of habeas corpus, 28 U.S.C. § 2254. Before us are Jenkins’s arguments\\n\\nthat he received ineffective assistance of counsel during the penalty phase of his\\n\\ntrial and that he is intellectually disabled and therefore ineligible for the death\\n\\npenalty. After careful consideration, and with the benefit of oral argument, we\\n\\naffirm the denial of Jenkins’s habeas petition on both grounds.\\n\\n I. BACKGROUND\\n\\n A. The Crime and Arrest\\n\\n We summarize the following background narrative from opinions of the\\n\\nAlabama Court of Criminal Appeals and our own review of the record. See Jenkins\\n\\nv. State, 627 So. 2d 1034, 1037–40 (Ala. Ct. Crim. App. 1992); Jenkins v. State,\\n\\n972 So. 2d 111, 119–20 (Ala. Ct. Crim. App. 2004). The events leading up to the\\n\\nmurder began the evening of April 17, 1989. Jenkins was at the home of his\\n\\nacquaintance Christine Nicholas. He had met her several months earlier at the\\n\\nOmelet Shoppe restaurant where she worked. Jenkins was very intoxicated and\\n\\ntried to seduce her. When she resisted, Jenkins got “real mad” and asked her\\n\\nseveral times what she would do if someone came up from behind her and grabbed\\n\\nher.\\n\\n At around 1 a.m., Jenkins and Nicholas drove to the Riverchase Omelet\\n\\nShoppe. Jenkins went inside and talked with one of the waitresses, Frieda Vines.\\n\\nThe manager, Douglas Thrash, recognized Jenkins as a regular customer who knew\\n\\n\\n 2\\n\\nall of the waitresses. Thrash overheard part of Jenkins’s conversation with Vines\\n\\nand heard mention of the Omelet Shoppe location near the Birmingham Airport.\\n\\nJenkins and Nicholas then returned to Nicholas’s home. Around 2 a.m., Nicholas’s\\n\\nmother asked Jenkins to leave. He did so, falling down some steps and ramming\\n\\nhis car into another vehicle in the process.\\n\\n Meanwhile, 23-year-old Tammy Hogeland was making her way to work at\\n\\nthe Omelet Shoppe. Her sister Wendy, along with Tammy’s young son, picked\\n\\nTammy up from her college classes and drove her to the Riverchase Omelet\\n\\nShoppe a little after 9:30 p.m. Sometime after 10:00, Tammy was sent to work at\\n\\nthe airport Omelet Shoppe on Tenth Avenue, since that location was unexpectedly\\n\\nshorthanded. Tammy was wearing her watch, a necklace with the words “special\\n\\nsister,” her class ring, and her diamond engagement ring. Tammy was working as a\\n\\ncook that night and was wearing a blue apron, black pants, a white shirt, and white\\n\\nshoes. She and Sarah Harris were the only employees working at the airport\\n\\nOmelet Shoppe that night.\\n\\n At about 2 a.m. on April 18, Harris saw Jenkins—whom she did not know—\\n\\ndrive up in a red sports car. His arrival was memorable to Harris because the car\\n\\nnearly jumped the curb and crashed through the restaurant’s glass wall. Jenkins got\\n\\nout of the car and came into the restaurant, appearing to be intoxicated. He walked\\n\\nover to Hogeland and began talking to her. Harris later saw Jenkins and Hogeland\\n\\n\\n 3\\n\\ndrive away together in the red sports car. That was the last time anyone who knew\\n\\nHogeland ever saw her. Hogeland left behind her cigarettes, lighter, purse, and\\n\\npaycheck. She left without telling anyone, which she had never done before.\\n\\nAlthough Harris saw them drive off, she could not later say whether Hogeland left\\n\\nwith Jenkins willingly or was instead abducted.\\n\\n At around 5 a.m. that morning, Bobby and Geraldine Coe had stopped to\\n\\nbuy gas at a Chevron station on I-59 northeast of Birmingham when they saw\\n\\nJenkins drive up in the red sports car. The Coes noticed a female who appeared to\\n\\nbe “passed out” in the front passenger seat, but they could not say whether she was\\n\\nalive or dead. While Bobby Coe was pumping gas, Jenkins asked him for some\\n\\ncigarettes and said, “Looks like it’s been a long night and it looks like it’s going to\\n\\nbe a long day.” Jenkins then said “God bless you” before asking directions to I-\\n\\n459. Coe gave him directions and got back in his car. As Coe pulled out onto the\\n\\ninterstate, he saw Jenkins follow him in the red sports car. Coe then saw the car\\n\\nflash its lights, slow down, and pull to the side of the road between mile markers\\n\\n151 and 152.\\n\\n At 8 a.m., Wendy Hogeland learned that Tammy was not at the restaurant,\\n\\nand their mother called the police. Meanwhile, Jenkins went to the home of Steve\\n\\nMusser, who noticed that he was wearing the same clothes as the day before.\\n\\nJenkins told Musser that his car had been stolen the night before and asked him if\\n\\n\\n 4\\n\\nhe would say that he had been with him all night. Musser refused. Christine\\n\\nNicholas also saw Jenkins in a grocery store that morning. He was looking at a\\n\\nnewspaper, making a phone call, and attempting to sell his Buick. Nicholas loaned\\n\\nhim $4 for gas. At around 10 o’clock that morning, Jenkins sold his car to Michael\\n\\nBrooks, a mechanic at a local Chevron station. Jenkins had explained that he\\n\\nneeded the money so he could visit his sick mother in California. Another\\n\\nmechanic at the service station drove Jenkins to the Greyhound bus station later\\n\\nthat morning.\\n\\n Jenkins awoke the next day on the Greyhound bus in Houston and was\\n\\nejected from the bus because his fare was used up. He then hitchhiked from\\n\\nHouston to Phoenix, then to San Diego and Los Angeles. Jenkins was first\\n\\nidentified as a suspect in Hogeland’s disappearance by Officer Mike Weems of the\\n\\nHoover Police Department on April 19. Weems ate dinner at the Riverchase\\n\\nOmelet Shoppe just about every day and had talked with Hogeland just as often.\\n\\nHe also knew Jenkins from the restaurant. Weems learned of Hogeland’s\\n\\ndisappearance from the other waitresses. Remembering how Jenkins often talked to\\n\\nHogeland and passed her notes,1 he gave Jenkins’s name to the missing persons\\n\\ninvestigator. The investigator also learned that a red Mazda RX-7 sports car, which\\n\\n\\n\\n1\\n The jury was not permitted to hear, and the record does not reveal, what exactly Jenkins had\\nsaid to Hogeland that made Officer Weems suspect him in her disappearance.\\n 5\\n\\nhad been reported stolen from the service station where Jenkins worked, had been\\n\\nrecovered on I-459.\\n\\n The afternoon of April 21, a truck driver who had happened to stop on I-59\\n\\nnear mile marker 151 discovered Hogeland’s body off the side of the road. She was\\n\\nnaked, wearing only a watch, and was so badly decomposed that the body had to\\n\\nbe identified by dental records. From the fractured hyoid bone, it was determined\\n\\nthat she was manually strangled to death. Also found at the scene were Hogeland’s\\n\\napron, shoes, bra, panties, pants, and hair net, as well as some beer cans, a Mazda\\n\\nRX-7 owner’s manual, and other items later determined to have come from the red\\n\\nMazda. Her necklace and rings were never found.\\n\\n Alabama authorities issued a warrant for Jenkins’s arrest and also obtained a\\n\\nfederal fugitive warrant. An FBI agent arrested Jenkins on May 1 in front of his\\n\\nuncle’s house in Wilmington, California. Jenkins’s uncle later gave the police three\\n\\nbags of Jenkins’s personal effects. Fibers recovered from Hogeland’s body and\\n\\nclothing matched those of the Mazda, and hairs matching Hogeland’s were found\\n\\nin the Mazda. Similarly, fibers from Jenkins’s clothing placed him in the Mazda,\\n\\nand fibers of his jeans were found on Hogeland’s apron. A bootprint near\\n\\nHogeland’s body matched a combat boot from among the belongings Jenkins’s\\n\\nuncle gave to the police, and Jenkins’s roommate said the boots looked just like\\n\\nthose that Jenkins wore.\\n\\n\\n 6\\n\\n A business card like those that had been in the Mazda’s glove box,\\n\\nbelonging to an owner of the Mazda, was recovered from Jenkins’s wallet after his\\n\\narrest in California. Later, Jenkins’s cellmate in the Alabama jail reported that\\n\\nJenkins told him “he had done the crime” and was worried that the couple from the\\n\\ngas station would identify him or that the police would find his fingerprint on a\\n\\nbeer can at the scene. A St. Clair County grand jury indicted Jenkins for capital\\n\\nmurder.\\n\\n B. The Trial and Sentencing\\n\\n Jenkins was represented at trial in 1991 by attorneys Douglas Scofield and\\n\\nStan Downey. Scofield, the lead attorney, had been practicing criminal defense in\\n\\nBirmingham since 1984 but had never before tried a capital murder case. Jenkins’s\\n\\nlandlord and his grandmother had arranged the referral and representation before\\n\\nthe court appointed him. Scofield agreed to the appointment on the condition that a\\n\\nSt. Clair County lawyer also be appointed. That lawyer was Downey, who had\\n\\nmore experience than Scofield but whose capital experience was limited to one\\n\\ncase that had not gone to trial. The two attorneys had agreed that Downey would\\n\\nbe responsible for jury selection, but Scofield eventually took over during the voir\\n\\ndire. Per their agreement, Scofield was responsible for the guilt-phase trial,\\n\\nalthough Downey participated by interviewing Jenkins several times. Scofield also\\n\\nmet with Jenkins around a dozen times. Jenkins had told him about his background\\n\\n\\n 7\\n\\nand childhood, including the fact that he had had a difficult childhood and had\\n\\nlived on the streets since age 11.\\n\\n Throughout the weeklong trial, Scofield pursued an outright acquittal. In\\n\\nparticular, he challenged Harris’s identification of Jenkins as the man who drove\\n\\naway with Hogeland, and he challenged the medical examiner’s opinions about the\\n\\ntime and cause of Hogeland’s death. In his closing argument, Scofield took a wide-\\n\\nranging approach to undermining the State’s wholly circumstantial case against\\n\\nJenkins. He attacked the State’s timeline of the crime, questioning whether Jenkins\\n\\nin his intoxicated state would have had the time and the ability to leave Nicholas’s\\n\\nhouse, steal the Mazda, and abduct and kill Hogeland. He reminded the jury of the\\n\\nweakness of Harris’s identification of Jenkins and the Coes’ identification of the\\n\\nwoman in the car. He pointed out that no physical evidence definitively linked\\n\\nJenkins to the Mazda, to Hogeland’s body, or to her jewelry, which was never\\n\\nrecovered. And he argued that the evidence provided no plausible motive for\\n\\nJenkins to have kidnapped, robbed, or killed Hogeland. In sum, he urged the jury\\n\\nto find reasonable doubt and acquit Jenkins. At the very least, he argued, they\\n\\nshould find that the State failed to prove Jenkins’s intent to rob and kidnap\\n\\nHogeland, which were necessary to convict Jenkins of capital murder. He further\\n\\nargued that, if the jury thought Jenkins killed Hogeland as a crime of passion, they\\n\\ncould convict him only of manslaughter. The jury deliberated for three hours and\\n\\n\\n 8\\n\\nfifteen minutes before convicting Jenkins of capital murder, Ala. Code § 13A-5-\\n\\n40(a)(1) & (2), i.e., murder committed in the course of kidnapping and robbery.\\n\\n The penalty phase took place that same afternoon. A week before trial,\\n\\nScofield and Downey had discussed the penalty phase, and Scofield understood\\n\\nthat Downey would be responsible for handling the penalty phase. Nonetheless, the\\n\\nwitness who was present to testify on Jenkins’s behalf during the penalty phase\\n\\nwas better acquainted with Scofield, so Downey suggested that afternoon that\\n\\nScofield conduct the direct examination. Scofield urged Downey to interview the\\n\\nwitness quickly and prepare to examine him, and Downey did so.\\n\\n Downey handled the penalty presentation to the jury. The State did not\\n\\npresent any additional aggravation evidence, resting on the evidence presented\\n\\nduring the guilt phase. One witness, Lonnie Seal, testified on Jenkins’s behalf in\\n\\norder, in Downey’s words, to “reveal another side of Mark Jenkins to you that you\\n\\ndon’t know anything about.” Seal had met Jenkins in Fontana, California, in 1988\\n\\nwhen they both worked at the same garage. Seal and his wife became friends with\\n\\nJenkins, and he visited in their home. When the Seals decided to move to Alabama\\n\\na few months later, Jenkins agreed to help with their move and to live with the\\n\\nSeals in Alabama for a time. It took Lonnie Seal several weeks to find a job, but\\n\\nJenkins found a job in two days and contributed toward the family’s rent and\\n\\n\\n\\n\\n 9\\n\\ngroceries. Jenkins continued to visit the Seals after he moved into his own home,\\n\\nand he would always bring a small gift for and play with the Seals’ infant son.\\n\\n Downey’s closing argument to the jury requested a sentence of life without\\n\\nparole. Downey reminded the jurors that they had agreed during voir dire that\\n\\ndeath was not always the appropriate punishment for homicide. Quoting the Bible,2\\n\\nhe reminded the jury that the Old Testament permitted capital punishment only\\n\\nwhen supported by the testimony of two or three eyewitnesses, whereas in this case\\n\\nthere had been only circumstantial evidence. He asserted from Seal’s testimony\\n\\nthat “Mark had a side of him that was generous, that was giving, and that was\\n\\nkind,” and argued that the deterrent and retributive purposes of capital punishment\\n\\nwould not be served in this case.\\n\\n The jury deliberated for 50 minutes before returning its nonbinding\\n\\nrecommendation of death by a vote of 10 to 2. See Ala. Code § 13A-5-46(e) (1981)\\n\\n(amended 2017 to make jury’s verdict binding). After a separate hearing, the court\\n\\nfound the two statutory aggravating factors for which the State had argued: that the\\n\\nmurder was committed during the commission of a robbery, and that it was\\n\\ncommitted during the commission of a kidnapping. See Ala. Code § 13A-5-49(4).\\n\\n\\n\\n\\n2\\n “If anyone kills a person, the murderer shall be put to death on the evidence of witnesses. But\\nno person shall be put to death on the testimony of one witness.” Numbers 35:30 (ESV). Downey\\nalso paraphrased Deuteronomy 17:6 as “On the evidence of two witnesses or three witnesses, he\\nis [sic] shall be put to death. He shall not be put to death on the evidence of one witness.”\\n 10\\n\\nThe court also found two statutory mitigating factors: that Jenkins had no\\n\\nsignificant history of prior criminal activity, and his age at the time of the crime\\n\\n(21). See id. § 13A-5-51(1), (7). The court considered but rejected the statutory\\n\\nmitigation that, due to his consumption of alcohol, Jenkins’s capacity to appreciate\\n\\nthe criminality of his conduct or to conform his conduct to the requirements of law\\n\\nwas substantially impaired. See id. § 13A-5-51(6). The court found that the\\n\\naggravating circumstances outweighed the mitigating circumstances and, in\\n\\naccordance with the jury’s recommendation, imposed a sentence of death.\\n\\n C. The State and Federal Post-Conviction Proceedings\\n\\n On direct appeal, the Alabama Court of Criminal Appeals affirmed Jenkins’s\\n\\nconviction and death sentence. Jenkins v. State, 627 So. 2d 1034 (Ala. Ct. Crim.\\n\\nApp. 1992), aff’d, 627 So. 2d 1054 (Ala. 1993), cert. denied, 511 U.S. 1012\\n\\n(1994). In 1995, Jenkins filed his first petition for state postconviction relief, Ala.\\n\\nR. Crim. P. 32,3 alleging, among many other things, that he received ineffective\\n\\nassistance of counsel during his penalty phase. The Rule 32 trial court conducted a\\n\\nthree-day hearing at which evidence was presented about the performance of\\n\\nJenkins’s attorneys.\\n\\n\\n\\n\\n3\\n “Subject to the limitations of Rule 32.2, any defendant who has been convicted of a criminal\\noffense may institute a proceeding in the court of original conviction to secure appropriate relief\\non the ground that . . . [t]he constitution of the United States or of the State of Alabama requires\\na new trial, a new sentence proceeding, or other relief. . . .” Ala. R. Crim. P. 32.1.\\n 11\\n\\n Scofield testified about the representation that he and Downey provided. In\\n\\nrelevant part, Scofield asserted that he and Downey had “decided that [developing\\n\\nmitigation for the penalty phase] would be something that Stan [Downey] would\\n\\nhandle.” Downey later told Scofield that he hadn’t done any such investigation, but\\n\\nScofield admitted that, “[a]part from what he has told me, I don’t know what he\\n\\nhas done.” Scofield testified that Downey “didn’t appear prepared” for the penalty\\n\\nphase, based on Downey’s last-minute suggestion that Scofield examine Seal since\\n\\nhe had talked with him before.\\n\\n In preparation for the penalty phase, Scofield himself did not interview any\\n\\nmember of Jenkins’s family, but he did not know whether Downey interviewed\\n\\nany family members. Scofield did interview Jenkins’s landlord and some other\\n\\nacquaintances but did not do so specifically for the purpose of developing\\n\\nmitigation. He had met with Jenkins around a dozen times, and Jenkins had told\\n\\nhim his life story including his experience of abuse as a child. In hindsight,\\n\\nScofield said he had a “lack of appreciation of mitigating circumstances.” He\\n\\nrepeatedly denied in his Rule 32 testimony that his omissions during the penalty\\n\\nphase were strategic: “I didn’t understand what I was doing. . . . I just didn’t\\n\\nunderstand what I needed to do.”\\n\\n For reasons that remain unclear, Downey—who is since deceased—did not\\n\\ntestify or submit an affidavit in the Rule 32 proceeding. Scofield answered,\\n\\n\\n 12\\n\\n“Obviously not,” when asked whether he had personal knowledge of everything\\n\\nDowney may or may not have done to prepare for the penalty phase. The record of\\n\\nthe extent of Downey’s pretrial preparation is thus limited to Downey’s detailed\\n\\nfee declaration, which he submitted to the State of Alabama in documentation of\\n\\nthe compensation due him as appointed counsel. See Ala. Code § 15-12-21. In it,\\n\\nDowney reported that he spent 71 hours4 on out-of-court preparation for trial. The\\n\\nreport details 35 hours of interviews with Jenkins in 18 separate visits to the jail\\n\\nover the course of 17 months, as well as one substantive conversation with\\n\\nJenkins’s grandmother.5\\n\\n The Rule 32 trial court also heard testimony from four family members and\\n\\none friend of Jenkins’s. The family members presented for the first time a stark\\n\\ndepiction of Jenkins as an unloved and abused child in California. Jenkins’s older\\n\\nbrother Michael Jenkins testified that all of his siblings have different fathers and\\n\\nthat his parents often used drugs and fought. Michael testified that Jenkins was\\n\\nbeaten frequently as a child by his stepfather, if not every day then every other day\\n\\n\\n\\n4\\n In Jenkins’s state postconviction appeal, the Alabama Court of Criminal Appeals erroneously\\nstated that Downey’s fee declaration reports 171 hours of pretrial preparation. See Jenkins, 972\\nSo. 2d at 137. We discuss the effect of this discrepancy infra note 7.\\n5\\n The state appellate court also noted that Downey had moved for a continuance in October 1989\\nso that he could conduct “[f]urther discovery and investigation (including a possible trip to\\nCalifornia).” Id. We note that Scofield had orally suggested to the trial court in September 1989\\nthat more time might be needed with respect to “evidence about the defendant” and witnesses in\\nCalifornia, as well as possible psychological evaluation.\\n 13\\n\\nor every three days. Jenkins was beaten all over his body with a variety of\\n\\nimplements. Michael was beaten too. Jenkins’s stepfather would lock Jenkins in\\n\\nhis bedroom for a couple of hours at a time, sometimes without dinner. Some of\\n\\nthe beatings were prompted by Jenkins’s bedwetting or soiling. Michael testified\\n\\nthat Jenkins’s stepfather sometimes made Jenkins wear his soiled underwear on his\\n\\nhead and even forced him to eat his own feces a few times. Michael testified that\\n\\nJenkins’s bladder and bowel control problems began after a camping trip he took\\n\\nwith his stepfather and stepgrandfather.\\n\\n Michael testified that, as children, he and Jenkins were made to work on car\\n\\nengines late into the night and to shovel horse manure daily. The family moved\\n\\nfrequently and the children attended many different schools as a result. Michael\\n\\ntestified that he was Jenkins’s only friend as a child and that Jenkins did poorly in\\n\\nschool. Jenkins began running away from home as a young teenager and lived with\\n\\nvarious other family members or on the streets. Michael said that Jenkins was a\\n\\ngood brother to him and had asked him to testify at his trial, but he never heard\\n\\nfrom his attorneys or found out when the trial was. Michael admitted to having\\n\\ncognitive and memory problems due to a head injury years earlier.\\n\\n Jenkins’s cousin Tammy Pitts testified that, as an infant, Jenkins’s mother\\n\\nbeat and neglected him, leaving him dirty and in soiled diapers. His stepfather also\\n\\nbeat Jenkins daily until he left home at age 13, sometimes bruising him so badly\\n\\n\\n 14\\n\\nthat he was laid up in bed for days or weeks. Jenkins had problems with\\n\\nbedwetting and soiling, and his stepfather made him wear soiled clothes to school.\\n\\nJenkins would be locked in his filthy bedroom around the clock except to attend\\n\\nschool and do chores. He was usually not allowed to eat dinner with the rest of the\\n\\nfamily and subsisted on scraps and dog food. The entire family belittled Jenkins\\n\\nbecause his biological father was Hispanic and his complexion was therefore\\n\\ndarker. Jenkins’s parents drank a lot and used drugs every day. Pitts knew these\\n\\nthings because she had daily contact with the family and lived with them off and\\n\\non, and Jenkins lived with her for a time after he ran away from home. No one had\\n\\ncontacted her about testifying at Jenkins’s trial.\\n\\n Jenkins’s second cousin Betty DeLavega testified that her relationship with\\n\\nJenkins was like that of an aunt. She lived with the Jenkins family for five months\\n\\nwhen Jenkins was 10 or 11. She saw Jenkins’s stepfather slap Jenkins “[q]uite a\\n\\nfew times,” and once, he made Jenkins eat his own feces. DeLavega testified that\\n\\nJenkins’s parents used drugs nearly every day and withheld affection from Jenkins\\n\\nand Michael. No one contacted her about Jenkins’s trial; Jenkins’s grandmother\\n\\nhad asked her to testify in the Rule 32 hearing about any abuse she had witnessed.\\n\\n Jenkins’s grandmother Doris Wagoner testified that while Jenkins’s mother\\n\\nwas pregnant with him, her husband was in prison and she drank and took drugs.\\n\\nWhen Jenkins was born, his mother initially put him up for adoption, then changed\\n\\n\\n 15\\n\\nher mind weeks later. Jenkins was a lethargic infant and his mother cared more\\n\\nabout partying than about her children. Wagoner testified that she heard of\\n\\nJenkins’s mistreatment as a child but didn’t see it. She could tell from the way that\\n\\nJenkins feared his stepfather that he was being abused, and she knew he slept in a\\n\\nfilthy bed. After Jenkins was arrested for murder, Wagoner spoke with Scofield,\\n\\nwho asked for money relating to his representation of Jenkins. She did not\\n\\nremember whether she talked with Scofield about Jenkins’s background. She could\\n\\nnot remember why she didn’t testify at the trial; she might have been ill, but she\\n\\nwas probably busy.\\n\\n Sherry Seal, the wife of Lonnie Seal, who had testified in the penalty phase\\n\\nof the trial, testified that Jenkins was respectful and generous and that she trusted\\n\\nhim with her children. She related how he supported her family after moving to\\n\\nAlabama, giving them “pretty much” his entire paycheck. She said she would still\\n\\nfeel safe around Jenkins. Two jailers also testified that Jenkins was a respectful,\\n\\ncourteous, model inmate who never complained or caused any problems.\\n\\n Jenkins also introduced medical, school, and juvenile court records from his\\n\\nchildhood. The medical records document his premature birth and his mother’s\\n\\nintention to give him up for adoption. The school records reflect a struggling\\n\\nstudent who was repeatedly promoted to the next grade because of his age, despite\\n\\nserious academic difficulties, chronic truancy, and frequent school changes. The\\n\\n\\n 16\\n\\njuvenile court records document Jenkins’s placement in juvenile hall starting at age\\n\\n14, following repeated acts of grand theft auto and running away.\\n\\n The Rule 32 hearing concluded with the testimony of two psychologists\\n\\nabout Jenkins’s history of psychological trauma stemming from his childhood\\n\\nabuse. The defense expert, Dr. David Lisak, had met with Jenkins three times and\\n\\ninterviewed a dozen of his family members, family friends, and associates. Dr.\\n\\nLisak testified that the pervasive abuse and neglect in Jenkins’s childhood impaired\\n\\nhis ability to cope with trauma as an adult. He opined that Jenkins’s childhood\\n\\nabuse adversely affected his cognitive and emotional development and other\\n\\nabilities. He described how Jenkins had been neglected as an infant and sexually\\n\\nabused by his grandfather at age four, which led to a lifetime of bladder and bowel\\n\\ndysfunction. He also discussed the hypothesis that children who are abused are at\\n\\ngreater risk for perpetrating violence later in life, but he acknowledged there would\\n\\nhave been other causative factors in Jenkins’s case. Dr. Lisak described Jenkins’s\\n\\nintellectual capacity as “somewhat borderline,” though he did not perform any\\n\\ntesting of his own.\\n\\n The State’s expert, Dr. Karl Kirkland, had met with Jenkins for several hours\\n\\nand administered several tests. On the Wechsler Adult Intelligence Scale, Jenkins\\n\\n“scored in the range of borderline intellectual functioning which is between mild\\n\\nmental retardation and low average intellectual functioning” with an overall IQ of\\n\\n\\n 17\\n\\n76. Dr. Kirkland agreed that Jenkins had been “reared in a chaotic dysfunctional\\n\\nfamily” and generally agreed with Dr. Lisak’s assessment of the negative effects of\\n\\nchildhood trauma.\\n\\n The court considered all of this evidence before denying Rule 32 relief in\\n\\nfull. As relevant here, it found that some of the evidence about the severity of\\n\\nabuse that Jenkins suffered as a child was not credible. It further found that Jenkins\\n\\nsuffered no prejudice from any deficient performance by his penalty-phase counsel\\n\\nbecause the aggravating circumstances outweighed any additional mitigating\\n\\nevidence Downey might have adduced.\\n\\n The Alabama Court of Criminal Appeals affirmed the denial of Rule 32\\n\\nrelief. Jenkins v. State, 972 So. 2d 111 (Ala. Ct. Crim. App. 2004). In relevant part,\\n\\nthe appeals court rejected Jenkins’s ineffective assistance claim, finding that\\n\\ncounsel reasonably chose to pursue a penalty-phase strategy of residual doubt and\\n\\ngood character. Id. at 147. It explained that the new evidence was self-\\n\\ncontradictory in that it asserted that childhood abuse would have made Jenkins a\\n\\nviolent adult yet also asserted that Jenkins was meek and mild. Id. The appeals\\n\\ncourt also agreed with the lower court’s conclusion that Jenkins suffered no\\n\\nprejudice from any error by his counsel because the aggravating circumstances still\\n\\noutweighed any new mitigating ones. Id. at 148. The court also noted that evidence\\n\\n\\n\\n\\n 18\\n\\nabout Jenkins’s good behavior in jail awaiting trial was “minimally mitigating” and\\n\\nwould not have affected the sentence. Id. at 149.\\n\\n Because the Supreme Court’s decision in Atkins v. Virginia6 had issued\\n\\nduring the pendency of the Rule 32 appeal, the Alabama appeals court ordered\\n\\nsupplemental briefing on the possible impact of Atkins upon Jenkins’s sentence.\\n\\nJenkins asked the court to stay the proceedings and conduct an evidentiary hearing\\n\\nafter the Alabama legislature enacted a statute about executing the mentally\\n\\nretarded or, in the alternative, to vacate his sentence and remand after the Alabama\\n\\nlegislature acted. He later asserted that he is mentally retarded and therefore\\n\\nineligible for the death penalty. The appeals court proceeded to reject Jenkins’s\\n\\nAtkins claim on the merits because the record established that his IQ was 76 and\\n\\nthat he maintained relationships and employment. Id. at 155. The Alabama\\n\\nSupreme Court summarily affirmed, in relevant part. Ex parte Jenkins, 972 So. 2d\\n\\n159, 165 (Ala. 2005), remanded to 972 So. 2d 165 (Ala. Ct. Crim. App. 2005)\\n\\n(juror misconduct claim), cert. denied, 552 U.S. 1167 (2008).\\n\\n In 2008, Jenkins filed his first petition for a writ of habeas corpus in federal\\n\\ncourt pursuant to 28 U.S.C. § 2254. The federal proceedings were stayed for\\n\\nseveral years while Jenkins pursued additional, unrelated Rule 32 relief in the state\\n\\n\\n6\\n 536 U.S. 304, 321 (2002) (holding that the execution of mentally retarded offenders violates the\\nEighth Amendment). The rule announced in Atkins is retroactive to cases on collateral review. In\\nre Holladay, 331 F.3d 1169, 1172–73 (11th Cir. 2003).\\n 19\\n\\ncourts, ultimately unsuccessfully. In the district court, Jenkins moved for an\\n\\nevidentiary hearing on his Atkins claim. After reviewing the state court record,\\n\\nwith a focus on the Rule 32 testimony of Drs. Lisak and Kirkland, the district court\\n\\ndenied the motion in 2015. Noting that our Circuit has accepted the State of\\n\\nAlabama’s definition of mental retardation following Atkins, the court found that\\n\\nJenkins is not intellectually disabled under that standard. In a 347-page order, the\\n\\ndistrict court also denied relief on all of Jenkins’s claims including ineffective\\n\\nassistance of counsel during the penalty phase and affirmed as reasonable the state\\n\\ncourt’s conclusion that Jenkins failed to establish prejudice from any error by his\\n\\ncounsel. The district court later denied reconsideration and a certificate of\\n\\nappealability.\\n\\n Jenkins now appeals. Our Court granted Jenkins a certificate of\\n\\nappealability, 28 U.S.C. § 2253(c), on two issues:\\n\\n 1) Whether the district court erred in denying Appellant’s claim\\n that his trial counsel rendered ineffective assistance of counsel in the\\n investigation and presentation of mitigating evidence during the\\n penalty phase of Appellant’s 1991 trial; and\\n\\n 2) Whether the district court erred in denying Appellant’s claim\\n that he is intellectually disabled and ineligible for the death penalty\\n under Atkins v. Virginia, 536 U.S. 304 (2002).\\n\\n II. STANDARDS OF REVIEW\\n\\n We review the denial of a petition for a writ of habeas corpus de novo.\\n\\nMorrow v. Warden, Ga. Diagnostic Prison, 886 F.3d 1138, 1146 (11th Cir. 2018).\\n\\n 20\\n\\nThe Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) instructs that\\n\\na writ of habeas corpus shall not be granted to a state prisoner unless the state\\n\\nadjudication of a claim “resulted in a decision that was contrary to, or involved an\\n\\nunreasonable application of, clearly established Federal law, as determined by the\\n\\nSupreme Court of the United States.” 28 U.S.C. § 2254(d)(1). “[C]learly\\n\\nestablished Federal law” means “the holdings, as opposed to the dicta, of [the\\n\\nSupreme] Court’s decisions as of the time of the relevant state-court decision.”\\n\\nWilliams v. Taylor, 529 U.S. 362, 412 (2000). “[W]hen the last state court to\\n\\ndecide a prisoner’s federal claim explains its decision on the merits in a reasoned\\n\\nopinion . . . a federal habeas court simply reviews the specific reasons given by the\\n\\nstate court and defers to those reasons if they are reasonable.” Wilson v. Sellers,\\n\\n138 S. Ct. 1188, 1192 (2018). A state court’s decision is reasonable “so long as\\n\\n‘fairminded jurists could disagree’ on the correctness of the state court’s decision.”\\n\\nHarrington v. Richter, 562 U.S. 86, 101 (2011) (quoting Yarborough v. Alvarado,\\n\\n541 U.S. 652, 664 (2004)). “[E]ven a strong case for relief does not mean the state\\n\\ncourt’s contrary conclusion was unreasonable.” Id. at 102. Rather, a prisoner must\\n\\nshow that the state court’s ruling “was so lacking in justification that there was an\\n\\nerror well understood and comprehended in existing law beyond any possibility for\\n\\nfairminded disagreement.” Id. at 103. That ruling must have been “‘objectively\\n\\nunreasonable,’ not merely wrong; even ‘clear error’ will not suffice.” White v.\\n\\n\\n 21\\n\\nWoodall, 572 U.S. 415, 419 (2014) (quoting Lockyer v. Andrade, 538 U.S. 63, 75\\n\\n(2003)).\\n\\n In addition, “a determination of a factual issue made by a State court shall be\\n\\npresumed to be correct,” and the prisoner bears “the burden of rebutting the\\n\\npresumption of correctness by clear and convincing evidence.” 28 U.S.C.\\n\\n§ 2254(e)(1). AEDPA further instructs that a writ of habeas corpus may not be\\n\\ngranted unless the state court adjudication “resulted in a decision that was based on\\n\\nan unreasonable determination of the facts in light of the evidence presented in the\\n\\nState court proceeding.” Id. § 2254(d)(2). When a federal habeas court makes a\\n\\nfactual finding as part of its habeas determination, we review for clear error, which\\n\\nis a “highly deferential” standard of review. Thomas v. Allen, 607 F.3d 749, 752\\n\\n(11th Cir. 2010) (quoting Holton v. City of Thomasville Sch. Dist., 425 F.3d 1325,\\n\\n1350 (11th Cir. 2005)). “A finding is ‘clearly erroneous’ when although there is\\n\\nevidence to support it, the reviewing court on the entire evidence is left with the\\n\\ndefinite and firm conviction that a mistake has been committed.” United States v.\\n\\nU.S. Gypsum Co., 333 U.S. 364, 395 (1948).\\n\\n III. DISCUSSION\\n\\n A. Ineffective Assistance of Counsel\\n\\n A prisoner alleging that he received ineffective assistance of counsel in\\n\\nviolation of the Sixth Amendment must establish two elements. Strickland v.\\n\\n\\n 22\\n\\nWashington, 466 U.S. 668, 687 (1984). “First, the defendant must show that\\n\\ncounsel’s performance was deficient.” Id. In other words, he must show “that\\n\\ncounsel made errors so serious that counsel was not functioning as the ‘counsel’\\n\\nguaranteed the defendant by the Sixth Amendment.” Id. “Second, the defendant\\n\\nmust show that the deficient performance prejudiced the defense.” Id. We discuss\\n\\neach of these elements in turn.\\n\\n 1. Deficient Performance\\n\\n A “doubly” deferential lens sits before our eyes as we review the\\n\\nperformance of Jenkins’s counsel during the penalty phase. Richter, 562 U.S. at\\n\\n105 (quoting Knowles v. Mirzayance, 556 U.S. 111, 123 (2009)). This redoubled\\n\\ndeference arises out of the tandem application of AEDPA’s “highly deferential”\\n\\nreview of a state court’s decision with Strickland’s “highly deferential” review of\\n\\nan attorney’s performance. Id. (quoting Lindh v. Murphy, 521 U.S. 320, 333 n.7\\n\\n(1997), and Strickland, 466 U.S. at 689). Under the Strickland standard, “counsel\\n\\nis strongly presumed to have rendered adequate assistance and made all significant\\n\\ndecisions in the exercise of reasonable professional judgment.” Strickland, 466\\n\\nU.S. at 690. And under AEPDA, we must deny habeas relief if “there is any\\n\\nreasonable argument that counsel satisfied Strickland’s deferential standard.”\\n\\nRichter, 562 U.S. at 105.\\n\\n\\n\\n\\n 23\\n\\n Jenkins argues that his counsel performed deficiently by failing to\\n\\n“investigate compelling mitigating evidence or develop a penalty-phase theory.”\\n\\nThe state court disagreed. It quoted decisions of our Court that hold that “there is\\n\\nno per se rule that evidence of a criminal defendant’s troubled childhood must\\n\\nalways be presented as mitigating evidence.” Jenkins, 972 So. 2d at 144 (quoting\\n\\nMarek v. Singletary, 62 F.3d 1295, 1300 (11th Cir. 1995)). Noting that “Downey\\n\\nwas in charge of the penalty phase” but “we do not have the benefit of Downey’s\\n\\ntestimony as to what occurred and why,” it reviewed the trial record to find that\\n\\n“counsel argued residual doubt and Jenkins’s good character at the penalty phase.”\\n\\nId. at 145–46. It concluded that “Downey’s decision to concentrate on reasonable\\n\\ndoubt and to portray Jenkins as a good person was reasonable under the\\n\\ncircumstances.” Id. at 147.\\n\\n Jenkins insists, however, that counsel’s failure to develop mitigating\\n\\nevidence about his childhood abuse was “not the product of strategy, but rather\\n\\narose from ingnorance and inexeperiance [sic].” The problem for Jenkins is that,\\n\\nlike the state courts, this Court also does not know what Downey did or why. The\\n\\nrecord is silent as to his thoughts and intentions as he prepared for the penalty\\n\\nphase. Given that Jenkins bore the burden of overcoming the strong presumption of\\n\\ncompetence and of proving Downey’s deficient performance, we are perplexed\\n\\nwhy Jenkins’s new counsel did not have Downey testify or submit an affidavit in\\n\\n\\n 24\\n\\nthe Rule 32 proceeding. See Strickland, 466 U.S. at 687 (“The defendant must\\n\\nshow that counsel’s performance was deficient.”). Nor did Jenkins testify to the\\n\\nsubstance of his conversations with his counsel. “The reasonableness of a trial\\n\\ncounsel’s acts, including lack of investigation or excluding character witnesses\\n\\nfrom the sentencing phase, depends ‘critically’ upon what information the client\\n\\ncommunicated to counsel.” Chandler v. United States, 218 F.3d 1305, 1324 (11th\\n\\nCir. 2000) (en banc) (quoting Strickland, 466 U.S. at 691). There is simply no\\n\\nrecord evidence to support Jenkins’s central assertion that Downey made no\\n\\ntactical choices in his preparation for the penalty phase.\\n\\n Jenkins instead relies on Scofield’s Rule 32 testimony about how he and\\n\\nDowney prepared for the trial and sentencing. We acknowledge that Scofield\\n\\nrepeatedly testified that he did not really know what he was doing and that he\\n\\nfailed to make strategic choices with respect to mitigating circumstances and\\n\\nsentencing. But Scofield was equally clear that Downey—the more experienced\\n\\nlawyer—was in fact responsible for the penalty phase. And Scofield also\\n\\nrepeatedly testified that he did not know what Downey may or may not have done\\n\\nto prepare. Furthermore, his lack of knowledge about Downey’s preparation was\\n\\nunderstandable. He and Downey were not colleagues who shared an office and\\n\\nmight reasonably be expected to know what the other was doing much of the time.\\n\\nScofield and Downey did not even work in the same city.\\n\\n\\n 25\\n\\n This case is unlike those in which we have found deficient performance\\n\\nwhere the record showed that counsel investigated insufficiently because he\\n\\n“waited until the eleventh hour” to begin preparing for the penalty phase while\\n\\nfaced with “overwhelming evidence of guilt.” See, e.g., Johnson v. Sec’y, D.O.C.,\\n\\n643 F.3d 907, 932 (11th Cir. 2011). The limited record that Jenkins developed\\n\\nestablishes that Downey spent at least 71 hours preparing for trial, spread out over\\n\\n17 months. The state court interpreted this sparse record as reflecting a penalty-\\n\\nphase strategy of residual doubt, the logical extension of Downey’s guilt-phase\\n\\ninvestigation of the evidence and Scofield’s pursuit at trial of an acquittal based on\\n\\nreasonable doubt. See Jenkins, 972 So. 2d at 147. In such a situation, “[a] lawyer’s\\n\\ntime and effort in preparing to defend his client in the guilt phase of a capital case\\n\\ncontinues to count at the sentencing phase.” Tarver v. Hopper, 169 F.3d 710, 715\\n\\n(11th Cir. 1999).\\n\\n We decline to dictate how many hours or how many sessions an attorney\\n\\nmust spend interviewing his counsel in order to provide effective assistance. “No\\n\\nabsolute rules dictate what is reasonable performance for lawyers.” Chandler, 218\\n\\nF.3d at 1317 (citing Strickland, 466 U.S. at 688–89). But even this limited record\\n\\nundermines Jenkins’s assertion that Downey prepared inadequately. Downey’s 71\\n\\nhours of preparation included 35 hours of jailhouse conversations with Jenkins\\n\\n\\n\\n\\n 26\\n\\nduring 18 separate meetings.7 According to Downey’s notes, Scofield was present\\n\\nfor only five of those meetings (12 hours’ worth), leaving 23 hours over 13\\n\\nmeetings between Jenkins and Downey alone. Downey’s notes reveal little about\\n\\nthe content of those conversations beyond broad notations such as “evidence,\\n\\nwitnesses, general status,” and discussions with Jenkins of the evidence and\\n\\ntestimony against him were surely relevant to Downey’s penalty-phase strategy of\\n\\nresidual doubt.\\n\\n But the record does not reveal that Downey knew or should have known that\\n\\nJenkins’s childhood merited investigation beyond that which he may have done.\\n\\nThe record shows that Downey developed enough of a rapport with Jenkins that,\\n\\naccording to his fee declaration, he “g[ave a] Christmas gift [and] discuss[ed] how\\n\\nmuch to tell his family.” But how much Downey knew about Jenkins’s family\\n\\nbackground can only be guessed from Scofield’s Rule 32 hearing testimony.\\n\\nScofield testified that he discussed Jenkins’s background during his dozen or so\\n\\njailhouse meetings with him, and that Downey was present for some of those\\n\\nmeetings. Scofield—and possibly Downey—learned that Jenkins experienced\\n\\nabuse as a child, that his stepfather was brutal, and that he lived on the streets and\\n\\n\\n\\n\\n7\\n We note that the state court clearly erroneously stated that Downey’s fee declaration\\ndocumented 171 hours of pretrial preparation. See Jenkins, 972 So. 2d at 137. However, the state\\ncourt accurately stated that “he spent over 25 hours talking with Jenkins in more than 10 visits to\\nthe jail.” Id. That statement was not an unreasonable determination of the facts.\\n 27\\n\\nin juvenile institutions because of his brutal home life. It is against this backdrop\\n\\nthat Downey’s presumptive “decision not to investigate must be directly assessed\\n\\nfor reasonableness in all the circumstances, applying a heavy measure of deference\\n\\nto counsel’s judgments.” Strickland, 466 U.S. at 691.\\n\\n That assessment depends a great deal on what Jenkins told his counsel about\\n\\nhis background. See id. But we do not know whether Downey was present when\\n\\nScofield learned about Jenkins’s childhood abuse, nor what else Downey may have\\n\\nlearned during his 13 solo discussions with Jenkins. Therefore, “given the absence\\n\\nof evidence in the record, we must assume counsel carried out his professional\\n\\nresponsibility and discussed mitigation with his client.” Chandler, 218 F.3d at\\n\\n1324. We will not “turn that presumption on its head by giving [Jenkins] the\\n\\nbenefit of the doubt where it is unclear what [Downey] did or did not do.” Williams\\n\\nv. Head, 185 F.3d 1223, 1235 (11th Cir. 1999). We will not assume that Jenkins\\n\\ntold Downey more than he told Scofield, and Scofield denied remembering that\\n\\nJenkins told him about the severity or frequency of his stepfather’s physical abuse.\\n\\nHe denied remembering Jenkins telling him about being beaten daily, being locked\\n\\nin his room 24/7, or being forced to eat his feces. And there is no suggestion in the\\n\\nrecord that counsel knew or had reason to investigate the allegation of sexual abuse\\n\\nby his stepgrandfather. The record simply does not reflect that Jenkins told either\\n\\nof his lawyers about the severe abuse that some of the Rule 32 witnesses later\\n\\n\\n 28\\n\\ndescribed. We will not fault those lawyers for failing to present as mitigation this\\n\\nlimited evidence that they may have reasonably believed to be unremarkable. See\\n\\nStrickland, 466 U.S. at 691.\\n\\n To be sure, counsel’s failure “to discover all reasonably available mitigating\\n\\nevidence” may constitute deficient performance in some cases. Wiggins v. Smith,\\n\\n539 U.S. 510, 524 (2003) (quoting ABA Guidelines for the Appointment and\\n\\nPerformance of Counsel in Death Penalty Cases § 11.4.1(C) (1989)). Our Court\\n\\nhas identified deficient performance in several instances where counsel had good\\n\\nreason to investigate the defendant’s background but failed to do so. See Morrow,\\n\\n886 F.3d at 1147 (collecting cases). But we assess “a particular decision not to\\n\\ninvestigate . . . for reasonableness in all the circumstances, applying a heavy\\n\\nmeasure of deference to counsel’s judgments.” Strickland, 466 U.S. at 691. Thus,\\n\\nwe have rejected a hard rule that counsel’s “strategic decisions can be considered\\n\\nreasonable only if they are preceded by a ‘thorough investigation.’” Williams, 185\\n\\nF.3d at 1237 (quoting Rogers v. Zant, 13 F.3d 384, 387 (11th Cir. 1994)). Because\\n\\nwe simply do not know why Downey chose not to investigate Jenkins’s childhood\\n\\nmore thoroughly, Downey is entitled to the presumption that this strategic decision\\n\\nwas reasonable.\\n\\n And though we have also rejected “a per se rule of ineffective assistance\\n\\nwhere counsel does not consult family members,” Holladay v. Haley, 209 F.3d\\n\\n\\n 29\\n\\n1243, 1252 (11th Cir. 2000), we note that Downey did not fail altogether to consult\\n\\nfamily members. Downey had at least one substantive conversation with Jenkins’s\\n\\ngrandmother, and we know from her Rule 32 hearing testimony that she never\\n\\nwitnessed Jenkins’s childhood abuse. Given the silence of the record and the strong\\n\\npresumption of competence to which Downey is entitled, we decline to fault him\\n\\nfor not doing more when Jenkins and his grandmother did not and could not,\\n\\nrespectively, provide a basis for doing so. Cf. United States v. Pease, 240 F.3d 938,\\n\\n941–92 (11th Cir. 2001) (“we cannot say as a matter of law that reliance on a\\n\\nclient’s statements is per se deficient performance”). Nor did Jenkins’s counsel\\n\\nhave an absolute duty to put on abusive-childhood evidence, even if they knew of\\n\\nit. See Chandler, 218 F.3d at 1319 (“Counsel is not required to present every\\n\\nnonfrivolous defense; nor is counsel required to present all mitigation evidence,\\n\\neven if the additional mitigation evidence would not have been incompatible with\\n\\ncounsel’s strategy.”); Waters v. Thomas, 46 F.3d 1506, 1511 (11th Cir. 1995) (en\\n\\nbanc) (“Our decisions are inconsistent with any notion that counsel must present\\n\\nall available mitigating circumstance evidence.”).\\n\\n Instead, we affirm as a reasonable application of Strickland the state court’s\\n\\nconclusion that Downey’s strategic decision to emphasize residual doubt was\\n\\nreasonable. See Jenkins, 972 So. 2d at 147. As we have noted, “[a] lawyer’s time\\n\\nand effort in preparing to defend his client in the guilt phase of a capital case\\n\\n\\n 30\\n\\ncontinues to count at the sentencing phase.” Tarver, 169 F.3d at 715. This\\n\\nreasoning is especially applicable where, as here, counsel pursued the reasonable\\n\\nstrategy of arguing for acquittal at the guilt phase and for life at the sentencing\\n\\nphase by highlighting doubt about the evidence. “[R]esidual doubt is perhaps the\\n\\nmost effective strategy to employ at sentencing.” Chandler, 218 F.3d at 1320 n.28.\\n\\n“[E]specially when—as in this case—the evidence of guilt was not overwhelming,\\n\\nwe expect that petitioners can rarely (if ever) prove a lawyer to be ineffective for\\n\\nrelying on this seemingly reasonable strategy to defend his client.” Id. at 1320\\n\\n(footnote omitted); see also Tarver, 169 F.3d at 715–16 (collecting law review\\n\\narticles on the merits of residual-doubt strategy). Not only can such a strategy be\\n\\nconstitutionally adequate, in some cases it may be “the most effective performance\\n\\nin defense to the death penalty.” Tarver, 169 F.3d at 716.\\n\\n We cannot say that Downey’s manifest strategy of residual doubt was\\n\\nobjectively unreasonable under the circumstances. Jenkins’s case is precisely the\\n\\nsort where reminding the jury of the misgivings it may have held about the State’s\\n\\ncase could have been effective. Scofield’s guilt-phase closing argument from that\\n\\nvery morning, which explicitly emphasized doubt and the weakness of the State’s\\n\\nevidence, would have been fresh in the minds of the jury.8 Although we are not\\n\\n\\n8\\n During the penalty phase of a capital case, “[e]vidence presented at the trial of the case may be\\nconsidered insofar as it is relevant to the aggravating and mitigating circumstances without the\\nnecessity of re-introducing that evidence at the sentence hearing.” Ala. Code § 13A-5-45(c).\\n 31\\n\\nquestioning the jury’s guilty verdict or the sufficiency of the evidence, which are\\n\\nbeyond the scope of our habeas review, we acknowledge that the State’s case\\n\\nagainst Jenkins was—as Downey reminded the jury during the penalty phase—\\n\\nentirely circumstantial. Indeed, Scofield maintained post-conviction that the case\\n\\nfor reasonable doubt was “[v]ery strong.” No one saw Jenkins commit either the\\n\\nmurder or the robbery and kidnapping that elevated Hogeland’s murder to a capital\\n\\ncrime. Scofield’s guilt-phase closing argument challenged the inferential leaps the\\n\\nState was asking the jury to make in finding that Jenkins committed these capital-\\n\\npredicate offenses and that he formed the intent to kill. This situation was not one\\n\\nwhere a guilt-phase strategy of acquittal—and therefore a penalty-phase strategy of\\n\\nresidual doubt—was unreasonable in light of the evidence. Cf. Johnson, 643 F.3d\\n\\nat 932–33 (deficient sentencing investigation where defendant had confessed to\\n\\ntwo murders).\\n\\n Thus, although Downey’s penalty-phase closing argument did not\\n\\nspecifically invoke lingering doubt, it clearly reminded the jury of the entirely\\n\\ncircumstantial nature of the State’s guilt-phase case. See Chandler, 218 F.3d at\\n\\n1320 & n.6 (a strategy of “focus[ing] on obtaining an acquittal and then, at\\n\\nsentencing, on lingering doubt” is reasonable, even if counsel does not use the\\n\\nwords “lingering doubt” or “residual doubt”). Downey closed this reminder by\\n\\nexplaining to the jury how a foundational ancient legal system proscribed capital\\n\\n\\n 32\\n\\npunishment in the absence of eyewitness testimony—the exact situation here. The\\n\\nstate court found that Downey’s penalty-phase closing argument reflected a\\n\\nstrategic “decision to concentrate on reasonable doubt,” see Jenkins, 972 So. 2d at\\n\\n147, a tactical choice entitled to a “strong presumption of correctness,” Marek, 62\\n\\nF.3d at 1300 & n.3.\\n\\n Jenkins suggests other theories of mitigation that Downey might have more\\n\\neffectively pursued,9 but our deferential review under Strickland does not ask\\n\\nwhether counsel could possibly or ideally have been more effective. “The test for\\n\\nineffectiveness is not whether counsel could have done more.” Waters, 46 F.3d at\\n\\n1518. We do not ask whether an attorney’s representation “deviated from best\\n\\npractices or common custom,” and we should resist the temptation to second-guess\\n\\nan attorney with the benefit of our hindsight. Richter, 562 U.S. at 105. We ask only\\n\\n“whether, in light of all the circumstances, the identified acts or omissions were\\n\\n\\n\\n\\n9\\n In particular, Jenkins argues that his counsel should have presented “model prisoner” evidence.\\nThe state court noted, “Good conduct during pretrial incarceration is not necessarily a mitigating\\ncircumstance,” before finding that Jenkins had not met his burden of establishing that his counsel\\neven had information about his conduct in jail. Jenkins, 972 So. 2d at 149. We agree, and we\\ndecline to impose a requirement of sua sponte investigation of this kind in every capital case. Cf.\\nWilliams, 529 U.S. at 396 (finding deficient performance based in part on not returning a call\\nfrom a prison ministry volunteer who offered to vouch for defendant’s good behavior in prison).\\n“Surely, counsel is not required to call a witness to testify to facts such as lack of violent nature\\nwhen the jury has rejected such an approach and has found that the defendant is guilty of\\nmurder.” Griffin v. Wainwright, 760 F.2d 1505, 1512 (11th Cir. 1985), vacated on other grounds,\\n476 U.S. 1112 (1986), reaff’d, 874 F.2d 1397 (11th Cir. 1989).\\n 33\\n\\noutside the range of professionally competent assistance.” Strickland, 466 U.S. at\\n\\n690.\\n\\n We conclude that the state court did not unreasonably determine that Jenkins\\n\\nfailed to establish objectively incompetent performance by his counsel during the\\n\\npenalty phase of his trial. Our inquiry is not “what the best lawyers would have\\n\\ndone. Nor is the test even what most good lawyers would have done. We ask only\\n\\nwhether some reasonable lawyer at the trial could have acted, in the circumstances,\\n\\nas defense counsel acted at trial. . . . We are not interested in grading lawyers’\\n\\nperformances.” White v. Singletary, 972 F.2d 1218, 1220–21 (11th Cir. 1992). And\\n\\n“[i]f fairminded jurists could reasonably disagree” about the reasonableness of\\n\\ncounsel’s performance, “then habeas relief is due to be denied.” Johnson, 643 F.3d\\n\\nat 932.\\n\\n 2. Prejudice\\n\\n Even if Jenkins had demonstrated that his counsel performed as no\\n\\nreasonable lawyer could have, he must also demonstrate a “reasonable probability\\n\\nthat, but for counsel’s unprofessional errors, the result of the proceeding would\\n\\nhave been different.” Strickland, 466 U.S. at 694. Given that the jury here\\n\\nrecommended a sentence of death by the narrowest possible vote, 10 to 2, Jenkins\\n\\nneed establish only “a reasonable probability that at least one juror would have\\n\\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537. “A\\n\\n\\n 34\\n\\nreasonable probability is a probability sufficient to undermine confidence in the\\n\\noutcome.” Strickland, 466 U.S. at 694. In measuring that probability here, we will\\n\\n“evaluate the totality of the available mitigation evidence—both that adduced at\\n\\ntrial, and the evidence adduced in the habeas proceeding in reweighing it against\\n\\nthe evidence in aggravation.” Williams, 529 U.S. at 397–98.\\n\\n We begin our analysis by clarifying what exactly Jenkins is arguing about\\n\\nprejudice. We do not know precisely which aggravating and mitigating factors the\\n\\njury may have implicitly found before making its recommendation of death. The\\n\\ntrial court found at sentencing two statutory aggravating factors: the robbery and\\n\\nthe kidnapping of the victim. See Ala. Code § 13A-5-49(4). Because none of the\\n\\nmitigation evidence offered in the Rule 32 hearing bears on those factors, we see\\n\\nno reasonable probability that the aggravating factors would have been determined\\n\\ndifferently by the jury or the court. The court also found at sentencing two\\n\\nstatutory mitigating factors: no significant history of prior criminal activity and\\n\\nage. See id. § 13A-5-51(1), (7). There was also strong evidence of Jenkins’s\\n\\nintoxication, which was a key part of Jenkins’s guilt-phase defense and could have\\n\\nsupported a third mitigating factor, impaired capacity. See id. § 13A-5-51(6).10\\n\\n\\n10\\n The jury had heard evidence about all three of these statutory mitigators during the guilt phase\\nof the trial. Downey’s penalty-phase opening statement reviewed for the jury all seven of the\\npossible statutory mitigating factors, and the court also recited them in its jury instructions,\\nreminding the jury that they could consider guilt-phase evidence as well as non-statutory\\nmitigation. Jenkins has not established that there is a substantial probability that the jury’s\\nrecommendation would have been different had it been more explicitly reminded of Jenkins’s\\n 35\\n\\nJenkins does not argue that the jury or the court would have found other statutory\\n\\nmitigators in light of the Rule 32 evidence. Rather, he contends that the jury might\\n\\nhave found non-statutory mitigation based on Jenkins’s abusive childhood, low\\n\\nintelligence, or mental health problems that would have outweighed the\\n\\naggravating factors.11\\n\\n The Rule 32 trial court found that Jenkins had not demonstrated prejudice\\n\\neven assuming that counsel had performed deficiently. See Jenkins, 972 So. 2d at\\n\\n138. It concluded that he had not shown a reasonable probability that the sentencer\\n\\n“would have concluded that a weighing of the aggravating and mitigating\\n\\ncircumstances did not warrant death” upon a reweighing of the evidence. See id. It\\n\\nemphasized the depravity of the crime and the aggravating circumstances and\\n\\nfound that the mitigating witnesses were not credible for various reasons. See id.\\n\\nThe state appellate court affirmed, declining to disturb the credibility\\n\\ndeterminations of the trial court. Id. at 142–44. Independently reweighing the\\n\\n\\nage, criminal history, and intoxication, given that the trial court, the Rule 32 trial court, and the\\nstate appellate court all explicitly considered those mitigators and found them outweighed by the\\naggravators.\\n11\\n Alabama law lists seven statutory mitigating circumstances, but states that mitigation “shall\\n. . . not be limited to” those seven circumstances. Ala. Code § 13A-5-51. Further, “mitigating\\ncircumstances shall include any aspect of a defendant’s character or record and any of the\\ncircumstances of the offense that the defendant offers as a basis for a sentence of life\\nimprisonment without parole instead of death, and any other relevant mitigating circumstances\\nwhich the defendant offers as a basis for a sentence of life imprisonment without parole instead\\nof death.” Id. § 13A-5-52. A jury may recommend a sentence of death only if it finds that “one or\\nmore aggravating circumstances . . . exist and that they outweigh the mitigating circumstances, if\\nany.” Id. § 13A-5-46(e)(3).\\n 36\\n\\nevidence, it concluded that, “[g]iven the aggravating circumstances that were\\n\\nproven by the State and the facts surrounding Hogeland’s murder, we, like the\\n\\ncircuit court, are confident that death was the appropriate punishment for Jenkins’s\\n\\nactions.” Id. at 148.\\n\\n Our review of this evidence does not show that it is likely that the jury\\n\\nwould have concluded that death was not the appropriate punishment if it had\\n\\nheard the Rule 32 evidence. Much of the proffered mitigation evidence is, in our\\n\\nview, much less mitigating than Jenkins asserts. In particular, the testimony of\\n\\nJenkins’s older brother Michael was likely to have been a double-edged sword. See\\n\\nEvans v. Sec’y, Dep’t of Corr., 703 F.3d 1316, 1327 (11th Cir. 2013) (en banc) (“it\\n\\nis reasonable to conclude that a defendant was not prejudiced when his mitigation\\n\\nevidence was a two-edged sword” (quoting Ponticelli v. Sec’y, Fla. Dep’t of Corr.,\\n\\n690 F.2d 1271, 1296 (11th Cir. 2012)). Michael was the closest relative to testify at\\n\\nthe Rule 32 hearing and his testimony about the abuse Jenkins experienced should\\n\\nhave been the most reliable, since he experienced much of the same physical abuse\\n\\nhimself. The brothers shared the same filthy bedroom, received beatings from the\\n\\nsame paddle, and were frequently absent from school together. But Michael’s\\n\\ntestimony contained some discrepancies, and he admitted to cognitive and memory\\n\\nproblems. Perhaps Michael’s testimony would have made Jenkins appear more\\n\\nsympathetic in the eyes of the jury—but perhaps not. While Jenkins ran away from\\n\\n\\n 37\\n\\nhome and eventually committed murder, Michael responded to the same\\n\\nbackground by running away from home, graduating from high school, and\\n\\nproviding health care for both the terminally ill and people with substance use\\n\\ndisorders. It is possible that the jury could even have found Jenkins to be more\\n\\nculpable for not overcoming his difficult upbringing and to pose more of a danger\\n\\nof future violence.\\n\\n We also find reasonable the state court’s identification of other problems\\n\\nwith the mitigating evidence Jenkins argues should have been presented. Hearing\\n\\nthat evidence firsthand in the Rule 32 proceeding, the trial court found the\\n\\ntestimony of the cousins less credible in light of the lack of physical abuse\\n\\ndocumented in the school health records. See Jenkins, 972 So. 2d at 139 (noting\\n\\nthat school records observed a rash and gingivitis). It faulted Pitts in particular for\\n\\npurporting to have witnessed ongoing, severe abuse but taking only minimal,\\n\\nineffective steps to obtain outside help. See id. It further noted that DeLavega and\\n\\nWagoner actually testified that they never saw Jenkins beaten, and it noted all of\\n\\nthe witnesses’ interest in getting Jenkins off death row. All of these factors would\\n\\nlikely have caused a reasonable jury to discount their testimony. See id. at 139–42.\\n\\nWe must accept the trial court’s views of the Rule 32 testimony. “In the absence of\\n\\nclear and convincing evidence, we have no power on federal habeas review to\\n\\nrevisit the state court’s credibility determinations.” Bishop v. Warden, 726 F.3d\\n\\n\\n 38\\n\\n1243, 1259 (11th Cir. 2013) (citing Marshall v. Lonberger, 459 U.S. 422, 434\\n\\n(1983)).\\n\\n The state court also noted that Wagoner testified that she was not available\\n\\nto testify at the trial, and it found that Seal’s testimony would have been\\n\\ncumulative to that of her husband. Thus, counsel’s failure to call those two\\n\\nwitnesses could not have been prejudicial to Jenkins. See id. at 141, 142. Neither of\\n\\nthese conclusions involves an unreasonable determination of the facts. We agree\\n\\nthat the Rule 32 testimony of Sherry Seal was substantially similar to the testimony\\n\\nof her husband Lonnie Seal that the jury heard during the penalty phase. Both\\n\\ntestified about meeting Jenkins in California, Jenkins’s assistance with their move\\n\\nto Alabama, and his residence with them in their Alabama home. Sherry added\\n\\nonly that Jenkins sometimes wet his bed. It was not unreasonable for the state court\\n\\nto conclude that Sherry’s testimony was cumulative of her husband’s, and “a\\n\\npetitioner cannot satisfy the prejudice prong of the Strickland test with evidence\\n\\nthat is merely cumulative of evidence already presented.” Rose v. McNeil, 634 F.3d\\n\\n1224, 1243 (11th Cir. 2011).\\n\\n Further, the school, medical, and juvenile court records that Jenkins\\n\\nintroduced in the Rule 32 proceeding have their own problems. As noted, they do\\n\\nnot document or even corroborate the severe level of abuse alleged by some of the\\n\\nwitnesses. Though these documents tell a sad story, they do not create a substantial\\n\\n\\n 39\\n\\nlikelihood that the jury would have recommended life. Although they portray\\n\\nJenkins as a victim of truly unfortunate circumstances, they also show his unsavory\\n\\ndelinquent behavior. That delinquency included a history of stealing cars, and\\n\\ngiven that the jury likely inferred that Jenkins stole the red Mazda, new\\n\\ninformation about his history of theft may not have been especially mitigating.\\n\\nSimilarly, we agree with the state court that evidence of Jenkins’s compliant\\n\\nconduct in pretrial detention is “minimally mitigating” where there had been no\\n\\nevidence of jail misconduct as an aggravator. See Jenkins, 972 So. 2d at 149; cf.\\n\\nSkipper v. South Carolina, 476 U.S. 1, 5 & n.1 (1986) (good jail behavior is\\n\\n“potentially mitigating” especially “[w]here the prosecution specifically relies on a\\n\\nprediction of future dangerousness in asking for the death penalty”).\\n\\n Jenkins argues that allowing the jury to see his experience of mental,\\n\\nemotional, and cognitive difficulties, compounded by years of abuse and neglect\\n\\nand considered alongside his mild character and good jail behavior as an adult,\\n\\nwould have convinced them that Hogeland’s murder was an “aberration” that\\n\\ndeserved a punishment less severe than death. That interpretation of Jenkins’s life\\n\\nstory may be conceivable. But more than mere conceivability is required to\\n\\nestablish prejudice: “The likelihood of a different result must be substantial, not\\n\\njust conceivable.” Richter, 562 U.S. at 112 (emphasis added). We do not find the\\n\\nstate court’s decision that a different result was not substantially likely to be an\\n\\n\\n 40\\n\\nunreasonable application of Strickland. As we have noted, we think it equally\\n\\nlikely that the jury could have rejected this alternate narrative out of hand in light\\n\\nof having just convicted him of the violent murder of a young mother. We also\\n\\nthink it possible that the jury could have viewed the murder as just one step in a\\n\\nlogical progression of degeneracy that began with behavioral problems at school,\\n\\ncontinued through the juvenile court system, and escalated past grand theft auto\\n\\ninto kidnapping and murder. Taken separately as well as together, the proffered\\n\\nevidence does not compel us to conclude that a verdict of life would have been\\n\\nsubstantially likely. Because fairminded jurists could disagree about the state\\n\\ncourt’s conclusion about Strickland prejudice, we affirm the denial of habeas relief\\n\\non this ground.\\n\\n B. Intellectual Disability\\n\\n Jenkins argues that the district court erred when it denied his claim that he is\\n\\nintellectually disabled and ineligible for the death penalty under Atkins v. Virginia,\\n\\n536 U.S. 304 (2002). The Supreme Court held in Atkins that executing an offender\\n\\nwho has mental retardation12 violates the Eighth Amendment’s ban on excessive\\n\\n\\n\\n\\n12\\n In 2002, the Atkins Court followed prevailing convention in referring to “mentally retarded”\\noffenders. We will use that now-disfavored term only when quoting or discussing older sources.\\nSince 2014, the Supreme Court has exclusively used the terms “intellectually disabled person”\\nand “person with intellectual disability.” See, e.g., Hall v. Florida, 572 U.S. 701, 704 (2014);\\nMoore v. Texas, 139 S. Ct. 666, 672 (2019).\\n 41\\n\\npunishments. Id. at 321. It defined mental retardation by quoting the American\\n\\nAssociation on Mental Retardation:\\n\\n Mental retardation refers to substantial limitations in present\\n functioning. It is characterized by significantly subaverage intellectual\\n functioning, existing concurrently with related limitations in two or\\n more of the following applicable adaptive skill areas: communication,\\n self-care, home living, social skills, community use, self-direction,\\n health and safety, functional academics, leisure, and work. Mental\\n retardation manifests before age 18.\\n\\nId. at 308 n.3. Alabama courts have followed that tripartite definition, adopting\\n\\nwhat they consider the “broadest” definition of mental retardation for Atkins\\n\\npurposes:\\n\\n [A] defendant, to be considered mentally retarded, must have\\n significantly subaverage intellectual functioning (an IQ of 70 or\\n below), and significant or substantial deficits in adaptive behavior.\\n Additionally, these problems must have manifested themselves during\\n the developmental period (i.e., before the defendant reached age 18).\\n\\nEx parte Perkins, 851 So. 2d 453, 456 (Ala. 2002).\\n\\n The clinical evidence adduced during the Rule 32 hearing with respect to\\n\\nJenkins’s intellectual ability was that his IQ was 76, reflecting “borderline\\n\\nintellectual functioning.” The record also included the report of the Lunacy\\n\\nCommission13 which found that Jenkins was competent to stand trial, as well as\\n\\n\\n\\n13\\n Alabama law requires a capital defendant to be “committed to the Department of Mental\\nHealth and Mental Retardation for examination” if there is a question about his capacity to stand\\ntrial or if there will be a mental-disease-or-defect defense. Ala. Code § 15-16-22(a). “The\\nassigned mental health professional(s) shall examine the defendant with respect to determining\\nthe presence of any mental disease or defect which, if determined to be present, would affect the\\n 42\\n\\nJenkins’s school and juvenile court records. The court also considered all the\\n\\ntestimony it had heard about Jenkins and reviewed the transcript of his murder trial\\n\\nbefore rejecting Jenkins’s claim of mental retardation.\\n\\n On appeal, after reciting the Perkins standard, the Alabama Court of\\n\\nCriminal Appeals explained that Jenkins had not established mental retardation:\\n\\n Dr. Kirkland testified that he performed psychological tests on Jenkins\\n and that Jenkins’s IQ was 76. There was evidence presented at\\n Jenkins’s trial indicating that Jenkins maintained relationships with\\n other individuals and that he had been employed by P.S. Edwards\\n Landscaping Company, Cotton Lowe 76 Service Station, and\\n Paramount Painting Company. The record fails to show that Jenkins\\n meets the most liberal view of mental retardation adopted by the\\n Alabama Supreme Court in Perkins.\\n\\nJenkins, 972 So. 2d at 155. In this habeas appeal, Jenkins raises arguments about\\n\\nall three components of the intellectual disability showing under Atkins and\\n\\nPerkins, which we address in turn. He also appeals the denial of an evidentiary\\n\\nhearing on this issue.\\n\\n 1. Intellectual Function\\n\\n The first component of intellectual disability under Perkins is “significantly\\n\\nsubaverage intellectual functioning (an IQ of 70 or below).” Perkins, 851 So. 2d at\\n\\n456. Jenkins argues that the use of a strict IQ cutoff of 70 was contrary to or an\\n\\nunreasonable application of Atkins.\\n\\n\\ncapacity of the defendant to proceed or continue to trial or which would affect the defendant’s\\ncriminal responsibility at the time of the commission of the crime.” Id.\\n 43\\n\\n We disagree. The state court’s decision that Jenkins failed to establish\\n\\nmental retardation in this respect was a reasonable application of the record and of\\n\\nAtkins. As a threshold matter, we note that the state court did not say it was\\n\\napplying a strict cutoff of 70; it merely stated that Dr. Kirkland had measured\\n\\nJenkins’s IQ as 76. Jenkins, 972 So. 2d at 155. Under our deferential review, the\\n\\nquestion is whether the decision that Jenkins had not established the intellectual\\n\\nfunctioning aspect of intellectual disability was an unreasonable application of\\n\\nAtkins. The Supreme Court explained in Atkins that it was leaving “to the State[s]\\n\\nthe task of developing appropriate ways to enforce the constitutional restriction\\n\\nupon [their] execution of sentences.” Atkins, 536 U.S. at 317 (quoting Ford v.\\n\\nWainwright, 477 U.S. 399, 416–17 (1986)). In line with that invitation, our Circuit\\n\\nhas accepted Alabama’s judicial definition of intellectual disability as laid out in\\n\\nPerkins.14 See Thomas, 607 F.3d at 752–53. Jenkins objects, however, that the\\n\\nPerkins standard, which he characterizes as articulating a “strict cutoff” of 70,\\n\\nconflicts with more recent Supreme Court precedent that instead requires a\\n\\n“clinical” definition that goes beyond a single IQ score.\\n\\n In particular, Jenkins points to Hall v. Florida, in which the Supreme Court\\n\\nheld that when an offender’s IQ is near but greater than 70, courts must take into\\n\\n\\n\\n14\\n Notwithstanding the urging of the Alabama Supreme Court, see Perkins, 851 So. 2d at 455\\nn.1, the Alabama legislature has not enacted a statutory definition of intellectual disability.\\n 44\\n\\naccount the IQ test’s standard error of measurement as well as non-IQ evidence of\\n\\nintellectual function. See Hall v. Florida, 572 U.S. 701, 723 (2014) (“This Court\\n\\nagrees with the medical experts that when a defendant’s IQ test score falls within\\n\\nthe test’s acknowledged and inherent margin of error, the defendant must be able\\n\\nto present additional evidence of intellectual disability, including testimony\\n\\nregarding adaptive deficits.”). But, as Jenkins concedes, our Circuit has\\n\\nspecifically held that Hall is not retroactive to cases on collateral review. See In re\\n\\nHenry, 757 F.3d 1151, 1161 (11th Cir. 2014) (explaining that Hall does not supply\\n\\na new substantive rule but rather “merely provides new procedures for ensuring\\n\\nthat States do not execute members of an already protected group”). We also note\\n\\nthat Alabama courts have found the Perkins standard to comport with Hall in that\\n\\nit does not prevent an offender from presenting the standard error of measurement\\n\\nor non-IQ evidence of intellectual disability. See Reeves v. State, 226 So. 3d 711,\\n\\n728–29 (Ala. Ct. Crim. App. 2016). Here, the state court did not unreasonably\\n\\napply Atkins when it analyzed the intellectual functioning component of\\n\\nintellectual disability.\\n\\n Jenkins also argues that the state court’s decision on this component in his\\n\\nRule 32 proceeding resulted from an unreasonable determination of the facts. He\\n\\nasserts that his IQ score of 76, when combined with the test’s standard error of\\n\\nmeasurement and considered alongside the Flynn effect (which we discuss below),\\n\\n\\n 45\\n\\ndoes show that he has significantly subaverage intellectual function. For the\\n\\nreasons that follow, we disagree.\\n\\n Most fundamentally, neither of the clinicians who testified about Jenkins’s\\n\\nintellectual ability in the Rule 32 hearing opined that he was mentally retarded.\\n\\nThe closest thing Jenkins can point to is the testimony of Dr. Kirkland that his IQ\\n\\nis two standard deviations below the mean. But his entire testimony, in context, did\\n\\nnot establish intellectual disability. To the contrary: Dr. Kirkland testified that\\n\\nJenkins “scored in the range of borderline intellectual functioning which is\\n\\nbetween mild mental retardation and low average intellectual functioning.”\\n\\nFurthermore, Jenkins’s own expert Dr. Lisak fully reviewed Dr. Kirkland’s clinical\\n\\nfindings. He agreed with those findings, concluding that the IQ results show\\n\\n“borderline intelligence.” It is tremendously significant that no clinical assessment\\n\\nin the entire record of these proceedings has found that Jenkins has mental\\n\\nretardation or intellectual disability.\\n\\n Jenkins nonetheless argues that his IQ score of 76 establishes the intellectual\\n\\ncomponent of intellectual ability when the standard error of measurement and the\\n\\nFlynn effect are considered alongside that score. It was not unreasonable for the\\n\\nstate court to disagree. As we have discussed, Jenkins is not retroactively entitled\\n\\nto have the state courts take into account the margin of error of the IQ test. That\\n\\nstatistical margin of error—5 points in this case—takes into account characteristics\\n\\n\\n 46\\n\\nof the test and variations due to chance in order to state how confident the tester is\\n\\nthat the measured result accurately states Jenkins’s true, unknowable IQ. A result\\n\\nof 76 with a standard error of measurement of 5 simply means that the tester is\\n\\n68% confident that Jenkins’s true IQ falls between 71 and 81, and 95% confident\\n\\nthat his true IQ falls between 66 and 86. See David H. Kaye & David A. Freedman,\\n\\nReference Guide on Statistics, in Reference Manual on Scientific Evidence 211,\\n\\n243–44 (3d ed. 2011); see also Hall, 572 U.S. at 738–39 (Alito, J., dissenting). Of\\n\\ncourse, Jenkins would like us to consider only the possibility that the test result\\n\\noverstated his IQ—but the standard error of measurement is a two-way street.\\n\\n“[T]he standard error of measurement is a bi-directional concept that does not\\n\\ncarry with it a presumption that an individual’s IQ falls to the bottom of his IQ\\n\\nrange.” Ledford v. Warden, Ga. Diag. & Clas. Prison, 818 F.3d 600, 641 (11th Cir.\\n\\n2016).\\n\\n Even if Jenkins were somehow entitled to subtract the 5-point margin of\\n\\nerror from his score and thereby lower his “true” IQ to 71, he still would not satisfy\\n\\nthe clinical intellectual component of intellectual disability. Thus, he sought in the\\n\\ndistrict court to introduce evidence about the Flynn effect in order to bring his IQ\\n\\ndown to 70 or below. As our Court has explained, the Flynn effect “recognizes the\\n\\nfact that IQ test scores have been increasing over time”:\\n\\n The Flynn effect acknowledges that as an intelligence test ages, or\\n moves farther from the date on which it was standardized, or normed,\\n 47\\n\\n the mean score of the population as a whole on that assessment\\n instrument increases, thereby artificially inflating the IQ scores of\\n individual test subjects. Therefore, the IQ test scores must be\\n recalibrated to keep all test subjects on a level playing field.\\n\\nThomas, 607 F.3d at 753. We have said that “[a]n evaluator may also consider the\\n\\n‘Flynn effect’” in assessing an offender’s possible intellectual disability, id.\\n\\n(emphasis added), but neither we nor the Alabama Supreme Court nor the U.S.\\n\\nSupreme Court have said that a court must consider it in order to reasonably apply\\n\\nAtkins. See, e.g., Reeves, 226 So. 3d at 739 (“This Court has repeatedly held that a\\n\\ncircuit court is not required to accept, consider, or apply the ‘Flynn Effect’ in\\n\\ndetermining intellectual disability.”).\\n\\n More problematically, Jenkins did not raise the Flynn effect issue in the\\n\\nAlabama state courts. Jenkins asserts that he implicitly raised it as part and parcel\\n\\nof his Atkins claim during his Rule 32 appeal, but we do not agree that the\\n\\nAlabama courts had a full and fair opportunity to consider this issue, because it\\n\\nwas not explicitly presented to them. See Kelley v. Sec’y for the Dep’t of Corr., 377\\n\\nF.3d 1317, 1349–50 (11th Cir. 2004) (holding that one ineffective-assistance\\n\\nargument was unexhausted despite the petitioner’s general presentation of a\\n\\nStrickland claim in the state courts). Thus, under AEDPA, Jenkins is procedurally\\n\\nbarred from raising this unexhausted claim in his federal habeas petition. See 28\\n\\nU.S.C. § 2254(b)(1)(A), (c); Mason v. Allen, 605 F.3d 1114, 1119 (11th Cir. 2010).\\n\\nWe will allow a prisoner to overcome a procedural bar if he establishes cause for\\n 48\\n\\nhis default and prejudice therefrom. See, e.g., Smith v. Jones, 256 F.3d 1135, 1138\\n\\n(11th Cir. 2001). But Jenkins cannot establish prejudice from being barred from\\n\\nasserting this claim now because the Supreme Court has not “clearly established”\\n\\nthat a court must consider the Flynn effect in assessing intellectual disability. Thus,\\n\\nunder AEDPA, it could not form the basis for a federal habeas claim. See 28\\n\\nU.S.C. § 2254(d)(1). We therefore conclude that the state court did not\\n\\nunreasonably determine the facts or unreasonably apply Atkins with respect to the\\n\\nintellectual component of intellectual disability.\\n\\n 2. Adaptive Function\\n\\n The second component of intellectual disability is “significant or substantial\\n\\ndeficits in adaptive behavior.” Perkins, 851 So. 2d at 456. Jenkins argues that the\\n\\nstate court, which mentioned only Jenkins’s successes in employment and social\\n\\nskills, see Jenkins, 972 So. 2d at 155, failed to consider his deficits. He adds that\\n\\neven the mention of employment success does not tell the whole story. Jenkins\\n\\ninsists that his jobs were menial, his employers took advantage of him, and he held\\n\\nthe jobs for only a few months each.\\n\\n But the state court’s terse discussion of Jenkins’s adaptive behavior did not\\n\\ninvolve an unreasonable determination of the facts or an unreasonable application\\n\\nof Atkins. To the contrary, its decision on this issue was reasonable in light of both\\n\\nthe record and Atkins.\\n\\n\\n 49\\n\\n Atkins elaborated that the adaptive-skill component of intellectual disability\\n\\nrequires substantial present limitation in at least two of the following areas:\\n\\n“communication, self-care, home living, social skills, community use, self-\\n\\ndirection, health and safety, functional academics, leisure, and work.” Atkins, 536\\n\\nU.S. at 308 n.3 (quoting definition of American Association on Mental\\n\\nRetardation). The record supports the state court’s judgment that Jenkins does not\\n\\nhave substantial deficits in at least two areas. To be sure, we will assume that\\n\\nJenkins may be substantially limited in functional academics; Dr. Kirkland\\n\\nassessed his reading, spelling, and arithmetic skills at a third-grade level. But the\\n\\nrecord does not compel a finding of substantial present deficits in the areas of\\n\\ncommunication, self-care, home living, and work. Although Jenkins emphasizes\\n\\nhis transient employment, the record shows he was able to find work quickly, to\\n\\nwork hard to support the Seals’ household until Lonnie Seal found a job, and, later,\\n\\nto move into his own home. That record evidence is inconsistent with a substantial\\n\\ndeficit in the area of work.\\n\\n As the State observes, the facts of the crime also fail to show significant\\n\\ndeficits in the areas of communication, self-care, community use, and self-\\n\\ndirection. Jenkins was able to communicate well enough to solicit an alibi and to\\n\\nsell his car, including writing out a bill of sale. He then was able to arrange for his\\n\\nflight from Alabama, including purchasing a bus ticket and hitchhiking across the\\n\\n\\n 50\\n\\ncountry. Indeed, throughout his trial and postconviction proceedings, the people\\n\\nwho knew Jenkins as an adult consistently described a man who did not have\\n\\nserious difficulties in communicating, forming relationships, working, and caring\\n\\nfor himself. No witness mentioned these kinds of difficulties in his or her\\n\\ntestimony about Jenkins. There was some evidence that Jenkins was “not\\n\\narticulate” and that his house was messy, but that testimony does not rise to the\\n\\nlevel of overwhelming evidence of substantial functional deficits that would\\n\\nsupport a conclusion that the state court unreasonably determined the facts.\\n\\n Of course, the record also contains evidence of Jenkins’s childhood\\n\\nacademic and social deficits, attributable at least in part to his parents’ neglect, but\\n\\nJenkins’s childhood is not directly relevant to our consideration of his present\\n\\nlimitations. Although we acknowledge that the Supreme Court has noted that such\\n\\n“risk factors” may support further exploration of the possibility of intellectual\\n\\ndisability, see Moore v. Texas, 137 S. Ct. 1039, 1051 (2017), the state court record\\n\\ncontains the results of that exploration. Overall, that record supports the state\\n\\ncourt’s conclusion that Jenkins does not have substantial deficits in adaptive\\n\\nbehaviors.\\n\\n 3. Juvenile Onset\\n\\n The third and final component of intellectual disability is manifestation of\\n\\nthe first and second components “before the defendant reached age 18.” Perkins,\\n\\n\\n 51\\n\\n851 So. 2d at 456. Jenkins asserts, the State assumes, and we agree that the state\\n\\ncourt did not rule upon this issue, so there is nothing to which we must defer under\\n\\nAEDPA. See, e.g., Brumfield v. Cain, 135 S. Ct. 2269, 2282 (2015). Thus, we will\\n\\nreview the district court’s legal conclusions on this issue de novo, and its factual\\n\\nfindings for clear error. See, e.g., Grossman v. McDonough, 466 F.3d 1325, 1335\\n\\n(11th Cir. 2006). Jenkins argues that he has satisfied the age-of-onset component\\n\\nby showing “a wide range of evidence of deficits in both intellectual and adaptive\\n\\nfunctioning throughout Jenkins’s childhood.”\\n\\n In fact, the ample record of Jenkins’s childhood, as developed in the Rule 32\\n\\nproceeding, does not point to intellectual disability before age 18. School\\n\\nassessments at age 12 reported an IQ score of 83, an adaptive behavior rating of\\n\\n“low,” and noted overall an intellectual capacity of “average,” a diagnosis of\\n\\ndyslexia, and “poor functional skills.” The psychologist attributed Jenkins’s\\n\\ndifficulties to a combination of “[f]requent moves, changes in schools and family\\n\\ninstability” with “[l]ow ability coupled with a specific learning disability” and\\n\\n“[l]ack of responsibility for self control and poor use of time.” Although Jenkins\\n\\nfailed several grades of school, teachers additionally attributed his difficulties to\\n\\nhis behavioral problems and his high number of absences from school.\\n\\n Similarly, IQ tests ordered by the juvenile court at age 14 gave scores of 81\\n\\nand 86 on different tests, in the “[d]ull–[n]ormal” range and reflecting “a definite\\n\\n\\n 52\\n\\nlearning disability.” These records all show a child with serious academic deficits\\n\\nand some intellectual and adaptive deficits, but they do not clearly show an\\n\\nintellectually disabled child. In particular, Jenkins’s childhood IQ scores of 83, 81,\\n\\nand 86 fail to point to intellectual disability or mental retardation. We are not left\\n\\nwith the definite and firm conviction that the district court erred when it found that\\n\\n“Jenkins’s [childhood] intelligence scores and adaptive functioning skills were\\n\\nattributable to his learning disability and other circumstances in his life, rather than\\n\\nmental retardation.”\\n\\n In sum, we agree with the district court that the decision of the Alabama\\n\\nCourt of Criminal Appeals that Jenkins does not have intellectual disability was\\n\\nnot contrary to or an unreasonable application of Atkins v. Virginia.\\n\\n 4. Evidentiary Hearing\\n\\n Finally, Jenkins argues that the district court erred when it denied his request\\n\\nfor an evidentiary hearing on his Atkins claim. If Jenkins diligently attempted to\\n\\npresent this claim in the state court, we review for abuse of discretion. Burgess v.\\n\\nComm’r, Ala. Dep’t of Corr., 723 F.3d 1308, 1320 (11th Cir. 2013). “The district\\n\\ncourt abuses its discretion where ‘such a hearing could enable an applicant to prove\\n\\nthe petition’s factual allegations, which, if true, would entitle the applicant to\\n\\nfederal habeas relief.’” Id. (quoting Schriro v. Landrigan, 550 U.S. 465, 474\\n\\n(2007)). If Jenkins did not diligently attempt to develop the factual basis for this\\n\\n\\n 53\\n\\nclaim in the state court, absent certain circumstances, he is not entitled to a hearing\\n\\nand we will affirm the district court’s denial. 28 U.S.C. § 2254(e)(2).\\n\\n The parties strenuously dispute whether Jenkins diligently attempted to\\n\\npresent this issue in the state court. On the one hand, the State argues that Jenkins\\n\\nfailed to allege that he was mentally retarded even after the U.S. Supreme Court\\n\\ndecided Atkins in 2002. On the other hand, Jenkins asserts that he raised the\\n\\npossibility of mental retardation as early as his original Rule 32 petition in 1995.\\n\\n As we have noted, shortly after the Supreme Court decided Atkins, the\\n\\nAlabama Court of Criminal Appeals ordered supplemental briefing on the possible\\n\\nimpact of Atkins on Jenkins. Jenkins’s Rule 32 appeal was pending at that point,\\n\\nhaving been fully briefed and orally argued in the state appellate court. Having\\n\\nreviewed the state court records, we agree with the State that Jenkins did not\\n\\ndiligently attempt to present his Atkins claim in the state courts.\\n\\n “Diligence will require in the usual case that the prisoner, at a minimum,\\n\\nseek an evidentiary hearing in state court.” Williams, 529 U.S. at 437. The record\\n\\nclearly shows that Jenkins did not do so. In the initial supplemental brief on Atkins\\n\\nhe filed in 2002, Jenkins merely requested a future hearing after the Alabama\\n\\nlegislature’s enactment of a statute on executing the mentally retarded. Seventeen\\n\\nyears later, that contingency is still in the future. Only in his reply supplemental\\n\\nbrief did Jenkins then assert, for the first time ever, that he is mentally retarded.\\n\\n\\n 54\\n\\n We acknowledge that Jenkins’s initial Rule 32 petition included in its\\n\\nallegations of ineffective assistance of counsel a list of unpresented mitigating\\n\\nevidence including his “developmental[] impair[ment]” and “learning disabilities,\\n\\nlow intelligence, poor comprehension and retarded socialization skills.” We\\n\\nconclude that these tangential Strickland references to his mental and social\\n\\nabilities do not amount to diligence. Thus, because Jenkins did not diligently\\n\\nattempt to develop the factual basis for his Atkins claim in the state courts, we\\n\\naffirm the district court’s denial of his request for an evidentiary hearing.\\n\\n Even if Jenkins had diligently attempted to present this claim in the state\\n\\ncourt, however, we conclude in the alternative that the district court did not abuse\\n\\nits discretion when it denied a hearing. We agree, in light of the ample evidence\\n\\nabout intellectual ability that Jenkins adduced in support of his ineffective-\\n\\nassistance claim, that such a hearing would not have enabled Jenkins to prove that\\n\\nhe is intellectually disabled. See Burgess, 723 F.3d at 1320.\\n\\n IV. CONCLUSION\\n\\n The denial of Jenkins’s petition for a writ of habeas corpus is\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 55\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434945/\", \"author_raw\": \"BRANCH, Circuit Judge:\"}, {\"author\": \"WILSON, Circuit Judge, dissenting\", \"type\": \"dissent\", \"text\": \"WILSON, Circuit Judge, dissenting:\\n\\n Mark Jenkins must prove that the Alabama state courts acted unreasonably\\n\\nin denying him post-conviction collateral relief. 28 U.S.C. § 2254(d)(1). Jenkins\\n\\nhas satisfied that burden as to his ineffective assistance of counsel claim. He has\\n\\nalso demonstrated that he is entitled to an evidentiary hearing regarding his Atkins\\n\\nv. Virginia claim. Because the majority came to the opposite conclusions, I\\n\\ndissent.\\n\\n I. Ineffective Assistance of Counsel at Penalty Phase\\n\\n Jenkins first asserts that his trial counsel’s failure to investigate Jenkins’\\n\\nabusive upbringing and to present any mitigating evidence at the penalty phase of\\n\\nhis trial constituted ineffective assistance of counsel. The majority rejects this\\n\\ncontention, concluding that the Alabama state courts did not act unreasonably in\\n\\ndenying Jenkins’ ineffective assistance of counsel claim on collateral review. I\\n\\ndisagree and maintain that Jenkins established both deficient performance and\\n\\nprejudice, entitling him to habeas relief. See Strickland v. Washington, 466 U.S.\\n\\n668, 687 (1984).\\n\\n A. Deficient Performance\\n\\n The Alabama state courts concluded that Jenkins’ counsel, Stan Downey,\\n\\ndid not perform deficiently at the penalty phase. The majority finds that\\n\\nconclusion reasonable primarily because Downey (1) was not obligated to\\n\\ninvestigate Jenkins’ abusive childhood and (2) reasonably pursued a residual doubt\\n\\nstrategy at the penalty phase. But further review of the record requires me to part\\n\\ncompany with the majority’s reasoning and, ultimately, persuades me that the\\n\\nAlabama courts’ conclusion was indeed unreasonable. See 28 U.S.C. § 2254(d)(1)\\n\\n(providing that a writ of habeas corpus should not be granted to a state prisoner\\n\\nunless the state court adjudication involved “a decision that was contrary to, or\\n\\ninvolved an unreasonable application of, clearly established Federal law, as\\n\\ndetermined by the Supreme Court of the United States”).\\n\\n Jenkins offers numerous examples of how his counsel performed deficiently\\n\\nat the penalty phase of trial. First, counsel failed to thoroughly investigate Jenkins’\\n\\nabusive childhood—an error that constitutes deficient performance. In Williams v.\\n\\nAllen, we held that counsel’s failure to investigate mitigating evidence at the\\n\\npenalty stage constituted ineffective assistance of counsel because counsel drew\\n\\nupon only three sources of information—a psychologist’s report, a presentence\\n\\ninvestigation report, and an interview with the appellant’s mother. 542 F.3d 1326,\\n\\n1340 (11th Cir. 2008); see also Debruce v. Comm’r, Ala. Dep’t of Corr., 758 F.3d\\n\\n1263, 1270, 1273–74 (11th Cir. 2014) (finding counsel’s performance deficient\\n\\nwhere counsel interviewed the petitioner and his mother, but failed to contact\\n\\nexpert witnesses or other family members). In that case, we concluded that\\n\\n\\n\\n\\n 57\\n\\ncounsel’s lack of investigation resulted in “an incomplete and misleading\\n\\nunderstanding of Williams’s life history.” Williams, 542 F.3d at 1340.\\n\\n Jenkins’ counsel did even less than the attorneys in Williams and Debruce.\\n\\nThat is, Jenkins’ counsel conducted no investigation. Downey did not consider\\n\\nJenkins’ poor school records, mention Jenkins’ juvenile history, produce\\n\\ndocumentary evidence that Jenkins suffered from severe depression, or pursue\\n\\nrecords demonstrating Jenkins’ significant deficits in cognitive functions. We have\\n\\npreviously found that a counsel’s failure to investigate these rudimentary examples\\n\\nof mitigating evidence constitutes ineffective assistance of counsel. See Daniel v.\\n\\nComm’r, Ala. Dep’t of Corr., 822 F.3d 1248 (2016) (reversing and remanding\\n\\ndistrict court’s dismissal of the ineffective assistance of counsel claim and noting\\n\\nhow counsel should have more thoroughly investigated Daniel’s history of\\n\\nphysical and sexual abuse, his poor school records, and evidence of a potential\\n\\nintellectual disability).\\n\\n Given the lack of preparation, it is not surprising that Downey failed to call\\n\\nany witnesses to testify at the penalty stage about Jenkins’ childhood, thus failing\\n\\nto give a complete depiction of Jenkins’ “excruciating life history.” Wiggins v.\\n\\nSmith, 539 U.S. 510, 537 (2003) (finding deficient performance where counsel did\\n\\nnot adequately investigate petitioner’s life history for mitigating evidence); see\\n\\nalso Debruce, 758 F.3d at 1270, 1273–74; Johnson v. Sec’y, Dep’t of Corr., 643\\n\\n\\n 58\\n\\nF.3d 907, 932 (11th Cir. 2011) (“No reasonable attorney, after being told by his\\n\\nclient that he had an abusive upbringing, would fail to interview members of his\\n\\nfamily who were readily available and could corroborate or refute the allegations\\n\\nof abuse.”).1 If Downey had effectively investigated—or investigated at all—he\\n\\nwould have been aware of several available and informed witnesses. Ultimately,\\n\\nonly one witness testified at the penalty phase: Lonnie Seal.2 Downey met with\\n\\nSeal for only fifteen minutes before the penalty phase began. That fifteen minutes\\n\\nrepresents the extent of Downey’s preparation for the penalty phase—a wholly\\n\\ninadequate amount of time under Supreme Court precedent. See Williams v.\\n\\nTaylor, 529 U.S. 362, 395 (2000) (finding mitigation investigation deficient where\\n\\n“counsel did not begin to prepare for that phase of the proceeding until a week\\n\\nbefore the trial”).\\n\\n Despite ample evidence of deficient performance, the Alabama Court of\\n\\nCriminal Appeals (CCA) concluded that Downey performed reasonably. The\\n\\nCCA’s decision rested on three explanations: (1) because Downey did not testify,\\n\\ncourts must presume he acted reasonably; (2) based on the record, Downey played\\n\\n\\n1\\n The majority goes to great pains to maintain that Downey did not know about Jenkins’ abusive\\nchildhood. This assertion is directly refuted by the record. Douglas Scofield, Jenkins’ lead\\nattorney, testified that Jenkins described his traumatic upbringing at meetings at which Downey\\nwas present.\\n2\\n It is unclear whether Downey contacted Seal and requested that he testify during the penalty\\nphase of trial. Given that Downey urged Scofield to conduct Seal’s direct examination because\\nScofield was better acquainted with Seal, it is likely that Scofield made the arrangements.\\n 59\\n\\nan active role investigating and preparing for the penalty phase of trial; and (3)\\n\\nDowney’s decision to pursue a residual doubt strategy was reasonable.3 Because\\n\\neach of these findings constituted an unreasonable determination of the facts or an\\n\\nunreasonable application of clearly established federal law, Jenkins is entitled to\\n\\nhabeas relief. See 28 U.S.C. § 2254(d)(1).\\n\\n The CCA’s first rationale—presuming Downey acted reasonably because\\n\\nDowney did not testify at the Rule 32 hearing—was based on both an unreasonable\\n\\ndetermination of facts and an unreasonable application of clearly established law.\\n\\nThe CCA stated that “the record is virtually silent as to what actions were or were\\n\\nnot taken or what was or was not done by Mr. Downey at trial and why” and\\n\\ntherefore assumed that Downey’s actions “could have been reasonable and\\n\\nstrategic under the circumstances.” The majority likewise assumes Downey acted\\n\\nreasonably. This conclusion is based, in large part, on Downey’s failure to testify\\n\\nat the Rule 32 hearing. According to the majority, “[t]he problem for Jenkins is\\n\\nthat, like the state courts, this Court also does not know what Downey did or why.”\\n\\n But the record clearly illustrates Downey’s failure to investigate and overall\\n\\nlack of preparation. First, Downey’s fee declarations—which the majority\\n\\nconcedes were “detailed”—showed no investigation of mitigating circumstances,\\n\\nand no preparation for the penalty phase. Importantly, Downey’s fee declarations\\n\\n\\n3\\n The majority opinion endorsed the CCA’s first and third explanations.\\n 60\\n\\nwere exceptionally detailed, recounting his work in five-minute increments. The\\n\\nsingle reference to the “mitigating” phase of trial in the declarations concerned a\\n\\nmeeting with Scofield that occurred only a week before trial. Other than that\\n\\nvague entry, no notation indicates that Downey did anything to prepare for the\\n\\npenalty phase of trial during the nearly two years he represented Jenkins.\\n\\n Second, and more persuasively, Douglas Scofield, Jenkins’ lead attorney,\\n\\nexplicitly testified that Downey told him that Downey had not investigated\\n\\npotential mitigating evidence. Both the Alabama courts and the majority ignore\\n\\nthat fact and instead focus on Scofield’s testimony that, “[a]part from what he has\\n\\ntold me, I don’t know what he has done.” This statement, to which the majority\\n\\nassigns great weight, is hardly illuminating. Of course Scofield knows only what\\n\\nDowney told him. And what Downey told him was this: Downey did not\\n\\ninvestigate potentially mitigating evidence before the penalty phase.\\n\\n In the face of this direct evidence of deficient performance, the Alabama\\n\\ncourts and the majority pivot to Downey’s failure to testify at the Rule 32 hearing.\\n\\nWithout his testimony, the majority reasons, we must assume that Downey’s\\n\\ndecision not to investigate was strategic and therefore his representation\\n\\nreasonable. I disagree, and I have deep reservations about the practical effects of\\n\\nthe majority’s understanding of the law. The majority suggests (if not holds) that\\n\\nthe absence of testimony from an allegedly deficient attorney per se means that the\\n\\n\\n 61\\n\\nattorney’s actions were reasonable. It is true that “counsel is strongly presumed to\\n\\nhave rendered adequate assistance,” Strickland, 466 U.S. at 690. But a petitioner\\n\\nmay overcome that burden with evidence of inadequate assistance. And that\\n\\nevidence may take various forms. To satisfy the burden, it is not necessary for the\\n\\nallegedly deficient counsel to testify and admit wrongdoing. Nor is it necessary for\\n\\nthe record to reflect counsel’s “thoughts and intentions as he prepared for the\\n\\npenalty phase,” as the majority seems to require. When evaluating whether a\\n\\ncounsel’s performance was deficient, we ask whether his performance “fell below\\n\\nan objective standard of reasonableness.” Strickland, 466 U.S. 687–88. We do not\\n\\ncontemplate what counsel thought or intended. We consider whether his\\n\\nperformance was objectively unreasonable. See id. Because the Alabama courts\\n\\ndid not apply the proper standard, their conclusion was an unreasonable application\\n\\nof clearly established law. See id.\\n\\n The CCA’s second reason for denying Jenkins relief was also based on an\\n\\nunreasonable application of clearly established federal law. The CCA concluded\\n\\nthat Downey had adequately investigated and prepared for the penalty phase of\\n\\ntrial, emphasizing that Downey met with Jenkins on several occasions and spoke\\n\\nwith Jenkins’ grandmother. Based on these findings, the CCA determined that\\n\\nDowney provided adequate assistance of counsel during the penalty phase. This is\\n\\nan unreasonable application of clearly established law that “[a]n attorney has a\\n\\n\\n 62\\n\\nduty to conduct a reasonable investigation, including an investigation of the\\n\\ndefendant’s background, for possible mitigating evidence.” Porter v. Singletary,\\n\\n14 F.3d 554, 557 (11th Cir. 1994).\\n\\n As previously discussed, it is established in the record that Downey knew of\\n\\nJenkins’ horrible upbringing yet still failed to investigate potentially mitigating\\n\\nevidence. “A reasonable investigation . . . should have included, at a minimum,\\n\\ninterviewing other family members who could corroborate the evidence of abuse\\n\\nand speak to the resulting impact on [Jenkins].” Williams, 542 F.3d at 1340. But\\n\\nDowney did not conduct a single interview—of a family member or otherwise—in\\n\\npreparation of the penalty phase.4 Because Downey did not conduct an\\n\\ninvestigation, let alone a reasonable investigation, the CCA’s conclusion that\\n\\nDowney adequately investigated potentially mitigating evidence was an\\n\\nunreasonable application of clearly established law.\\n\\n The CCA’s third rationale for its decision—that Jenkins’ counsel reasonably\\n\\nand strategically decided to focus on residual doubt—was both an unreasonable\\n\\napplication of clearly established law and an unreasonable determination of the\\n\\nfacts. In Williams v. Taylor, the Supreme Court held that otherwise reasonable\\n\\ntrial strategies cannot be justified unless counsel first conducted a constitutionally\\n\\n\\n4\\n Downey’s fee declaration reflects one telephone call he had with Jenkins’ grandmother, Doris\\nWagoner. But that call occurred 18 months before trial and, according to Wagoner, concerned a\\nrequest for money.\\n 63\\n\\nadequate investigation. 529 U.S. at 395–96. And here, Downey conducted no\\n\\nsuch investigation. It cannot be said, therefore, that Downey’s decision—if one\\n\\ncan even call it that—to pursue a residual doubt strategy was strategically\\n\\nreasonable. See id. The CCA’s conclusion to the contrary was an unreasonable\\n\\napplication of clearly established law.\\n\\n The conclusion was likewise an unreasonable determination of the facts. On\\n\\ndirect appeal, the Alabama CCA noted that “[t]he evidence overwhelmingly\\n\\npointed to the guilt of the appellant.” Jenkins v. State, 627 So. 2d 1054 (Ala. Crim.\\n\\nApp. Feb. 27, 1992). Then, on appeal from Jenkins’ Rule 32 petition, the CCA\\n\\nagain acknowledged how “[t]he evidence establishing Jenkins’s guilt was\\n\\noverwhelming.” Jenkins v. State, 972 So. 2d 111, 157 (Ala. Crim. App. Feb. 27,\\n\\n2004); see also id. at 156 (“The amount of evidence incriminating Jenkins . . . was\\n\\noverwhelming.”). Despite the “overwhelming” evidence of Jenkins’ guilt, the\\n\\nCCA concluded that Jenkins’ counsel acted reasonably in presenting a residual\\n\\ndoubt defense during the penalty phase of trial. But no reasonable attorney would\\n\\nemploy a residual doubt strategy in the face of such overwhelming evidence. See\\n\\nJohnson, 643 F.3d at 932–33. The Alabama courts’ contrary conclusion, therefore,\\n\\nwas an unreasonable determination of the facts.\\n\\n\\n\\n\\n 64\\n\\n B. Prejudice\\n\\n Jenkins’ Rule 32 petition also pleaded sufficient facts to show that his\\n\\ncounsel’s deficient performance prejudiced his defense, satisfying the second\\n\\nStrickland prong. To establish prejudice, a petitioner must show a “reasonable\\n\\nprobability that, but for counsel’s unprofessional errors, the result of the\\n\\nproceeding would have been different.” Strickland, 466 U.S. at 694. Here, the\\n\\njury voted in favor of the death penalty by the slimmest margin permitted under\\n\\nAlabama law: ten votes in favor and two votes against. Jenkins need only\\n\\nestablish, then, “a reasonable probability that at least one juror would have struck\\n\\na different balance.” Wiggins, 539 U.S. at 537 (emphasis added). In measuring\\n\\nthat probability, we “evaluate the totality of the available mitigation evidence—\\n\\nboth that adduced at trial, and the evidence adduced in the habeas proceeding in\\n\\nreweighing it against the evidence in aggravation.” Williams, 529 U.S. at 397.\\n\\n The testimony elicited at the Rule 32 hearing was horrendous. Four family\\n\\nmembers offered detailed, consistent testimony about Jenkins’ traumatic\\n\\nchildhood. The majority describes that testimony, highlighting some of the most\\n\\nappalling facts elicited at the Rule 32 hearing. For example, Jenkins’ stepfather\\n\\nwould beat Jenkins on a semi-daily basis, would lock him in a room for hours\\n\\nwithout food, and even made Jenkins eat his own feces. In addition to the\\n\\nmitigating facts detailed by the majority, the following came to light during the\\n\\n\\n 65\\n\\nRule 32 hearing. Jenkins’ mother had an affair while her husband, Jenkins’\\n\\nstepfather, was in prison. When Jenkins’ stepfather was released, he took out his\\n\\nfrustrations about the affair on Jenkins. Because Jenkins’ biological father was\\n\\nJenkins’ stepfather referred to Jenkins in demeaning names like “Mexican trash,”\\n\\nand “Puerto Rican puke.” The stepfather also called him “piece of shit” and\\n\\n“bastard.”\\n\\n The Rule 32 testimony also detailed the alcohol substance abuse, drug\\n\\naddiction, and incessant domestic violence that surrounded Jenkins’ childhood.\\n\\nAnd, most shockingly, the testimony described how Jenkins’ grandfather sexually\\n\\nabused Jenkins when Jenkins was just four years old. As a result of that sexual\\n\\nabuse, Jenkins, suffered bladder and bowel movement issues. Jenkins’\\n\\nincontinence became a trigger for his stepfather: when Jenkins had an accident, his\\n\\nstepfather would beat him or subject him to severe degradation. For example,\\n\\nJenkins’ stepfather would hang Jenkins’ soiled sheets in the front yard for all the\\n\\nneighbors to see, make Jenkins wear his soiled clothes on his head, and, most\\n\\ndisturbingly, make Jenkins eat his own feces. The only way to escape the abuse\\n\\nwas by running away. So, by age twelve, Jenkins lived on the streets.5\\n\\n\\n\\n5\\n The CCA found parts of the family members’ testimony not credible, though it did not\\nexplicitly discredit testimony describing some of the worst abuse. Given the grotesque nature of\\nthe family members’ testimony, and the fact that their testimony was consistent, there is\\nreasonable probability that at least one more juror would have been compelled to vote against a\\ndeath sentence.\\n 66\\n\\n Jenkins also introduced medical, school, and juvenile court records from his\\n\\nchildhood, which shed further light on his difficult childhood and intellectual\\n\\ndeficiencies. Juvenile records show Jenkins as “an emotionally disturbed child”\\n\\nwith “psychological problems” and expressed “grave concerns for the emotional\\n\\nwill being of this adolescent”. Moreover, two psychologists testified about\\n\\nJenkins’ history of psychological trauma stemming from his childhood abuse—\\n\\nweighty evidence that is not even considered in the majority’s discussion. Finally,\\n\\ntwo prison guards testified that Jenkins was the ideal prisoner during pretrial\\n\\ndetention. One of the guards even described Jenkins as the best inmate he had\\n\\nevery supervised.\\n\\n Given this compelling mitigating evidence, it is not difficult to conclude that\\n\\nJenkins established “a reasonable probability that at least one juror would have\\n\\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537; see\\n\\nalso Daniel, 822 F.3d at 1276 (listing several factors that support a prejudice\\n\\nfinding, including (1) evidence of impaired intellectual functioning; (2) evidence of\\n\\nchildhood sexual abuse; and (3) the fact that the jury voted only 10–2 in\\n\\nrecommending a death sentence).\\n\\n I. Evidentiary Hearing\\n\\n Jenkins also argues that he is entitled to habeas relief under Atkins v.\\n\\nVirginia, 563 U.S. 304 (2002). He contends that he satisfies the three requirements\\n\\n\\n 67\\n\\nnecessary to establish intellectual disability—(1) subaverage intellectual\\n\\nfunctioning; (2) deficits in adaptive behavior; and (3) onset before the age of 18,\\n\\nsee Ex parte Perkins, 851 So. 2d 453, 456 (Ala. 2002)—thus, exempting him from\\n\\nimplementation of the death penalty. Alternatively, Jenkins argues that we should\\n\\nremand for an evidentiary hearing on the Atkins issue.\\n\\n The majority dismisses each argument. First, it concludes that the CCA did\\n\\nnot act unreasonably in concluding that Jenkins failed to satisfy any of the Perkins\\n\\nprongs. Second, it determines that Jenkins is not entitled to an evidentiary hearing\\n\\nbecause he “did not diligently attempt to develop the factual basis for this claim in\\n\\nthe state court.” I disagree.\\n\\n First, it is not possible to undergo a substantive Atkins analysis based on the\\n\\nrecord before us. For one, the CCA’s evaluation of Jenkins’ Atkins claim was a\\n\\nmere three sentences. And, more significantly, Jenkins never had an opportunity to\\n\\npresent evidence on his Atkins claim. To deny Jenkins’ claim, therefore, the\\n\\nmajority relies on evidence from unrelated parts of the record. Such an approach is\\n\\nimproper because, “[p]rior to Atkins, [the petitioner] could not have been expected\\n\\nto necessarily present evidence sufficient to support an Atkins claim because such\\n\\nevidence constituted a double edged sword.” Burgess v. Comm’r, Ala. Dep’t of\\n\\nCorr., 723 F.3d 1308, 1320 (11th Cir. 2013) (internal citation omitted). The\\n\\nevidence was a double-edged sword, we explained, because prior to Atkins,\\n\\n\\n 68\\n\\nevidence of intellectual disability could suggest that a defendant would be\\n\\ndangerous in the future; therefore, “a defendant could reasonably decide not to\\n\\nhighlight his mental retardation.” Id. at 1318. For that reason, courts should not\\n\\nrely on pre-Atkins evidence—evidence that was presented in a wholly distinct\\n\\ncontext—to support a finding that a petitioner is not intellectually disabled. So,\\n\\nhere, instead of attempting to make a medical diagnosis based on an insufficient\\n\\nrecord, we should remand for an evidentiary hearing. The majority refuses to do\\n\\nso, reasoning that “Jenkins did not diligently attempt to present his Atkins claim in\\n\\nthe state courts.” I disagree.\\n\\n “[I]f the petitioner was diligent in developing the record in the state habeas\\n\\nproceedings, ‘a federal court may grant an evidentiary hearing without further\\n\\nregard for the provisions of § 2254(3)(2).” Burgess, 723 F.3d at 1320 (citation\\n\\nomitted). “‘Diligence [for purposes of § 2254(d)] depends on whether the prisoner\\n\\nmade a reasonable attempt, in light of the information available at the time, to\\n\\ninvestigate and pursue claims in state court,’ and ‘will require in the usual case that\\n\\nprisoner, at a minimum, seek an evidentiary hearing in state court.’” Id. (quoting\\n\\nWilliams, 529 U.S. at 435 (emphasis added)).\\n\\n The majority concludes that Jenkins failed to diligently pursue his Atkins\\n\\nclaim in state court. But again, based on binding precedent and the record, I must\\n\\npart company with the majority. In conducting its analysis, the majority does not\\n\\n\\n 69\\n\\nadequately consider whether Jenkins “made a reasonable attempt, in light of the\\n\\ncircumstances available at the time, to investigate and pursue claims in state\\n\\ncourt.” Id. (emphasis added) (quoting Williams, 529 U.S. at 435). Specifically, the\\n\\nmajority did not consider the unusual procedural posture of Jenkins’ claim.\\n\\n The Supreme Court issued Atkins v. Virginia while Jenkins’ petition for\\n\\npost-conviction collateral relief was pending before the CCA. The CCA, therefore,\\n\\nrequested supplemental briefing on the effect, if any, Atkins had on Jenkins’\\n\\npetition. At that point, neither the Alabama Supreme Court nor the Alabama\\n\\nlegislature had interpreted the groundbreaking case, making Jenkins’ counsel’s\\n\\nnext step unclear. Jenkins’ counsel nonetheless made a reasonable attempt to\\n\\npursue Jenkins’ Atkins claim—he requested that the CCA “vacate his death\\n\\nsentence and remand the case back to the lower court for further proceedings.”\\n\\nThe request appears even more reasonable when considered alongside procedurally\\n\\nanalogous cases decided around the same time.\\n\\n At the time Jenkins’ counsel was preparing the supplemental briefing, there\\n\\nhad been three other post-conviction petitions for relief that shared Jenkins’\\n\\nunusual procedural posture. See Clemons v. State, 55 So. 3d 314 (Ala. Crim. App.\\n\\nAug. 29, 2003); Wood v. State, 891 So. 2d 398 (Ala. Crim. App. Apr. 25, 2003);\\n\\nTarver v. State, 940 So. 2d 312 (Ala. Crim. App. Feb. 27, 2004). These three cases\\n\\nare similar to Jenkins’ petition in that (1) the petitioners were sentenced to death,\\n\\n\\n 70\\n\\n(2) the petitioners filed Rule 32 petitions, (3) their Rule 32 petitions were\\n\\ndismissed at the trial court level, (4) while their appeals before the CCA were\\n\\npending, Atkins was decided, and (5) given the Atkins decision, each petitioner\\n\\nsubmitted supplemental briefing. In their supplemental briefs, Clemons, Wood,\\n\\nand Tarver each argued that he was intellectually disabled under Atkins and thus it\\n\\nwould be unconstitutional to execute him. None of the petitioners asked for an\\n\\nevidentiary hearing on their Atkins claim or requested a remand for further\\n\\nproceedings. Despite the lack of request for a hearing, the CCA remanded for an\\n\\nevidentiary hearing in each case.\\n\\n With these cases in mind, it would have been reasonable for Jenkins to think\\n\\nthat broadly stating his Atkins claim would constitute “diligent” pursuit of his\\n\\nclaim. Nonetheless, Jenkins did more than these other petitioners: he specifically\\n\\nrequested that the CCA “vacate his death sentence and remand the case back to the\\n\\nlower court for further proceedings.” Given “the information available at the\\n\\ntime,” Jenkins’ made a “reasonable attempt” to pursue his claims.\\n\\n I also cannot come to terms with the majority’s ultimate conclusion. The\\n\\nmajority determines that Jenkins did not act diligently in pursuing his Atkins claim\\n\\nbecause “[t]he record clearly shows that Jenkins did not [seek an evidentiary\\n\\nhearing].” But Jenkins did ask the CCA to “remand his case for further\\n\\nproceedings.” And, in Alabama, post-conviction relief petitioners with factually\\n\\n\\n 71\\n\\nsufficient claims are entitled to an evidentiary hearing. Ala. R. Crim. P. 32.9. So,\\n\\nif the CCA had granted Jenkins’ request to remand “for further proceedings,”\\n\\nJenkins would have received an evidentiary hearing. Jenkins’ request that the\\n\\nCCA “remand for further proceedings,” then, was effectively equivalent to a\\n\\nrequest for an evidentiary hearing. The majority’s superficial conclusion to the\\n\\ncontrary misses the point.\\n\\n II. Conclusion\\n\\n Jenkins’ Rule 32 evidentiary hearing spanned three days. Four family\\n\\nmembers testified about his horrific upbringing. Two psychiatrists testified about\\n\\nhow that upbringing affected Jenkins as an adult. Two prison guards testified\\n\\nabout Jenkins’ exemplary behavior while in pretrial detention. And one friend\\n\\ntestified about Jenkins’ character. Evidence introduced also documented Jenkins’\\n\\nacademic failings and his juvenile record. In short, Jenkins presented an\\n\\nabundance of compelling mitigating evidence. But the jury never heard any of that\\n\\nevidence. Downey failed to introduce it at the penalty phase. In fact, Downey\\n\\nfailed to investigate any mitigating evidence whatsoever. Downey’s failures\\n\\ncannot be attributed to strategic decisions. Nor can they be called reasonable. His\\n\\nperformance can only be described as deficient. And that deficient performance\\n\\nwas surely prejudicial—there is more than a reasonable probability that the\\n\\nhorrifying and detailed Rule 32 testimony would have persuaded one juror to vote\\n\\n\\n 72\\n\\nagainst sentencing Jenkins to death. Jenkins is therefore entitled to habeas relief\\n\\non his ineffective assistance of counsel claim.\\n\\n Jenkins is also entitled to a hearing on his Atkins claim. Though Jenkins\\n\\npresented ample mitigating evidence at the Rule 32 hearing, he never had an\\n\\nopportunity to present evidence concerning (1) his intellectual functioning, (2) his\\n\\nadaptive deficits, or (3) whether his potential intellectual disability manifested\\n\\nbefore Jenkins reached 18—the three requirements necessary to establish\\n\\nintellectual disability and thus exempt an individual from a death sentence. The\\n\\nCCA nonetheless decided to evaluate Jenkins’ Atkins claim, apparently believing\\n\\nthat the record, which included no direct evidence concerning whether Jenkins was\\n\\nintellectually disabled, was sufficient to resolve the complex medical issue. In just\\n\\nthree sentences, the CCA arrived at a diagnosis: not intellectually disabled.\\n\\n Despite the insufficient factual record, the majority declines to remand with\\n\\ninstructions to conduct an evidentiary hearing—what I view as the proper\\n\\nrecourse—because Jenkins did not use the phrase “evidentiary hearing” when\\n\\nrequesting further proceedings on his Atkins claim in the Alabama courts. Such a\\n\\nrigid understanding of the standard should not bar a death row petitioner from\\n\\nbringing a constitutional, and potentially lifesaving, claim. I therefore dissent.\\n\\n\\n\\n\\n 73\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434945/\", \"author_raw\": \"WILSON, Circuit Judge, dissenting\"}]}"},"judge1":{"kind":"string","value":"TJOFLAT"},"judge2":{"kind":"string","value":"WILSON"},"judge3":{"kind":"string","value":"BRANCH"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"TJOFLAT\": \", Circuit\", \"WILSON\": \", Circuit\", \"BRANCH\": \", Circuit\"}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":1,"string":"1"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4657692/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Mark Jenkins v. Commissioner, Alabama Department of Corrections\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-30\nOpinion type: majority\nHas dissenting opinion: Yes\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\n[PUBLISH]\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n\n\n No. 17-12524\n\n\n\n D.C. Docket No. 4:08-cv-00869-VEH\n\nMARK ALLEN JENKINS,\n\n Petitioner - Appellant,\n\n versus\n\n\n\nCOMMISSIONER, ALABAMA DEPARTMENT OF CORRECTIONS,\n\n Respondent - Appellee.\n\n\n\n Appeal from the United States District Court\n for the Northern District of Alabama\n\n\n (August 30, 2019)\n\nBefore TJOFLAT, WILSON, and BRANCH, Circuit Judges.\n\nBRANCH, Circuit Judge:\n\n Mark Allen Jenkins, an Alabama prisoner sentenced to death for the 1989\n\nmurder of Tammy Ruth Hogeland, appeals the district court’s denial of his petition\n\nfor a writ of habeas corpus, 28 U.S.C. § 2254. Before us are Jenkins’s arguments\n\nthat he received ineffective assistance of counsel during the penalty phase of his\n\ntrial and that he is intellectually disabled and therefore ineligible for the death\n\npenalty. After careful consideration, and with the benefit of oral argument, we\n\naffirm the denial of Jenkins’s habeas petition on both grounds.\n\n I. BACKGROUND\n\n A. The Crime and Arrest\n\n We summarize the following background narrative from opinions of the\n\nAlabama Court of Criminal Appeals and our own review of the record. See Jenkins\n\nv. State, 627 So. 2d 1034, 1037–40 (Ala. Ct. Crim. App. 1992); Jenkins v. State,\n\n972 So. 2d 111, 119–20 (Ala. Ct. Crim. App. 2004). The events leading up to the\n\nmurder began the evening of April 17, 1989. Jenkins was at the home of his\n\nacquaintance Christine Nicholas. He had met her several months earlier at the\n\nOmelet Shoppe restaurant where she worked. Jenkins was very intoxicated and\n\ntried to seduce her. When she resisted, Jenkins got “real mad” and asked her\n\nseveral times what she would do if someone came up from behind her and grabbed\n\nher.\n\n At around 1 a.m., Jenkins and Nicholas drove to the Riverchase Omelet\n\nShoppe. Jenkins went inside and talked with one of the waitresses, Frieda Vines.\n\nThe manager, Douglas Thrash, recognized Jenkins as a regular customer who knew\n\n\n 2\n\nall of the waitresses. Thrash overheard part of Jenkins’s conversation with Vines\n\nand heard mention of the Omelet Shoppe location near the Birmingham Airport.\n\nJenkins and Nicholas then returned to Nicholas’s home. Around 2 a.m., Nicholas’s\n\nmother asked Jenkins to leave. He did so, falling down some steps and ramming\n\nhis car into another vehicle in the process.\n\n Meanwhile, 23-year-old Tammy Hogeland was making her way to work at\n\nthe Omelet Shoppe. Her sister Wendy, along with Tammy’s young son, picked\n\nTammy up from her college classes and drove her to the Riverchase Omelet\n\nShoppe a little after 9:30 p.m. Sometime after 10:00, Tammy was sent to work at\n\nthe airport Omelet Shoppe on Tenth Avenue, since that location was unexpectedly\n\nshorthanded. Tammy was wearing her watch, a necklace with the words “special\n\nsister,” her class ring, and her diamond engagement ring. Tammy was working as a\n\ncook that night and was wearing a blue apron, black pants, a white shirt, and white\n\nshoes. She and Sarah Harris were the only employees working at the airport\n\nOmelet Shoppe that night.\n\n At about 2 a.m. on April 18, Harris saw Jenkins—whom she did not know—\n\ndrive up in a red sports car. His arrival was memorable to Harris because the car\n\nnearly jumped the curb and crashed through the restaurant’s glass wall. Jenkins got\n\nout of the car and came into the restaurant, appearing to be intoxicated. He walked\n\nover to Hogeland and began talking to her. Harris later saw Jenkins and Hogeland\n\n\n 3\n\ndrive away together in the red sports car. That was the last time anyone who knew\n\nHogeland ever saw her. Hogeland left behind her cigarettes, lighter, purse, and\n\npaycheck. She left without telling anyone, which she had never done before.\n\nAlthough Harris saw them drive off, she could not later say whether Hogeland left\n\nwith Jenkins willingly or was instead abducted.\n\n At around 5 a.m. that morning, Bobby and Geraldine Coe had stopped to\n\nbuy gas at a Chevron station on I-59 northeast of Birmingham when they saw\n\nJenkins drive up in the red sports car. The Coes noticed a female who appeared to\n\nbe “passed out” in the front passenger seat, but they could not say whether she was\n\nalive or dead. While Bobby Coe was pumping gas, Jenkins asked him for some\n\ncigarettes and said, “Looks like it’s been a long night and it looks like it’s going to\n\nbe a long day.” Jenkins then said “God bless you” before asking directions to I-\n\n459. Coe gave him directions and got back in his car. As Coe pulled out onto the\n\ninterstate, he saw Jenkins follow him in the red sports car. Coe then saw the car\n\nflash its lights, slow down, and pull to the side of the road between mile markers\n\n151 and 152.\n\n At 8 a.m., Wendy Hogeland learned that Tammy was not at the restaurant,\n\nand their mother called the police. Meanwhile, Jenkins went to the home of Steve\n\nMusser, who noticed that he was wearing the same clothes as the day before.\n\nJenkins told Musser that his car had been stolen the night before and asked him if\n\n\n 4\n\nhe would say that he had been with him all night. Musser refused. Christine\n\nNicholas also saw Jenkins in a grocery store that morning. He was looking at a\n\nnewspaper, making a phone call, and attempting to sell his Buick. Nicholas loaned\n\nhim $4 for gas. At around 10 o’clock that morning, Jenkins sold his car to Michael\n\nBrooks, a mechanic at a local Chevron station. Jenkins had explained that he\n\nneeded the money so he could visit his sick mother in California. Another\n\nmechanic at the service station drove Jenkins to the Greyhound bus station later\n\nthat morning.\n\n Jenkins awoke the next day on the Greyhound bus in Houston and was\n\nejected from the bus because his fare was used up. He then hitchhiked from\n\nHouston to Phoenix, then to San Diego and Los Angeles. Jenkins was first\n\nidentified as a suspect in Hogeland’s disappearance by Officer Mike Weems of the\n\nHoover Police Department on April 19. Weems ate dinner at the Riverchase\n\nOmelet Shoppe just about every day and had talked with Hogeland just as often.\n\nHe also knew Jenkins from the restaurant. Weems learned of Hogeland’s\n\ndisappearance from the other waitresses. Remembering how Jenkins often talked to\n\nHogeland and passed her notes,1 he gave Jenkins’s name to the missing persons\n\ninvestigator. The investigator also learned that a red Mazda RX-7 sports car, which\n\n\n\n1\n The jury was not permitted to hear, and the record does not reveal, what exactly Jenkins had\nsaid to Hogeland that made Officer Weems suspect him in her disappearance.\n 5\n\nhad been reported stolen from the service station where Jenkins worked, had been\n\nrecovered on I-459.\n\n The afternoon of April 21, a truck driver who had happened to stop on I-59\n\nnear mile marker 151 discovered Hogeland’s body off the side of the road. She was\n\nnaked, wearing only a watch, and was so badly decomposed that the body had to\n\nbe identified by dental records. From the fractured hyoid bone, it was determined\n\nthat she was manually strangled to death. Also found at the scene were Hogeland’s\n\napron, shoes, bra, panties, pants, and hair net, as well as some beer cans, a Mazda\n\nRX-7 owner’s manual, and other items later determined to have come from the red\n\nMazda. Her necklace and rings were never found.\n\n Alabama authorities issued a warrant for Jenkins’s arrest and also obtained a\n\nfederal fugitive warrant. An FBI agent arrested Jenkins on May 1 in front of his\n\nuncle’s house in Wilmington, California. Jenkins’s uncle later gave the police three\n\nbags of Jenkins’s personal effects. Fibers recovered from Hogeland’s body and\n\nclothing matched those of the Mazda, and hairs matching Hogeland’s were found\n\nin the Mazda. Similarly, fibers from Jenkins’s clothing placed him in the Mazda,\n\nand fibers of his jeans were found on Hogeland’s apron. A bootprint near\n\nHogeland’s body matched a combat boot from among the belongings Jenkins’s\n\nuncle gave to the police, and Jenkins’s roommate said the boots looked just like\n\nthose that Jenkins wore.\n\n\n 6\n\n A business card like those that had been in the Mazda’s glove box,\n\nbelonging to an owner of the Mazda, was recovered from Jenkins’s wallet after his\n\narrest in California. Later, Jenkins’s cellmate in the Alabama jail reported that\n\nJenkins told him “he had done the crime” and was worried that the couple from the\n\ngas station would identify him or that the police would find his fingerprint on a\n\nbeer can at the scene. A St. Clair County grand jury indicted Jenkins for capital\n\nmurder.\n\n B. The Trial and Sentencing\n\n Jenkins was represented at trial in 1991 by attorneys Douglas Scofield and\n\nStan Downey. Scofield, the lead attorney, had been practicing criminal defense in\n\nBirmingham since 1984 but had never before tried a capital murder case. Jenkins’s\n\nlandlord and his grandmother had arranged the referral and representation before\n\nthe court appointed him. Scofield agreed to the appointment on the condition that a\n\nSt. Clair County lawyer also be appointed. That lawyer was Downey, who had\n\nmore experience than Scofield but whose capital experience was limited to one\n\ncase that had not gone to trial. The two attorneys had agreed that Downey would\n\nbe responsible for jury selection, but Scofield eventually took over during the voir\n\ndire. Per their agreement, Scofield was responsible for the guilt-phase trial,\n\nalthough Downey participated by interviewing Jenkins several times. Scofield also\n\nmet with Jenkins around a dozen times. Jenkins had told him about his background\n\n\n 7\n\nand childhood, including the fact that he had had a difficult childhood and had\n\nlived on the streets since age 11.\n\n Throughout the weeklong trial, Scofield pursued an outright acquittal. In\n\nparticular, he challenged Harris’s identification of Jenkins as the man who drove\n\naway with Hogeland, and he challenged the medical examiner’s opinions about the\n\ntime and cause of Hogeland’s death. In his closing argument, Scofield took a wide-\n\nranging approach to undermining the State’s wholly circumstantial case against\n\nJenkins. He attacked the State’s timeline of the crime, questioning whether Jenkins\n\nin his intoxicated state would have had the time and the ability to leave Nicholas’s\n\nhouse, steal the Mazda, and abduct and kill Hogeland. He reminded the jury of the\n\nweakness of Harris’s identification of Jenkins and the Coes’ identification of the\n\nwoman in the car. He pointed out that no physical evidence definitively linked\n\nJenkins to the Mazda, to Hogeland’s body, or to her jewelry, which was never\n\nrecovered. And he argued that the evidence provided no plausible motive for\n\nJenkins to have kidnapped, robbed, or killed Hogeland. In sum, he urged the jury\n\nto find reasonable doubt and acquit Jenkins. At the very least, he argued, they\n\nshould find that the State failed to prove Jenkins’s intent to rob and kidnap\n\nHogeland, which were necessary to convict Jenkins of capital murder. He further\n\nargued that, if the jury thought Jenkins killed Hogeland as a crime of passion, they\n\ncould convict him only of manslaughter. The jury deliberated for three hours and\n\n\n 8\n\nfifteen minutes before convicting Jenkins of capital murder, Ala. Code § 13A-5-\n\n40(a)(1) & (2), i.e., murder committed in the course of kidnapping and robbery.\n\n The penalty phase took place that same afternoon. A week before trial,\n\nScofield and Downey had discussed the penalty phase, and Scofield understood\n\nthat Downey would be responsible for handling the penalty phase. Nonetheless, the\n\nwitness who was present to testify on Jenkins’s behalf during the penalty phase\n\nwas better acquainted with Scofield, so Downey suggested that afternoon that\n\nScofield conduct the direct examination. Scofield urged Downey to interview the\n\nwitness quickly and prepare to examine him, and Downey did so.\n\n Downey handled the penalty presentation to the jury. The State did not\n\npresent any additional aggravation evidence, resting on the evidence presented\n\nduring the guilt phase. One witness, Lonnie Seal, testified on Jenkins’s behalf in\n\norder, in Downey’s words, to “reveal another side of Mark Jenkins to you that you\n\ndon’t know anything about.” Seal had met Jenkins in Fontana, California, in 1988\n\nwhen they both worked at the same garage. Seal and his wife became friends with\n\nJenkins, and he visited in their home. When the Seals decided to move to Alabama\n\na few months later, Jenkins agreed to help with their move and to live with the\n\nSeals in Alabama for a time. It took Lonnie Seal several weeks to find a job, but\n\nJenkins found a job in two days and contributed toward the family’s rent and\n\n\n\n\n 9\n\ngroceries. Jenkins continued to visit the Seals after he moved into his own home,\n\nand he would always bring a small gift for and play with the Seals’ infant son.\n\n Downey’s closing argument to the jury requested a sentence of life without\n\nparole. Downey reminded the jurors that they had agreed during voir dire that\n\ndeath was not always the appropriate punishment for homicide. Quoting the Bible,2\n\nhe reminded the jury that the Old Testament permitted capital punishment only\n\nwhen supported by the testimony of two or three eyewitnesses, whereas in this case\n\nthere had been only circumstantial evidence. He asserted from Seal’s testimony\n\nthat “Mark had a side of him that was generous, that was giving, and that was\n\nkind,” and argued that the deterrent and retributive purposes of capital punishment\n\nwould not be served in this case.\n\n The jury deliberated for 50 minutes before returning its nonbinding\n\nrecommendation of death by a vote of 10 to 2. See Ala. Code § 13A-5-46(e) (1981)\n\n(amended 2017 to make jury’s verdict binding). After a separate hearing, the court\n\nfound the two statutory aggravating factors for which the State had argued: that the\n\nmurder was committed during the commission of a robbery, and that it was\n\ncommitted during the commission of a kidnapping. See Ala. Code § 13A-5-49(4).\n\n\n\n\n2\n “If anyone kills a person, the murderer shall be put to death on the evidence of witnesses. But\nno person shall be put to death on the testimony of one witness.” Numbers 35:30 (ESV). Downey\nalso paraphrased Deuteronomy 17:6 as “On the evidence of two witnesses or three witnesses, he\nis [sic] shall be put to death. He shall not be put to death on the evidence of one witness.”\n 10\n\nThe court also found two statutory mitigating factors: that Jenkins had no\n\nsignificant history of prior criminal activity, and his age at the time of the crime\n\n(21). See id. § 13A-5-51(1), (7). The court considered but rejected the statutory\n\nmitigation that, due to his consumption of alcohol, Jenkins’s capacity to appreciate\n\nthe criminality of his conduct or to conform his conduct to the requirements of law\n\nwas substantially impaired. See id. § 13A-5-51(6). The court found that the\n\naggravating circumstances outweighed the mitigating circumstances and, in\n\naccordance with the jury’s recommendation, imposed a sentence of death.\n\n C. The State and Federal Post-Conviction Proceedings\n\n On direct appeal, the Alabama Court of Criminal Appeals affirmed Jenkins’s\n\nconviction and death sentence. Jenkins v. State, 627 So. 2d 1034 (Ala. Ct. Crim.\n\nApp. 1992), aff’d, 627 So. 2d 1054 (Ala. 1993), cert. denied, 511 U.S. 1012\n\n(1994). In 1995, Jenkins filed his first petition for state postconviction relief, Ala.\n\nR. Crim. P. 32,3 alleging, among many other things, that he received ineffective\n\nassistance of counsel during his penalty phase. The Rule 32 trial court conducted a\n\nthree-day hearing at which evidence was presented about the performance of\n\nJenkins’s attorneys.\n\n\n\n\n3\n “Subject to the limitations of Rule 32.2, any defendant who has been convicted of a criminal\noffense may institute a proceeding in the court of original conviction to secure appropriate relief\non the ground that . . . [t]he constitution of the United States or of the State of Alabama requires\na new trial, a new sentence proceeding, or other relief. . . .” Ala. R. Crim. P. 32.1.\n 11\n\n Scofield testified about the representation that he and Downey provided. In\n\nrelevant part, Scofield asserted that he and Downey had “decided that [developing\n\nmitigation for the penalty phase] would be something that Stan [Downey] would\n\nhandle.” Downey later told Scofield that he hadn’t done any such investigation, but\n\nScofield admitted that, “[a]part from what he has told me, I don’t know what he\n\nhas done.” Scofield testified that Downey “didn’t appear prepared” for the penalty\n\nphase, based on Downey’s last-minute suggestion that Scofield examine Seal since\n\nhe had talked with him before.\n\n In preparation for the penalty phase, Scofield himself did not interview any\n\nmember of Jenkins’s family, but he did not know whether Downey interviewed\n\nany family members. Scofield did interview Jenkins’s landlord and some other\n\nacquaintances but did not do so specifically for the purpose of developing\n\nmitigation. He had met with Jenkins around a dozen times, and Jenkins had told\n\nhim his life story including his experience of abuse as a child. In hindsight,\n\nScofield said he had a “lack of appreciation of mitigating circumstances.” He\n\nrepeatedly denied in his Rule 32 testimony that his omissions during the penalty\n\nphase were strategic: “I didn’t understand what I was doing. . . . I just didn’t\n\nunderstand what I needed to do.”\n\n For reasons that remain unclear, Downey—who is since deceased—did not\n\ntestify or submit an affidavit in the Rule 32 proceeding. Scofield answered,\n\n\n 12\n\n“Obviously not,” when asked whether he had personal knowledge of everything\n\nDowney may or may not have done to prepare for the penalty phase. The record of\n\nthe extent of Downey’s pretrial preparation is thus limited to Downey’s detailed\n\nfee declaration, which he submitted to the State of Alabama in documentation of\n\nthe compensation due him as appointed counsel. See Ala. Code § 15-12-21. In it,\n\nDowney reported that he spent 71 hours4 on out-of-court preparation for trial. The\n\nreport details 35 hours of interviews with Jenkins in 18 separate visits to the jail\n\nover the course of 17 months, as well as one substantive conversation with\n\nJenkins’s grandmother.5\n\n The Rule 32 trial court also heard testimony from four family members and\n\none friend of Jenkins’s. The family members presented for the first time a stark\n\ndepiction of Jenkins as an unloved and abused child in California. Jenkins’s older\n\nbrother Michael Jenkins testified that all of his siblings have different fathers and\n\nthat his parents often used drugs and fought. Michael testified that Jenkins was\n\nbeaten frequently as a child by his stepfather, if not every day then every other day\n\n\n\n4\n In Jenkins’s state postconviction appeal, the Alabama Court of Criminal Appeals erroneously\nstated that Downey’s fee declaration reports 171 hours of pretrial preparation. See Jenkins, 972\nSo. 2d at 137. We discuss the effect of this discrepancy infra note 7.\n5\n The state appellate court also noted that Downey had moved for a continuance in October 1989\nso that he could conduct “[f]urther discovery and investigation (including a possible trip to\nCalifornia).” Id. We note that Scofield had orally suggested to the trial court in September 1989\nthat more time might be needed with respect to “evidence about the defendant” and witnesses in\nCalifornia, as well as possible psychological evaluation.\n 13\n\nor every three days. Jenkins was beaten all over his body with a variety of\n\nimplements. Michael was beaten too. Jenkins’s stepfather would lock Jenkins in\n\nhis bedroom for a couple of hours at a time, sometimes without dinner. Some of\n\nthe beatings were prompted by Jenkins’s bedwetting or soiling. Michael testified\n\nthat Jenkins’s stepfather sometimes made Jenkins wear his soiled underwear on his\n\nhead and even forced him to eat his own feces a few times. Michael testified that\n\nJenkins’s bladder and bowel control problems began after a camping trip he took\n\nwith his stepfather and stepgrandfather.\n\n Michael testified that, as children, he and Jenkins were made to work on car\n\nengines late into the night and to shovel horse manure daily. The family moved\n\nfrequently and the children attended many different schools as a result. Michael\n\ntestified that he was Jenkins’s only friend as a child and that Jenkins did poorly in\n\nschool. Jenkins began running away from home as a young teenager and lived with\n\nvarious other family members or on the streets. Michael said that Jenkins was a\n\ngood brother to him and had asked him to testify at his trial, but he never heard\n\nfrom his attorneys or found out when the trial was. Michael admitted to having\n\ncognitive and memory problems due to a head injury years earlier.\n\n Jenkins’s cousin Tammy Pitts testified that, as an infant, Jenkins’s mother\n\nbeat and neglected him, leaving him dirty and in soiled diapers. His stepfather also\n\nbeat Jenkins daily until he left home at age 13, sometimes bruising him so badly\n\n\n 14\n\nthat he was laid up in bed for days or weeks. Jenkins had problems with\n\nbedwetting and soiling, and his stepfather made him wear soiled clothes to school.\n\nJenkins would be locked in his filthy bedroom around the clock except to attend\n\nschool and do chores. He was usually not allowed to eat dinner with the rest of the\n\nfamily and subsisted on scraps and dog food. The entire family belittled Jenkins\n\nbecause his biological father was Hispanic and his complexion was therefore\n\ndarker. Jenkins’s parents drank a lot and used drugs every day. Pitts knew these\n\nthings because she had daily contact with the family and lived with them off and\n\non, and Jenkins lived with her for a time after he ran away from home. No one had\n\ncontacted her about testifying at Jenkins’s trial.\n\n Jenkins’s second cousin Betty DeLavega testified that her relationship with\n\nJenkins was like that of an aunt. She lived with the Jenkins family for five months\n\nwhen Jenkins was 10 or 11. She saw Jenkins’s stepfather slap Jenkins “[q]uite a\n\nfew times,” and once, he made Jenkins eat his own feces. DeLavega testified that\n\nJenkins’s parents used drugs nearly every day and withheld affection from Jenkins\n\nand Michael. No one contacted her about Jenkins’s trial; Jenkins’s grandmother\n\nhad asked her to testify in the Rule 32 hearing about any abuse she had witnessed.\n\n Jenkins’s grandmother Doris Wagoner testified that while Jenkins’s mother\n\nwas pregnant with him, her husband was in prison and she drank and took drugs.\n\nWhen Jenkins was born, his mother initially put him up for adoption, then changed\n\n\n 15\n\nher mind weeks later. Jenkins was a lethargic infant and his mother cared more\n\nabout partying than about her children. Wagoner testified that she heard of\n\nJenkins’s mistreatment as a child but didn’t see it. She could tell from the way that\n\nJenkins feared his stepfather that he was being abused, and she knew he slept in a\n\nfilthy bed. After Jenkins was arrested for murder, Wagoner spoke with Scofield,\n\nwho asked for money relating to his representation of Jenkins. She did not\n\nremember whether she talked with Scofield about Jenkins’s background. She could\n\nnot remember why she didn’t testify at the trial; she might have been ill, but she\n\nwas probably busy.\n\n Sherry Seal, the wife of Lonnie Seal, who had testified in the penalty phase\n\nof the trial, testified that Jenkins was respectful and generous and that she trusted\n\nhim with her children. She related how he supported her family after moving to\n\nAlabama, giving them “pretty much” his entire paycheck. She said she would still\n\nfeel safe around Jenkins. Two jailers also testified that Jenkins was a respectful,\n\ncourteous, model inmate who never complained or caused any problems.\n\n Jenkins also introduced medical, school, and juvenile court records from his\n\nchildhood. The medical records document his premature birth and his mother’s\n\nintention to give him up for adoption. The school records reflect a struggling\n\nstudent who was repeatedly promoted to the next grade because of his age, despite\n\nserious academic difficulties, chronic truancy, and frequent school changes. The\n\n\n 16\n\njuvenile court records document Jenkins’s placement in juvenile hall starting at age\n\n14, following repeated acts of grand theft auto and running away.\n\n The Rule 32 hearing concluded with the testimony of two psychologists\n\nabout Jenkins’s history of psychological trauma stemming from his childhood\n\nabuse. The defense expert, Dr. David Lisak, had met with Jenkins three times and\n\ninterviewed a dozen of his family members, family friends, and associates. Dr.\n\nLisak testified that the pervasive abuse and neglect in Jenkins’s childhood impaired\n\nhis ability to cope with trauma as an adult. He opined that Jenkins’s childhood\n\nabuse adversely affected his cognitive and emotional development and other\n\nabilities. He described how Jenkins had been neglected as an infant and sexually\n\nabused by his grandfather at age four, which led to a lifetime of bladder and bowel\n\ndysfunction. He also discussed the hypothesis that children who are abused are at\n\ngreater risk for perpetrating violence later in life, but he acknowledged there would\n\nhave been other causative factors in Jenkins’s case. Dr. Lisak described Jenkins’s\n\nintellectual capacity as “somewhat borderline,” though he did not perform any\n\ntesting of his own.\n\n The State’s expert, Dr. Karl Kirkland, had met with Jenkins for several hours\n\nand administered several tests. On the Wechsler Adult Intelligence Scale, Jenkins\n\n“scored in the range of borderline intellectual functioning which is between mild\n\nmental retardation and low average intellectual functioning” with an overall IQ of\n\n\n 17\n\n76. Dr. Kirkland agreed that Jenkins had been “reared in a chaotic dysfunctional\n\nfamily” and generally agreed with Dr. Lisak’s assessment of the negative effects of\n\nchildhood trauma.\n\n The court considered all of this evidence before denying Rule 32 relief in\n\nfull. As relevant here, it found that some of the evidence about the severity of\n\nabuse that Jenkins suffered as a child was not credible. It further found that Jenkins\n\nsuffered no prejudice from any deficient performance by his penalty-phase counsel\n\nbecause the aggravating circumstances outweighed any additional mitigating\n\nevidence Downey might have adduced.\n\n The Alabama Court of Criminal Appeals affirmed the denial of Rule 32\n\nrelief. Jenkins v. State, 972 So. 2d 111 (Ala. Ct. Crim. App. 2004). In relevant part,\n\nthe appeals court rejected Jenkins’s ineffective assistance claim, finding that\n\ncounsel reasonably chose to pursue a penalty-phase strategy of residual doubt and\n\ngood character. Id. at 147. It explained that the new evidence was self-\n\ncontradictory in that it asserted that childhood abuse would have made Jenkins a\n\nviolent adult yet also asserted that Jenkins was meek and mild. Id. The appeals\n\ncourt also agreed with the lower court’s conclusion that Jenkins suffered no\n\nprejudice from any error by his counsel because the aggravating circumstances still\n\noutweighed any new mitigating ones. Id. at 148. The court also noted that evidence\n\n\n\n\n 18\n\nabout Jenkins’s good behavior in jail awaiting trial was “minimally mitigating” and\n\nwould not have affected the sentence. Id. at 149.\n\n Because the Supreme Court’s decision in Atkins v. Virginia6 had issued\n\nduring the pendency of the Rule 32 appeal, the Alabama appeals court ordered\n\nsupplemental briefing on the possible impact of Atkins upon Jenkins’s sentence.\n\nJenkins asked the court to stay the proceedings and conduct an evidentiary hearing\n\nafter the Alabama legislature enacted a statute about executing the mentally\n\nretarded or, in the alternative, to vacate his sentence and remand after the Alabama\n\nlegislature acted. He later asserted that he is mentally retarded and therefore\n\nineligible for the death penalty. The appeals court proceeded to reject Jenkins’s\n\nAtkins claim on the merits because the record established that his IQ was 76 and\n\nthat he maintained relationships and employment. Id. at 155. The Alabama\n\nSupreme Court summarily affirmed, in relevant part. Ex parte Jenkins, 972 So. 2d\n\n159, 165 (Ala. 2005), remanded to 972 So. 2d 165 (Ala. Ct. Crim. App. 2005)\n\n(juror misconduct claim), cert. denied, 552 U.S. 1167 (2008).\n\n In 2008, Jenkins filed his first petition for a writ of habeas corpus in federal\n\ncourt pursuant to 28 U.S.C. § 2254. The federal proceedings were stayed for\n\nseveral years while Jenkins pursued additional, unrelated Rule 32 relief in the state\n\n\n6\n 536 U.S. 304, 321 (2002) (holding that the execution of mentally retarded offenders violates the\nEighth Amendment). The rule announced in Atkins is retroactive to cases on collateral review. In\nre Holladay, 331 F.3d 1169, 1172–73 (11th Cir. 2003).\n 19\n\ncourts, ultimately unsuccessfully. In the district court, Jenkins moved for an\n\nevidentiary hearing on his Atkins claim. After reviewing the state court record,\n\nwith a focus on the Rule 32 testimony of Drs. Lisak and Kirkland, the district court\n\ndenied the motion in 2015. Noting that our Circuit has accepted the State of\n\nAlabama’s definition of mental retardation following Atkins, the court found that\n\nJenkins is not intellectually disabled under that standard. In a 347-page order, the\n\ndistrict court also denied relief on all of Jenkins’s claims including ineffective\n\nassistance of counsel during the penalty phase and affirmed as reasonable the state\n\ncourt’s conclusion that Jenkins failed to establish prejudice from any error by his\n\ncounsel. The district court later denied reconsideration and a certificate of\n\nappealability.\n\n Jenkins now appeals. Our Court granted Jenkins a certificate of\n\nappealability, 28 U.S.C. § 2253(c), on two issues:\n\n 1) Whether the district court erred in denying Appellant’s claim\n that his trial counsel rendered ineffective assistance of counsel in the\n investigation and presentation of mitigating evidence during the\n penalty phase of Appellant’s 1991 trial; and\n\n 2) Whether the district court erred in denying Appellant’s claim\n that he is intellectually disabled and ineligible for the death penalty\n under Atkins v. Virginia, 536 U.S. 304 (2002).\n\n II. STANDARDS OF REVIEW\n\n We review the denial of a petition for a writ of habeas corpus de novo.\n\nMorrow v. Warden, Ga. Diagnostic Prison, 886 F.3d 1138, 1146 (11th Cir. 2018).\n\n 20\n\nThe Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) instructs that\n\na writ of habeas corpus shall not be granted to a state prisoner unless the state\n\nadjudication of a claim “resulted in a decision that was contrary to, or involved an\n\nunreasonable application of, clearly established Federal law, as determined by the\n\nSupreme Court of the United States.” 28 U.S.C. § 2254(d)(1). “[C]learly\n\nestablished Federal law” means “the holdings, as opposed to the dicta, of [the\n\nSupreme] Court’s decisions as of the time of the relevant state-court decision.”\n\nWilliams v. Taylor, 529 U.S. 362, 412 (2000). “[W]hen the last state court to\n\ndecide a prisoner’s federal claim explains its decision on the merits in a reasoned\n\nopinion . . . a federal habeas court simply reviews the specific reasons given by the\n\nstate court and defers to those reasons if they are reasonable.” Wilson v. Sellers,\n\n138 S. Ct. 1188, 1192 (2018). A state court’s decision is reasonable “so long as\n\n‘fairminded jurists could disagree’ on the correctness of the state court’s decision.”\n\nHarrington v. Richter, 562 U.S. 86, 101 (2011) (quoting Yarborough v. Alvarado,\n\n541 U.S. 652, 664 (2004)). “[E]ven a strong case for relief does not mean the state\n\ncourt’s contrary conclusion was unreasonable.” Id. at 102. Rather, a prisoner must\n\nshow that the state court’s ruling “was so lacking in justification that there was an\n\nerror well understood and comprehended in existing law beyond any possibility for\n\nfairminded disagreement.” Id. at 103. That ruling must have been “‘objectively\n\nunreasonable,’ not merely wrong; even ‘clear error’ will not suffice.” White v.\n\n\n 21\n\nWoodall, 572 U.S. 415, 419 (2014) (quoting Lockyer v. Andrade, 538 U.S. 63, 75\n\n(2003)).\n\n In addition, “a determination of a factual issue made by a State court shall be\n\npresumed to be correct,” and the prisoner bears “the burden of rebutting the\n\npresumption of correctness by clear and convincing evidence.” 28 U.S.C.\n\n§ 2254(e)(1). AEDPA further instructs that a writ of habeas corpus may not be\n\ngranted unless the state court adjudication “resulted in a decision that was based on\n\nan unreasonable determination of the facts in light of the evidence presented in the\n\nState court proceeding.” Id. § 2254(d)(2). When a federal habeas court makes a\n\nfactual finding as part of its habeas determination, we review for clear error, which\n\nis a “highly deferential” standard of review. Thomas v. Allen, 607 F.3d 749, 752\n\n(11th Cir. 2010) (quoting Holton v. City of Thomasville Sch. Dist., 425 F.3d 1325,\n\n1350 (11th Cir. 2005)). “A finding is ‘clearly erroneous’ when although there is\n\nevidence to support it, the reviewing court on the entire evidence is left with the\n\ndefinite and firm conviction that a mistake has been committed.” United States v.\n\nU.S. Gypsum Co., 333 U.S. 364, 395 (1948).\n\n III. DISCUSSION\n\n A. Ineffective Assistance of Counsel\n\n A prisoner alleging that he received ineffective assistance of counsel in\n\nviolation of the Sixth Amendment must establish two elements. Strickland v.\n\n\n 22\n\nWashington, 466 U.S. 668, 687 (1984). “First, the defendant must show that\n\ncounsel’s performance was deficient.” Id. In other words, he must show “that\n\ncounsel made errors so serious that counsel was not functioning as the ‘counsel’\n\nguaranteed the defendant by the Sixth Amendment.” Id. “Second, the defendant\n\nmust show that the deficient performance prejudiced the defense.” Id. We discuss\n\neach of these elements in turn.\n\n 1. Deficient Performance\n\n A “doubly” deferential lens sits before our eyes as we review the\n\nperformance of Jenkins’s counsel during the penalty phase. Richter, 562 U.S. at\n\n105 (quoting Knowles v. Mirzayance, 556 U.S. 111, 123 (2009)). This redoubled\n\ndeference arises out of the tandem application of AEDPA’s “highly deferential”\n\nreview of a state court’s decision with Strickland’s “highly deferential” review of\n\nan attorney’s performance. Id. (quoting Lindh v. Murphy, 521 U.S. 320, 333 n.7\n\n(1997), and Strickland, 466 U.S. at 689). Under the Strickland standard, “counsel\n\nis strongly presumed to have rendered adequate assistance and made all significant\n\ndecisions in the exercise of reasonable professional judgment.” Strickland, 466\n\nU.S. at 690. And under AEPDA, we must deny habeas relief if “there is any\n\nreasonable argument that counsel satisfied Strickland’s deferential standard.”\n\nRichter, 562 U.S. at 105.\n\n\n\n\n 23\n\n Jenkins argues that his counsel performed deficiently by failing to\n\n“investigate compelling mitigating evidence or develop a penalty-phase theory.”\n\nThe state court disagreed. It quoted decisions of our Court that hold that “there is\n\nno per se rule that evidence of a criminal defendant’s troubled childhood must\n\nalways be presented as mitigating evidence.” Jenkins, 972 So. 2d at 144 (quoting\n\nMarek v. Singletary, 62 F.3d 1295, 1300 (11th Cir. 1995)). Noting that “Downey\n\nwas in charge of the penalty phase” but “we do not have the benefit of Downey’s\n\ntestimony as to what occurred and why,” it reviewed the trial record to find that\n\n“counsel argued residual doubt and Jenkins’s good character at the penalty phase.”\n\nId. at 145–46. It concluded that “Downey’s decision to concentrate on reasonable\n\ndoubt and to portray Jenkins as a good person was reasonable under the\n\ncircumstances.” Id. at 147.\n\n Jenkins insists, however, that counsel’s failure to develop mitigating\n\nevidence about his childhood abuse was “not the product of strategy, but rather\n\narose from ingnorance and inexeperiance [sic].” The problem for Jenkins is that,\n\nlike the state courts, this Court also does not know what Downey did or why. The\n\nrecord is silent as to his thoughts and intentions as he prepared for the penalty\n\nphase. Given that Jenkins bore the burden of overcoming the strong presumption of\n\ncompetence and of proving Downey’s deficient performance, we are perplexed\n\nwhy Jenkins’s new counsel did not have Downey testify or submit an affidavit in\n\n\n 24\n\nthe Rule 32 proceeding. See Strickland, 466 U.S. at 687 (“The defendant must\n\nshow that counsel’s performance was deficient.”). Nor did Jenkins testify to the\n\nsubstance of his conversations with his counsel. “The reasonableness of a trial\n\ncounsel’s acts, including lack of investigation or excluding character witnesses\n\nfrom the sentencing phase, depends ‘critically’ upon what information the client\n\ncommunicated to counsel.” Chandler v. United States, 218 F.3d 1305, 1324 (11th\n\nCir. 2000) (en banc) (quoting Strickland, 466 U.S. at 691). There is simply no\n\nrecord evidence to support Jenkins’s central assertion that Downey made no\n\ntactical choices in his preparation for the penalty phase.\n\n Jenkins instead relies on Scofield’s Rule 32 testimony about how he and\n\nDowney prepared for the trial and sentencing. We acknowledge that Scofield\n\nrepeatedly testified that he did not really know what he was doing and that he\n\nfailed to make strategic choices with respect to mitigating circumstances and\n\nsentencing. But Scofield was equally clear that Downey—the more experienced\n\nlawyer—was in fact responsible for the penalty phase. And Scofield also\n\nrepeatedly testified that he did not know what Downey may or may not have done\n\nto prepare. Furthermore, his lack of knowledge about Downey’s preparation was\n\nunderstandable. He and Downey were not colleagues who shared an office and\n\nmight reasonably be expected to know what the other was doing much of the time.\n\nScofield and Downey did not even work in the same city.\n\n\n 25\n\n This case is unlike those in which we have found deficient performance\n\nwhere the record showed that counsel investigated insufficiently because he\n\n“waited until the eleventh hour” to begin preparing for the penalty phase while\n\nfaced with “overwhelming evidence of guilt.” See, e.g., Johnson v. Sec’y, D.O.C.,\n\n643 F.3d 907, 932 (11th Cir. 2011). The limited record that Jenkins developed\n\nestablishes that Downey spent at least 71 hours preparing for trial, spread out over\n\n17 months. The state court interpreted this sparse record as reflecting a penalty-\n\nphase strategy of residual doubt, the logical extension of Downey’s guilt-phase\n\ninvestigation of the evidence and Scofield’s pursuit at trial of an acquittal based on\n\nreasonable doubt. See Jenkins, 972 So. 2d at 147. In such a situation, “[a] lawyer’s\n\ntime and effort in preparing to defend his client in the guilt phase of a capital case\n\ncontinues to count at the sentencing phase.” Tarver v. Hopper, 169 F.3d 710, 715\n\n(11th Cir. 1999).\n\n We decline to dictate how many hours or how many sessions an attorney\n\nmust spend interviewing his counsel in order to provide effective assistance. “No\n\nabsolute rules dictate what is reasonable performance for lawyers.” Chandler, 218\n\nF.3d at 1317 (citing Strickland, 466 U.S. at 688–89). But even this limited record\n\nundermines Jenkins’s assertion that Downey prepared inadequately. Downey’s 71\n\nhours of preparation included 35 hours of jailhouse conversations with Jenkins\n\n\n\n\n 26\n\nduring 18 separate meetings.7 According to Downey’s notes, Scofield was present\n\nfor only five of those meetings (12 hours’ worth), leaving 23 hours over 13\n\nmeetings between Jenkins and Downey alone. Downey’s notes reveal little about\n\nthe content of those conversations beyond broad notations such as “evidence,\n\nwitnesses, general status,” and discussions with Jenkins of the evidence and\n\ntestimony against him were surely relevant to Downey’s penalty-phase strategy of\n\nresidual doubt.\n\n But the record does not reveal that Downey knew or should have known that\n\nJenkins’s childhood merited investigation beyond that which he may have done.\n\nThe record shows that Downey developed enough of a rapport with Jenkins that,\n\naccording to his fee declaration, he “g[ave a] Christmas gift [and] discuss[ed] how\n\nmuch to tell his family.” But how much Downey knew about Jenkins’s family\n\nbackground can only be guessed from Scofield’s Rule 32 hearing testimony.\n\nScofield testified that he discussed Jenkins’s background during his dozen or so\n\njailhouse meetings with him, and that Downey was present for some of those\n\nmeetings. Scofield—and possibly Downey—learned that Jenkins experienced\n\nabuse as a child, that his stepfather was brutal, and that he lived on the streets and\n\n\n\n\n7\n We note that the state court clearly erroneously stated that Downey’s fee declaration\ndocumented 171 hours of pretrial preparation. See Jenkins, 972 So. 2d at 137. However, the state\ncourt accurately stated that “he spent over 25 hours talking with Jenkins in more than 10 visits to\nthe jail.” Id. That statement was not an unreasonable determination of the facts.\n 27\n\nin juvenile institutions because of his brutal home life. It is against this backdrop\n\nthat Downey’s presumptive “decision not to investigate must be directly assessed\n\nfor reasonableness in all the circumstances, applying a heavy measure of deference\n\nto counsel’s judgments.” Strickland, 466 U.S. at 691.\n\n That assessment depends a great deal on what Jenkins told his counsel about\n\nhis background. See id. But we do not know whether Downey was present when\n\nScofield learned about Jenkins’s childhood abuse, nor what else Downey may have\n\nlearned during his 13 solo discussions with Jenkins. Therefore, “given the absence\n\nof evidence in the record, we must assume counsel carried out his professional\n\nresponsibility and discussed mitigation with his client.” Chandler, 218 F.3d at\n\n1324. We will not “turn that presumption on its head by giving [Jenkins] the\n\nbenefit of the doubt where it is unclear what [Downey] did or did not do.” Williams\n\nv. Head, 185 F.3d 1223, 1235 (11th Cir. 1999). We will not assume that Jenkins\n\ntold Downey more than he told Scofield, and Scofield denied remembering that\n\nJenkins told him about the severity or frequency of his stepfather’s physical abuse.\n\nHe denied remembering Jenkins telling him about being beaten daily, being locked\n\nin his room 24/7, or being forced to eat his feces. And there is no suggestion in the\n\nrecord that counsel knew or had reason to investigate the allegation of sexual abuse\n\nby his stepgrandfather. The record simply does not reflect that Jenkins told either\n\nof his lawyers about the severe abuse that some of the Rule 32 witnesses later\n\n\n 28\n\ndescribed. We will not fault those lawyers for failing to present as mitigation this\n\nlimited evidence that they may have reasonably believed to be unremarkable. See\n\nStrickland, 466 U.S. at 691.\n\n To be sure, counsel’s failure “to discover all reasonably available mitigating\n\nevidence” may constitute deficient performance in some cases. Wiggins v. Smith,\n\n539 U.S. 510, 524 (2003) (quoting ABA Guidelines for the Appointment and\n\nPerformance of Counsel in Death Penalty Cases § 11.4.1(C) (1989)). Our Court\n\nhas identified deficient performance in several instances where counsel had good\n\nreason to investigate the defendant’s background but failed to do so. See Morrow,\n\n886 F.3d at 1147 (collecting cases). But we assess “a particular decision not to\n\ninvestigate . . . for reasonableness in all the circumstances, applying a heavy\n\nmeasure of deference to counsel’s judgments.” Strickland, 466 U.S. at 691. Thus,\n\nwe have rejected a hard rule that counsel’s “strategic decisions can be considered\n\nreasonable only if they are preceded by a ‘thorough investigation.’” Williams, 185\n\nF.3d at 1237 (quoting Rogers v. Zant, 13 F.3d 384, 387 (11th Cir. 1994)). Because\n\nwe simply do not know why Downey chose not to investigate Jenkins’s childhood\n\nmore thoroughly, Downey is entitled to the presumption that this strategic decision\n\nwas reasonable.\n\n And though we have also rejected “a per se rule of ineffective assistance\n\nwhere counsel does not consult family members,” Holladay v. Haley, 209 F.3d\n\n\n 29\n\n1243, 1252 (11th Cir. 2000), we note that Downey did not fail altogether to consult\n\nfamily members. Downey had at least one substantive conversation with Jenkins’s\n\ngrandmother, and we know from her Rule 32 hearing testimony that she never\n\nwitnessed Jenkins’s childhood abuse. Given the silence of the record and the strong\n\npresumption of competence to which Downey is entitled, we decline to fault him\n\nfor not doing more when Jenkins and his grandmother did not and could not,\n\nrespectively, provide a basis for doing so. Cf. United States v. Pease, 240 F.3d 938,\n\n941–92 (11th Cir. 2001) (“we cannot say as a matter of law that reliance on a\n\nclient’s statements is per se deficient performance”). Nor did Jenkins’s counsel\n\nhave an absolute duty to put on abusive-childhood evidence, even if they knew of\n\nit. See Chandler, 218 F.3d at 1319 (“Counsel is not required to present every\n\nnonfrivolous defense; nor is counsel required to present all mitigation evidence,\n\neven if the additional mitigation evidence would not have been incompatible with\n\ncounsel’s strategy.”); Waters v. Thomas, 46 F.3d 1506, 1511 (11th Cir. 1995) (en\n\nbanc) (“Our decisions are inconsistent with any notion that counsel must present\n\nall available mitigating circumstance evidence.”).\n\n Instead, we affirm as a reasonable application of Strickland the state court’s\n\nconclusion that Downey’s strategic decision to emphasize residual doubt was\n\nreasonable. See Jenkins, 972 So. 2d at 147. As we have noted, “[a] lawyer’s time\n\nand effort in preparing to defend his client in the guilt phase of a capital case\n\n\n 30\n\ncontinues to count at the sentencing phase.” Tarver, 169 F.3d at 715. This\n\nreasoning is especially applicable where, as here, counsel pursued the reasonable\n\nstrategy of arguing for acquittal at the guilt phase and for life at the sentencing\n\nphase by highlighting doubt about the evidence. “[R]esidual doubt is perhaps the\n\nmost effective strategy to employ at sentencing.” Chandler, 218 F.3d at 1320 n.28.\n\n“[E]specially when—as in this case—the evidence of guilt was not overwhelming,\n\nwe expect that petitioners can rarely (if ever) prove a lawyer to be ineffective for\n\nrelying on this seemingly reasonable strategy to defend his client.” Id. at 1320\n\n(footnote omitted); see also Tarver, 169 F.3d at 715–16 (collecting law review\n\narticles on the merits of residual-doubt strategy). Not only can such a strategy be\n\nconstitutionally adequate, in some cases it may be “the most effective performance\n\nin defense to the death penalty.” Tarver, 169 F.3d at 716.\n\n We cannot say that Downey’s manifest strategy of residual doubt was\n\nobjectively unreasonable under the circumstances. Jenkins’s case is precisely the\n\nsort where reminding the jury of the misgivings it may have held about the State’s\n\ncase could have been effective. Scofield’s guilt-phase closing argument from that\n\nvery morning, which explicitly emphasized doubt and the weakness of the State’s\n\nevidence, would have been fresh in the minds of the jury.8 Although we are not\n\n\n8\n During the penalty phase of a capital case, “[e]vidence presented at the trial of the case may be\nconsidered insofar as it is relevant to the aggravating and mitigating circumstances without the\nnecessity of re-introducing that evidence at the sentence hearing.” Ala. Code § 13A-5-45(c).\n 31\n\nquestioning the jury’s guilty verdict or the sufficiency of the evidence, which are\n\nbeyond the scope of our habeas review, we acknowledge that the State’s case\n\nagainst Jenkins was—as Downey reminded the jury during the penalty phase—\n\nentirely circumstantial. Indeed, Scofield maintained post-conviction that the case\n\nfor reasonable doubt was “[v]ery strong.” No one saw Jenkins commit either the\n\nmurder or the robbery and kidnapping that elevated Hogeland’s murder to a capital\n\ncrime. Scofield’s guilt-phase closing argument challenged the inferential leaps the\n\nState was asking the jury to make in finding that Jenkins committed these capital-\n\npredicate offenses and that he formed the intent to kill. This situation was not one\n\nwhere a guilt-phase strategy of acquittal—and therefore a penalty-phase strategy of\n\nresidual doubt—was unreasonable in light of the evidence. Cf. Johnson, 643 F.3d\n\nat 932–33 (deficient sentencing investigation where defendant had confessed to\n\ntwo murders).\n\n Thus, although Downey’s penalty-phase closing argument did not\n\nspecifically invoke lingering doubt, it clearly reminded the jury of the entirely\n\ncircumstantial nature of the State’s guilt-phase case. See Chandler, 218 F.3d at\n\n1320 & n.6 (a strategy of “focus[ing] on obtaining an acquittal and then, at\n\nsentencing, on lingering doubt” is reasonable, even if counsel does not use the\n\nwords “lingering doubt” or “residual doubt”). Downey closed this reminder by\n\nexplaining to the jury how a foundational ancient legal system proscribed capital\n\n\n 32\n\npunishment in the absence of eyewitness testimony—the exact situation here. The\n\nstate court found that Downey’s penalty-phase closing argument reflected a\n\nstrategic “decision to concentrate on reasonable doubt,” see Jenkins, 972 So. 2d at\n\n147, a tactical choice entitled to a “strong presumption of correctness,” Marek, 62\n\nF.3d at 1300 & n.3.\n\n Jenkins suggests other theories of mitigation that Downey might have more\n\neffectively pursued,9 but our deferential review under Strickland does not ask\n\nwhether counsel could possibly or ideally have been more effective. “The test for\n\nineffectiveness is not whether counsel could have done more.” Waters, 46 F.3d at\n\n1518. We do not ask whether an attorney’s representation “deviated from best\n\npractices or common custom,” and we should resist the temptation to second-guess\n\nan attorney with the benefit of our hindsight. Richter, 562 U.S. at 105. We ask only\n\n“whether, in light of all the circumstances, the identified acts or omissions were\n\n\n\n\n9\n In particular, Jenkins argues that his counsel should have presented “model prisoner” evidence.\nThe state court noted, “Good conduct during pretrial incarceration is not necessarily a mitigating\ncircumstance,” before finding that Jenkins had not met his burden of establishing that his counsel\neven had information about his conduct in jail. Jenkins, 972 So. 2d at 149. We agree, and we\ndecline to impose a requirement of sua sponte investigation of this kind in every capital case. Cf.\nWilliams, 529 U.S. at 396 (finding deficient performance based in part on not returning a call\nfrom a prison ministry volunteer who offered to vouch for defendant’s good behavior in prison).\n“Surely, counsel is not required to call a witness to testify to facts such as lack of violent nature\nwhen the jury has rejected such an approach and has found that the defendant is guilty of\nmurder.” Griffin v. Wainwright, 760 F.2d 1505, 1512 (11th Cir. 1985), vacated on other grounds,\n476 U.S. 1112 (1986), reaff’d, 874 F.2d 1397 (11th Cir. 1989).\n 33\n\noutside the range of professionally competent assistance.” Strickland, 466 U.S. at\n\n690.\n\n We conclude that the state court did not unreasonably determine that Jenkins\n\nfailed to establish objectively incompetent performance by his counsel during the\n\npenalty phase of his trial. Our inquiry is not “what the best lawyers would have\n\ndone. Nor is the test even what most good lawyers would have done. We ask only\n\nwhether some reasonable lawyer at the trial could have acted, in the circumstances,\n\nas defense counsel acted at trial. . . . We are not interested in grading lawyers’\n\nperformances.” White v. Singletary, 972 F.2d 1218, 1220–21 (11th Cir. 1992). And\n\n“[i]f fairminded jurists could reasonably disagree” about the reasonableness of\n\ncounsel’s performance, “then habeas relief is due to be denied.” Johnson, 643 F.3d\n\nat 932.\n\n 2. Prejudice\n\n Even if Jenkins had demonstrated that his counsel performed as no\n\nreasonable lawyer could have, he must also demonstrate a “reasonable probability\n\nthat, but for counsel’s unprofessional errors, the result of the proceeding would\n\nhave been different.” Strickland, 466 U.S. at 694. Given that the jury here\n\nrecommended a sentence of death by the narrowest possible vote, 10 to 2, Jenkins\n\nneed establish only “a reasonable probability that at least one juror would have\n\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537. “A\n\n\n 34\n\nreasonable probability is a probability sufficient to undermine confidence in the\n\noutcome.” Strickland, 466 U.S. at 694. In measuring that probability here, we will\n\n“evaluate the totality of the available mitigation evidence—both that adduced at\n\ntrial, and the evidence adduced in the habeas proceeding in reweighing it against\n\nthe evidence in aggravation.” Williams, 529 U.S. at 397–98.\n\n We begin our analysis by clarifying what exactly Jenkins is arguing about\n\nprejudice. We do not know precisely which aggravating and mitigating factors the\n\njury may have implicitly found before making its recommendation of death. The\n\ntrial court found at sentencing two statutory aggravating factors: the robbery and\n\nthe kidnapping of the victim. See Ala. Code § 13A-5-49(4). Because none of the\n\nmitigation evidence offered in the Rule 32 hearing bears on those factors, we see\n\nno reasonable probability that the aggravating factors would have been determined\n\ndifferently by the jury or the court. The court also found at sentencing two\n\nstatutory mitigating factors: no significant history of prior criminal activity and\n\nage. See id. § 13A-5-51(1), (7). There was also strong evidence of Jenkins’s\n\nintoxication, which was a key part of Jenkins’s guilt-phase defense and could have\n\nsupported a third mitigating factor, impaired capacity. See id. § 13A-5-51(6).10\n\n\n10\n The jury had heard evidence about all three of these statutory mitigators during the guilt phase\nof the trial. Downey’s penalty-phase opening statement reviewed for the jury all seven of the\npossible statutory mitigating factors, and the court also recited them in its jury instructions,\nreminding the jury that they could consider guilt-phase evidence as well as non-statutory\nmitigation. Jenkins has not established that there is a substantial probability that the jury’s\nrecommendation would have been different had it been more explicitly reminded of Jenkins’s\n 35\n\nJenkins does not argue that the jury or the court would have found other statutory\n\nmitigators in light of the Rule 32 evidence. Rather, he contends that the jury might\n\nhave found non-statutory mitigation based on Jenkins’s abusive childhood, low\n\nintelligence, or mental health problems that would have outweighed the\n\naggravating factors.11\n\n The Rule 32 trial court found that Jenkins had not demonstrated prejudice\n\neven assuming that counsel had performed deficiently. See Jenkins, 972 So. 2d at\n\n138. It concluded that he had not shown a reasonable probability that the sentencer\n\n“would have concluded that a weighing of the aggravating and mitigating\n\ncircumstances did not warrant death” upon a reweighing of the evidence. See id. It\n\nemphasized the depravity of the crime and the aggravating circumstances and\n\nfound that the mitigating witnesses were not credible for various reasons. See id.\n\nThe state appellate court affirmed, declining to disturb the credibility\n\ndeterminations of the trial court. Id. at 142–44. Independently reweighing the\n\n\nage, criminal history, and intoxication, given that the trial court, the Rule 32 trial court, and the\nstate appellate court all explicitly considered those mitigators and found them outweighed by the\naggravators.\n11\n Alabama law lists seven statutory mitigating circumstances, but states that mitigation “shall\n. . . not be limited to” those seven circumstances. Ala. Code § 13A-5-51. Further, “mitigating\ncircumstances shall include any aspect of a defendant’s character or record and any of the\ncircumstances of the offense that the defendant offers as a basis for a sentence of life\nimprisonment without parole instead of death, and any other relevant mitigating circumstances\nwhich the defendant offers as a basis for a sentence of life imprisonment without parole instead\nof death.” Id. § 13A-5-52. A jury may recommend a sentence of death only if it finds that “one or\nmore aggravating circumstances . . . exist and that they outweigh the mitigating circumstances, if\nany.” Id. § 13A-5-46(e)(3).\n 36\n\nevidence, it concluded that, “[g]iven the aggravating circumstances that were\n\nproven by the State and the facts surrounding Hogeland’s murder, we, like the\n\ncircuit court, are confident that death was the appropriate punishment for Jenkins’s\n\nactions.” Id. at 148.\n\n Our review of this evidence does not show that it is likely that the jury\n\nwould have concluded that death was not the appropriate punishment if it had\n\nheard the Rule 32 evidence. Much of the proffered mitigation evidence is, in our\n\nview, much less mitigating than Jenkins asserts. In particular, the testimony of\n\nJenkins’s older brother Michael was likely to have been a double-edged sword. See\n\nEvans v. Sec’y, Dep’t of Corr., 703 F.3d 1316, 1327 (11th Cir. 2013) (en banc) (“it\n\nis reasonable to conclude that a defendant was not prejudiced when his mitigation\n\nevidence was a two-edged sword” (quoting Ponticelli v. Sec’y, Fla. Dep’t of Corr.,\n\n690 F.2d 1271, 1296 (11th Cir. 2012)). Michael was the closest relative to testify at\n\nthe Rule 32 hearing and his testimony about the abuse Jenkins experienced should\n\nhave been the most reliable, since he experienced much of the same physical abuse\n\nhimself. The brothers shared the same filthy bedroom, received beatings from the\n\nsame paddle, and were frequently absent from school together. But Michael’s\n\ntestimony contained some discrepancies, and he admitted to cognitive and memory\n\nproblems. Perhaps Michael’s testimony would have made Jenkins appear more\n\nsympathetic in the eyes of the jury—but perhaps not. While Jenkins ran away from\n\n\n 37\n\nhome and eventually committed murder, Michael responded to the same\n\nbackground by running away from home, graduating from high school, and\n\nproviding health care for both the terminally ill and people with substance use\n\ndisorders. It is possible that the jury could even have found Jenkins to be more\n\nculpable for not overcoming his difficult upbringing and to pose more of a danger\n\nof future violence.\n\n We also find reasonable the state court’s identification of other problems\n\nwith the mitigating evidence Jenkins argues should have been presented. Hearing\n\nthat evidence firsthand in the Rule 32 proceeding, the trial court found the\n\ntestimony of the cousins less credible in light of the lack of physical abuse\n\ndocumented in the school health records. See Jenkins, 972 So. 2d at 139 (noting\n\nthat school records observed a rash and gingivitis). It faulted Pitts in particular for\n\npurporting to have witnessed ongoing, severe abuse but taking only minimal,\n\nineffective steps to obtain outside help. See id. It further noted that DeLavega and\n\nWagoner actually testified that they never saw Jenkins beaten, and it noted all of\n\nthe witnesses’ interest in getting Jenkins off death row. All of these factors would\n\nlikely have caused a reasonable jury to discount their testimony. See id. at 139–42.\n\nWe must accept the trial court’s views of the Rule 32 testimony. “In the absence of\n\nclear and convincing evidence, we have no power on federal habeas review to\n\nrevisit the state court’s credibility determinations.” Bishop v. Warden, 726 F.3d\n\n\n 38\n\n1243, 1259 (11th Cir. 2013) (citing Marshall v. Lonberger, 459 U.S. 422, 434\n\n(1983)).\n\n The state court also noted that Wagoner testified that she was not available\n\nto testify at the trial, and it found that Seal’s testimony would have been\n\ncumulative to that of her husband. Thus, counsel’s failure to call those two\n\nwitnesses could not have been prejudicial to Jenkins. See id. at 141, 142. Neither of\n\nthese conclusions involves an unreasonable determination of the facts. We agree\n\nthat the Rule 32 testimony of Sherry Seal was substantially similar to the testimony\n\nof her husband Lonnie Seal that the jury heard during the penalty phase. Both\n\ntestified about meeting Jenkins in California, Jenkins’s assistance with their move\n\nto Alabama, and his residence with them in their Alabama home. Sherry added\n\nonly that Jenkins sometimes wet his bed. It was not unreasonable for the state court\n\nto conclude that Sherry’s testimony was cumulative of her husband’s, and “a\n\npetitioner cannot satisfy the prejudice prong of the Strickland test with evidence\n\nthat is merely cumulative of evidence already presented.” Rose v. McNeil, 634 F.3d\n\n1224, 1243 (11th Cir. 2011).\n\n Further, the school, medical, and juvenile court records that Jenkins\n\nintroduced in the Rule 32 proceeding have their own problems. As noted, they do\n\nnot document or even corroborate the severe level of abuse alleged by some of the\n\nwitnesses. Though these documents tell a sad story, they do not create a substantial\n\n\n 39\n\nlikelihood that the jury would have recommended life. Although they portray\n\nJenkins as a victim of truly unfortunate circumstances, they also show his unsavory\n\ndelinquent behavior. That delinquency included a history of stealing cars, and\n\ngiven that the jury likely inferred that Jenkins stole the red Mazda, new\n\ninformation about his history of theft may not have been especially mitigating.\n\nSimilarly, we agree with the state court that evidence of Jenkins’s compliant\n\nconduct in pretrial detention is “minimally mitigating” where there had been no\n\nevidence of jail misconduct as an aggravator. See Jenkins, 972 So. 2d at 149; cf.\n\nSkipper v. South Carolina, 476 U.S. 1, 5 & n.1 (1986) (good jail behavior is\n\n“potentially mitigating” especially “[w]here the prosecution specifically relies on a\n\nprediction of future dangerousness in asking for the death penalty”).\n\n Jenkins argues that allowing the jury to see his experience of mental,\n\nemotional, and cognitive difficulties, compounded by years of abuse and neglect\n\nand considered alongside his mild character and good jail behavior as an adult,\n\nwould have convinced them that Hogeland’s murder was an “aberration” that\n\ndeserved a punishment less severe than death. That interpretation of Jenkins’s life\n\nstory may be conceivable. But more than mere conceivability is required to\n\nestablish prejudice: “The likelihood of a different result must be substantial, not\n\njust conceivable.” Richter, 562 U.S. at 112 (emphasis added). We do not find the\n\nstate court’s decision that a different result was not substantially likely to be an\n\n\n 40\n\nunreasonable application of Strickland. As we have noted, we think it equally\n\nlikely that the jury could have rejected this alternate narrative out of hand in light\n\nof having just convicted him of the violent murder of a young mother. We also\n\nthink it possible that the jury could have viewed the murder as just one step in a\n\nlogical progression of degeneracy that began with behavioral problems at school,\n\ncontinued through the juvenile court system, and escalated past grand theft auto\n\ninto kidnapping and murder. Taken separately as well as together, the proffered\n\nevidence does not compel us to conclude that a verdict of life would have been\n\nsubstantially likely. Because fairminded jurists could disagree about the state\n\ncourt’s conclusion about Strickland prejudice, we affirm the denial of habeas relief\n\non this ground.\n\n B. Intellectual Disability\n\n Jenkins argues that the district court erred when it denied his claim that he is\n\nintellectually disabled and ineligible for the death penalty under Atkins v. Virginia,\n\n536 U.S. 304 (2002). The Supreme Court held in Atkins that executing an offender\n\nwho has mental retardation12 violates the Eighth Amendment’s ban on excessive\n\n\n\n\n12\n In 2002, the Atkins Court followed prevailing convention in referring to “mentally retarded”\noffenders. We will use that now-disfavored term only when quoting or discussing older sources.\nSince 2014, the Supreme Court has exclusively used the terms “intellectually disabled person”\nand “person with intellectual disability.” See, e.g., Hall v. Florida, 572 U.S. 701, 704 (2014);\nMoore v. Texas, 139 S. Ct. 666, 672 (2019).\n 41\n\npunishments. Id. at 321. It defined mental retardation by quoting the American\n\nAssociation on Mental Retardation:\n\n Mental retardation refers to substantial limitations in present\n functioning. It is characterized by significantly subaverage intellectual\n functioning, existing concurrently with related limitations in two or\n more of the following applicable adaptive skill areas: communication,\n self-care, home living, social skills, community use, self-direction,\n health and safety, functional academics, leisure, and work. Mental\n retardation manifests before age 18.\n\nId. at 308 n.3. Alabama courts have followed that tripartite definition, adopting\n\nwhat they consider the “broadest” definition of mental retardation for Atkins\n\npurposes:\n\n [A] defendant, to be considered mentally retarded, must have\n significantly subaverage intellectual functioning (an IQ of 70 or\n below), and significant or substantial deficits in adaptive behavior.\n Additionally, these problems must have manifested themselves during\n the developmental period (i.e., before the defendant reached age 18).\n\nEx parte Perkins, 851 So. 2d 453, 456 (Ala. 2002).\n\n The clinical evidence adduced during the Rule 32 hearing with respect to\n\nJenkins’s intellectual ability was that his IQ was 76, reflecting “borderline\n\nintellectual functioning.” The record also included the report of the Lunacy\n\nCommission13 which found that Jenkins was competent to stand trial, as well as\n\n\n\n13\n Alabama law requires a capital defendant to be “committed to the Department of Mental\nHealth and Mental Retardation for examination” if there is a question about his capacity to stand\ntrial or if there will be a mental-disease-or-defect defense. Ala. Code § 15-16-22(a). “The\nassigned mental health professional(s) shall examine the defendant with respect to determining\nthe presence of any mental disease or defect which, if determined to be present, would affect the\n 42\n\nJenkins’s school and juvenile court records. The court also considered all the\n\ntestimony it had heard about Jenkins and reviewed the transcript of his murder trial\n\nbefore rejecting Jenkins’s claim of mental retardation.\n\n On appeal, after reciting the Perkins standard, the Alabama Court of\n\nCriminal Appeals explained that Jenkins had not established mental retardation:\n\n Dr. Kirkland testified that he performed psychological tests on Jenkins\n and that Jenkins’s IQ was 76. There was evidence presented at\n Jenkins’s trial indicating that Jenkins maintained relationships with\n other individuals and that he had been employed by P.S. Edwards\n Landscaping Company, Cotton Lowe 76 Service Station, and\n Paramount Painting Company. The record fails to show that Jenkins\n meets the most liberal view of mental retardation adopted by the\n Alabama Supreme Court in Perkins.\n\nJenkins, 972 So. 2d at 155. In this habeas appeal, Jenkins raises arguments about\n\nall three components of the intellectual disability showing under Atkins and\n\nPerkins, which we address in turn. He also appeals the denial of an evidentiary\n\nhearing on this issue.\n\n 1. Intellectual Function\n\n The first component of intellectual disability under Perkins is “significantly\n\nsubaverage intellectual functioning (an IQ of 70 or below).” Perkins, 851 So. 2d at\n\n456. Jenkins argues that the use of a strict IQ cutoff of 70 was contrary to or an\n\nunreasonable application of Atkins.\n\n\ncapacity of the defendant to proceed or continue to trial or which would affect the defendant’s\ncriminal responsibility at the time of the commission of the crime.” Id.\n 43\n\n We disagree. The state court’s decision that Jenkins failed to establish\n\nmental retardation in this respect was a reasonable application of the record and of\n\nAtkins. As a threshold matter, we note that the state court did not say it was\n\napplying a strict cutoff of 70; it merely stated that Dr. Kirkland had measured\n\nJenkins’s IQ as 76. Jenkins, 972 So. 2d at 155. Under our deferential review, the\n\nquestion is whether the decision that Jenkins had not established the intellectual\n\nfunctioning aspect of intellectual disability was an unreasonable application of\n\nAtkins. The Supreme Court explained in Atkins that it was leaving “to the State[s]\n\nthe task of developing appropriate ways to enforce the constitutional restriction\n\nupon [their] execution of sentences.” Atkins, 536 U.S. at 317 (quoting Ford v.\n\nWainwright, 477 U.S. 399, 416–17 (1986)). In line with that invitation, our Circuit\n\nhas accepted Alabama’s judicial definition of intellectual disability as laid out in\n\nPerkins.14 See Thomas, 607 F.3d at 752–53. Jenkins objects, however, that the\n\nPerkins standard, which he characterizes as articulating a “strict cutoff” of 70,\n\nconflicts with more recent Supreme Court precedent that instead requires a\n\n“clinical” definition that goes beyond a single IQ score.\n\n In particular, Jenkins points to Hall v. Florida, in which the Supreme Court\n\nheld that when an offender’s IQ is near but greater than 70, courts must take into\n\n\n\n14\n Notwithstanding the urging of the Alabama Supreme Court, see Perkins, 851 So. 2d at 455\nn.1, the Alabama legislature has not enacted a statutory definition of intellectual disability.\n 44\n\naccount the IQ test’s standard error of measurement as well as non-IQ evidence of\n\nintellectual function. See Hall v. Florida, 572 U.S. 701, 723 (2014) (“This Court\n\nagrees with the medical experts that when a defendant’s IQ test score falls within\n\nthe test’s acknowledged and inherent margin of error, the defendant must be able\n\nto present additional evidence of intellectual disability, including testimony\n\nregarding adaptive deficits.”). But, as Jenkins concedes, our Circuit has\n\nspecifically held that Hall is not retroactive to cases on collateral review. See In re\n\nHenry, 757 F.3d 1151, 1161 (11th Cir. 2014) (explaining that Hall does not supply\n\na new substantive rule but rather “merely provides new procedures for ensuring\n\nthat States do not execute members of an already protected group”). We also note\n\nthat Alabama courts have found the Perkins standard to comport with Hall in that\n\nit does not prevent an offender from presenting the standard error of measurement\n\nor non-IQ evidence of intellectual disability. See Reeves v. State, 226 So. 3d 711,\n\n728–29 (Ala. Ct. Crim. App. 2016). Here, the state court did not unreasonably\n\napply Atkins when it analyzed the intellectual functioning component of\n\nintellectual disability.\n\n Jenkins also argues that the state court’s decision on this component in his\n\nRule 32 proceeding resulted from an unreasonable determination of the facts. He\n\nasserts that his IQ score of 76, when combined with the test’s standard error of\n\nmeasurement and considered alongside the Flynn effect (which we discuss below),\n\n\n 45\n\ndoes show that he has significantly subaverage intellectual function. For the\n\nreasons that follow, we disagree.\n\n Most fundamentally, neither of the clinicians who testified about Jenkins’s\n\nintellectual ability in the Rule 32 hearing opined that he was mentally retarded.\n\nThe closest thing Jenkins can point to is the testimony of Dr. Kirkland that his IQ\n\nis two standard deviations below the mean. But his entire testimony, in context, did\n\nnot establish intellectual disability. To the contrary: Dr. Kirkland testified that\n\nJenkins “scored in the range of borderline intellectual functioning which is\n\nbetween mild mental retardation and low average intellectual functioning.”\n\nFurthermore, Jenkins’s own expert Dr. Lisak fully reviewed Dr. Kirkland’s clinical\n\nfindings. He agreed with those findings, concluding that the IQ results show\n\n“borderline intelligence.” It is tremendously significant that no clinical assessment\n\nin the entire record of these proceedings has found that Jenkins has mental\n\nretardation or intellectual disability.\n\n Jenkins nonetheless argues that his IQ score of 76 establishes the intellectual\n\ncomponent of intellectual ability when the standard error of measurement and the\n\nFlynn effect are considered alongside that score. It was not unreasonable for the\n\nstate court to disagree. As we have discussed, Jenkins is not retroactively entitled\n\nto have the state courts take into account the margin of error of the IQ test. That\n\nstatistical margin of error—5 points in this case—takes into account characteristics\n\n\n 46\n\nof the test and variations due to chance in order to state how confident the tester is\n\nthat the measured result accurately states Jenkins’s true, unknowable IQ. A result\n\nof 76 with a standard error of measurement of 5 simply means that the tester is\n\n68% confident that Jenkins’s true IQ falls between 71 and 81, and 95% confident\n\nthat his true IQ falls between 66 and 86. See David H. Kaye & David A. Freedman,\n\nReference Guide on Statistics, in Reference Manual on Scientific Evidence 211,\n\n243–44 (3d ed. 2011); see also Hall, 572 U.S. at 738–39 (Alito, J., dissenting). Of\n\ncourse, Jenkins would like us to consider only the possibility that the test result\n\noverstated his IQ—but the standard error of measurement is a two-way street.\n\n“[T]he standard error of measurement is a bi-directional concept that does not\n\ncarry with it a presumption that an individual’s IQ falls to the bottom of his IQ\n\nrange.” Ledford v. Warden, Ga. Diag. & Clas. Prison, 818 F.3d 600, 641 (11th Cir.\n\n2016).\n\n Even if Jenkins were somehow entitled to subtract the 5-point margin of\n\nerror from his score and thereby lower his “true” IQ to 71, he still would not satisfy\n\nthe clinical intellectual component of intellectual disability. Thus, he sought in the\n\ndistrict court to introduce evidence about the Flynn effect in order to bring his IQ\n\ndown to 70 or below. As our Court has explained, the Flynn effect “recognizes the\n\nfact that IQ test scores have been increasing over time”:\n\n The Flynn effect acknowledges that as an intelligence test ages, or\n moves farther from the date on which it was standardized, or normed,\n 47\n\n the mean score of the population as a whole on that assessment\n instrument increases, thereby artificially inflating the IQ scores of\n individual test subjects. Therefore, the IQ test scores must be\n recalibrated to keep all test subjects on a level playing field.\n\nThomas, 607 F.3d at 753. We have said that “[a]n evaluator may also consider the\n\n‘Flynn effect’” in assessing an offender’s possible intellectual disability, id.\n\n(emphasis added), but neither we nor the Alabama Supreme Court nor the U.S.\n\nSupreme Court have said that a court must consider it in order to reasonably apply\n\nAtkins. See, e.g., Reeves, 226 So. 3d at 739 (“This Court has repeatedly held that a\n\ncircuit court is not required to accept, consider, or apply the ‘Flynn Effect’ in\n\ndetermining intellectual disability.”).\n\n More problematically, Jenkins did not raise the Flynn effect issue in the\n\nAlabama state courts. Jenkins asserts that he implicitly raised it as part and parcel\n\nof his Atkins claim during his Rule 32 appeal, but we do not agree that the\n\nAlabama courts had a full and fair opportunity to consider this issue, because it\n\nwas not explicitly presented to them. See Kelley v. Sec’y for the Dep’t of Corr., 377\n\nF.3d 1317, 1349–50 (11th Cir. 2004) (holding that one ineffective-assistance\n\nargument was unexhausted despite the petitioner’s general presentation of a\n\nStrickland claim in the state courts). Thus, under AEDPA, Jenkins is procedurally\n\nbarred from raising this unexhausted claim in his federal habeas petition. See 28\n\nU.S.C. § 2254(b)(1)(A), (c); Mason v. Allen, 605 F.3d 1114, 1119 (11th Cir. 2010).\n\nWe will allow a prisoner to overcome a procedural bar if he establishes cause for\n 48\n\nhis default and prejudice therefrom. See, e.g., Smith v. Jones, 256 F.3d 1135, 1138\n\n(11th Cir. 2001). But Jenkins cannot establish prejudice from being barred from\n\nasserting this claim now because the Supreme Court has not “clearly established”\n\nthat a court must consider the Flynn effect in assessing intellectual disability. Thus,\n\nunder AEDPA, it could not form the basis for a federal habeas claim. See 28\n\nU.S.C. § 2254(d)(1). We therefore conclude that the state court did not\n\nunreasonably determine the facts or unreasonably apply Atkins with respect to the\n\nintellectual component of intellectual disability.\n\n 2. Adaptive Function\n\n The second component of intellectual disability is “significant or substantial\n\ndeficits in adaptive behavior.” Perkins, 851 So. 2d at 456. Jenkins argues that the\n\nstate court, which mentioned only Jenkins’s successes in employment and social\n\nskills, see Jenkins, 972 So. 2d at 155, failed to consider his deficits. He adds that\n\neven the mention of employment success does not tell the whole story. Jenkins\n\ninsists that his jobs were menial, his employers took advantage of him, and he held\n\nthe jobs for only a few months each.\n\n But the state court’s terse discussion of Jenkins’s adaptive behavior did not\n\ninvolve an unreasonable determination of the facts or an unreasonable application\n\nof Atkins. To the contrary, its decision on this issue was reasonable in light of both\n\nthe record and Atkins.\n\n\n 49\n\n Atkins elaborated that the adaptive-skill component of intellectual disability\n\nrequires substantial present limitation in at least two of the following areas:\n\n“communication, self-care, home living, social skills, community use, self-\n\ndirection, health and safety, functional academics, leisure, and work.” Atkins, 536\n\nU.S. at 308 n.3 (quoting definition of American Association on Mental\n\nRetardation). The record supports the state court’s judgment that Jenkins does not\n\nhave substantial deficits in at least two areas. To be sure, we will assume that\n\nJenkins may be substantially limited in functional academics; Dr. Kirkland\n\nassessed his reading, spelling, and arithmetic skills at a third-grade level. But the\n\nrecord does not compel a finding of substantial present deficits in the areas of\n\ncommunication, self-care, home living, and work. Although Jenkins emphasizes\n\nhis transient employment, the record shows he was able to find work quickly, to\n\nwork hard to support the Seals’ household until Lonnie Seal found a job, and, later,\n\nto move into his own home. That record evidence is inconsistent with a substantial\n\ndeficit in the area of work.\n\n As the State observes, the facts of the crime also fail to show significant\n\ndeficits in the areas of communication, self-care, community use, and self-\n\ndirection. Jenkins was able to communicate well enough to solicit an alibi and to\n\nsell his car, including writing out a bill of sale. He then was able to arrange for his\n\nflight from Alabama, including purchasing a bus ticket and hitchhiking across the\n\n\n 50\n\ncountry. Indeed, throughout his trial and postconviction proceedings, the people\n\nwho knew Jenkins as an adult consistently described a man who did not have\n\nserious difficulties in communicating, forming relationships, working, and caring\n\nfor himself. No witness mentioned these kinds of difficulties in his or her\n\ntestimony about Jenkins. There was some evidence that Jenkins was “not\n\narticulate” and that his house was messy, but that testimony does not rise to the\n\nlevel of overwhelming evidence of substantial functional deficits that would\n\nsupport a conclusion that the state court unreasonably determined the facts.\n\n Of course, the record also contains evidence of Jenkins’s childhood\n\nacademic and social deficits, attributable at least in part to his parents’ neglect, but\n\nJenkins’s childhood is not directly relevant to our consideration of his present\n\nlimitations. Although we acknowledge that the Supreme Court has noted that such\n\n“risk factors” may support further exploration of the possibility of intellectual\n\ndisability, see Moore v. Texas, 137 S. Ct. 1039, 1051 (2017), the state court record\n\ncontains the results of that exploration. Overall, that record supports the state\n\ncourt’s conclusion that Jenkins does not have substantial deficits in adaptive\n\nbehaviors.\n\n 3. Juvenile Onset\n\n The third and final component of intellectual disability is manifestation of\n\nthe first and second components “before the defendant reached age 18.” Perkins,\n\n\n 51\n\n851 So. 2d at 456. Jenkins asserts, the State assumes, and we agree that the state\n\ncourt did not rule upon this issue, so there is nothing to which we must defer under\n\nAEDPA. See, e.g., Brumfield v. Cain, 135 S. Ct. 2269, 2282 (2015). Thus, we will\n\nreview the district court’s legal conclusions on this issue de novo, and its factual\n\nfindings for clear error. See, e.g., Grossman v. McDonough, 466 F.3d 1325, 1335\n\n(11th Cir. 2006). Jenkins argues that he has satisfied the age-of-onset component\n\nby showing “a wide range of evidence of deficits in both intellectual and adaptive\n\nfunctioning throughout Jenkins’s childhood.”\n\n In fact, the ample record of Jenkins’s childhood, as developed in the Rule 32\n\nproceeding, does not point to intellectual disability before age 18. School\n\nassessments at age 12 reported an IQ score of 83, an adaptive behavior rating of\n\n“low,” and noted overall an intellectual capacity of “average,” a diagnosis of\n\ndyslexia, and “poor functional skills.” The psychologist attributed Jenkins’s\n\ndifficulties to a combination of “[f]requent moves, changes in schools and family\n\ninstability” with “[l]ow ability coupled with a specific learning disability” and\n\n“[l]ack of responsibility for self control and poor use of time.” Although Jenkins\n\nfailed several grades of school, teachers additionally attributed his difficulties to\n\nhis behavioral problems and his high number of absences from school.\n\n Similarly, IQ tests ordered by the juvenile court at age 14 gave scores of 81\n\nand 86 on different tests, in the “[d]ull–[n]ormal” range and reflecting “a definite\n\n\n 52\n\nlearning disability.” These records all show a child with serious academic deficits\n\nand some intellectual and adaptive deficits, but they do not clearly show an\n\nintellectually disabled child. In particular, Jenkins’s childhood IQ scores of 83, 81,\n\nand 86 fail to point to intellectual disability or mental retardation. We are not left\n\nwith the definite and firm conviction that the district court erred when it found that\n\n“Jenkins’s [childhood] intelligence scores and adaptive functioning skills were\n\nattributable to his learning disability and other circumstances in his life, rather than\n\nmental retardation.”\n\n In sum, we agree with the district court that the decision of the Alabama\n\nCourt of Criminal Appeals that Jenkins does not have intellectual disability was\n\nnot contrary to or an unreasonable application of Atkins v. Virginia.\n\n 4. Evidentiary Hearing\n\n Finally, Jenkins argues that the district court erred when it denied his request\n\nfor an evidentiary hearing on his Atkins claim. If Jenkins diligently attempted to\n\npresent this claim in the state court, we review for abuse of discretion. Burgess v.\n\nComm’r, Ala. Dep’t of Corr., 723 F.3d 1308, 1320 (11th Cir. 2013). “The district\n\ncourt abuses its discretion where ‘such a hearing could enable an applicant to prove\n\nthe petition’s factual allegations, which, if true, would entitle the applicant to\n\nfederal habeas relief.’” Id. (quoting Schriro v. Landrigan, 550 U.S. 465, 474\n\n(2007)). If Jenkins did not diligently attempt to develop the factual basis for this\n\n\n 53\n\nclaim in the state court, absent certain circumstances, he is not entitled to a hearing\n\nand we will affirm the district court’s denial. 28 U.S.C. § 2254(e)(2).\n\n The parties strenuously dispute whether Jenkins diligently attempted to\n\npresent this issue in the state court. On the one hand, the State argues that Jenkins\n\nfailed to allege that he was mentally retarded even after the U.S. Supreme Court\n\ndecided Atkins in 2002. On the other hand, Jenkins asserts that he raised the\n\npossibility of mental retardation as early as his original Rule 32 petition in 1995.\n\n As we have noted, shortly after the Supreme Court decided Atkins, the\n\nAlabama Court of Criminal Appeals ordered supplemental briefing on the possible\n\nimpact of Atkins on Jenkins. Jenkins’s Rule 32 appeal was pending at that point,\n\nhaving been fully briefed and orally argued in the state appellate court. Having\n\nreviewed the state court records, we agree with the State that Jenkins did not\n\ndiligently attempt to present his Atkins claim in the state courts.\n\n “Diligence will require in the usual case that the prisoner, at a minimum,\n\nseek an evidentiary hearing in state court.” Williams, 529 U.S. at 437. The record\n\nclearly shows that Jenkins did not do so. In the initial supplemental brief on Atkins\n\nhe filed in 2002, Jenkins merely requested a future hearing after the Alabama\n\nlegislature’s enactment of a statute on executing the mentally retarded. Seventeen\n\nyears later, that contingency is still in the future. Only in his reply supplemental\n\nbrief did Jenkins then assert, for the first time ever, that he is mentally retarded.\n\n\n 54\n\n We acknowledge that Jenkins’s initial Rule 32 petition included in its\n\nallegations of ineffective assistance of counsel a list of unpresented mitigating\n\nevidence including his “developmental[] impair[ment]” and “learning disabilities,\n\nlow intelligence, poor comprehension and retarded socialization skills.” We\n\nconclude that these tangential Strickland references to his mental and social\n\nabilities do not amount to diligence. Thus, because Jenkins did not diligently\n\nattempt to develop the factual basis for his Atkins claim in the state courts, we\n\naffirm the district court’s denial of his request for an evidentiary hearing.\n\n Even if Jenkins had diligently attempted to present this claim in the state\n\ncourt, however, we conclude in the alternative that the district court did not abuse\n\nits discretion when it denied a hearing. We agree, in light of the ample evidence\n\nabout intellectual ability that Jenkins adduced in support of his ineffective-\n\nassistance claim, that such a hearing would not have enabled Jenkins to prove that\n\nhe is intellectually disabled. See Burgess, 723 F.3d at 1320.\n\n IV. CONCLUSION\n\n The denial of Jenkins’s petition for a writ of habeas corpus is\n\n AFFIRMED.\n\n\n\n\n 55\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Mark Jenkins v. Commissioner, Alabama Department of Corrections\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-30\\nOpinion type: majority\\nHas dissenting opinion: Yes\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\n[PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n\\n\\n No. 17-12524\\n\\n\\n\\n D.C. Docket No. 4:08-cv-00869-VEH\\n\\nMARK ALLEN JENKINS,\\n\\n Petitioner - Appellant,\\n\\n versus\\n\\n\\n\\nCOMMISSIONER, ALABAMA DEPARTMENT OF CORRECTIONS,\\n\\n Respondent - Appellee.\\n\\n\\n\\n Appeal from the United States District Court\\n for the Northern District of Alabama\\n\\n\\n (August 30, 2019)\\n\\nBefore TJOFLAT, WILSON, and BRANCH, Circuit Judges.\\n\\nBRANCH, Circuit Judge:\\n\\n Mark Allen Jenkins, an Alabama prisoner sentenced to death for the 1989\\n\\nmurder of Tammy Ruth Hogeland, appeals the district court’s denial of his petition\\n\\nfor a writ of habeas corpus, 28 U.S.C. § 2254. Before us are Jenkins’s arguments\\n\\nthat he received ineffective assistance of counsel during the penalty phase of his\\n\\ntrial and that he is intellectually disabled and therefore ineligible for the death\\n\\npenalty. After careful consideration, and with the benefit of oral argument, we\\n\\naffirm the denial of Jenkins’s habeas petition on both grounds.\\n\\n I. BACKGROUND\\n\\n A. The Crime and Arrest\\n\\n We summarize the following background narrative from opinions of the\\n\\nAlabama Court of Criminal Appeals and our own review of the record. See Jenkins\\n\\nv. State, 627 So. 2d 1034, 1037–40 (Ala. Ct. Crim. App. 1992); Jenkins v. State,\\n\\n972 So. 2d 111, 119–20 (Ala. Ct. Crim. App. 2004). The events leading up to the\\n\\nmurder began the evening of April 17, 1989. Jenkins was at the home of his\\n\\nacquaintance Christine Nicholas. He had met her several months earlier at the\\n\\nOmelet Shoppe restaurant where she worked. Jenkins was very intoxicated and\\n\\ntried to seduce her. When she resisted, Jenkins got “real mad” and asked her\\n\\nseveral times what she would do if someone came up from behind her and grabbed\\n\\nher.\\n\\n At around 1 a.m., Jenkins and Nicholas drove to the Riverchase Omelet\\n\\nShoppe. Jenkins went inside and talked with one of the waitresses, Frieda Vines.\\n\\nThe manager, Douglas Thrash, recognized Jenkins as a regular customer who knew\\n\\n\\n 2\\n\\nall of the waitresses. Thrash overheard part of Jenkins’s conversation with Vines\\n\\nand heard mention of the Omelet Shoppe location near the Birmingham Airport.\\n\\nJenkins and Nicholas then returned to Nicholas’s home. Around 2 a.m., Nicholas’s\\n\\nmother asked Jenkins to leave. He did so, falling down some steps and ramming\\n\\nhis car into another vehicle in the process.\\n\\n Meanwhile, 23-year-old Tammy Hogeland was making her way to work at\\n\\nthe Omelet Shoppe. Her sister Wendy, along with Tammy’s young son, picked\\n\\nTammy up from her college classes and drove her to the Riverchase Omelet\\n\\nShoppe a little after 9:30 p.m. Sometime after 10:00, Tammy was sent to work at\\n\\nthe airport Omelet Shoppe on Tenth Avenue, since that location was unexpectedly\\n\\nshorthanded. Tammy was wearing her watch, a necklace with the words “special\\n\\nsister,” her class ring, and her diamond engagement ring. Tammy was working as a\\n\\ncook that night and was wearing a blue apron, black pants, a white shirt, and white\\n\\nshoes. She and Sarah Harris were the only employees working at the airport\\n\\nOmelet Shoppe that night.\\n\\n At about 2 a.m. on April 18, Harris saw Jenkins—whom she did not know—\\n\\ndrive up in a red sports car. His arrival was memorable to Harris because the car\\n\\nnearly jumped the curb and crashed through the restaurant’s glass wall. Jenkins got\\n\\nout of the car and came into the restaurant, appearing to be intoxicated. He walked\\n\\nover to Hogeland and began talking to her. Harris later saw Jenkins and Hogeland\\n\\n\\n 3\\n\\ndrive away together in the red sports car. That was the last time anyone who knew\\n\\nHogeland ever saw her. Hogeland left behind her cigarettes, lighter, purse, and\\n\\npaycheck. She left without telling anyone, which she had never done before.\\n\\nAlthough Harris saw them drive off, she could not later say whether Hogeland left\\n\\nwith Jenkins willingly or was instead abducted.\\n\\n At around 5 a.m. that morning, Bobby and Geraldine Coe had stopped to\\n\\nbuy gas at a Chevron station on I-59 northeast of Birmingham when they saw\\n\\nJenkins drive up in the red sports car. The Coes noticed a female who appeared to\\n\\nbe “passed out” in the front passenger seat, but they could not say whether she was\\n\\nalive or dead. While Bobby Coe was pumping gas, Jenkins asked him for some\\n\\ncigarettes and said, “Looks like it’s been a long night and it looks like it’s going to\\n\\nbe a long day.” Jenkins then said “God bless you” before asking directions to I-\\n\\n459. Coe gave him directions and got back in his car. As Coe pulled out onto the\\n\\ninterstate, he saw Jenkins follow him in the red sports car. Coe then saw the car\\n\\nflash its lights, slow down, and pull to the side of the road between mile markers\\n\\n151 and 152.\\n\\n At 8 a.m., Wendy Hogeland learned that Tammy was not at the restaurant,\\n\\nand their mother called the police. Meanwhile, Jenkins went to the home of Steve\\n\\nMusser, who noticed that he was wearing the same clothes as the day before.\\n\\nJenkins told Musser that his car had been stolen the night before and asked him if\\n\\n\\n 4\\n\\nhe would say that he had been with him all night. Musser refused. Christine\\n\\nNicholas also saw Jenkins in a grocery store that morning. He was looking at a\\n\\nnewspaper, making a phone call, and attempting to sell his Buick. Nicholas loaned\\n\\nhim $4 for gas. At around 10 o’clock that morning, Jenkins sold his car to Michael\\n\\nBrooks, a mechanic at a local Chevron station. Jenkins had explained that he\\n\\nneeded the money so he could visit his sick mother in California. Another\\n\\nmechanic at the service station drove Jenkins to the Greyhound bus station later\\n\\nthat morning.\\n\\n Jenkins awoke the next day on the Greyhound bus in Houston and was\\n\\nejected from the bus because his fare was used up. He then hitchhiked from\\n\\nHouston to Phoenix, then to San Diego and Los Angeles. Jenkins was first\\n\\nidentified as a suspect in Hogeland’s disappearance by Officer Mike Weems of the\\n\\nHoover Police Department on April 19. Weems ate dinner at the Riverchase\\n\\nOmelet Shoppe just about every day and had talked with Hogeland just as often.\\n\\nHe also knew Jenkins from the restaurant. Weems learned of Hogeland’s\\n\\ndisappearance from the other waitresses. Remembering how Jenkins often talked to\\n\\nHogeland and passed her notes,1 he gave Jenkins’s name to the missing persons\\n\\ninvestigator. The investigator also learned that a red Mazda RX-7 sports car, which\\n\\n\\n\\n1\\n The jury was not permitted to hear, and the record does not reveal, what exactly Jenkins had\\nsaid to Hogeland that made Officer Weems suspect him in her disappearance.\\n 5\\n\\nhad been reported stolen from the service station where Jenkins worked, had been\\n\\nrecovered on I-459.\\n\\n The afternoon of April 21, a truck driver who had happened to stop on I-59\\n\\nnear mile marker 151 discovered Hogeland’s body off the side of the road. She was\\n\\nnaked, wearing only a watch, and was so badly decomposed that the body had to\\n\\nbe identified by dental records. From the fractured hyoid bone, it was determined\\n\\nthat she was manually strangled to death. Also found at the scene were Hogeland’s\\n\\napron, shoes, bra, panties, pants, and hair net, as well as some beer cans, a Mazda\\n\\nRX-7 owner’s manual, and other items later determined to have come from the red\\n\\nMazda. Her necklace and rings were never found.\\n\\n Alabama authorities issued a warrant for Jenkins’s arrest and also obtained a\\n\\nfederal fugitive warrant. An FBI agent arrested Jenkins on May 1 in front of his\\n\\nuncle’s house in Wilmington, California. Jenkins’s uncle later gave the police three\\n\\nbags of Jenkins’s personal effects. Fibers recovered from Hogeland’s body and\\n\\nclothing matched those of the Mazda, and hairs matching Hogeland’s were found\\n\\nin the Mazda. Similarly, fibers from Jenkins’s clothing placed him in the Mazda,\\n\\nand fibers of his jeans were found on Hogeland’s apron. A bootprint near\\n\\nHogeland’s body matched a combat boot from among the belongings Jenkins’s\\n\\nuncle gave to the police, and Jenkins’s roommate said the boots looked just like\\n\\nthose that Jenkins wore.\\n\\n\\n 6\\n\\n A business card like those that had been in the Mazda’s glove box,\\n\\nbelonging to an owner of the Mazda, was recovered from Jenkins’s wallet after his\\n\\narrest in California. Later, Jenkins’s cellmate in the Alabama jail reported that\\n\\nJenkins told him “he had done the crime” and was worried that the couple from the\\n\\ngas station would identify him or that the police would find his fingerprint on a\\n\\nbeer can at the scene. A St. Clair County grand jury indicted Jenkins for capital\\n\\nmurder.\\n\\n B. The Trial and Sentencing\\n\\n Jenkins was represented at trial in 1991 by attorneys Douglas Scofield and\\n\\nStan Downey. Scofield, the lead attorney, had been practicing criminal defense in\\n\\nBirmingham since 1984 but had never before tried a capital murder case. Jenkins’s\\n\\nlandlord and his grandmother had arranged the referral and representation before\\n\\nthe court appointed him. Scofield agreed to the appointment on the condition that a\\n\\nSt. Clair County lawyer also be appointed. That lawyer was Downey, who had\\n\\nmore experience than Scofield but whose capital experience was limited to one\\n\\ncase that had not gone to trial. The two attorneys had agreed that Downey would\\n\\nbe responsible for jury selection, but Scofield eventually took over during the voir\\n\\ndire. Per their agreement, Scofield was responsible for the guilt-phase trial,\\n\\nalthough Downey participated by interviewing Jenkins several times. Scofield also\\n\\nmet with Jenkins around a dozen times. Jenkins had told him about his background\\n\\n\\n 7\\n\\nand childhood, including the fact that he had had a difficult childhood and had\\n\\nlived on the streets since age 11.\\n\\n Throughout the weeklong trial, Scofield pursued an outright acquittal. In\\n\\nparticular, he challenged Harris’s identification of Jenkins as the man who drove\\n\\naway with Hogeland, and he challenged the medical examiner’s opinions about the\\n\\ntime and cause of Hogeland’s death. In his closing argument, Scofield took a wide-\\n\\nranging approach to undermining the State’s wholly circumstantial case against\\n\\nJenkins. He attacked the State’s timeline of the crime, questioning whether Jenkins\\n\\nin his intoxicated state would have had the time and the ability to leave Nicholas’s\\n\\nhouse, steal the Mazda, and abduct and kill Hogeland. He reminded the jury of the\\n\\nweakness of Harris’s identification of Jenkins and the Coes’ identification of the\\n\\nwoman in the car. He pointed out that no physical evidence definitively linked\\n\\nJenkins to the Mazda, to Hogeland’s body, or to her jewelry, which was never\\n\\nrecovered. And he argued that the evidence provided no plausible motive for\\n\\nJenkins to have kidnapped, robbed, or killed Hogeland. In sum, he urged the jury\\n\\nto find reasonable doubt and acquit Jenkins. At the very least, he argued, they\\n\\nshould find that the State failed to prove Jenkins’s intent to rob and kidnap\\n\\nHogeland, which were necessary to convict Jenkins of capital murder. He further\\n\\nargued that, if the jury thought Jenkins killed Hogeland as a crime of passion, they\\n\\ncould convict him only of manslaughter. The jury deliberated for three hours and\\n\\n\\n 8\\n\\nfifteen minutes before convicting Jenkins of capital murder, Ala. Code § 13A-5-\\n\\n40(a)(1) & (2), i.e., murder committed in the course of kidnapping and robbery.\\n\\n The penalty phase took place that same afternoon. A week before trial,\\n\\nScofield and Downey had discussed the penalty phase, and Scofield understood\\n\\nthat Downey would be responsible for handling the penalty phase. Nonetheless, the\\n\\nwitness who was present to testify on Jenkins’s behalf during the penalty phase\\n\\nwas better acquainted with Scofield, so Downey suggested that afternoon that\\n\\nScofield conduct the direct examination. Scofield urged Downey to interview the\\n\\nwitness quickly and prepare to examine him, and Downey did so.\\n\\n Downey handled the penalty presentation to the jury. The State did not\\n\\npresent any additional aggravation evidence, resting on the evidence presented\\n\\nduring the guilt phase. One witness, Lonnie Seal, testified on Jenkins’s behalf in\\n\\norder, in Downey’s words, to “reveal another side of Mark Jenkins to you that you\\n\\ndon’t know anything about.” Seal had met Jenkins in Fontana, California, in 1988\\n\\nwhen they both worked at the same garage. Seal and his wife became friends with\\n\\nJenkins, and he visited in their home. When the Seals decided to move to Alabama\\n\\na few months later, Jenkins agreed to help with their move and to live with the\\n\\nSeals in Alabama for a time. It took Lonnie Seal several weeks to find a job, but\\n\\nJenkins found a job in two days and contributed toward the family’s rent and\\n\\n\\n\\n\\n 9\\n\\ngroceries. Jenkins continued to visit the Seals after he moved into his own home,\\n\\nand he would always bring a small gift for and play with the Seals’ infant son.\\n\\n Downey’s closing argument to the jury requested a sentence of life without\\n\\nparole. Downey reminded the jurors that they had agreed during voir dire that\\n\\ndeath was not always the appropriate punishment for homicide. Quoting the Bible,2\\n\\nhe reminded the jury that the Old Testament permitted capital punishment only\\n\\nwhen supported by the testimony of two or three eyewitnesses, whereas in this case\\n\\nthere had been only circumstantial evidence. He asserted from Seal’s testimony\\n\\nthat “Mark had a side of him that was generous, that was giving, and that was\\n\\nkind,” and argued that the deterrent and retributive purposes of capital punishment\\n\\nwould not be served in this case.\\n\\n The jury deliberated for 50 minutes before returning its nonbinding\\n\\nrecommendation of death by a vote of 10 to 2. See Ala. Code § 13A-5-46(e) (1981)\\n\\n(amended 2017 to make jury’s verdict binding). After a separate hearing, the court\\n\\nfound the two statutory aggravating factors for which the State had argued: that the\\n\\nmurder was committed during the commission of a robbery, and that it was\\n\\ncommitted during the commission of a kidnapping. See Ala. Code § 13A-5-49(4).\\n\\n\\n\\n\\n2\\n “If anyone kills a person, the murderer shall be put to death on the evidence of witnesses. But\\nno person shall be put to death on the testimony of one witness.” Numbers 35:30 (ESV). Downey\\nalso paraphrased Deuteronomy 17:6 as “On the evidence of two witnesses or three witnesses, he\\nis [sic] shall be put to death. He shall not be put to death on the evidence of one witness.”\\n 10\\n\\nThe court also found two statutory mitigating factors: that Jenkins had no\\n\\nsignificant history of prior criminal activity, and his age at the time of the crime\\n\\n(21). See id. § 13A-5-51(1), (7). The court considered but rejected the statutory\\n\\nmitigation that, due to his consumption of alcohol, Jenkins’s capacity to appreciate\\n\\nthe criminality of his conduct or to conform his conduct to the requirements of law\\n\\nwas substantially impaired. See id. § 13A-5-51(6). The court found that the\\n\\naggravating circumstances outweighed the mitigating circumstances and, in\\n\\naccordance with the jury’s recommendation, imposed a sentence of death.\\n\\n C. The State and Federal Post-Conviction Proceedings\\n\\n On direct appeal, the Alabama Court of Criminal Appeals affirmed Jenkins’s\\n\\nconviction and death sentence. Jenkins v. State, 627 So. 2d 1034 (Ala. Ct. Crim.\\n\\nApp. 1992), aff’d, 627 So. 2d 1054 (Ala. 1993), cert. denied, 511 U.S. 1012\\n\\n(1994). In 1995, Jenkins filed his first petition for state postconviction relief, Ala.\\n\\nR. Crim. P. 32,3 alleging, among many other things, that he received ineffective\\n\\nassistance of counsel during his penalty phase. The Rule 32 trial court conducted a\\n\\nthree-day hearing at which evidence was presented about the performance of\\n\\nJenkins’s attorneys.\\n\\n\\n\\n\\n3\\n “Subject to the limitations of Rule 32.2, any defendant who has been convicted of a criminal\\noffense may institute a proceeding in the court of original conviction to secure appropriate relief\\non the ground that . . . [t]he constitution of the United States or of the State of Alabama requires\\na new trial, a new sentence proceeding, or other relief. . . .” Ala. R. Crim. P. 32.1.\\n 11\\n\\n Scofield testified about the representation that he and Downey provided. In\\n\\nrelevant part, Scofield asserted that he and Downey had “decided that [developing\\n\\nmitigation for the penalty phase] would be something that Stan [Downey] would\\n\\nhandle.” Downey later told Scofield that he hadn’t done any such investigation, but\\n\\nScofield admitted that, “[a]part from what he has told me, I don’t know what he\\n\\nhas done.” Scofield testified that Downey “didn’t appear prepared” for the penalty\\n\\nphase, based on Downey’s last-minute suggestion that Scofield examine Seal since\\n\\nhe had talked with him before.\\n\\n In preparation for the penalty phase, Scofield himself did not interview any\\n\\nmember of Jenkins’s family, but he did not know whether Downey interviewed\\n\\nany family members. Scofield did interview Jenkins’s landlord and some other\\n\\nacquaintances but did not do so specifically for the purpose of developing\\n\\nmitigation. He had met with Jenkins around a dozen times, and Jenkins had told\\n\\nhim his life story including his experience of abuse as a child. In hindsight,\\n\\nScofield said he had a “lack of appreciation of mitigating circumstances.” He\\n\\nrepeatedly denied in his Rule 32 testimony that his omissions during the penalty\\n\\nphase were strategic: “I didn’t understand what I was doing. . . . I just didn’t\\n\\nunderstand what I needed to do.”\\n\\n For reasons that remain unclear, Downey—who is since deceased—did not\\n\\ntestify or submit an affidavit in the Rule 32 proceeding. Scofield answered,\\n\\n\\n 12\\n\\n“Obviously not,” when asked whether he had personal knowledge of everything\\n\\nDowney may or may not have done to prepare for the penalty phase. The record of\\n\\nthe extent of Downey’s pretrial preparation is thus limited to Downey’s detailed\\n\\nfee declaration, which he submitted to the State of Alabama in documentation of\\n\\nthe compensation due him as appointed counsel. See Ala. Code § 15-12-21. In it,\\n\\nDowney reported that he spent 71 hours4 on out-of-court preparation for trial. The\\n\\nreport details 35 hours of interviews with Jenkins in 18 separate visits to the jail\\n\\nover the course of 17 months, as well as one substantive conversation with\\n\\nJenkins’s grandmother.5\\n\\n The Rule 32 trial court also heard testimony from four family members and\\n\\none friend of Jenkins’s. The family members presented for the first time a stark\\n\\ndepiction of Jenkins as an unloved and abused child in California. Jenkins’s older\\n\\nbrother Michael Jenkins testified that all of his siblings have different fathers and\\n\\nthat his parents often used drugs and fought. Michael testified that Jenkins was\\n\\nbeaten frequently as a child by his stepfather, if not every day then every other day\\n\\n\\n\\n4\\n In Jenkins’s state postconviction appeal, the Alabama Court of Criminal Appeals erroneously\\nstated that Downey’s fee declaration reports 171 hours of pretrial preparation. See Jenkins, 972\\nSo. 2d at 137. We discuss the effect of this discrepancy infra note 7.\\n5\\n The state appellate court also noted that Downey had moved for a continuance in October 1989\\nso that he could conduct “[f]urther discovery and investigation (including a possible trip to\\nCalifornia).” Id. We note that Scofield had orally suggested to the trial court in September 1989\\nthat more time might be needed with respect to “evidence about the defendant” and witnesses in\\nCalifornia, as well as possible psychological evaluation.\\n 13\\n\\nor every three days. Jenkins was beaten all over his body with a variety of\\n\\nimplements. Michael was beaten too. Jenkins’s stepfather would lock Jenkins in\\n\\nhis bedroom for a couple of hours at a time, sometimes without dinner. Some of\\n\\nthe beatings were prompted by Jenkins’s bedwetting or soiling. Michael testified\\n\\nthat Jenkins’s stepfather sometimes made Jenkins wear his soiled underwear on his\\n\\nhead and even forced him to eat his own feces a few times. Michael testified that\\n\\nJenkins’s bladder and bowel control problems began after a camping trip he took\\n\\nwith his stepfather and stepgrandfather.\\n\\n Michael testified that, as children, he and Jenkins were made to work on car\\n\\nengines late into the night and to shovel horse manure daily. The family moved\\n\\nfrequently and the children attended many different schools as a result. Michael\\n\\ntestified that he was Jenkins’s only friend as a child and that Jenkins did poorly in\\n\\nschool. Jenkins began running away from home as a young teenager and lived with\\n\\nvarious other family members or on the streets. Michael said that Jenkins was a\\n\\ngood brother to him and had asked him to testify at his trial, but he never heard\\n\\nfrom his attorneys or found out when the trial was. Michael admitted to having\\n\\ncognitive and memory problems due to a head injury years earlier.\\n\\n Jenkins’s cousin Tammy Pitts testified that, as an infant, Jenkins’s mother\\n\\nbeat and neglected him, leaving him dirty and in soiled diapers. His stepfather also\\n\\nbeat Jenkins daily until he left home at age 13, sometimes bruising him so badly\\n\\n\\n 14\\n\\nthat he was laid up in bed for days or weeks. Jenkins had problems with\\n\\nbedwetting and soiling, and his stepfather made him wear soiled clothes to school.\\n\\nJenkins would be locked in his filthy bedroom around the clock except to attend\\n\\nschool and do chores. He was usually not allowed to eat dinner with the rest of the\\n\\nfamily and subsisted on scraps and dog food. The entire family belittled Jenkins\\n\\nbecause his biological father was Hispanic and his complexion was therefore\\n\\ndarker. Jenkins’s parents drank a lot and used drugs every day. Pitts knew these\\n\\nthings because she had daily contact with the family and lived with them off and\\n\\non, and Jenkins lived with her for a time after he ran away from home. No one had\\n\\ncontacted her about testifying at Jenkins’s trial.\\n\\n Jenkins’s second cousin Betty DeLavega testified that her relationship with\\n\\nJenkins was like that of an aunt. She lived with the Jenkins family for five months\\n\\nwhen Jenkins was 10 or 11. She saw Jenkins’s stepfather slap Jenkins “[q]uite a\\n\\nfew times,” and once, he made Jenkins eat his own feces. DeLavega testified that\\n\\nJenkins’s parents used drugs nearly every day and withheld affection from Jenkins\\n\\nand Michael. No one contacted her about Jenkins’s trial; Jenkins’s grandmother\\n\\nhad asked her to testify in the Rule 32 hearing about any abuse she had witnessed.\\n\\n Jenkins’s grandmother Doris Wagoner testified that while Jenkins’s mother\\n\\nwas pregnant with him, her husband was in prison and she drank and took drugs.\\n\\nWhen Jenkins was born, his mother initially put him up for adoption, then changed\\n\\n\\n 15\\n\\nher mind weeks later. Jenkins was a lethargic infant and his mother cared more\\n\\nabout partying than about her children. Wagoner testified that she heard of\\n\\nJenkins’s mistreatment as a child but didn’t see it. She could tell from the way that\\n\\nJenkins feared his stepfather that he was being abused, and she knew he slept in a\\n\\nfilthy bed. After Jenkins was arrested for murder, Wagoner spoke with Scofield,\\n\\nwho asked for money relating to his representation of Jenkins. She did not\\n\\nremember whether she talked with Scofield about Jenkins’s background. She could\\n\\nnot remember why she didn’t testify at the trial; she might have been ill, but she\\n\\nwas probably busy.\\n\\n Sherry Seal, the wife of Lonnie Seal, who had testified in the penalty phase\\n\\nof the trial, testified that Jenkins was respectful and generous and that she trusted\\n\\nhim with her children. She related how he supported her family after moving to\\n\\nAlabama, giving them “pretty much” his entire paycheck. She said she would still\\n\\nfeel safe around Jenkins. Two jailers also testified that Jenkins was a respectful,\\n\\ncourteous, model inmate who never complained or caused any problems.\\n\\n Jenkins also introduced medical, school, and juvenile court records from his\\n\\nchildhood. The medical records document his premature birth and his mother’s\\n\\nintention to give him up for adoption. The school records reflect a struggling\\n\\nstudent who was repeatedly promoted to the next grade because of his age, despite\\n\\nserious academic difficulties, chronic truancy, and frequent school changes. The\\n\\n\\n 16\\n\\njuvenile court records document Jenkins’s placement in juvenile hall starting at age\\n\\n14, following repeated acts of grand theft auto and running away.\\n\\n The Rule 32 hearing concluded with the testimony of two psychologists\\n\\nabout Jenkins’s history of psychological trauma stemming from his childhood\\n\\nabuse. The defense expert, Dr. David Lisak, had met with Jenkins three times and\\n\\ninterviewed a dozen of his family members, family friends, and associates. Dr.\\n\\nLisak testified that the pervasive abuse and neglect in Jenkins’s childhood impaired\\n\\nhis ability to cope with trauma as an adult. He opined that Jenkins’s childhood\\n\\nabuse adversely affected his cognitive and emotional development and other\\n\\nabilities. He described how Jenkins had been neglected as an infant and sexually\\n\\nabused by his grandfather at age four, which led to a lifetime of bladder and bowel\\n\\ndysfunction. He also discussed the hypothesis that children who are abused are at\\n\\ngreater risk for perpetrating violence later in life, but he acknowledged there would\\n\\nhave been other causative factors in Jenkins’s case. Dr. Lisak described Jenkins’s\\n\\nintellectual capacity as “somewhat borderline,” though he did not perform any\\n\\ntesting of his own.\\n\\n The State’s expert, Dr. Karl Kirkland, had met with Jenkins for several hours\\n\\nand administered several tests. On the Wechsler Adult Intelligence Scale, Jenkins\\n\\n“scored in the range of borderline intellectual functioning which is between mild\\n\\nmental retardation and low average intellectual functioning” with an overall IQ of\\n\\n\\n 17\\n\\n76. Dr. Kirkland agreed that Jenkins had been “reared in a chaotic dysfunctional\\n\\nfamily” and generally agreed with Dr. Lisak’s assessment of the negative effects of\\n\\nchildhood trauma.\\n\\n The court considered all of this evidence before denying Rule 32 relief in\\n\\nfull. As relevant here, it found that some of the evidence about the severity of\\n\\nabuse that Jenkins suffered as a child was not credible. It further found that Jenkins\\n\\nsuffered no prejudice from any deficient performance by his penalty-phase counsel\\n\\nbecause the aggravating circumstances outweighed any additional mitigating\\n\\nevidence Downey might have adduced.\\n\\n The Alabama Court of Criminal Appeals affirmed the denial of Rule 32\\n\\nrelief. Jenkins v. State, 972 So. 2d 111 (Ala. Ct. Crim. App. 2004). In relevant part,\\n\\nthe appeals court rejected Jenkins’s ineffective assistance claim, finding that\\n\\ncounsel reasonably chose to pursue a penalty-phase strategy of residual doubt and\\n\\ngood character. Id. at 147. It explained that the new evidence was self-\\n\\ncontradictory in that it asserted that childhood abuse would have made Jenkins a\\n\\nviolent adult yet also asserted that Jenkins was meek and mild. Id. The appeals\\n\\ncourt also agreed with the lower court’s conclusion that Jenkins suffered no\\n\\nprejudice from any error by his counsel because the aggravating circumstances still\\n\\noutweighed any new mitigating ones. Id. at 148. The court also noted that evidence\\n\\n\\n\\n\\n 18\\n\\nabout Jenkins’s good behavior in jail awaiting trial was “minimally mitigating” and\\n\\nwould not have affected the sentence. Id. at 149.\\n\\n Because the Supreme Court’s decision in Atkins v. Virginia6 had issued\\n\\nduring the pendency of the Rule 32 appeal, the Alabama appeals court ordered\\n\\nsupplemental briefing on the possible impact of Atkins upon Jenkins’s sentence.\\n\\nJenkins asked the court to stay the proceedings and conduct an evidentiary hearing\\n\\nafter the Alabama legislature enacted a statute about executing the mentally\\n\\nretarded or, in the alternative, to vacate his sentence and remand after the Alabama\\n\\nlegislature acted. He later asserted that he is mentally retarded and therefore\\n\\nineligible for the death penalty. The appeals court proceeded to reject Jenkins’s\\n\\nAtkins claim on the merits because the record established that his IQ was 76 and\\n\\nthat he maintained relationships and employment. Id. at 155. The Alabama\\n\\nSupreme Court summarily affirmed, in relevant part. Ex parte Jenkins, 972 So. 2d\\n\\n159, 165 (Ala. 2005), remanded to 972 So. 2d 165 (Ala. Ct. Crim. App. 2005)\\n\\n(juror misconduct claim), cert. denied, 552 U.S. 1167 (2008).\\n\\n In 2008, Jenkins filed his first petition for a writ of habeas corpus in federal\\n\\ncourt pursuant to 28 U.S.C. § 2254. The federal proceedings were stayed for\\n\\nseveral years while Jenkins pursued additional, unrelated Rule 32 relief in the state\\n\\n\\n6\\n 536 U.S. 304, 321 (2002) (holding that the execution of mentally retarded offenders violates the\\nEighth Amendment). The rule announced in Atkins is retroactive to cases on collateral review. In\\nre Holladay, 331 F.3d 1169, 1172–73 (11th Cir. 2003).\\n 19\\n\\ncourts, ultimately unsuccessfully. In the district court, Jenkins moved for an\\n\\nevidentiary hearing on his Atkins claim. After reviewing the state court record,\\n\\nwith a focus on the Rule 32 testimony of Drs. Lisak and Kirkland, the district court\\n\\ndenied the motion in 2015. Noting that our Circuit has accepted the State of\\n\\nAlabama’s definition of mental retardation following Atkins, the court found that\\n\\nJenkins is not intellectually disabled under that standard. In a 347-page order, the\\n\\ndistrict court also denied relief on all of Jenkins’s claims including ineffective\\n\\nassistance of counsel during the penalty phase and affirmed as reasonable the state\\n\\ncourt’s conclusion that Jenkins failed to establish prejudice from any error by his\\n\\ncounsel. The district court later denied reconsideration and a certificate of\\n\\nappealability.\\n\\n Jenkins now appeals. Our Court granted Jenkins a certificate of\\n\\nappealability, 28 U.S.C. § 2253(c), on two issues:\\n\\n 1) Whether the district court erred in denying Appellant’s claim\\n that his trial counsel rendered ineffective assistance of counsel in the\\n investigation and presentation of mitigating evidence during the\\n penalty phase of Appellant’s 1991 trial; and\\n\\n 2) Whether the district court erred in denying Appellant’s claim\\n that he is intellectually disabled and ineligible for the death penalty\\n under Atkins v. Virginia, 536 U.S. 304 (2002).\\n\\n II. STANDARDS OF REVIEW\\n\\n We review the denial of a petition for a writ of habeas corpus de novo.\\n\\nMorrow v. Warden, Ga. Diagnostic Prison, 886 F.3d 1138, 1146 (11th Cir. 2018).\\n\\n 20\\n\\nThe Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) instructs that\\n\\na writ of habeas corpus shall not be granted to a state prisoner unless the state\\n\\nadjudication of a claim “resulted in a decision that was contrary to, or involved an\\n\\nunreasonable application of, clearly established Federal law, as determined by the\\n\\nSupreme Court of the United States.” 28 U.S.C. § 2254(d)(1). “[C]learly\\n\\nestablished Federal law” means “the holdings, as opposed to the dicta, of [the\\n\\nSupreme] Court’s decisions as of the time of the relevant state-court decision.”\\n\\nWilliams v. Taylor, 529 U.S. 362, 412 (2000). “[W]hen the last state court to\\n\\ndecide a prisoner’s federal claim explains its decision on the merits in a reasoned\\n\\nopinion . . . a federal habeas court simply reviews the specific reasons given by the\\n\\nstate court and defers to those reasons if they are reasonable.” Wilson v. Sellers,\\n\\n138 S. Ct. 1188, 1192 (2018). A state court’s decision is reasonable “so long as\\n\\n‘fairminded jurists could disagree’ on the correctness of the state court’s decision.”\\n\\nHarrington v. Richter, 562 U.S. 86, 101 (2011) (quoting Yarborough v. Alvarado,\\n\\n541 U.S. 652, 664 (2004)). “[E]ven a strong case for relief does not mean the state\\n\\ncourt’s contrary conclusion was unreasonable.” Id. at 102. Rather, a prisoner must\\n\\nshow that the state court’s ruling “was so lacking in justification that there was an\\n\\nerror well understood and comprehended in existing law beyond any possibility for\\n\\nfairminded disagreement.” Id. at 103. That ruling must have been “‘objectively\\n\\nunreasonable,’ not merely wrong; even ‘clear error’ will not suffice.” White v.\\n\\n\\n 21\\n\\nWoodall, 572 U.S. 415, 419 (2014) (quoting Lockyer v. Andrade, 538 U.S. 63, 75\\n\\n(2003)).\\n\\n In addition, “a determination of a factual issue made by a State court shall be\\n\\npresumed to be correct,” and the prisoner bears “the burden of rebutting the\\n\\npresumption of correctness by clear and convincing evidence.” 28 U.S.C.\\n\\n§ 2254(e)(1). AEDPA further instructs that a writ of habeas corpus may not be\\n\\ngranted unless the state court adjudication “resulted in a decision that was based on\\n\\nan unreasonable determination of the facts in light of the evidence presented in the\\n\\nState court proceeding.” Id. § 2254(d)(2). When a federal habeas court makes a\\n\\nfactual finding as part of its habeas determination, we review for clear error, which\\n\\nis a “highly deferential” standard of review. Thomas v. Allen, 607 F.3d 749, 752\\n\\n(11th Cir. 2010) (quoting Holton v. City of Thomasville Sch. Dist., 425 F.3d 1325,\\n\\n1350 (11th Cir. 2005)). “A finding is ‘clearly erroneous’ when although there is\\n\\nevidence to support it, the reviewing court on the entire evidence is left with the\\n\\ndefinite and firm conviction that a mistake has been committed.” United States v.\\n\\nU.S. Gypsum Co., 333 U.S. 364, 395 (1948).\\n\\n III. DISCUSSION\\n\\n A. Ineffective Assistance of Counsel\\n\\n A prisoner alleging that he received ineffective assistance of counsel in\\n\\nviolation of the Sixth Amendment must establish two elements. Strickland v.\\n\\n\\n 22\\n\\nWashington, 466 U.S. 668, 687 (1984). “First, the defendant must show that\\n\\ncounsel’s performance was deficient.” Id. In other words, he must show “that\\n\\ncounsel made errors so serious that counsel was not functioning as the ‘counsel’\\n\\nguaranteed the defendant by the Sixth Amendment.” Id. “Second, the defendant\\n\\nmust show that the deficient performance prejudiced the defense.” Id. We discuss\\n\\neach of these elements in turn.\\n\\n 1. Deficient Performance\\n\\n A “doubly” deferential lens sits before our eyes as we review the\\n\\nperformance of Jenkins’s counsel during the penalty phase. Richter, 562 U.S. at\\n\\n105 (quoting Knowles v. Mirzayance, 556 U.S. 111, 123 (2009)). This redoubled\\n\\ndeference arises out of the tandem application of AEDPA’s “highly deferential”\\n\\nreview of a state court’s decision with Strickland’s “highly deferential” review of\\n\\nan attorney’s performance. Id. (quoting Lindh v. Murphy, 521 U.S. 320, 333 n.7\\n\\n(1997), and Strickland, 466 U.S. at 689). Under the Strickland standard, “counsel\\n\\nis strongly presumed to have rendered adequate assistance and made all significant\\n\\ndecisions in the exercise of reasonable professional judgment.” Strickland, 466\\n\\nU.S. at 690. And under AEPDA, we must deny habeas relief if “there is any\\n\\nreasonable argument that counsel satisfied Strickland’s deferential standard.”\\n\\nRichter, 562 U.S. at 105.\\n\\n\\n\\n\\n 23\\n\\n Jenkins argues that his counsel performed deficiently by failing to\\n\\n“investigate compelling mitigating evidence or develop a penalty-phase theory.”\\n\\nThe state court disagreed. It quoted decisions of our Court that hold that “there is\\n\\nno per se rule that evidence of a criminal defendant’s troubled childhood must\\n\\nalways be presented as mitigating evidence.” Jenkins, 972 So. 2d at 144 (quoting\\n\\nMarek v. Singletary, 62 F.3d 1295, 1300 (11th Cir. 1995)). Noting that “Downey\\n\\nwas in charge of the penalty phase” but “we do not have the benefit of Downey’s\\n\\ntestimony as to what occurred and why,” it reviewed the trial record to find that\\n\\n“counsel argued residual doubt and Jenkins’s good character at the penalty phase.”\\n\\nId. at 145–46. It concluded that “Downey’s decision to concentrate on reasonable\\n\\ndoubt and to portray Jenkins as a good person was reasonable under the\\n\\ncircumstances.” Id. at 147.\\n\\n Jenkins insists, however, that counsel’s failure to develop mitigating\\n\\nevidence about his childhood abuse was “not the product of strategy, but rather\\n\\narose from ingnorance and inexeperiance [sic].” The problem for Jenkins is that,\\n\\nlike the state courts, this Court also does not know what Downey did or why. The\\n\\nrecord is silent as to his thoughts and intentions as he prepared for the penalty\\n\\nphase. Given that Jenkins bore the burden of overcoming the strong presumption of\\n\\ncompetence and of proving Downey’s deficient performance, we are perplexed\\n\\nwhy Jenkins’s new counsel did not have Downey testify or submit an affidavit in\\n\\n\\n 24\\n\\nthe Rule 32 proceeding. See Strickland, 466 U.S. at 687 (“The defendant must\\n\\nshow that counsel’s performance was deficient.”). Nor did Jenkins testify to the\\n\\nsubstance of his conversations with his counsel. “The reasonableness of a trial\\n\\ncounsel’s acts, including lack of investigation or excluding character witnesses\\n\\nfrom the sentencing phase, depends ‘critically’ upon what information the client\\n\\ncommunicated to counsel.” Chandler v. United States, 218 F.3d 1305, 1324 (11th\\n\\nCir. 2000) (en banc) (quoting Strickland, 466 U.S. at 691). There is simply no\\n\\nrecord evidence to support Jenkins’s central assertion that Downey made no\\n\\ntactical choices in his preparation for the penalty phase.\\n\\n Jenkins instead relies on Scofield’s Rule 32 testimony about how he and\\n\\nDowney prepared for the trial and sentencing. We acknowledge that Scofield\\n\\nrepeatedly testified that he did not really know what he was doing and that he\\n\\nfailed to make strategic choices with respect to mitigating circumstances and\\n\\nsentencing. But Scofield was equally clear that Downey—the more experienced\\n\\nlawyer—was in fact responsible for the penalty phase. And Scofield also\\n\\nrepeatedly testified that he did not know what Downey may or may not have done\\n\\nto prepare. Furthermore, his lack of knowledge about Downey’s preparation was\\n\\nunderstandable. He and Downey were not colleagues who shared an office and\\n\\nmight reasonably be expected to know what the other was doing much of the time.\\n\\nScofield and Downey did not even work in the same city.\\n\\n\\n 25\\n\\n This case is unlike those in which we have found deficient performance\\n\\nwhere the record showed that counsel investigated insufficiently because he\\n\\n“waited until the eleventh hour” to begin preparing for the penalty phase while\\n\\nfaced with “overwhelming evidence of guilt.” See, e.g., Johnson v. Sec’y, D.O.C.,\\n\\n643 F.3d 907, 932 (11th Cir. 2011). The limited record that Jenkins developed\\n\\nestablishes that Downey spent at least 71 hours preparing for trial, spread out over\\n\\n17 months. The state court interpreted this sparse record as reflecting a penalty-\\n\\nphase strategy of residual doubt, the logical extension of Downey’s guilt-phase\\n\\ninvestigation of the evidence and Scofield’s pursuit at trial of an acquittal based on\\n\\nreasonable doubt. See Jenkins, 972 So. 2d at 147. In such a situation, “[a] lawyer’s\\n\\ntime and effort in preparing to defend his client in the guilt phase of a capital case\\n\\ncontinues to count at the sentencing phase.” Tarver v. Hopper, 169 F.3d 710, 715\\n\\n(11th Cir. 1999).\\n\\n We decline to dictate how many hours or how many sessions an attorney\\n\\nmust spend interviewing his counsel in order to provide effective assistance. “No\\n\\nabsolute rules dictate what is reasonable performance for lawyers.” Chandler, 218\\n\\nF.3d at 1317 (citing Strickland, 466 U.S. at 688–89). But even this limited record\\n\\nundermines Jenkins’s assertion that Downey prepared inadequately. Downey’s 71\\n\\nhours of preparation included 35 hours of jailhouse conversations with Jenkins\\n\\n\\n\\n\\n 26\\n\\nduring 18 separate meetings.7 According to Downey’s notes, Scofield was present\\n\\nfor only five of those meetings (12 hours’ worth), leaving 23 hours over 13\\n\\nmeetings between Jenkins and Downey alone. Downey’s notes reveal little about\\n\\nthe content of those conversations beyond broad notations such as “evidence,\\n\\nwitnesses, general status,” and discussions with Jenkins of the evidence and\\n\\ntestimony against him were surely relevant to Downey’s penalty-phase strategy of\\n\\nresidual doubt.\\n\\n But the record does not reveal that Downey knew or should have known that\\n\\nJenkins’s childhood merited investigation beyond that which he may have done.\\n\\nThe record shows that Downey developed enough of a rapport with Jenkins that,\\n\\naccording to his fee declaration, he “g[ave a] Christmas gift [and] discuss[ed] how\\n\\nmuch to tell his family.” But how much Downey knew about Jenkins’s family\\n\\nbackground can only be guessed from Scofield’s Rule 32 hearing testimony.\\n\\nScofield testified that he discussed Jenkins’s background during his dozen or so\\n\\njailhouse meetings with him, and that Downey was present for some of those\\n\\nmeetings. Scofield—and possibly Downey—learned that Jenkins experienced\\n\\nabuse as a child, that his stepfather was brutal, and that he lived on the streets and\\n\\n\\n\\n\\n7\\n We note that the state court clearly erroneously stated that Downey’s fee declaration\\ndocumented 171 hours of pretrial preparation. See Jenkins, 972 So. 2d at 137. However, the state\\ncourt accurately stated that “he spent over 25 hours talking with Jenkins in more than 10 visits to\\nthe jail.” Id. That statement was not an unreasonable determination of the facts.\\n 27\\n\\nin juvenile institutions because of his brutal home life. It is against this backdrop\\n\\nthat Downey’s presumptive “decision not to investigate must be directly assessed\\n\\nfor reasonableness in all the circumstances, applying a heavy measure of deference\\n\\nto counsel’s judgments.” Strickland, 466 U.S. at 691.\\n\\n That assessment depends a great deal on what Jenkins told his counsel about\\n\\nhis background. See id. But we do not know whether Downey was present when\\n\\nScofield learned about Jenkins’s childhood abuse, nor what else Downey may have\\n\\nlearned during his 13 solo discussions with Jenkins. Therefore, “given the absence\\n\\nof evidence in the record, we must assume counsel carried out his professional\\n\\nresponsibility and discussed mitigation with his client.” Chandler, 218 F.3d at\\n\\n1324. We will not “turn that presumption on its head by giving [Jenkins] the\\n\\nbenefit of the doubt where it is unclear what [Downey] did or did not do.” Williams\\n\\nv. Head, 185 F.3d 1223, 1235 (11th Cir. 1999). We will not assume that Jenkins\\n\\ntold Downey more than he told Scofield, and Scofield denied remembering that\\n\\nJenkins told him about the severity or frequency of his stepfather’s physical abuse.\\n\\nHe denied remembering Jenkins telling him about being beaten daily, being locked\\n\\nin his room 24/7, or being forced to eat his feces. And there is no suggestion in the\\n\\nrecord that counsel knew or had reason to investigate the allegation of sexual abuse\\n\\nby his stepgrandfather. The record simply does not reflect that Jenkins told either\\n\\nof his lawyers about the severe abuse that some of the Rule 32 witnesses later\\n\\n\\n 28\\n\\ndescribed. We will not fault those lawyers for failing to present as mitigation this\\n\\nlimited evidence that they may have reasonably believed to be unremarkable. See\\n\\nStrickland, 466 U.S. at 691.\\n\\n To be sure, counsel’s failure “to discover all reasonably available mitigating\\n\\nevidence” may constitute deficient performance in some cases. Wiggins v. Smith,\\n\\n539 U.S. 510, 524 (2003) (quoting ABA Guidelines for the Appointment and\\n\\nPerformance of Counsel in Death Penalty Cases § 11.4.1(C) (1989)). Our Court\\n\\nhas identified deficient performance in several instances where counsel had good\\n\\nreason to investigate the defendant’s background but failed to do so. See Morrow,\\n\\n886 F.3d at 1147 (collecting cases). But we assess “a particular decision not to\\n\\ninvestigate . . . for reasonableness in all the circumstances, applying a heavy\\n\\nmeasure of deference to counsel’s judgments.” Strickland, 466 U.S. at 691. Thus,\\n\\nwe have rejected a hard rule that counsel’s “strategic decisions can be considered\\n\\nreasonable only if they are preceded by a ‘thorough investigation.’” Williams, 185\\n\\nF.3d at 1237 (quoting Rogers v. Zant, 13 F.3d 384, 387 (11th Cir. 1994)). Because\\n\\nwe simply do not know why Downey chose not to investigate Jenkins’s childhood\\n\\nmore thoroughly, Downey is entitled to the presumption that this strategic decision\\n\\nwas reasonable.\\n\\n And though we have also rejected “a per se rule of ineffective assistance\\n\\nwhere counsel does not consult family members,” Holladay v. Haley, 209 F.3d\\n\\n\\n 29\\n\\n1243, 1252 (11th Cir. 2000), we note that Downey did not fail altogether to consult\\n\\nfamily members. Downey had at least one substantive conversation with Jenkins’s\\n\\ngrandmother, and we know from her Rule 32 hearing testimony that she never\\n\\nwitnessed Jenkins’s childhood abuse. Given the silence of the record and the strong\\n\\npresumption of competence to which Downey is entitled, we decline to fault him\\n\\nfor not doing more when Jenkins and his grandmother did not and could not,\\n\\nrespectively, provide a basis for doing so. Cf. United States v. Pease, 240 F.3d 938,\\n\\n941–92 (11th Cir. 2001) (“we cannot say as a matter of law that reliance on a\\n\\nclient’s statements is per se deficient performance”). Nor did Jenkins’s counsel\\n\\nhave an absolute duty to put on abusive-childhood evidence, even if they knew of\\n\\nit. See Chandler, 218 F.3d at 1319 (“Counsel is not required to present every\\n\\nnonfrivolous defense; nor is counsel required to present all mitigation evidence,\\n\\neven if the additional mitigation evidence would not have been incompatible with\\n\\ncounsel’s strategy.”); Waters v. Thomas, 46 F.3d 1506, 1511 (11th Cir. 1995) (en\\n\\nbanc) (“Our decisions are inconsistent with any notion that counsel must present\\n\\nall available mitigating circumstance evidence.”).\\n\\n Instead, we affirm as a reasonable application of Strickland the state court’s\\n\\nconclusion that Downey’s strategic decision to emphasize residual doubt was\\n\\nreasonable. See Jenkins, 972 So. 2d at 147. As we have noted, “[a] lawyer’s time\\n\\nand effort in preparing to defend his client in the guilt phase of a capital case\\n\\n\\n 30\\n\\ncontinues to count at the sentencing phase.” Tarver, 169 F.3d at 715. This\\n\\nreasoning is especially applicable where, as here, counsel pursued the reasonable\\n\\nstrategy of arguing for acquittal at the guilt phase and for life at the sentencing\\n\\nphase by highlighting doubt about the evidence. “[R]esidual doubt is perhaps the\\n\\nmost effective strategy to employ at sentencing.” Chandler, 218 F.3d at 1320 n.28.\\n\\n“[E]specially when—as in this case—the evidence of guilt was not overwhelming,\\n\\nwe expect that petitioners can rarely (if ever) prove a lawyer to be ineffective for\\n\\nrelying on this seemingly reasonable strategy to defend his client.” Id. at 1320\\n\\n(footnote omitted); see also Tarver, 169 F.3d at 715–16 (collecting law review\\n\\narticles on the merits of residual-doubt strategy). Not only can such a strategy be\\n\\nconstitutionally adequate, in some cases it may be “the most effective performance\\n\\nin defense to the death penalty.” Tarver, 169 F.3d at 716.\\n\\n We cannot say that Downey’s manifest strategy of residual doubt was\\n\\nobjectively unreasonable under the circumstances. Jenkins’s case is precisely the\\n\\nsort where reminding the jury of the misgivings it may have held about the State’s\\n\\ncase could have been effective. Scofield’s guilt-phase closing argument from that\\n\\nvery morning, which explicitly emphasized doubt and the weakness of the State’s\\n\\nevidence, would have been fresh in the minds of the jury.8 Although we are not\\n\\n\\n8\\n During the penalty phase of a capital case, “[e]vidence presented at the trial of the case may be\\nconsidered insofar as it is relevant to the aggravating and mitigating circumstances without the\\nnecessity of re-introducing that evidence at the sentence hearing.” Ala. Code § 13A-5-45(c).\\n 31\\n\\nquestioning the jury’s guilty verdict or the sufficiency of the evidence, which are\\n\\nbeyond the scope of our habeas review, we acknowledge that the State’s case\\n\\nagainst Jenkins was—as Downey reminded the jury during the penalty phase—\\n\\nentirely circumstantial. Indeed, Scofield maintained post-conviction that the case\\n\\nfor reasonable doubt was “[v]ery strong.” No one saw Jenkins commit either the\\n\\nmurder or the robbery and kidnapping that elevated Hogeland’s murder to a capital\\n\\ncrime. Scofield’s guilt-phase closing argument challenged the inferential leaps the\\n\\nState was asking the jury to make in finding that Jenkins committed these capital-\\n\\npredicate offenses and that he formed the intent to kill. This situation was not one\\n\\nwhere a guilt-phase strategy of acquittal—and therefore a penalty-phase strategy of\\n\\nresidual doubt—was unreasonable in light of the evidence. Cf. Johnson, 643 F.3d\\n\\nat 932–33 (deficient sentencing investigation where defendant had confessed to\\n\\ntwo murders).\\n\\n Thus, although Downey’s penalty-phase closing argument did not\\n\\nspecifically invoke lingering doubt, it clearly reminded the jury of the entirely\\n\\ncircumstantial nature of the State’s guilt-phase case. See Chandler, 218 F.3d at\\n\\n1320 & n.6 (a strategy of “focus[ing] on obtaining an acquittal and then, at\\n\\nsentencing, on lingering doubt” is reasonable, even if counsel does not use the\\n\\nwords “lingering doubt” or “residual doubt”). Downey closed this reminder by\\n\\nexplaining to the jury how a foundational ancient legal system proscribed capital\\n\\n\\n 32\\n\\npunishment in the absence of eyewitness testimony—the exact situation here. The\\n\\nstate court found that Downey’s penalty-phase closing argument reflected a\\n\\nstrategic “decision to concentrate on reasonable doubt,” see Jenkins, 972 So. 2d at\\n\\n147, a tactical choice entitled to a “strong presumption of correctness,” Marek, 62\\n\\nF.3d at 1300 & n.3.\\n\\n Jenkins suggests other theories of mitigation that Downey might have more\\n\\neffectively pursued,9 but our deferential review under Strickland does not ask\\n\\nwhether counsel could possibly or ideally have been more effective. “The test for\\n\\nineffectiveness is not whether counsel could have done more.” Waters, 46 F.3d at\\n\\n1518. We do not ask whether an attorney’s representation “deviated from best\\n\\npractices or common custom,” and we should resist the temptation to second-guess\\n\\nan attorney with the benefit of our hindsight. Richter, 562 U.S. at 105. We ask only\\n\\n“whether, in light of all the circumstances, the identified acts or omissions were\\n\\n\\n\\n\\n9\\n In particular, Jenkins argues that his counsel should have presented “model prisoner” evidence.\\nThe state court noted, “Good conduct during pretrial incarceration is not necessarily a mitigating\\ncircumstance,” before finding that Jenkins had not met his burden of establishing that his counsel\\neven had information about his conduct in jail. Jenkins, 972 So. 2d at 149. We agree, and we\\ndecline to impose a requirement of sua sponte investigation of this kind in every capital case. Cf.\\nWilliams, 529 U.S. at 396 (finding deficient performance based in part on not returning a call\\nfrom a prison ministry volunteer who offered to vouch for defendant’s good behavior in prison).\\n“Surely, counsel is not required to call a witness to testify to facts such as lack of violent nature\\nwhen the jury has rejected such an approach and has found that the defendant is guilty of\\nmurder.” Griffin v. Wainwright, 760 F.2d 1505, 1512 (11th Cir. 1985), vacated on other grounds,\\n476 U.S. 1112 (1986), reaff’d, 874 F.2d 1397 (11th Cir. 1989).\\n 33\\n\\noutside the range of professionally competent assistance.” Strickland, 466 U.S. at\\n\\n690.\\n\\n We conclude that the state court did not unreasonably determine that Jenkins\\n\\nfailed to establish objectively incompetent performance by his counsel during the\\n\\npenalty phase of his trial. Our inquiry is not “what the best lawyers would have\\n\\ndone. Nor is the test even what most good lawyers would have done. We ask only\\n\\nwhether some reasonable lawyer at the trial could have acted, in the circumstances,\\n\\nas defense counsel acted at trial. . . . We are not interested in grading lawyers’\\n\\nperformances.” White v. Singletary, 972 F.2d 1218, 1220–21 (11th Cir. 1992). And\\n\\n“[i]f fairminded jurists could reasonably disagree” about the reasonableness of\\n\\ncounsel’s performance, “then habeas relief is due to be denied.” Johnson, 643 F.3d\\n\\nat 932.\\n\\n 2. Prejudice\\n\\n Even if Jenkins had demonstrated that his counsel performed as no\\n\\nreasonable lawyer could have, he must also demonstrate a “reasonable probability\\n\\nthat, but for counsel’s unprofessional errors, the result of the proceeding would\\n\\nhave been different.” Strickland, 466 U.S. at 694. Given that the jury here\\n\\nrecommended a sentence of death by the narrowest possible vote, 10 to 2, Jenkins\\n\\nneed establish only “a reasonable probability that at least one juror would have\\n\\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537. “A\\n\\n\\n 34\\n\\nreasonable probability is a probability sufficient to undermine confidence in the\\n\\noutcome.” Strickland, 466 U.S. at 694. In measuring that probability here, we will\\n\\n“evaluate the totality of the available mitigation evidence—both that adduced at\\n\\ntrial, and the evidence adduced in the habeas proceeding in reweighing it against\\n\\nthe evidence in aggravation.” Williams, 529 U.S. at 397–98.\\n\\n We begin our analysis by clarifying what exactly Jenkins is arguing about\\n\\nprejudice. We do not know precisely which aggravating and mitigating factors the\\n\\njury may have implicitly found before making its recommendation of death. The\\n\\ntrial court found at sentencing two statutory aggravating factors: the robbery and\\n\\nthe kidnapping of the victim. See Ala. Code § 13A-5-49(4). Because none of the\\n\\nmitigation evidence offered in the Rule 32 hearing bears on those factors, we see\\n\\nno reasonable probability that the aggravating factors would have been determined\\n\\ndifferently by the jury or the court. The court also found at sentencing two\\n\\nstatutory mitigating factors: no significant history of prior criminal activity and\\n\\nage. See id. § 13A-5-51(1), (7). There was also strong evidence of Jenkins’s\\n\\nintoxication, which was a key part of Jenkins’s guilt-phase defense and could have\\n\\nsupported a third mitigating factor, impaired capacity. See id. § 13A-5-51(6).10\\n\\n\\n10\\n The jury had heard evidence about all three of these statutory mitigators during the guilt phase\\nof the trial. Downey’s penalty-phase opening statement reviewed for the jury all seven of the\\npossible statutory mitigating factors, and the court also recited them in its jury instructions,\\nreminding the jury that they could consider guilt-phase evidence as well as non-statutory\\nmitigation. Jenkins has not established that there is a substantial probability that the jury’s\\nrecommendation would have been different had it been more explicitly reminded of Jenkins’s\\n 35\\n\\nJenkins does not argue that the jury or the court would have found other statutory\\n\\nmitigators in light of the Rule 32 evidence. Rather, he contends that the jury might\\n\\nhave found non-statutory mitigation based on Jenkins’s abusive childhood, low\\n\\nintelligence, or mental health problems that would have outweighed the\\n\\naggravating factors.11\\n\\n The Rule 32 trial court found that Jenkins had not demonstrated prejudice\\n\\neven assuming that counsel had performed deficiently. See Jenkins, 972 So. 2d at\\n\\n138. It concluded that he had not shown a reasonable probability that the sentencer\\n\\n“would have concluded that a weighing of the aggravating and mitigating\\n\\ncircumstances did not warrant death” upon a reweighing of the evidence. See id. It\\n\\nemphasized the depravity of the crime and the aggravating circumstances and\\n\\nfound that the mitigating witnesses were not credible for various reasons. See id.\\n\\nThe state appellate court affirmed, declining to disturb the credibility\\n\\ndeterminations of the trial court. Id. at 142–44. Independently reweighing the\\n\\n\\nage, criminal history, and intoxication, given that the trial court, the Rule 32 trial court, and the\\nstate appellate court all explicitly considered those mitigators and found them outweighed by the\\naggravators.\\n11\\n Alabama law lists seven statutory mitigating circumstances, but states that mitigation “shall\\n. . . not be limited to” those seven circumstances. Ala. Code § 13A-5-51. Further, “mitigating\\ncircumstances shall include any aspect of a defendant’s character or record and any of the\\ncircumstances of the offense that the defendant offers as a basis for a sentence of life\\nimprisonment without parole instead of death, and any other relevant mitigating circumstances\\nwhich the defendant offers as a basis for a sentence of life imprisonment without parole instead\\nof death.” Id. § 13A-5-52. A jury may recommend a sentence of death only if it finds that “one or\\nmore aggravating circumstances . . . exist and that they outweigh the mitigating circumstances, if\\nany.” Id. § 13A-5-46(e)(3).\\n 36\\n\\nevidence, it concluded that, “[g]iven the aggravating circumstances that were\\n\\nproven by the State and the facts surrounding Hogeland’s murder, we, like the\\n\\ncircuit court, are confident that death was the appropriate punishment for Jenkins’s\\n\\nactions.” Id. at 148.\\n\\n Our review of this evidence does not show that it is likely that the jury\\n\\nwould have concluded that death was not the appropriate punishment if it had\\n\\nheard the Rule 32 evidence. Much of the proffered mitigation evidence is, in our\\n\\nview, much less mitigating than Jenkins asserts. In particular, the testimony of\\n\\nJenkins’s older brother Michael was likely to have been a double-edged sword. See\\n\\nEvans v. Sec’y, Dep’t of Corr., 703 F.3d 1316, 1327 (11th Cir. 2013) (en banc) (“it\\n\\nis reasonable to conclude that a defendant was not prejudiced when his mitigation\\n\\nevidence was a two-edged sword” (quoting Ponticelli v. Sec’y, Fla. Dep’t of Corr.,\\n\\n690 F.2d 1271, 1296 (11th Cir. 2012)). Michael was the closest relative to testify at\\n\\nthe Rule 32 hearing and his testimony about the abuse Jenkins experienced should\\n\\nhave been the most reliable, since he experienced much of the same physical abuse\\n\\nhimself. The brothers shared the same filthy bedroom, received beatings from the\\n\\nsame paddle, and were frequently absent from school together. But Michael’s\\n\\ntestimony contained some discrepancies, and he admitted to cognitive and memory\\n\\nproblems. Perhaps Michael’s testimony would have made Jenkins appear more\\n\\nsympathetic in the eyes of the jury—but perhaps not. While Jenkins ran away from\\n\\n\\n 37\\n\\nhome and eventually committed murder, Michael responded to the same\\n\\nbackground by running away from home, graduating from high school, and\\n\\nproviding health care for both the terminally ill and people with substance use\\n\\ndisorders. It is possible that the jury could even have found Jenkins to be more\\n\\nculpable for not overcoming his difficult upbringing and to pose more of a danger\\n\\nof future violence.\\n\\n We also find reasonable the state court’s identification of other problems\\n\\nwith the mitigating evidence Jenkins argues should have been presented. Hearing\\n\\nthat evidence firsthand in the Rule 32 proceeding, the trial court found the\\n\\ntestimony of the cousins less credible in light of the lack of physical abuse\\n\\ndocumented in the school health records. See Jenkins, 972 So. 2d at 139 (noting\\n\\nthat school records observed a rash and gingivitis). It faulted Pitts in particular for\\n\\npurporting to have witnessed ongoing, severe abuse but taking only minimal,\\n\\nineffective steps to obtain outside help. See id. It further noted that DeLavega and\\n\\nWagoner actually testified that they never saw Jenkins beaten, and it noted all of\\n\\nthe witnesses’ interest in getting Jenkins off death row. All of these factors would\\n\\nlikely have caused a reasonable jury to discount their testimony. See id. at 139–42.\\n\\nWe must accept the trial court’s views of the Rule 32 testimony. “In the absence of\\n\\nclear and convincing evidence, we have no power on federal habeas review to\\n\\nrevisit the state court’s credibility determinations.” Bishop v. Warden, 726 F.3d\\n\\n\\n 38\\n\\n1243, 1259 (11th Cir. 2013) (citing Marshall v. Lonberger, 459 U.S. 422, 434\\n\\n(1983)).\\n\\n The state court also noted that Wagoner testified that she was not available\\n\\nto testify at the trial, and it found that Seal’s testimony would have been\\n\\ncumulative to that of her husband. Thus, counsel’s failure to call those two\\n\\nwitnesses could not have been prejudicial to Jenkins. See id. at 141, 142. Neither of\\n\\nthese conclusions involves an unreasonable determination of the facts. We agree\\n\\nthat the Rule 32 testimony of Sherry Seal was substantially similar to the testimony\\n\\nof her husband Lonnie Seal that the jury heard during the penalty phase. Both\\n\\ntestified about meeting Jenkins in California, Jenkins’s assistance with their move\\n\\nto Alabama, and his residence with them in their Alabama home. Sherry added\\n\\nonly that Jenkins sometimes wet his bed. It was not unreasonable for the state court\\n\\nto conclude that Sherry’s testimony was cumulative of her husband’s, and “a\\n\\npetitioner cannot satisfy the prejudice prong of the Strickland test with evidence\\n\\nthat is merely cumulative of evidence already presented.” Rose v. McNeil, 634 F.3d\\n\\n1224, 1243 (11th Cir. 2011).\\n\\n Further, the school, medical, and juvenile court records that Jenkins\\n\\nintroduced in the Rule 32 proceeding have their own problems. As noted, they do\\n\\nnot document or even corroborate the severe level of abuse alleged by some of the\\n\\nwitnesses. Though these documents tell a sad story, they do not create a substantial\\n\\n\\n 39\\n\\nlikelihood that the jury would have recommended life. Although they portray\\n\\nJenkins as a victim of truly unfortunate circumstances, they also show his unsavory\\n\\ndelinquent behavior. That delinquency included a history of stealing cars, and\\n\\ngiven that the jury likely inferred that Jenkins stole the red Mazda, new\\n\\ninformation about his history of theft may not have been especially mitigating.\\n\\nSimilarly, we agree with the state court that evidence of Jenkins’s compliant\\n\\nconduct in pretrial detention is “minimally mitigating” where there had been no\\n\\nevidence of jail misconduct as an aggravator. See Jenkins, 972 So. 2d at 149; cf.\\n\\nSkipper v. South Carolina, 476 U.S. 1, 5 & n.1 (1986) (good jail behavior is\\n\\n“potentially mitigating” especially “[w]here the prosecution specifically relies on a\\n\\nprediction of future dangerousness in asking for the death penalty”).\\n\\n Jenkins argues that allowing the jury to see his experience of mental,\\n\\nemotional, and cognitive difficulties, compounded by years of abuse and neglect\\n\\nand considered alongside his mild character and good jail behavior as an adult,\\n\\nwould have convinced them that Hogeland’s murder was an “aberration” that\\n\\ndeserved a punishment less severe than death. That interpretation of Jenkins’s life\\n\\nstory may be conceivable. But more than mere conceivability is required to\\n\\nestablish prejudice: “The likelihood of a different result must be substantial, not\\n\\njust conceivable.” Richter, 562 U.S. at 112 (emphasis added). We do not find the\\n\\nstate court’s decision that a different result was not substantially likely to be an\\n\\n\\n 40\\n\\nunreasonable application of Strickland. As we have noted, we think it equally\\n\\nlikely that the jury could have rejected this alternate narrative out of hand in light\\n\\nof having just convicted him of the violent murder of a young mother. We also\\n\\nthink it possible that the jury could have viewed the murder as just one step in a\\n\\nlogical progression of degeneracy that began with behavioral problems at school,\\n\\ncontinued through the juvenile court system, and escalated past grand theft auto\\n\\ninto kidnapping and murder. Taken separately as well as together, the proffered\\n\\nevidence does not compel us to conclude that a verdict of life would have been\\n\\nsubstantially likely. Because fairminded jurists could disagree about the state\\n\\ncourt’s conclusion about Strickland prejudice, we affirm the denial of habeas relief\\n\\non this ground.\\n\\n B. Intellectual Disability\\n\\n Jenkins argues that the district court erred when it denied his claim that he is\\n\\nintellectually disabled and ineligible for the death penalty under Atkins v. Virginia,\\n\\n536 U.S. 304 (2002). The Supreme Court held in Atkins that executing an offender\\n\\nwho has mental retardation12 violates the Eighth Amendment’s ban on excessive\\n\\n\\n\\n\\n12\\n In 2002, the Atkins Court followed prevailing convention in referring to “mentally retarded”\\noffenders. We will use that now-disfavored term only when quoting or discussing older sources.\\nSince 2014, the Supreme Court has exclusively used the terms “intellectually disabled person”\\nand “person with intellectual disability.” See, e.g., Hall v. Florida, 572 U.S. 701, 704 (2014);\\nMoore v. Texas, 139 S. Ct. 666, 672 (2019).\\n 41\\n\\npunishments. Id. at 321. It defined mental retardation by quoting the American\\n\\nAssociation on Mental Retardation:\\n\\n Mental retardation refers to substantial limitations in present\\n functioning. It is characterized by significantly subaverage intellectual\\n functioning, existing concurrently with related limitations in two or\\n more of the following applicable adaptive skill areas: communication,\\n self-care, home living, social skills, community use, self-direction,\\n health and safety, functional academics, leisure, and work. Mental\\n retardation manifests before age 18.\\n\\nId. at 308 n.3. Alabama courts have followed that tripartite definition, adopting\\n\\nwhat they consider the “broadest” definition of mental retardation for Atkins\\n\\npurposes:\\n\\n [A] defendant, to be considered mentally retarded, must have\\n significantly subaverage intellectual functioning (an IQ of 70 or\\n below), and significant or substantial deficits in adaptive behavior.\\n Additionally, these problems must have manifested themselves during\\n the developmental period (i.e., before the defendant reached age 18).\\n\\nEx parte Perkins, 851 So. 2d 453, 456 (Ala. 2002).\\n\\n The clinical evidence adduced during the Rule 32 hearing with respect to\\n\\nJenkins’s intellectual ability was that his IQ was 76, reflecting “borderline\\n\\nintellectual functioning.” The record also included the report of the Lunacy\\n\\nCommission13 which found that Jenkins was competent to stand trial, as well as\\n\\n\\n\\n13\\n Alabama law requires a capital defendant to be “committed to the Department of Mental\\nHealth and Mental Retardation for examination” if there is a question about his capacity to stand\\ntrial or if there will be a mental-disease-or-defect defense. Ala. Code § 15-16-22(a). “The\\nassigned mental health professional(s) shall examine the defendant with respect to determining\\nthe presence of any mental disease or defect which, if determined to be present, would affect the\\n 42\\n\\nJenkins’s school and juvenile court records. The court also considered all the\\n\\ntestimony it had heard about Jenkins and reviewed the transcript of his murder trial\\n\\nbefore rejecting Jenkins’s claim of mental retardation.\\n\\n On appeal, after reciting the Perkins standard, the Alabama Court of\\n\\nCriminal Appeals explained that Jenkins had not established mental retardation:\\n\\n Dr. Kirkland testified that he performed psychological tests on Jenkins\\n and that Jenkins’s IQ was 76. There was evidence presented at\\n Jenkins’s trial indicating that Jenkins maintained relationships with\\n other individuals and that he had been employed by P.S. Edwards\\n Landscaping Company, Cotton Lowe 76 Service Station, and\\n Paramount Painting Company. The record fails to show that Jenkins\\n meets the most liberal view of mental retardation adopted by the\\n Alabama Supreme Court in Perkins.\\n\\nJenkins, 972 So. 2d at 155. In this habeas appeal, Jenkins raises arguments about\\n\\nall three components of the intellectual disability showing under Atkins and\\n\\nPerkins, which we address in turn. He also appeals the denial of an evidentiary\\n\\nhearing on this issue.\\n\\n 1. Intellectual Function\\n\\n The first component of intellectual disability under Perkins is “significantly\\n\\nsubaverage intellectual functioning (an IQ of 70 or below).” Perkins, 851 So. 2d at\\n\\n456. Jenkins argues that the use of a strict IQ cutoff of 70 was contrary to or an\\n\\nunreasonable application of Atkins.\\n\\n\\ncapacity of the defendant to proceed or continue to trial or which would affect the defendant’s\\ncriminal responsibility at the time of the commission of the crime.” Id.\\n 43\\n\\n We disagree. The state court’s decision that Jenkins failed to establish\\n\\nmental retardation in this respect was a reasonable application of the record and of\\n\\nAtkins. As a threshold matter, we note that the state court did not say it was\\n\\napplying a strict cutoff of 70; it merely stated that Dr. Kirkland had measured\\n\\nJenkins’s IQ as 76. Jenkins, 972 So. 2d at 155. Under our deferential review, the\\n\\nquestion is whether the decision that Jenkins had not established the intellectual\\n\\nfunctioning aspect of intellectual disability was an unreasonable application of\\n\\nAtkins. The Supreme Court explained in Atkins that it was leaving “to the State[s]\\n\\nthe task of developing appropriate ways to enforce the constitutional restriction\\n\\nupon [their] execution of sentences.” Atkins, 536 U.S. at 317 (quoting Ford v.\\n\\nWainwright, 477 U.S. 399, 416–17 (1986)). In line with that invitation, our Circuit\\n\\nhas accepted Alabama’s judicial definition of intellectual disability as laid out in\\n\\nPerkins.14 See Thomas, 607 F.3d at 752–53. Jenkins objects, however, that the\\n\\nPerkins standard, which he characterizes as articulating a “strict cutoff” of 70,\\n\\nconflicts with more recent Supreme Court precedent that instead requires a\\n\\n“clinical” definition that goes beyond a single IQ score.\\n\\n In particular, Jenkins points to Hall v. Florida, in which the Supreme Court\\n\\nheld that when an offender’s IQ is near but greater than 70, courts must take into\\n\\n\\n\\n14\\n Notwithstanding the urging of the Alabama Supreme Court, see Perkins, 851 So. 2d at 455\\nn.1, the Alabama legislature has not enacted a statutory definition of intellectual disability.\\n 44\\n\\naccount the IQ test’s standard error of measurement as well as non-IQ evidence of\\n\\nintellectual function. See Hall v. Florida, 572 U.S. 701, 723 (2014) (“This Court\\n\\nagrees with the medical experts that when a defendant’s IQ test score falls within\\n\\nthe test’s acknowledged and inherent margin of error, the defendant must be able\\n\\nto present additional evidence of intellectual disability, including testimony\\n\\nregarding adaptive deficits.”). But, as Jenkins concedes, our Circuit has\\n\\nspecifically held that Hall is not retroactive to cases on collateral review. See In re\\n\\nHenry, 757 F.3d 1151, 1161 (11th Cir. 2014) (explaining that Hall does not supply\\n\\na new substantive rule but rather “merely provides new procedures for ensuring\\n\\nthat States do not execute members of an already protected group”). We also note\\n\\nthat Alabama courts have found the Perkins standard to comport with Hall in that\\n\\nit does not prevent an offender from presenting the standard error of measurement\\n\\nor non-IQ evidence of intellectual disability. See Reeves v. State, 226 So. 3d 711,\\n\\n728–29 (Ala. Ct. Crim. App. 2016). Here, the state court did not unreasonably\\n\\napply Atkins when it analyzed the intellectual functioning component of\\n\\nintellectual disability.\\n\\n Jenkins also argues that the state court’s decision on this component in his\\n\\nRule 32 proceeding resulted from an unreasonable determination of the facts. He\\n\\nasserts that his IQ score of 76, when combined with the test’s standard error of\\n\\nmeasurement and considered alongside the Flynn effect (which we discuss below),\\n\\n\\n 45\\n\\ndoes show that he has significantly subaverage intellectual function. For the\\n\\nreasons that follow, we disagree.\\n\\n Most fundamentally, neither of the clinicians who testified about Jenkins’s\\n\\nintellectual ability in the Rule 32 hearing opined that he was mentally retarded.\\n\\nThe closest thing Jenkins can point to is the testimony of Dr. Kirkland that his IQ\\n\\nis two standard deviations below the mean. But his entire testimony, in context, did\\n\\nnot establish intellectual disability. To the contrary: Dr. Kirkland testified that\\n\\nJenkins “scored in the range of borderline intellectual functioning which is\\n\\nbetween mild mental retardation and low average intellectual functioning.”\\n\\nFurthermore, Jenkins’s own expert Dr. Lisak fully reviewed Dr. Kirkland’s clinical\\n\\nfindings. He agreed with those findings, concluding that the IQ results show\\n\\n“borderline intelligence.” It is tremendously significant that no clinical assessment\\n\\nin the entire record of these proceedings has found that Jenkins has mental\\n\\nretardation or intellectual disability.\\n\\n Jenkins nonetheless argues that his IQ score of 76 establishes the intellectual\\n\\ncomponent of intellectual ability when the standard error of measurement and the\\n\\nFlynn effect are considered alongside that score. It was not unreasonable for the\\n\\nstate court to disagree. As we have discussed, Jenkins is not retroactively entitled\\n\\nto have the state courts take into account the margin of error of the IQ test. That\\n\\nstatistical margin of error—5 points in this case—takes into account characteristics\\n\\n\\n 46\\n\\nof the test and variations due to chance in order to state how confident the tester is\\n\\nthat the measured result accurately states Jenkins’s true, unknowable IQ. A result\\n\\nof 76 with a standard error of measurement of 5 simply means that the tester is\\n\\n68% confident that Jenkins’s true IQ falls between 71 and 81, and 95% confident\\n\\nthat his true IQ falls between 66 and 86. See David H. Kaye & David A. Freedman,\\n\\nReference Guide on Statistics, in Reference Manual on Scientific Evidence 211,\\n\\n243–44 (3d ed. 2011); see also Hall, 572 U.S. at 738–39 (Alito, J., dissenting). Of\\n\\ncourse, Jenkins would like us to consider only the possibility that the test result\\n\\noverstated his IQ—but the standard error of measurement is a two-way street.\\n\\n“[T]he standard error of measurement is a bi-directional concept that does not\\n\\ncarry with it a presumption that an individual’s IQ falls to the bottom of his IQ\\n\\nrange.” Ledford v. Warden, Ga. Diag. & Clas. Prison, 818 F.3d 600, 641 (11th Cir.\\n\\n2016).\\n\\n Even if Jenkins were somehow entitled to subtract the 5-point margin of\\n\\nerror from his score and thereby lower his “true” IQ to 71, he still would not satisfy\\n\\nthe clinical intellectual component of intellectual disability. Thus, he sought in the\\n\\ndistrict court to introduce evidence about the Flynn effect in order to bring his IQ\\n\\ndown to 70 or below. As our Court has explained, the Flynn effect “recognizes the\\n\\nfact that IQ test scores have been increasing over time”:\\n\\n The Flynn effect acknowledges that as an intelligence test ages, or\\n moves farther from the date on which it was standardized, or normed,\\n 47\\n\\n the mean score of the population as a whole on that assessment\\n instrument increases, thereby artificially inflating the IQ scores of\\n individual test subjects. Therefore, the IQ test scores must be\\n recalibrated to keep all test subjects on a level playing field.\\n\\nThomas, 607 F.3d at 753. We have said that “[a]n evaluator may also consider the\\n\\n‘Flynn effect’” in assessing an offender’s possible intellectual disability, id.\\n\\n(emphasis added), but neither we nor the Alabama Supreme Court nor the U.S.\\n\\nSupreme Court have said that a court must consider it in order to reasonably apply\\n\\nAtkins. See, e.g., Reeves, 226 So. 3d at 739 (“This Court has repeatedly held that a\\n\\ncircuit court is not required to accept, consider, or apply the ‘Flynn Effect’ in\\n\\ndetermining intellectual disability.”).\\n\\n More problematically, Jenkins did not raise the Flynn effect issue in the\\n\\nAlabama state courts. Jenkins asserts that he implicitly raised it as part and parcel\\n\\nof his Atkins claim during his Rule 32 appeal, but we do not agree that the\\n\\nAlabama courts had a full and fair opportunity to consider this issue, because it\\n\\nwas not explicitly presented to them. See Kelley v. Sec’y for the Dep’t of Corr., 377\\n\\nF.3d 1317, 1349–50 (11th Cir. 2004) (holding that one ineffective-assistance\\n\\nargument was unexhausted despite the petitioner’s general presentation of a\\n\\nStrickland claim in the state courts). Thus, under AEDPA, Jenkins is procedurally\\n\\nbarred from raising this unexhausted claim in his federal habeas petition. See 28\\n\\nU.S.C. § 2254(b)(1)(A), (c); Mason v. Allen, 605 F.3d 1114, 1119 (11th Cir. 2010).\\n\\nWe will allow a prisoner to overcome a procedural bar if he establishes cause for\\n 48\\n\\nhis default and prejudice therefrom. See, e.g., Smith v. Jones, 256 F.3d 1135, 1138\\n\\n(11th Cir. 2001). But Jenkins cannot establish prejudice from being barred from\\n\\nasserting this claim now because the Supreme Court has not “clearly established”\\n\\nthat a court must consider the Flynn effect in assessing intellectual disability. Thus,\\n\\nunder AEDPA, it could not form the basis for a federal habeas claim. See 28\\n\\nU.S.C. § 2254(d)(1). We therefore conclude that the state court did not\\n\\nunreasonably determine the facts or unreasonably apply Atkins with respect to the\\n\\nintellectual component of intellectual disability.\\n\\n 2. Adaptive Function\\n\\n The second component of intellectual disability is “significant or substantial\\n\\ndeficits in adaptive behavior.” Perkins, 851 So. 2d at 456. Jenkins argues that the\\n\\nstate court, which mentioned only Jenkins’s successes in employment and social\\n\\nskills, see Jenkins, 972 So. 2d at 155, failed to consider his deficits. He adds that\\n\\neven the mention of employment success does not tell the whole story. Jenkins\\n\\ninsists that his jobs were menial, his employers took advantage of him, and he held\\n\\nthe jobs for only a few months each.\\n\\n But the state court’s terse discussion of Jenkins’s adaptive behavior did not\\n\\ninvolve an unreasonable determination of the facts or an unreasonable application\\n\\nof Atkins. To the contrary, its decision on this issue was reasonable in light of both\\n\\nthe record and Atkins.\\n\\n\\n 49\\n\\n Atkins elaborated that the adaptive-skill component of intellectual disability\\n\\nrequires substantial present limitation in at least two of the following areas:\\n\\n“communication, self-care, home living, social skills, community use, self-\\n\\ndirection, health and safety, functional academics, leisure, and work.” Atkins, 536\\n\\nU.S. at 308 n.3 (quoting definition of American Association on Mental\\n\\nRetardation). The record supports the state court’s judgment that Jenkins does not\\n\\nhave substantial deficits in at least two areas. To be sure, we will assume that\\n\\nJenkins may be substantially limited in functional academics; Dr. Kirkland\\n\\nassessed his reading, spelling, and arithmetic skills at a third-grade level. But the\\n\\nrecord does not compel a finding of substantial present deficits in the areas of\\n\\ncommunication, self-care, home living, and work. Although Jenkins emphasizes\\n\\nhis transient employment, the record shows he was able to find work quickly, to\\n\\nwork hard to support the Seals’ household until Lonnie Seal found a job, and, later,\\n\\nto move into his own home. That record evidence is inconsistent with a substantial\\n\\ndeficit in the area of work.\\n\\n As the State observes, the facts of the crime also fail to show significant\\n\\ndeficits in the areas of communication, self-care, community use, and self-\\n\\ndirection. Jenkins was able to communicate well enough to solicit an alibi and to\\n\\nsell his car, including writing out a bill of sale. He then was able to arrange for his\\n\\nflight from Alabama, including purchasing a bus ticket and hitchhiking across the\\n\\n\\n 50\\n\\ncountry. Indeed, throughout his trial and postconviction proceedings, the people\\n\\nwho knew Jenkins as an adult consistently described a man who did not have\\n\\nserious difficulties in communicating, forming relationships, working, and caring\\n\\nfor himself. No witness mentioned these kinds of difficulties in his or her\\n\\ntestimony about Jenkins. There was some evidence that Jenkins was “not\\n\\narticulate” and that his house was messy, but that testimony does not rise to the\\n\\nlevel of overwhelming evidence of substantial functional deficits that would\\n\\nsupport a conclusion that the state court unreasonably determined the facts.\\n\\n Of course, the record also contains evidence of Jenkins’s childhood\\n\\nacademic and social deficits, attributable at least in part to his parents’ neglect, but\\n\\nJenkins’s childhood is not directly relevant to our consideration of his present\\n\\nlimitations. Although we acknowledge that the Supreme Court has noted that such\\n\\n“risk factors” may support further exploration of the possibility of intellectual\\n\\ndisability, see Moore v. Texas, 137 S. Ct. 1039, 1051 (2017), the state court record\\n\\ncontains the results of that exploration. Overall, that record supports the state\\n\\ncourt’s conclusion that Jenkins does not have substantial deficits in adaptive\\n\\nbehaviors.\\n\\n 3. Juvenile Onset\\n\\n The third and final component of intellectual disability is manifestation of\\n\\nthe first and second components “before the defendant reached age 18.” Perkins,\\n\\n\\n 51\\n\\n851 So. 2d at 456. Jenkins asserts, the State assumes, and we agree that the state\\n\\ncourt did not rule upon this issue, so there is nothing to which we must defer under\\n\\nAEDPA. See, e.g., Brumfield v. Cain, 135 S. Ct. 2269, 2282 (2015). Thus, we will\\n\\nreview the district court’s legal conclusions on this issue de novo, and its factual\\n\\nfindings for clear error. See, e.g., Grossman v. McDonough, 466 F.3d 1325, 1335\\n\\n(11th Cir. 2006). Jenkins argues that he has satisfied the age-of-onset component\\n\\nby showing “a wide range of evidence of deficits in both intellectual and adaptive\\n\\nfunctioning throughout Jenkins’s childhood.”\\n\\n In fact, the ample record of Jenkins’s childhood, as developed in the Rule 32\\n\\nproceeding, does not point to intellectual disability before age 18. School\\n\\nassessments at age 12 reported an IQ score of 83, an adaptive behavior rating of\\n\\n“low,” and noted overall an intellectual capacity of “average,” a diagnosis of\\n\\ndyslexia, and “poor functional skills.” The psychologist attributed Jenkins’s\\n\\ndifficulties to a combination of “[f]requent moves, changes in schools and family\\n\\ninstability” with “[l]ow ability coupled with a specific learning disability” and\\n\\n“[l]ack of responsibility for self control and poor use of time.” Although Jenkins\\n\\nfailed several grades of school, teachers additionally attributed his difficulties to\\n\\nhis behavioral problems and his high number of absences from school.\\n\\n Similarly, IQ tests ordered by the juvenile court at age 14 gave scores of 81\\n\\nand 86 on different tests, in the “[d]ull–[n]ormal” range and reflecting “a definite\\n\\n\\n 52\\n\\nlearning disability.” These records all show a child with serious academic deficits\\n\\nand some intellectual and adaptive deficits, but they do not clearly show an\\n\\nintellectually disabled child. In particular, Jenkins’s childhood IQ scores of 83, 81,\\n\\nand 86 fail to point to intellectual disability or mental retardation. We are not left\\n\\nwith the definite and firm conviction that the district court erred when it found that\\n\\n“Jenkins’s [childhood] intelligence scores and adaptive functioning skills were\\n\\nattributable to his learning disability and other circumstances in his life, rather than\\n\\nmental retardation.”\\n\\n In sum, we agree with the district court that the decision of the Alabama\\n\\nCourt of Criminal Appeals that Jenkins does not have intellectual disability was\\n\\nnot contrary to or an unreasonable application of Atkins v. Virginia.\\n\\n 4. Evidentiary Hearing\\n\\n Finally, Jenkins argues that the district court erred when it denied his request\\n\\nfor an evidentiary hearing on his Atkins claim. If Jenkins diligently attempted to\\n\\npresent this claim in the state court, we review for abuse of discretion. Burgess v.\\n\\nComm’r, Ala. Dep’t of Corr., 723 F.3d 1308, 1320 (11th Cir. 2013). “The district\\n\\ncourt abuses its discretion where ‘such a hearing could enable an applicant to prove\\n\\nthe petition’s factual allegations, which, if true, would entitle the applicant to\\n\\nfederal habeas relief.’” Id. (quoting Schriro v. Landrigan, 550 U.S. 465, 474\\n\\n(2007)). If Jenkins did not diligently attempt to develop the factual basis for this\\n\\n\\n 53\\n\\nclaim in the state court, absent certain circumstances, he is not entitled to a hearing\\n\\nand we will affirm the district court’s denial. 28 U.S.C. § 2254(e)(2).\\n\\n The parties strenuously dispute whether Jenkins diligently attempted to\\n\\npresent this issue in the state court. On the one hand, the State argues that Jenkins\\n\\nfailed to allege that he was mentally retarded even after the U.S. Supreme Court\\n\\ndecided Atkins in 2002. On the other hand, Jenkins asserts that he raised the\\n\\npossibility of mental retardation as early as his original Rule 32 petition in 1995.\\n\\n As we have noted, shortly after the Supreme Court decided Atkins, the\\n\\nAlabama Court of Criminal Appeals ordered supplemental briefing on the possible\\n\\nimpact of Atkins on Jenkins. Jenkins’s Rule 32 appeal was pending at that point,\\n\\nhaving been fully briefed and orally argued in the state appellate court. Having\\n\\nreviewed the state court records, we agree with the State that Jenkins did not\\n\\ndiligently attempt to present his Atkins claim in the state courts.\\n\\n “Diligence will require in the usual case that the prisoner, at a minimum,\\n\\nseek an evidentiary hearing in state court.” Williams, 529 U.S. at 437. The record\\n\\nclearly shows that Jenkins did not do so. In the initial supplemental brief on Atkins\\n\\nhe filed in 2002, Jenkins merely requested a future hearing after the Alabama\\n\\nlegislature’s enactment of a statute on executing the mentally retarded. Seventeen\\n\\nyears later, that contingency is still in the future. Only in his reply supplemental\\n\\nbrief did Jenkins then assert, for the first time ever, that he is mentally retarded.\\n\\n\\n 54\\n\\n We acknowledge that Jenkins’s initial Rule 32 petition included in its\\n\\nallegations of ineffective assistance of counsel a list of unpresented mitigating\\n\\nevidence including his “developmental[] impair[ment]” and “learning disabilities,\\n\\nlow intelligence, poor comprehension and retarded socialization skills.” We\\n\\nconclude that these tangential Strickland references to his mental and social\\n\\nabilities do not amount to diligence. Thus, because Jenkins did not diligently\\n\\nattempt to develop the factual basis for his Atkins claim in the state courts, we\\n\\naffirm the district court’s denial of his request for an evidentiary hearing.\\n\\n Even if Jenkins had diligently attempted to present this claim in the state\\n\\ncourt, however, we conclude in the alternative that the district court did not abuse\\n\\nits discretion when it denied a hearing. We agree, in light of the ample evidence\\n\\nabout intellectual ability that Jenkins adduced in support of his ineffective-\\n\\nassistance claim, that such a hearing would not have enabled Jenkins to prove that\\n\\nhe is intellectually disabled. See Burgess, 723 F.3d at 1320.\\n\\n IV. CONCLUSION\\n\\n The denial of Jenkins’s petition for a writ of habeas corpus is\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 55\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4162,"cells":{"id":{"kind":"number","value":4657746,"string":"4,657,746"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"United States v. Hi-Tech Pharmaceuticals, Inc."},"decision_date":{"kind":"string","value":"2019-08-30"},"docket_number":{"kind":"string","value":"17-13376"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before TJOFLAT and JORDAN, Circuit Judges, and HINKLE, * District Judge.\", \"parties\": \"\", \"opinions\": [{\"author\": \"HINKLE, District Judge:\", \"type\": \"010combined\", \"text\": \"Case: 17-13376 Date Filed: 08/30/2019 Page: 1 of 27\\n\\n\\n [PUBLISH]\\n\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 17-13376\\n ________________________\\n\\n D.C. Docket No. 1:13-cv-03675-WBH\\n\\n\\nUNITED STATES OF AMERICA,\\n\\n Plaintiff - Appellee,\\n\\nversus\\n\\nUNDETERMINED QUANTITIES OF ALL ARTICLES OF\\nFINISHED AND IN-PROCESS FOODS,\\nraw ingredients (bulk powders, bulk capsules), with any lot number,\\nsize, or type container, whether labeled or unlabeled, et al.,\\n\\n Defendants,\\n\\nHI-TECH PHARMACEUTICALS, INC.,\\nJARED WHEAT,\\n\\n Claimants - Appellants.\\n\\n\\n ________________________\\n\\n Appeal from the United States District Court\\n for the Northern District of Georgia\\n ________________________\\n\\n (August 30, 2019)\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 2 of 27\\n\\n\\nBefore TJOFLAT and JORDAN, Circuit Judges, and HINKLE, * District Judge.\\n\\nHINKLE, District Judge:\\n\\n The Dietary Supplement Health and Education Act of 1994 provides\\n\\nfavorable treatment for “dietary supplements,” defined to include any “botanical”\\n\\nor “constituent” of a botanical. This case presents the question whether these terms\\n\\napply to a substance that was invented in a laboratory and is artificially produced\\n\\nfor commercial sale but that, entirely coincidentally, may be found in trace\\n\\namounts in a plant. We hold that the terms do not extend this far.\\n\\n I. Proceedings\\n\\n The Food and Drug Administration seized from Hi-Tech Pharmaceuticals,\\n\\nInc. a substantial quantity of products containing 1,3-dimethylamylamine or\\n\\n“DMAA.” DMAA is used in fitness products aimed at bodybuilders and other\\n\\nathletes.\\n\\n The seizure led to two actions that were consolidated in the district court.\\n\\nOne was a forfeiture action filed by the United States against the products. Hi-Tech\\n\\nand its chief executive officer, Jared Wheat, intervened as claimants. Hi-Tech filed\\n\\nthe other action against the FDA and other governmental defendants.\\n\\n\\n\\n\\n *\\n Honorable Robert L. Hinkle, United States District Judge for the Northern District of\\nFlorida, sitting by designation.\\n 2\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 3 of 27\\n\\n\\n Hi-Tech asserted that DMAA is a dietary supplement; that under the\\n\\nAdministrative Procedure Act the FDA can properly ban DMAA, if at all, only\\n\\nthrough rulemaking; and that the seizure of Hi-Tech’s DMAA violated the Fifth\\n\\nAmendment’s Due Process Clause.\\n\\n The parties filed cross-motions for summary judgment. The district court\\n\\ngranted the FDA’s motion, holding the seizure proper both substantively and\\n\\nprocedurally. The district court denied a motion to reconsider that included a\\n\\nrequest to reopen discovery. Hi-Tech and Mr. Wheat have appealed. The appeal\\n\\nhas been fully briefed and orally argued.\\n\\n II. Standard of Review\\n\\n We review de novo the district court’s grant of summary judgment. See, e.g.,\\n\\nPrice v. Comm’r, Dep’t of Corr., 920 F.3d 1317, 1323 (11th Cir. 2019). We review\\n\\nfor abuse of discretion the district court’s denial of the motion for reconsideration\\n\\nand refusal to reopen discovery. See, e.g., Corwin v. Walt Disney Co., 475 F.3d\\n\\n1239, 1254 (11th Cir. 2007) (reconsideration); Artistic Entm’t, Inc. v. City of\\n\\nWarner Robins, 331 F.3d 1196, 1202 (11th Cir. 2003) (reopening discovery).\\n\\n III. The Statute and the Issues\\n\\n The Federal Food, Drug, and Cosmetic Act prohibits the introduction of\\n\\nadulterated foods into interstate commerce. 21 U.S.C. § 331(a). The FDA enforces\\n\\nthe Act. Id. § 393(b)(2)(A). The agency may bring an in rem forfeiture action in\\n\\n 3\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 4 of 27\\n\\n\\ndistrict court to condemn adulterated foods. Id. § 334(a)(1). Hi-Tech’s DMAA\\n\\nproducts were adulterated foods if they were “food additives” but not if they were\\n\\n“dietary supplements.”\\n\\n The background is this. The Dietary Supplement Health and Education Act\\n\\nof 1994, commonly referred to as “DSHEA,” amended the Federal Food, Drug,\\n\\nand Cosmetic Act to provide favorable treatment for dietary supplements. The\\n\\nstatute’s definition of “dietary supplement” includes multiple parts. 21 U.S.C.\\n\\n§ 321(ff). The only part relevant to Hi-Tech’s DMAA is this: a product that is\\n\\nintended to supplement the diet—this includes DMAA—is a dietary supplement if\\n\\nit contains “an herb or other botanical” or “a concentrate, metabolite, constituent,\\n\\nextract, or combination of” an herb or other botanical. Id. § 321(ff)(1)(C) & (F).\\n\\nThe statute describes these—as well as other substances not at issue here—as\\n\\n“dietary ingredients.”\\n\\n Under DSHEA, and subject to exceptions not relevant here, a dietary\\n\\nsupplement can be condemned as adulterated only if the FDA carries the burden of\\n\\nproving that the substance presents a “significant or unreasonable risk of illness or\\n\\ninjury” under recommended, suggested, or ordinary conditions of use. Id.\\n\\n§ 342(f)(1)(A). The FDA did not attempt to make that showing for the DMAA\\n\\nproducts it seized from Hi-Tech. A ruling that DMAA is a dietary supplement thus\\n\\nwould resolve this appeal in Hi-Tech’s favor.\\n\\n 4\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 5 of 27\\n\\n\\n On the other hand, a ruling that DMAA is a “food additive” would resolve\\n\\nthe dispute in the FDA’s favor. A substance intended for human consumption is a\\n\\nfood additive if it is not a dietary supplement and is not “generally recognized,\\n\\namong experts qualified by scientific training and experience to evaluate its safety,\\n\\nas having been adequately shown through scientific procedures . . . to be safe under\\n\\nthe conditions of its intended use.” Id. § 321(s). For a substance in common use in\\n\\nfood prior to January 1, 1958—this does not include DMAA—the adequate\\n\\nshowing of safety can be made not only by scientific procedures but also by\\n\\nexperience. There are other exceptions to this definition of “food additive,” but\\n\\nnone applies here.\\n\\n The FDA asserts that DMAA is not a dietary supplement, is not generally\\n\\nrecognized as safe, does not meet any other exception, and is therefore a food\\n\\nadditive. Hi-Tech insists that DMAA is a dietary supplement and thus is not a food\\n\\nadditive, but that even if DMAA is not a dietary supplement, DMAA is generally\\n\\nrecognized as safe and thus still is not a food additive.\\n\\n The issues thus are first, whether Hi-Tech’s DMAA products are “an herb or\\n\\nother botanical” or “a concentrate, metabolite, constituent, extract, or combination\\n\\nof” an herb or other botanical, and second, if not, whether the products are\\n\\ngenerally recognized as safe. Secondary issues are whether the FDA was entitled to\\n\\n\\n\\n\\n 5\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 6 of 27\\n\\n\\nseize and forfeit the products without engaging in rulemaking and whether the\\n\\ndistrict court should have reopened discovery.\\n\\n IV. DMAA\\n\\n The earliest known identification or use of DMAA occurred in 1944. In that\\n\\nyear Eli Lilly & Co. synthesized and patented DMAA for use as a nasal\\n\\ndecongestant. For marketing reasons, Eli Lilly asked the FDA to withdraw its\\n\\napproval of this use in 1983. At least insofar as shown by this record, DMAA was\\n\\nnot used as a dietary supplement or food additive at that time, and no health\\n\\nconcerns had been noted.\\n\\n DMAA eventually made a resurgence, this time in fitness products aimed at\\n\\nbodybuilders and other athletes. Because of DMAA’s noticeable stimulant effect,\\n\\nthe compound made its way into pre-workout energy and fat-burner products\\n\\naround the world.\\n\\n The FDA eventually adopted the position that DMAA is not a dietary\\n\\nsupplement but an unsafe food additive. The FDA issued cease-and-desist letters to\\n\\nat least some entities marketing DMAA products. Perhaps unaware of Hi-Tech’s\\n\\nmarketing of DMAA products, the FDA did not issue a cease-and-desist letter to\\n\\nHi-Tech.\\n\\n Around the same time, researchers began to find trace amounts of DMAA in\\n\\ngeraniums of the genus pelargonium. A 2013 survey concluded that overall, the\\n\\n 6\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 7 of 27\\n\\n\\nstudies showed that DMAA “is found naturally in some, but not all, geranium\\n\\nplants and extracted geranium oils.” Thomas D. Gauthier, Evidence for the\\n\\nPresence of 1,3-Dimethylamylamine (1,3-DMAA) in Geranium Plant Materials, 8\\n\\nAnalytical Chemistry Insights 29-40 (2013), available at\\n\\nhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC3682735/. Indeed, the FDA’s own\\n\\nexpert had previously participated in a study that found trace amounts of DMAA in\\n\\ngeraniums.\\n\\n Even so, this record presents a genuine factual dispute over whether trace\\n\\namounts of DMAA are naturally contained in geraniums. On the one hand, studies\\n\\nhave found trace amounts of DMAA in geraniums. On the other hand, some\\n\\nfertilizers contain DMAA that could be a source of trace amounts of DMAA in\\n\\ngeraniums, and the record includes competent testimony that there is no known\\n\\npathway by which geraniums could produce DMAA. Either way, it is clear that\\n\\nDMAA is not contained in geraniums in amounts greater than could reasonably be\\n\\ncharacterized as trace amounts. No study has found a greater amount.\\n\\n V. “Herb or Other Botanical”\\n\\n The first rule of statutory construction is to apply the plain meaning of the\\n\\nstatutory language. See, e.g., Bankston v. Then, 615 F.3d 1364, 1367 (11th Cir.\\n\\n2010). Here the meaning is not completely clear.\\n\\n\\n\\n\\n 7\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 8 of 27\\n\\n\\n Hi-Tech says DSHEA uses “botanical” to mean all plant life, nothing more\\n\\nand nothing less—that is, to mean flora, without limitation. The suggestion is\\n\\nsensible enough—“botany” is the study of plants. On the other hand, it would be\\n\\npassing strange for a writer wishing to cover the universe of plant life—to mean all\\n\\nflora—to achieve that result through the term “herb or other botanical.” Moreover,\\n\\nthe usual connotation of “botanical” when used as a noun, as recognized in\\n\\ndictionaries in use when DSHEA was enacted as well as those is use today, is a\\n\\nsubstance derived from a plant used for a limited category of purposes.\\n\\n In 1993, a year before DSHEA became law, Merriam-Webster’s defined the\\n\\nnoun “botanical” as a “a plant part or extract used esp[ecially] in skin and hair care\\n\\nproducts.” “Botanical,” Merriam-Webster’s Collegiate Dictionary 134 (10th ed.\\n\\n1993). The current edition defines the noun “botanical” as a “substance obtained or\\n\\nderived from a plant[,] such as . . . a plant part or extract used especially in skin\\n\\nand hair care products[,] a medicinal preparation derived from a plant[, or] plant\\n\\nmaterial used as a flavoring agent.” “Botanical,” Merriam-Webster’s Online\\n\\nDictionary 2019, available at https://www.merriam-\\n\\nwebster.com/dictionary/botanical. Neither definition suggests that the noun\\n\\n“botanical” includes an artificially produced substance that, entirely coincidentally,\\n\\nmay be found in trace amounts in a plant. Nor do they suggest that “botanical”\\n\\nincludes all flora.\\n\\n 8\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 9 of 27\\n\\n\\n Both the Supreme Court and the Eleventh Circuit have relied on Merriam-\\n\\nWebster’s as an aid in construing statutes. See, e.g., Octane Fitness, LLC v. ICON\\n\\nHealth & Fitness, Inc., 572 U.S. 545, 553 (2014); Burlington N. & Santa Fe Ry.\\n\\nCo. v. United States, 556 U.S. 599, 611 (2009); United States v. Zuniga-Arteaga,\\n\\n681 F.3d 1220, 1224 (11th Cir. 2012); Arriaga v. Fla. Pac. Farms, LLC, 305 F.3d\\n\\n1228, 1242 (11th Cir. 2002). This does not make these cited definitions of\\n\\n“botanical” dispositive; dictionaries are not controlling and in any event give\\n\\nexamples to convey a term’s most common uses, not necessarily to suggest limits.\\n\\nBut the narrower connotation suggested by the dictionaries is consistent with\\n\\nDSHEA’s use of the term “herb or other botanical” rather than a broader term\\n\\nplainly encompassing all plant life.\\n\\n To be sure, the difference between Hi-Tech’s broad view—all flora—and\\n\\nthe narrower dictionary definitions is not as stark as might appear at first blush.\\n\\nThat a substance derived from a plant is used in a dietary product brings it close to\\n\\nthe current dictionary definition, which includes a medicinal preparation derived\\n\\nfrom a plant.\\n\\n Still, the use of “herb or other botanical” in the statute, together with the\\n\\ndictionary definitions of a botanical as “derived from a plant,” supports a much\\n\\nnarrower construction than Hi-Tech proposes. Had Congress meant all plants and\\n\\nanything contained in them, it could have said so. It did not. At the least, the\\n\\n 9\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 10 of 27\\n\\n\\nstatutory language and dictionary definitions support a conclusion that would be\\n\\nreasonable anyway: it is unlikely that Congress used the term “herb or other\\n\\nbotanical” to mean a substance invented in a laboratory and artificially produced,\\n\\nthat can be found in a plant only in trace amounts, only coincidentally, and that has\\n\\nnever been derived from a plant for use in any medicinal, cosmetic, or dietary\\n\\nproduct.\\n\\n VI. “Constituent”\\n\\n The statutory definition of a dietary supplement extends not only to an “herb\\n\\nor other botanical” but also to “a concentrate, metabolite, constituent, extract, or\\n\\ncombination of” an herb or other botanical. 21 U.S.C. § 321(ff)(1)(F). Hi-Tech\\n\\nasserts “constituent” means anything naturally contained in. The word could be\\n\\ngiven that meaning, but the connotation is usually not so broad. Indeed, both the\\n\\n1993 edition and the current edition of Merriam-Webster’s define “constituent” as\\n\\n“an essential part.” “Constituent,” Merriam-Webster’s Collegiate Dictionary 248\\n\\n(10th ed. 1993); “Constituent,” Merriam-Webster Online Dictionary 2019,\\n\\navailable at https://www.merriam-webster.com/dictionary/constituent. This\\n\\ndefinition suggests a connotation much narrower than proposed by Hi-Tech and\\n\\ntoo narrow to include the DMAA—if there is any—contained in geraniums.\\n\\n For its part, the FDA says Hi-Tech’s proposed definition of “constituent”\\n\\nwould render superfluous the statute’s inclusion of the word “extract.” The FDA\\n\\n 10\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 11 of 27\\n\\n\\nsays the meaning of “constituent” must be informed by the other words in the\\n\\nstatutory list, under the canon noscitur a sociis. See In re Piazza, 719 F.3d 1253,\\n\\n1263 n.4 (11th Cir. 2013) (discussing this canon); Antonin Scalia & Bryan A.\\n\\nGarner, Reading Law: The Interpretation of Legal Texts (“Reading Law”) 195-98\\n\\n(2012) (same). The “most common effect of the canon is not to establish which of\\n\\ntwo totally different meanings applies but rather to limit a general term to a subset\\n\\nof all the things or actions that it covers—but only according to its ordinary\\n\\nmeaning.” Reading Law at 196.\\n\\n A concentrate or extract of a product is derived from the product in usable\\n\\nform or amount. So is a combination of the product with another substance. A\\n\\nmetabolite, too, is physically derived from a product. If, as Hi-Tech asserts,\\n\\nconstituent means anything contained in, the word is both markedly different from\\n\\nthe others in the list and awkwardly placed—the broadest term in a five-item list\\n\\nbut placed not first or last but in the center.\\n\\n None of this is dispositive. The safest conclusion is this: it is unlikely that\\n\\nCongress used the term “constituent” to mean a substance that is present in a plant\\n\\nin only trace amounts and that has never been derived from a plant for use in any\\n\\nmedicinal, cosmetic, or dietary product.\\n\\n VII. The Reason for the Statutory Presumption\\n\\n\\n\\n\\n 11\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 12 of 27\\n\\n\\n As set out above, DSHEA gives a preference to dietary supplements. The\\n\\nFDA can condemn a dietary supplement as adulterated only on a showing that it\\n\\npresents a “significant or unreasonable risk of illness or injury” under\\n\\nrecommended, suggested, or ordinary conditions of use. 21 U.S.C. § 342(f)(1)(A).\\n\\nThis is, in effect, a rebuttable presumption that the product is safe when used as\\n\\nintended.\\n\\n A principal reason for rebuttable presumptions, whether in statutes or other\\n\\nlegal constructs, is administrative convenience. When a proposition is usually true,\\n\\nit sometimes makes sense to presume it is true, subject only to rebuttal in the\\n\\noccasional instance when it is not true. Perhaps more importantly, at least in the\\n\\nregulatory context, a presumption can avoid unnecessary expense and delay—a\\n\\nperson or entity can go forward with proposed action without awaiting regulatory\\n\\napproval. This approach works best when a proposition is usually true and when\\n\\nthe rebuttable presumption is clear and easily applied—otherwise the unnecessary\\n\\nexpense and delay is not likely to be avoided.\\n\\n DSHEA well illustrates this approach. Congress thought it better to have a\\n\\nclear, administrable rule—dietary supplements are presumed safe, subject only to a\\n\\ncontrary showing—than to require a particularized inquiry in every case. See S.\\n\\nRep. No. 103-410, at 21-22 (1994). A fair inference is that herbs and other\\n\\n\\n\\n\\n 12\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 13 of 27\\n\\n\\nbotanicals and their constituents made the list of favored dietary ingredients\\n\\nbecause consuming them is ordinarily safe.\\n\\n Consuming most herbs or other botanicals, though surely not all, is safe. The\\n\\nsame is true even for most plants, and people have been consuming plants for as\\n\\nlong as there have been people. Congress reasonably could choose to treat any\\n\\nproduct derived from a plant as adulterated only on a showing that it is unsafe. A\\n\\nrebuttable presumption for anything derived from a plant would serve\\n\\nadministrative convenience and avoid delay in introducing a product to the market.\\n\\n It is a stretch, though, to apply the same reasoning to a substance invented in\\n\\na laboratory and artificially produced, that can be found in a plant, if at all, only in\\n\\ntrace amounts, only coincidentally, and that has never been derived from a plant\\n\\nfor use in any medicinal, cosmetic, or dietary product. The fact that DMAA can be\\n\\nfound in trace amounts in geraniums, if true, says absolutely nothing about whether\\n\\nconsuming the substance is safe.\\n\\n Nor does applying a rebuttable presumption to a substance of this kind serve\\n\\nadministrative convenience. It is easy enough to identify plants or substances\\n\\nactually derived from plants. But as this case illustrates, it is not always easy to\\n\\ndetermine whether a product invented in a laboratory and artificially manufactured\\n\\ncan be found in trace amounts in some plant somewhere in the world.\\n\\n\\n\\n\\n 13\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 14 of 27\\n\\n\\n There is no reason to believe that when it adopted DSHEA, Congress\\n\\nintended to put in place a rebuttable presumption that such a product is safe. We\\n\\nhold that DSHEA does not go that far.\\n\\n This does not mean that DSHEA applies only to products actually derived\\n\\nfrom plants, not those artificially manufactured. If a product is indeed a dietary\\n\\nsupplement because it contains a qualifying dietary ingredient—including, for\\n\\nexample, an herb or other botanical—a manufacturer may take the dietary\\n\\ningredient from nature or produce it artificially. But there must be a qualifying\\n\\ndietary ingredient. The ability to create a substance in a laboratory and\\n\\nmanufacture it artificially does not give a substance that status. Nor does\\n\\ncoincidentally identifying the substance in trace amounts in some plant somewhere\\n\\nin the world.\\n\\n VIII. Generally Recognized as Safe\\n\\n Hi-Tech says DMAA is generally recognized as safe—or, to quote the\\n\\nstatute’s more exacting standard, DMAA is “generally recognized, among experts\\n\\nqualified by scientific training and experience to evaluate its safety, as having been\\n\\nadequately shown through scientific procedures . . . to be safe under the conditions\\n\\nof its intended use.” 21 U.S.C. § 321(s). The FDA’s rule on this concludes that a\\n\\nsubstance meets this standard only when, based on “common knowledge\\n\\nthroughout the scientific community knowledgeable about the safety of substances\\n\\n 14\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 15 of 27\\n\\n\\ndirectly or indirectly added to food,” there is “reasonable certainty that the\\n\\nsubstance is not harmful under the conditions of its intended use.” 21 C.F.R.\\n\\n§ 170.30(a).\\n\\n As the statutory requirement for general recognition makes clear, the issue is\\n\\nnot whether, as an original matter, the factfinder in a legal proceeding would\\n\\nevaluate the evidence and conclude that a substance is safe. The issue is only\\n\\nwhether the substance is generally recognized as safe among qualified experts\\n\\nbased on adequate studies. To establish the contrary, the FDA “need only show the\\n\\nlack of the proper reputation . . . for safety of the [substance] among the\\n\\nappropriate experts, or that what reputation there is, is not based on adequate\\n\\nstudies.” United States v. Articles of Food & Drug Consisting of Coli-Trol 80,\\n\\nF4C-60 Feed Grade, Entrol-S Medicated, Entrol-P, 518 F.2d 743, 746 (5th Cir.\\n\\n1975). As a pre-Bonner decision of the Fifth Circuit, Coli-Trol remains binding in\\n\\nthis court. See Bonner v. City of Prichard, 661 F.2d 1206, 1207 (11th Cir. 1981)\\n\\n(en banc). Other circuits, too, have enforced the requirement for general\\n\\nrecognition. See United States v. Article of Food, 752 F.2d 11, 15 n.4 (1st Cir.\\n\\n1985); Premo Pharmaceutical Labs., Inc. v. United States, 629 F.2d 795, 803-05\\n\\n(2d Cir. 1980).\\n\\n The FDA made the required showing. Multiple sources, including in peer-\\n\\nreviewed publications, call into question DMAA’s safety. Among their\\n\\n 15\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 16 of 27\\n\\n\\nconclusions: DMAA may cause increases in blood pressure and hemorrhagic\\n\\nstroke; individuals with blood pressure of 120/80 mmHg or higher (much of the\\n\\nAmerican population) should avoid DMAA; use of DMAA has been associated\\n\\nwith multiple adverse events, including deaths; and DMAA may inhibit activity of\\n\\nliver enzymes and cause liver toxicity.\\n\\n After four soldiers died with DMAA in their systems, the Department of\\n\\nDefense removed all DMAA products from military exchanges and commissioned\\n\\na Safety Review Panel. The Panel issued a report finding that “deaths, hepatic\\n\\nfailure, myocardial infarction, heat stroke and rhabdomyolysis, seizure and stroke”\\n\\nwere temporally associated with service members’ “use of [DMAA-containing]\\n\\nproducts.” U.S. Dep’t of Def., Report of the Department of Defense 1,3\\n\\nDimethylamylamine (DMAA) Safety Review Panel 9 (2013). The report said this\\n\\nsuggested that some individuals “may be predisposed to severe health\\n\\nconsequences after using DMAA.” Id. The report said there appeared to be\\n\\n“significant association of DMAA use, particularly high frequency DMAA use,\\n\\nand multiple adverse events.” Id. And the report concluded that “the available\\n\\nevidence supports an elevated health risk associated with the use of DMAA-\\n\\ncontaining products.” Id. The Department continued its ban on DMAA products at\\n\\nmilitary exchanges. Id. at 10-11.\\n\\n\\n\\n\\n 16\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 17 of 27\\n\\n\\n With this track record, it is hardly surprising that the FDA’s expert in food\\n\\nchemical risk management determined that DMAA is not generally recognized as\\n\\nsafe by qualified experts.\\n\\n Hi-Tech asserts, though, that the studies and reports on which the FDA relies\\n\\ninvolve DMAA use in doses greater than Hi-Tech recommends. Hi-Tech says that\\n\\nuse of DMAA as intended does not present the same risks. Hi-Tech cites studies\\n\\nand presents expert testimony concluding that DMAA is safe at the recommended\\n\\ndoses.\\n\\n Hi-Tech’s submissions are far from conclusive. The studies use small\\n\\nsample sizes and look at short-term results. None measure the effect of DMAA in\\n\\nhigh-risk populations or on individuals with elevated blood pressure. And while\\n\\nsome but not all of the FDA’s cited studies involve high doses of DMAA, it seems\\n\\nunlikely that all the adverse events suffered by military personnel and others\\n\\nresulted from abnormal or unintended use. Correlation is not causation, but neither\\n\\nmust correlation be ignored.\\n\\n If the issue was whether DMAA is safe, Hi-Tech’s evidence would create a\\n\\ngenuine issue of fact precluding summary judgment; neither side’s evidence is\\n\\nconclusive. See Sparling v. Doyle, No. EP-13-CV-323-DCG, 2015 WL 4528759 at\\n\\n*20 (W.D. Tex. July 27, 2015) (“It is clear . . . that the scientific literature on\\n\\nDMAA presents insufficient data to conclude that DMAA is safe or that DMAA\\n\\n 17\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 18 of 27\\n\\n\\ncauses harm because the sample sizes are too small.”). But the issue is not whether\\n\\nDMAA is safe; the issue is only whether DMAA is generally recognized as safe. It\\n\\nplainly is not. On the issue of general recognition, the FDA was entitled to\\n\\nsummary judgment.\\n\\n IX. The Motion to Reopen Discovery\\n\\n The district court provided ample time for discovery—the full amount the\\n\\nparties requested. The parties submitted cross-motions for summary judgment\\n\\nwithout asking for more time or asserting that any further discovery was needed.\\n\\nBut after the court granted summary judgment for the FDA, Hi-Tech moved to\\n\\nreconsider, taking issue with the court’s legal analysis and asserting the court\\n\\nshould reopen discovery. Hi-Tech said it needed more discovery because the\\n\\ncourt’s legal analysis did not match up with the position argued by either side.\\n\\n Ours is an adversary system. When, as here, there are two sides, each side is\\n\\nafforded the opportunity to argue its position. But the court is not limited to\\n\\nchoosing one side’s position or the other’s. The court’s role is to get it right, not to\\n\\nchoose which side’s argument is better and adopt it lock, stock, and barrel. See,\\n\\ne.g., United States v. Baston, 818 F.3d 651, 663 (11th Cir. 2016) (concluding that\\n\\non a disputed legal issue, “[n]either party is correct,” and applying the correct\\n\\nstandard that neither party advocated); see also Colburn v. Odom, 911 F.3d 1110\\n\\n(11th Cir. 2018) (resolving an appeal on a ground not addressed in either side’s\\n\\n 18\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 19 of 27\\n\\n\\nbrief but essential to proper resolution of the dispute). Were it otherwise, there\\n\\nwould be no plain-error doctrine.\\n\\n Thousands of cases could be cited illustrating this principle. Indeed, the\\n\\nprinciple is so well settled that it is rarely mentioned. When a court adopts a view\\n\\nof the law that is not precisely in line with either side’s argument, the court usually\\n\\nsets out its view of the law without citing authority for the proposition that it may\\n\\ndo so. The Supreme Court has explained it this way: “[w]hen an issue or claim is\\n\\nproperly before the court, the court is not limited to the particular legal theories\\n\\nadvanced by the parties, but rather retains the independent power to identify and\\n\\napply the proper construction of governing law.” U.S. Nat’l Bk. of Oregon v.\\n\\nIndependent Ins. Agents of Am., Inc., 508 U.S. 439, 446 (1993) (quoting Kamen v.\\n\\nKemper Fin. Servs., Inc., 500 U.S. 90, 99 (1991)).\\n\\n Hi-Tech is correct that the district court did not simply accept either side’s\\n\\nview of the facts and law. Nor should the court have done so; neither side had it\\n\\njust right. Similarly, on appeal, we have not simply chosen one side’s view or the\\n\\nother’s; we have considered the arguments and provided the analysis we believe is\\n\\ncorrect. One would expect nothing less.\\n\\n Hi-Tech says, though, that it was blindsided when the district court\\n\\nemphasized that DMAA has never actually been derived from geraniums for use in\\n\\nany product. Hi-Tech says it needs more discovery to fully present its position on\\n\\n 19\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 20 of 27\\n\\n\\nthis issue—to attempt to find evidence that DMAA has in fact been derived from\\n\\ngeraniums.\\n\\n The assertion misses the mark for two reasons, either of which would be\\n\\nsufficient standing alone.\\n\\n First, Hi-Tech could not have been surprised that the court considered\\n\\nwhether DMAA has actually been derived from geraniums. The question leaps off\\n\\nthe page at anyone first considering the issues in this case. Hi-Tech asserts it does\\n\\nnot matter whether DMAA has actually been derived from geraniums—a colorable\\n\\nposition—but Hi-Tech could not have missed the possibility that a court would\\n\\ndisagree.\\n\\n Second, regardless of whether Hi-Tech recognized or should have\\n\\nrecognized that a court might find actual derivation critical, Hi-Tech had every\\n\\nincentive to fully develop the facts on this during the original discovery period. An\\n\\nintensely disputed issue was whether DMAA was contained in geraniums. Hi-Tech\\n\\nsaid yes; the FDA said no. The best support Hi-Tech could have garnered for its\\n\\nposition on this issue—as Hi-Tech surely knew—was evidence that DMAA had\\n\\nactually been derived from geraniums. The reason one can’t get blood from a\\n\\nturnip is that there is no blood in a turnip. The reason one can get juice from an\\n\\norange is that oranges are full of juice. The reason Hi-Tech found no evidence\\n\\nduring the original discovery period that DMAA had actually been derived from\\n\\n 20\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 21 of 27\\n\\n\\ngeraniums was not because Hi-Tech didn’t know to look; it was because no such\\n\\nevidence existed. Or perhaps because, despite every incentive to do so, Hi-Tech\\n\\ncouldn’t find it in the ample time it requested—and the court provided—for\\n\\ndiscovery. Hi-Tech is not entitled to more time.\\n\\n The district court did not abuse its discretion when it declined to reopen\\n\\ndiscovery.\\n\\n X. The Absence of Rulemaking\\n\\n Hi-Tech faults the FDA for bringing a forfeiture action rather than\\n\\nproceeding through rulemaking. But it is “well established” that “agencies have\\n\\ndiscretion to choose whether to proceed by rulemaking or adjudication.” RTC\\n\\nTransp. Inc. v. ICC, 731 F.2d 1502, 1505 (11th Cir. 1984). Not surprisingly, then,\\n\\nwe have upheld a forfeiture judgment in favor of the FDA against a food additive\\n\\nwithout requiring rulemaking. See United States v. Articles of Food & Drug\\n\\nConsisting of Coli-Trol 80, F4C-60 Feed Grade, Entrol-S Medicated, Entrol-P,\\n\\n518 F.2d 743, 746 (5th Cir. 1975); see also United States v. Article of Food, 752\\n\\nF.2d 11, 15-16 (1st Cir. 1985). The FDA was not required to engage in rulemaking\\n\\nbut could elect instead to proceed through a forfeiture action against Hi-Tech’s\\n\\nDMAA products.\\n\\n Proceeding in this manner did not violate the Constitution. The governing\\n\\nstatute provides notice that unapproved food additives are subject to forfeiture. 21\\n\\n 21\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 22 of 27\\n\\n\\nU.S.C. § 334(a)(1). The statute is not unconstitutionally vague, and Hi-Tech\\n\\ndoesn’t claim it is. As part of the forfeiture proceeding, Hi-Tech was afforded the\\n\\nfull range of procedural due process available in a federal court. The issues were\\n\\njoined and fully adjudicated on the merits. Due process requires nothing more.\\n\\n XI. Conclusion\\n\\n DMAA is not an “herb or other botanical.” It is not a “constituent” of an\\n\\nherb or other botanical. And it is not generally recognized by qualified experts, as\\n\\nadequately shown through scientific procedures, to be safe under the conditions of\\n\\nits intended use. The district court properly so ruled. The decision is\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 22\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 23 of 27\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434999/\", \"author_raw\": \"HINKLE, District Judge:\"}, {\"author\": \"JORDAN, Circuit Judge, concurring\", \"type\": \"concurrence\", \"text\": \"JORDAN, Circuit Judge, concurring in part and dissenting in part.\\n\\n This is a difficult case, and in my opinion there is no “right” or “wrong”\\n\\nanswer to the principal statutory question we confront. The majority opinion sets\\n\\nout one plausible interpretation of 21 U.S.C. § 321(ff)(1)(C) & (F), but I read the\\n\\nstatute differently. So, although I join Parts I–IV and VIII–X of the majority opinion,\\n\\nI respectfully dissent from Parts V–VII.\\n\\n *****\\n\\n As relevant here, § 321(ff)(1)(C) & (F) provides that a product is a “dietary\\n\\ningredient”—and therefore can be marketed without FDA pre-approval—if it\\n\\ncontains “an herb or other botanical” or a “concentrate, metabolite, constituent,\\n\\nextract, or combination of any ingredient” in an “herb or other botanical.” Hi-Tech\\n\\ncontends that DMAA satisfies these definitions because it is a “constituent” of a\\n\\ngeranium plant and therefore a “constituent” of a “botanical.” See Br. for Appellant\\n\\nat 8. So we need to figure out what the words “herb,” “botanical,” and “constituent”\\n\\nmean.\\n\\n The principal dictionary definition for the word “herb” concerns its status as\\n\\nflora: a plant whose stem is not woody and persistent, and which generally dies at\\n\\nthe end of its flowering or growing season. See The American Heritage Dictionary\\n\\nof the English Language 820 (4th ed. 2009); Webster’s Third New International\\n\\nDictionary of the English Language Unabridged 1058 (2002); 1 Shorter Oxford\\n\\n 23\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 24 of 27\\n\\n\\nEnglish Dictionary 1228 (5th ed. 2002). It is also, but secondarily, defined as a part\\n\\nof a plant that is useful for food or medicine. See id. (“A . . . plant used for flavoring\\n\\nor scent, in medicine, etc.”).\\n\\n Some dictionary definitions of the noun “botanical” refer to a drug, medicinal\\n\\npreparation, or similar substance obtained or derived from a plant or several plants.\\n\\nSee The American Heritage Dictionary of the English Language 215 (4th ed. 2009);\\n\\nThe Random House College Dictionary 157–58 (1973). Some even refer to the drug\\n\\nor preparation as crude, or maintaining the ingredient more or less in its natural state.\\n\\nSee Webster’s Third New International Dictionary of the English Language\\n\\nUnabridged 258 (2002); McGraw-Hill Dictionary of Scientific and Technical Terms\\n\\n272 (6th ed. 2003). But as the FDA concedes, see Br. for Appellee at 16, “botanical”\\n\\nalso is defined as the plant (or part of the plant) itself. See, e.g., Merriam-Webster’s\\n\\nCollegiate Dictionary 134 (10th ed. 1994) (“a plant part or extract used sp. in skin\\n\\nand hair care products”); The American Heritage Dictionary of the English\\n\\nLanguage 298 (3d ed. 1993) (“of or relating to plants or plant life”).\\n\\n The statute uses “other botanical” in conjunction with “herb.” It therefore\\n\\nseems to me that the word “botanical” contextually refers to a plant or part of a plant,\\n\\nand not a drug or medicinal preparation derived from a plant. See generally Dole v.\\n\\nUnited Steelworkers of America, 494 U.S. 26, 36 (1990) (explaining that “words\\n\\n\\n\\n\\n 24\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 25 of 27\\n\\n\\ngrouped in a list should be given related meaning”) (internal quotation marks and\\n\\ncitation omitted). And a geranium is certainly a plant.\\n\\n That leaves the word “constituent.” It means a component or element of a\\n\\nwhole, and—significantly—not all dictionaries require the component or element to\\n\\nbe “essential.” See, e.g., 1 Shorter Oxford English Dictionary 496 (5th ed. 2002)\\n\\n(“an element of a complex whole”); The American Heritage Dictionary of the\\n\\nEnglish Language 394 (4th ed. 2009) (“[s]erving as part of a whole; component”);\\n\\nWebster’s Third New International Dictionary of the English Language Unabridged\\n\\n258 (2002) (“a thing, person, or organism that along with others serves in making\\n\\nup a complete whole or unit”).\\n\\n As the majority acknowledges, there is evidence that geraniums contain a\\n\\ntrace amount of DMAA. See Maj. Op. at 7–8. There is also evidence, however, that\\n\\nsome fertilizers contain DMAA—which could be the source of trace amounts in\\n\\ngeraniums—and that geraniums have no known pathways of producing DMAA. Id.\\n\\nViewing the record in the light most favorable to Hi-Tech, there is a genuine issue\\n\\nof material fact as to whether DMAA—even in trace amounts—is a “constituent”\\n\\n(i.e., a component or element) of geraniums.\\n\\n *****\\n\\n In my view, the statutory text does not provide a basis for the district court’s\\n\\nconclusion that a “constituent” of a “botanical” must have a history of being\\n\\n 25\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 26 of 27\\n\\n\\nextracted in usable quantities, or for the majority’s holding that to be a “constituent”\\n\\nan ingredient must have been derived from a plant for use in a medicinal, cosmetic,\\n\\nor dietary product. Indeed, reading “constituent” to mean something that has been\\n\\ntaken out of a plant in usable amounts may make “extract”—another statutory\\n\\nterm—surplusage.\\n\\n The statute lists “constituent” among several other words: “a concentrate,\\n\\nmetabolite, constituent, extract, or combination thereof.” § 321(ff)(1)(F). When\\n\\nCongress uses “or” to separate several words in a list, that term’s “ordinary use is\\n\\nalmost always disjunctive, that is, the words it connects are to be given separate\\n\\nmeanings.” Loughrin v. United States, 573 U.S. 351, 357 (2014). As a noun, the\\n\\nterm “extract” means “something extracted . . . a preparation obtained by\\n\\nevaporation (as of a solution of a drug or the juice of a plant).” Webster’s Third\\n\\nNew International Dictionary of the English Language Unabridged 806 (2002).\\n\\nAgain, “constituent” is broadly defined as a part of something else, and ascribing a\\n\\nmore narrow definition would eliminate any independent meaning Congress\\n\\nintended by using “extract.” See Yates v. United States, 135 S. Ct. 1074, 1085 (2015)\\n\\n(explaining that courts should “avoid ascribing to one word a meaning so broad that\\n\\nit is inconsistent with its accompanying words”) (internal quotation omitted).\\n\\n The majority’s contrary interpretation of § 321(ff)(1)(C) & (F) seems\\n\\ninfluenced by policy reasons which call for a narrower reading of the statutory text.\\n\\n 26\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 27 of 27\\n\\n\\nSee Maj. Op. at 13–14. I do not challenge those reasons, but believe they are not\\n\\nours to consider. See Sturges v. Crowninshield, 17 U.S. 122, 202 (1819) (we should\\n\\nnot “infer from extrinsic circumstances, that a case for which the words of an\\n\\ninstrument expressly provide, shall be exempted from its operation”). Although the\\n\\nstatutory reading advocated by Hi-Tech is expansive, that reading squares with the\\n\\nbroad language Congress chose. As the Supreme Court has told us, “the fact that a\\n\\nstatute can be applied in situations not expressly anticipated by Congress does not\\n\\ndemonstrate ambiguity. It demonstrates breadth.” Pa. Dept. of Corrections v.\\n\\nYeskey, 524 U.S. 206, 212 (1998) (internal quotation marks and citation omitted).\\n\\n *****\\n\\n As I read the statute and the record, the FDA was not entitled to summary\\n\\njudgment. I would remand for a trial on whether DMAA is a “constituent” of\\n\\ngeraniums.\\n\\n\\n\\n\\n 27\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4434999/\", \"author_raw\": \"JORDAN, Circuit Judge, concurring\"}]}"},"judge1":{"kind":"string","value":"TJOFLAT"},"judge2":{"kind":"string","value":"JORDAN"},"judge3":{"kind":"string","value":"HINKLE"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"TJOFLAT\": \", Circuit\", \"JORDAN\": \", Circuit\", \"HINKLE\": \", District\"}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":1,"string":"1"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4657746/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: United States v. Hi-Tech Pharmaceuticals, Inc.\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-30\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: Yes\nAdditional notes: \n\n[CASE_TEXT]\nCase: 17-13376 Date Filed: 08/30/2019 Page: 1 of 27\n\n\n [PUBLISH]\n\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n ________________________\n\n No. 17-13376\n ________________________\n\n D.C. Docket No. 1:13-cv-03675-WBH\n\n\nUNITED STATES OF AMERICA,\n\n Plaintiff - Appellee,\n\nversus\n\nUNDETERMINED QUANTITIES OF ALL ARTICLES OF\nFINISHED AND IN-PROCESS FOODS,\nraw ingredients (bulk powders, bulk capsules), with any lot number,\nsize, or type container, whether labeled or unlabeled, et al.,\n\n Defendants,\n\nHI-TECH PHARMACEUTICALS, INC.,\nJARED WHEAT,\n\n Claimants - Appellants.\n\n\n ________________________\n\n Appeal from the United States District Court\n for the Northern District of Georgia\n ________________________\n\n (August 30, 2019)\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 2 of 27\n\n\nBefore TJOFLAT and JORDAN, Circuit Judges, and HINKLE, * District Judge.\n\nHINKLE, District Judge:\n\n The Dietary Supplement Health and Education Act of 1994 provides\n\nfavorable treatment for “dietary supplements,” defined to include any “botanical”\n\nor “constituent” of a botanical. This case presents the question whether these terms\n\napply to a substance that was invented in a laboratory and is artificially produced\n\nfor commercial sale but that, entirely coincidentally, may be found in trace\n\namounts in a plant. We hold that the terms do not extend this far.\n\n I. Proceedings\n\n The Food and Drug Administration seized from Hi-Tech Pharmaceuticals,\n\nInc. a substantial quantity of products containing 1,3-dimethylamylamine or\n\n“DMAA.” DMAA is used in fitness products aimed at bodybuilders and other\n\nathletes.\n\n The seizure led to two actions that were consolidated in the district court.\n\nOne was a forfeiture action filed by the United States against the products. Hi-Tech\n\nand its chief executive officer, Jared Wheat, intervened as claimants. Hi-Tech filed\n\nthe other action against the FDA and other governmental defendants.\n\n\n\n\n *\n Honorable Robert L. Hinkle, United States District Judge for the Northern District of\nFlorida, sitting by designation.\n 2\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 3 of 27\n\n\n Hi-Tech asserted that DMAA is a dietary supplement; that under the\n\nAdministrative Procedure Act the FDA can properly ban DMAA, if at all, only\n\nthrough rulemaking; and that the seizure of Hi-Tech’s DMAA violated the Fifth\n\nAmendment’s Due Process Clause.\n\n The parties filed cross-motions for summary judgment. The district court\n\ngranted the FDA’s motion, holding the seizure proper both substantively and\n\nprocedurally. The district court denied a motion to reconsider that included a\n\nrequest to reopen discovery. Hi-Tech and Mr. Wheat have appealed. The appeal\n\nhas been fully briefed and orally argued.\n\n II. Standard of Review\n\n We review de novo the district court’s grant of summary judgment. See, e.g.,\n\nPrice v. Comm’r, Dep’t of Corr., 920 F.3d 1317, 1323 (11th Cir. 2019). We review\n\nfor abuse of discretion the district court’s denial of the motion for reconsideration\n\nand refusal to reopen discovery. See, e.g., Corwin v. Walt Disney Co., 475 F.3d\n\n1239, 1254 (11th Cir. 2007) (reconsideration); Artistic Entm’t, Inc. v. City of\n\nWarner Robins, 331 F.3d 1196, 1202 (11th Cir. 2003) (reopening discovery).\n\n III. The Statute and the Issues\n\n The Federal Food, Drug, and Cosmetic Act prohibits the introduction of\n\nadulterated foods into interstate commerce. 21 U.S.C. § 331(a). The FDA enforces\n\nthe Act. Id. § 393(b)(2)(A). The agency may bring an in rem forfeiture action in\n\n 3\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 4 of 27\n\n\ndistrict court to condemn adulterated foods. Id. § 334(a)(1). Hi-Tech’s DMAA\n\nproducts were adulterated foods if they were “food additives” but not if they were\n\n“dietary supplements.”\n\n The background is this. The Dietary Supplement Health and Education Act\n\nof 1994, commonly referred to as “DSHEA,” amended the Federal Food, Drug,\n\nand Cosmetic Act to provide favorable treatment for dietary supplements. The\n\nstatute’s definition of “dietary supplement” includes multiple parts. 21 U.S.C.\n\n§ 321(ff). The only part relevant to Hi-Tech’s DMAA is this: a product that is\n\nintended to supplement the diet—this includes DMAA—is a dietary supplement if\n\nit contains “an herb or other botanical” or “a concentrate, metabolite, constituent,\n\nextract, or combination of” an herb or other botanical. Id. § 321(ff)(1)(C) & (F).\n\nThe statute describes these—as well as other substances not at issue here—as\n\n“dietary ingredients.”\n\n Under DSHEA, and subject to exceptions not relevant here, a dietary\n\nsupplement can be condemned as adulterated only if the FDA carries the burden of\n\nproving that the substance presents a “significant or unreasonable risk of illness or\n\ninjury” under recommended, suggested, or ordinary conditions of use. Id.\n\n§ 342(f)(1)(A). The FDA did not attempt to make that showing for the DMAA\n\nproducts it seized from Hi-Tech. A ruling that DMAA is a dietary supplement thus\n\nwould resolve this appeal in Hi-Tech’s favor.\n\n 4\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 5 of 27\n\n\n On the other hand, a ruling that DMAA is a “food additive” would resolve\n\nthe dispute in the FDA’s favor. A substance intended for human consumption is a\n\nfood additive if it is not a dietary supplement and is not “generally recognized,\n\namong experts qualified by scientific training and experience to evaluate its safety,\n\nas having been adequately shown through scientific procedures . . . to be safe under\n\nthe conditions of its intended use.” Id. § 321(s). For a substance in common use in\n\nfood prior to January 1, 1958—this does not include DMAA—the adequate\n\nshowing of safety can be made not only by scientific procedures but also by\n\nexperience. There are other exceptions to this definition of “food additive,” but\n\nnone applies here.\n\n The FDA asserts that DMAA is not a dietary supplement, is not generally\n\nrecognized as safe, does not meet any other exception, and is therefore a food\n\nadditive. Hi-Tech insists that DMAA is a dietary supplement and thus is not a food\n\nadditive, but that even if DMAA is not a dietary supplement, DMAA is generally\n\nrecognized as safe and thus still is not a food additive.\n\n The issues thus are first, whether Hi-Tech’s DMAA products are “an herb or\n\nother botanical” or “a concentrate, metabolite, constituent, extract, or combination\n\nof” an herb or other botanical, and second, if not, whether the products are\n\ngenerally recognized as safe. Secondary issues are whether the FDA was entitled to\n\n\n\n\n 5\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 6 of 27\n\n\nseize and forfeit the products without engaging in rulemaking and whether the\n\ndistrict court should have reopened discovery.\n\n IV. DMAA\n\n The earliest known identification or use of DMAA occurred in 1944. In that\n\nyear Eli Lilly & Co. synthesized and patented DMAA for use as a nasal\n\ndecongestant. For marketing reasons, Eli Lilly asked the FDA to withdraw its\n\napproval of this use in 1983. At least insofar as shown by this record, DMAA was\n\nnot used as a dietary supplement or food additive at that time, and no health\n\nconcerns had been noted.\n\n DMAA eventually made a resurgence, this time in fitness products aimed at\n\nbodybuilders and other athletes. Because of DMAA’s noticeable stimulant effect,\n\nthe compound made its way into pre-workout energy and fat-burner products\n\naround the world.\n\n The FDA eventually adopted the position that DMAA is not a dietary\n\nsupplement but an unsafe food additive. The FDA issued cease-and-desist letters to\n\nat least some entities marketing DMAA products. Perhaps unaware of Hi-Tech’s\n\nmarketing of DMAA products, the FDA did not issue a cease-and-desist letter to\n\nHi-Tech.\n\n Around the same time, researchers began to find trace amounts of DMAA in\n\ngeraniums of the genus pelargonium. A 2013 survey concluded that overall, the\n\n 6\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 7 of 27\n\n\nstudies showed that DMAA “is found naturally in some, but not all, geranium\n\nplants and extracted geranium oils.” Thomas D. Gauthier, Evidence for the\n\nPresence of 1,3-Dimethylamylamine (1,3-DMAA) in Geranium Plant Materials, 8\n\nAnalytical Chemistry Insights 29-40 (2013), available at\n\nhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC3682735/. Indeed, the FDA’s own\n\nexpert had previously participated in a study that found trace amounts of DMAA in\n\ngeraniums.\n\n Even so, this record presents a genuine factual dispute over whether trace\n\namounts of DMAA are naturally contained in geraniums. On the one hand, studies\n\nhave found trace amounts of DMAA in geraniums. On the other hand, some\n\nfertilizers contain DMAA that could be a source of trace amounts of DMAA in\n\ngeraniums, and the record includes competent testimony that there is no known\n\npathway by which geraniums could produce DMAA. Either way, it is clear that\n\nDMAA is not contained in geraniums in amounts greater than could reasonably be\n\ncharacterized as trace amounts. No study has found a greater amount.\n\n V. “Herb or Other Botanical”\n\n The first rule of statutory construction is to apply the plain meaning of the\n\nstatutory language. See, e.g., Bankston v. Then, 615 F.3d 1364, 1367 (11th Cir.\n\n2010). Here the meaning is not completely clear.\n\n\n\n\n 7\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 8 of 27\n\n\n Hi-Tech says DSHEA uses “botanical” to mean all plant life, nothing more\n\nand nothing less—that is, to mean flora, without limitation. The suggestion is\n\nsensible enough—“botany” is the study of plants. On the other hand, it would be\n\npassing strange for a writer wishing to cover the universe of plant life—to mean all\n\nflora—to achieve that result through the term “herb or other botanical.” Moreover,\n\nthe usual connotation of “botanical” when used as a noun, as recognized in\n\ndictionaries in use when DSHEA was enacted as well as those is use today, is a\n\nsubstance derived from a plant used for a limited category of purposes.\n\n In 1993, a year before DSHEA became law, Merriam-Webster’s defined the\n\nnoun “botanical” as a “a plant part or extract used esp[ecially] in skin and hair care\n\nproducts.” “Botanical,” Merriam-Webster’s Collegiate Dictionary 134 (10th ed.\n\n1993). The current edition defines the noun “botanical” as a “substance obtained or\n\nderived from a plant[,] such as . . . a plant part or extract used especially in skin\n\nand hair care products[,] a medicinal preparation derived from a plant[, or] plant\n\nmaterial used as a flavoring agent.” “Botanical,” Merriam-Webster’s Online\n\nDictionary 2019, available at https://www.merriam-\n\nwebster.com/dictionary/botanical. Neither definition suggests that the noun\n\n“botanical” includes an artificially produced substance that, entirely coincidentally,\n\nmay be found in trace amounts in a plant. Nor do they suggest that “botanical”\n\nincludes all flora.\n\n 8\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 9 of 27\n\n\n Both the Supreme Court and the Eleventh Circuit have relied on Merriam-\n\nWebster’s as an aid in construing statutes. See, e.g., Octane Fitness, LLC v. ICON\n\nHealth & Fitness, Inc., 572 U.S. 545, 553 (2014); Burlington N. & Santa Fe Ry.\n\nCo. v. United States, 556 U.S. 599, 611 (2009); United States v. Zuniga-Arteaga,\n\n681 F.3d 1220, 1224 (11th Cir. 2012); Arriaga v. Fla. Pac. Farms, LLC, 305 F.3d\n\n1228, 1242 (11th Cir. 2002). This does not make these cited definitions of\n\n“botanical” dispositive; dictionaries are not controlling and in any event give\n\nexamples to convey a term’s most common uses, not necessarily to suggest limits.\n\nBut the narrower connotation suggested by the dictionaries is consistent with\n\nDSHEA’s use of the term “herb or other botanical” rather than a broader term\n\nplainly encompassing all plant life.\n\n To be sure, the difference between Hi-Tech’s broad view—all flora—and\n\nthe narrower dictionary definitions is not as stark as might appear at first blush.\n\nThat a substance derived from a plant is used in a dietary product brings it close to\n\nthe current dictionary definition, which includes a medicinal preparation derived\n\nfrom a plant.\n\n Still, the use of “herb or other botanical” in the statute, together with the\n\ndictionary definitions of a botanical as “derived from a plant,” supports a much\n\nnarrower construction than Hi-Tech proposes. Had Congress meant all plants and\n\nanything contained in them, it could have said so. It did not. At the least, the\n\n 9\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 10 of 27\n\n\nstatutory language and dictionary definitions support a conclusion that would be\n\nreasonable anyway: it is unlikely that Congress used the term “herb or other\n\nbotanical” to mean a substance invented in a laboratory and artificially produced,\n\nthat can be found in a plant only in trace amounts, only coincidentally, and that has\n\nnever been derived from a plant for use in any medicinal, cosmetic, or dietary\n\nproduct.\n\n VI. “Constituent”\n\n The statutory definition of a dietary supplement extends not only to an “herb\n\nor other botanical” but also to “a concentrate, metabolite, constituent, extract, or\n\ncombination of” an herb or other botanical. 21 U.S.C. § 321(ff)(1)(F). Hi-Tech\n\nasserts “constituent” means anything naturally contained in. The word could be\n\ngiven that meaning, but the connotation is usually not so broad. Indeed, both the\n\n1993 edition and the current edition of Merriam-Webster’s define “constituent” as\n\n“an essential part.” “Constituent,” Merriam-Webster’s Collegiate Dictionary 248\n\n(10th ed. 1993); “Constituent,” Merriam-Webster Online Dictionary 2019,\n\navailable at https://www.merriam-webster.com/dictionary/constituent. This\n\ndefinition suggests a connotation much narrower than proposed by Hi-Tech and\n\ntoo narrow to include the DMAA—if there is any—contained in geraniums.\n\n For its part, the FDA says Hi-Tech’s proposed definition of “constituent”\n\nwould render superfluous the statute’s inclusion of the word “extract.” The FDA\n\n 10\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 11 of 27\n\n\nsays the meaning of “constituent” must be informed by the other words in the\n\nstatutory list, under the canon noscitur a sociis. See In re Piazza, 719 F.3d 1253,\n\n1263 n.4 (11th Cir. 2013) (discussing this canon); Antonin Scalia & Bryan A.\n\nGarner, Reading Law: The Interpretation of Legal Texts (“Reading Law”) 195-98\n\n(2012) (same). The “most common effect of the canon is not to establish which of\n\ntwo totally different meanings applies but rather to limit a general term to a subset\n\nof all the things or actions that it covers—but only according to its ordinary\n\nmeaning.” Reading Law at 196.\n\n A concentrate or extract of a product is derived from the product in usable\n\nform or amount. So is a combination of the product with another substance. A\n\nmetabolite, too, is physically derived from a product. If, as Hi-Tech asserts,\n\nconstituent means anything contained in, the word is both markedly different from\n\nthe others in the list and awkwardly placed—the broadest term in a five-item list\n\nbut placed not first or last but in the center.\n\n None of this is dispositive. The safest conclusion is this: it is unlikely that\n\nCongress used the term “constituent” to mean a substance that is present in a plant\n\nin only trace amounts and that has never been derived from a plant for use in any\n\nmedicinal, cosmetic, or dietary product.\n\n VII. The Reason for the Statutory Presumption\n\n\n\n\n 11\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 12 of 27\n\n\n As set out above, DSHEA gives a preference to dietary supplements. The\n\nFDA can condemn a dietary supplement as adulterated only on a showing that it\n\npresents a “significant or unreasonable risk of illness or injury” under\n\nrecommended, suggested, or ordinary conditions of use. 21 U.S.C. § 342(f)(1)(A).\n\nThis is, in effect, a rebuttable presumption that the product is safe when used as\n\nintended.\n\n A principal reason for rebuttable presumptions, whether in statutes or other\n\nlegal constructs, is administrative convenience. When a proposition is usually true,\n\nit sometimes makes sense to presume it is true, subject only to rebuttal in the\n\noccasional instance when it is not true. Perhaps more importantly, at least in the\n\nregulatory context, a presumption can avoid unnecessary expense and delay—a\n\nperson or entity can go forward with proposed action without awaiting regulatory\n\napproval. This approach works best when a proposition is usually true and when\n\nthe rebuttable presumption is clear and easily applied—otherwise the unnecessary\n\nexpense and delay is not likely to be avoided.\n\n DSHEA well illustrates this approach. Congress thought it better to have a\n\nclear, administrable rule—dietary supplements are presumed safe, subject only to a\n\ncontrary showing—than to require a particularized inquiry in every case. See S.\n\nRep. No. 103-410, at 21-22 (1994). A fair inference is that herbs and other\n\n\n\n\n 12\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 13 of 27\n\n\nbotanicals and their constituents made the list of favored dietary ingredients\n\nbecause consuming them is ordinarily safe.\n\n Consuming most herbs or other botanicals, though surely not all, is safe. The\n\nsame is true even for most plants, and people have been consuming plants for as\n\nlong as there have been people. Congress reasonably could choose to treat any\n\nproduct derived from a plant as adulterated only on a showing that it is unsafe. A\n\nrebuttable presumption for anything derived from a plant would serve\n\nadministrative convenience and avoid delay in introducing a product to the market.\n\n It is a stretch, though, to apply the same reasoning to a substance invented in\n\na laboratory and artificially produced, that can be found in a plant, if at all, only in\n\ntrace amounts, only coincidentally, and that has never been derived from a plant\n\nfor use in any medicinal, cosmetic, or dietary product. The fact that DMAA can be\n\nfound in trace amounts in geraniums, if true, says absolutely nothing about whether\n\nconsuming the substance is safe.\n\n Nor does applying a rebuttable presumption to a substance of this kind serve\n\nadministrative convenience. It is easy enough to identify plants or substances\n\nactually derived from plants. But as this case illustrates, it is not always easy to\n\ndetermine whether a product invented in a laboratory and artificially manufactured\n\ncan be found in trace amounts in some plant somewhere in the world.\n\n\n\n\n 13\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 14 of 27\n\n\n There is no reason to believe that when it adopted DSHEA, Congress\n\nintended to put in place a rebuttable presumption that such a product is safe. We\n\nhold that DSHEA does not go that far.\n\n This does not mean that DSHEA applies only to products actually derived\n\nfrom plants, not those artificially manufactured. If a product is indeed a dietary\n\nsupplement because it contains a qualifying dietary ingredient—including, for\n\nexample, an herb or other botanical—a manufacturer may take the dietary\n\ningredient from nature or produce it artificially. But there must be a qualifying\n\ndietary ingredient. The ability to create a substance in a laboratory and\n\nmanufacture it artificially does not give a substance that status. Nor does\n\ncoincidentally identifying the substance in trace amounts in some plant somewhere\n\nin the world.\n\n VIII. Generally Recognized as Safe\n\n Hi-Tech says DMAA is generally recognized as safe—or, to quote the\n\nstatute’s more exacting standard, DMAA is “generally recognized, among experts\n\nqualified by scientific training and experience to evaluate its safety, as having been\n\nadequately shown through scientific procedures . . . to be safe under the conditions\n\nof its intended use.” 21 U.S.C. § 321(s). The FDA’s rule on this concludes that a\n\nsubstance meets this standard only when, based on “common knowledge\n\nthroughout the scientific community knowledgeable about the safety of substances\n\n 14\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 15 of 27\n\n\ndirectly or indirectly added to food,” there is “reasonable certainty that the\n\nsubstance is not harmful under the conditions of its intended use.” 21 C.F.R.\n\n§ 170.30(a).\n\n As the statutory requirement for general recognition makes clear, the issue is\n\nnot whether, as an original matter, the factfinder in a legal proceeding would\n\nevaluate the evidence and conclude that a substance is safe. The issue is only\n\nwhether the substance is generally recognized as safe among qualified experts\n\nbased on adequate studies. To establish the contrary, the FDA “need only show the\n\nlack of the proper reputation . . . for safety of the [substance] among the\n\nappropriate experts, or that what reputation there is, is not based on adequate\n\nstudies.” United States v. Articles of Food & Drug Consisting of Coli-Trol 80,\n\nF4C-60 Feed Grade, Entrol-S Medicated, Entrol-P, 518 F.2d 743, 746 (5th Cir.\n\n1975). As a pre-Bonner decision of the Fifth Circuit, Coli-Trol remains binding in\n\nthis court. See Bonner v. City of Prichard, 661 F.2d 1206, 1207 (11th Cir. 1981)\n\n(en banc). Other circuits, too, have enforced the requirement for general\n\nrecognition. See United States v. Article of Food, 752 F.2d 11, 15 n.4 (1st Cir.\n\n1985); Premo Pharmaceutical Labs., Inc. v. United States, 629 F.2d 795, 803-05\n\n(2d Cir. 1980).\n\n The FDA made the required showing. Multiple sources, including in peer-\n\nreviewed publications, call into question DMAA’s safety. Among their\n\n 15\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 16 of 27\n\n\nconclusions: DMAA may cause increases in blood pressure and hemorrhagic\n\nstroke; individuals with blood pressure of 120/80 mmHg or higher (much of the\n\nAmerican population) should avoid DMAA; use of DMAA has been associated\n\nwith multiple adverse events, including deaths; and DMAA may inhibit activity of\n\nliver enzymes and cause liver toxicity.\n\n After four soldiers died with DMAA in their systems, the Department of\n\nDefense removed all DMAA products from military exchanges and commissioned\n\na Safety Review Panel. The Panel issued a report finding that “deaths, hepatic\n\nfailure, myocardial infarction, heat stroke and rhabdomyolysis, seizure and stroke”\n\nwere temporally associated with service members’ “use of [DMAA-containing]\n\nproducts.” U.S. Dep’t of Def., Report of the Department of Defense 1,3\n\nDimethylamylamine (DMAA) Safety Review Panel 9 (2013). The report said this\n\nsuggested that some individuals “may be predisposed to severe health\n\nconsequences after using DMAA.” Id. The report said there appeared to be\n\n“significant association of DMAA use, particularly high frequency DMAA use,\n\nand multiple adverse events.” Id. And the report concluded that “the available\n\nevidence supports an elevated health risk associated with the use of DMAA-\n\ncontaining products.” Id. The Department continued its ban on DMAA products at\n\nmilitary exchanges. Id. at 10-11.\n\n\n\n\n 16\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 17 of 27\n\n\n With this track record, it is hardly surprising that the FDA’s expert in food\n\nchemical risk management determined that DMAA is not generally recognized as\n\nsafe by qualified experts.\n\n Hi-Tech asserts, though, that the studies and reports on which the FDA relies\n\ninvolve DMAA use in doses greater than Hi-Tech recommends. Hi-Tech says that\n\nuse of DMAA as intended does not present the same risks. Hi-Tech cites studies\n\nand presents expert testimony concluding that DMAA is safe at the recommended\n\ndoses.\n\n Hi-Tech’s submissions are far from conclusive. The studies use small\n\nsample sizes and look at short-term results. None measure the effect of DMAA in\n\nhigh-risk populations or on individuals with elevated blood pressure. And while\n\nsome but not all of the FDA’s cited studies involve high doses of DMAA, it seems\n\nunlikely that all the adverse events suffered by military personnel and others\n\nresulted from abnormal or unintended use. Correlation is not causation, but neither\n\nmust correlation be ignored.\n\n If the issue was whether DMAA is safe, Hi-Tech’s evidence would create a\n\ngenuine issue of fact precluding summary judgment; neither side’s evidence is\n\nconclusive. See Sparling v. Doyle, No. EP-13-CV-323-DCG, 2015 WL 4528759 at\n\n*20 (W.D. Tex. July 27, 2015) (“It is clear . . . that the scientific literature on\n\nDMAA presents insufficient data to conclude that DMAA is safe or that DMAA\n\n 17\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 18 of 27\n\n\ncauses harm because the sample sizes are too small.”). But the issue is not whether\n\nDMAA is safe; the issue is only whether DMAA is generally recognized as safe. It\n\nplainly is not. On the issue of general recognition, the FDA was entitled to\n\nsummary judgment.\n\n IX. The Motion to Reopen Discovery\n\n The district court provided ample time for discovery—the full amount the\n\nparties requested. The parties submitted cross-motions for summary judgment\n\nwithout asking for more time or asserting that any further discovery was needed.\n\nBut after the court granted summary judgment for the FDA, Hi-Tech moved to\n\nreconsider, taking issue with the court’s legal analysis and asserting the court\n\nshould reopen discovery. Hi-Tech said it needed more discovery because the\n\ncourt’s legal analysis did not match up with the position argued by either side.\n\n Ours is an adversary system. When, as here, there are two sides, each side is\n\nafforded the opportunity to argue its position. But the court is not limited to\n\nchoosing one side’s position or the other’s. The court’s role is to get it right, not to\n\nchoose which side’s argument is better and adopt it lock, stock, and barrel. See,\n\ne.g., United States v. Baston, 818 F.3d 651, 663 (11th Cir. 2016) (concluding that\n\non a disputed legal issue, “[n]either party is correct,” and applying the correct\n\nstandard that neither party advocated); see also Colburn v. Odom, 911 F.3d 1110\n\n(11th Cir. 2018) (resolving an appeal on a ground not addressed in either side’s\n\n 18\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 19 of 27\n\n\nbrief but essential to proper resolution of the dispute). Were it otherwise, there\n\nwould be no plain-error doctrine.\n\n Thousands of cases could be cited illustrating this principle. Indeed, the\n\nprinciple is so well settled that it is rarely mentioned. When a court adopts a view\n\nof the law that is not precisely in line with either side’s argument, the court usually\n\nsets out its view of the law without citing authority for the proposition that it may\n\ndo so. The Supreme Court has explained it this way: “[w]hen an issue or claim is\n\nproperly before the court, the court is not limited to the particular legal theories\n\nadvanced by the parties, but rather retains the independent power to identify and\n\napply the proper construction of governing law.” U.S. Nat’l Bk. of Oregon v.\n\nIndependent Ins. Agents of Am., Inc., 508 U.S. 439, 446 (1993) (quoting Kamen v.\n\nKemper Fin. Servs., Inc., 500 U.S. 90, 99 (1991)).\n\n Hi-Tech is correct that the district court did not simply accept either side’s\n\nview of the facts and law. Nor should the court have done so; neither side had it\n\njust right. Similarly, on appeal, we have not simply chosen one side’s view or the\n\nother’s; we have considered the arguments and provided the analysis we believe is\n\ncorrect. One would expect nothing less.\n\n Hi-Tech says, though, that it was blindsided when the district court\n\nemphasized that DMAA has never actually been derived from geraniums for use in\n\nany product. Hi-Tech says it needs more discovery to fully present its position on\n\n 19\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 20 of 27\n\n\nthis issue—to attempt to find evidence that DMAA has in fact been derived from\n\ngeraniums.\n\n The assertion misses the mark for two reasons, either of which would be\n\nsufficient standing alone.\n\n First, Hi-Tech could not have been surprised that the court considered\n\nwhether DMAA has actually been derived from geraniums. The question leaps off\n\nthe page at anyone first considering the issues in this case. Hi-Tech asserts it does\n\nnot matter whether DMAA has actually been derived from geraniums—a colorable\n\nposition—but Hi-Tech could not have missed the possibility that a court would\n\ndisagree.\n\n Second, regardless of whether Hi-Tech recognized or should have\n\nrecognized that a court might find actual derivation critical, Hi-Tech had every\n\nincentive to fully develop the facts on this during the original discovery period. An\n\nintensely disputed issue was whether DMAA was contained in geraniums. Hi-Tech\n\nsaid yes; the FDA said no. The best support Hi-Tech could have garnered for its\n\nposition on this issue—as Hi-Tech surely knew—was evidence that DMAA had\n\nactually been derived from geraniums. The reason one can’t get blood from a\n\nturnip is that there is no blood in a turnip. The reason one can get juice from an\n\norange is that oranges are full of juice. The reason Hi-Tech found no evidence\n\nduring the original discovery period that DMAA had actually been derived from\n\n 20\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 21 of 27\n\n\ngeraniums was not because Hi-Tech didn’t know to look; it was because no such\n\nevidence existed. Or perhaps because, despite every incentive to do so, Hi-Tech\n\ncouldn’t find it in the ample time it requested—and the court provided—for\n\ndiscovery. Hi-Tech is not entitled to more time.\n\n The district court did not abuse its discretion when it declined to reopen\n\ndiscovery.\n\n X. The Absence of Rulemaking\n\n Hi-Tech faults the FDA for bringing a forfeiture action rather than\n\nproceeding through rulemaking. But it is “well established” that “agencies have\n\ndiscretion to choose whether to proceed by rulemaking or adjudication.” RTC\n\nTransp. Inc. v. ICC, 731 F.2d 1502, 1505 (11th Cir. 1984). Not surprisingly, then,\n\nwe have upheld a forfeiture judgment in favor of the FDA against a food additive\n\nwithout requiring rulemaking. See United States v. Articles of Food & Drug\n\nConsisting of Coli-Trol 80, F4C-60 Feed Grade, Entrol-S Medicated, Entrol-P,\n\n518 F.2d 743, 746 (5th Cir. 1975); see also United States v. Article of Food, 752\n\nF.2d 11, 15-16 (1st Cir. 1985). The FDA was not required to engage in rulemaking\n\nbut could elect instead to proceed through a forfeiture action against Hi-Tech’s\n\nDMAA products.\n\n Proceeding in this manner did not violate the Constitution. The governing\n\nstatute provides notice that unapproved food additives are subject to forfeiture. 21\n\n 21\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 22 of 27\n\n\nU.S.C. § 334(a)(1). The statute is not unconstitutionally vague, and Hi-Tech\n\ndoesn’t claim it is. As part of the forfeiture proceeding, Hi-Tech was afforded the\n\nfull range of procedural due process available in a federal court. The issues were\n\njoined and fully adjudicated on the merits. Due process requires nothing more.\n\n XI. Conclusion\n\n DMAA is not an “herb or other botanical.” It is not a “constituent” of an\n\nherb or other botanical. And it is not generally recognized by qualified experts, as\n\nadequately shown through scientific procedures, to be safe under the conditions of\n\nits intended use. The district court properly so ruled. The decision is\n\n AFFIRMED.\n\n\n\n\n 22\n\n Case: 17-13376 Date Filed: 08/30/2019 Page: 23 of 27\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: United States v. Hi-Tech Pharmaceuticals, Inc.\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-30\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: Yes\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 17-13376 Date Filed: 08/30/2019 Page: 1 of 27\\n\\n\\n [PUBLISH]\\n\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 17-13376\\n ________________________\\n\\n D.C. Docket No. 1:13-cv-03675-WBH\\n\\n\\nUNITED STATES OF AMERICA,\\n\\n Plaintiff - Appellee,\\n\\nversus\\n\\nUNDETERMINED QUANTITIES OF ALL ARTICLES OF\\nFINISHED AND IN-PROCESS FOODS,\\nraw ingredients (bulk powders, bulk capsules), with any lot number,\\nsize, or type container, whether labeled or unlabeled, et al.,\\n\\n Defendants,\\n\\nHI-TECH PHARMACEUTICALS, INC.,\\nJARED WHEAT,\\n\\n Claimants - Appellants.\\n\\n\\n ________________________\\n\\n Appeal from the United States District Court\\n for the Northern District of Georgia\\n ________________________\\n\\n (August 30, 2019)\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 2 of 27\\n\\n\\nBefore TJOFLAT and JORDAN, Circuit Judges, and HINKLE, * District Judge.\\n\\nHINKLE, District Judge:\\n\\n The Dietary Supplement Health and Education Act of 1994 provides\\n\\nfavorable treatment for “dietary supplements,” defined to include any “botanical”\\n\\nor “constituent” of a botanical. This case presents the question whether these terms\\n\\napply to a substance that was invented in a laboratory and is artificially produced\\n\\nfor commercial sale but that, entirely coincidentally, may be found in trace\\n\\namounts in a plant. We hold that the terms do not extend this far.\\n\\n I. Proceedings\\n\\n The Food and Drug Administration seized from Hi-Tech Pharmaceuticals,\\n\\nInc. a substantial quantity of products containing 1,3-dimethylamylamine or\\n\\n“DMAA.” DMAA is used in fitness products aimed at bodybuilders and other\\n\\nathletes.\\n\\n The seizure led to two actions that were consolidated in the district court.\\n\\nOne was a forfeiture action filed by the United States against the products. Hi-Tech\\n\\nand its chief executive officer, Jared Wheat, intervened as claimants. Hi-Tech filed\\n\\nthe other action against the FDA and other governmental defendants.\\n\\n\\n\\n\\n *\\n Honorable Robert L. Hinkle, United States District Judge for the Northern District of\\nFlorida, sitting by designation.\\n 2\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 3 of 27\\n\\n\\n Hi-Tech asserted that DMAA is a dietary supplement; that under the\\n\\nAdministrative Procedure Act the FDA can properly ban DMAA, if at all, only\\n\\nthrough rulemaking; and that the seizure of Hi-Tech’s DMAA violated the Fifth\\n\\nAmendment’s Due Process Clause.\\n\\n The parties filed cross-motions for summary judgment. The district court\\n\\ngranted the FDA’s motion, holding the seizure proper both substantively and\\n\\nprocedurally. The district court denied a motion to reconsider that included a\\n\\nrequest to reopen discovery. Hi-Tech and Mr. Wheat have appealed. The appeal\\n\\nhas been fully briefed and orally argued.\\n\\n II. Standard of Review\\n\\n We review de novo the district court’s grant of summary judgment. See, e.g.,\\n\\nPrice v. Comm’r, Dep’t of Corr., 920 F.3d 1317, 1323 (11th Cir. 2019). We review\\n\\nfor abuse of discretion the district court’s denial of the motion for reconsideration\\n\\nand refusal to reopen discovery. See, e.g., Corwin v. Walt Disney Co., 475 F.3d\\n\\n1239, 1254 (11th Cir. 2007) (reconsideration); Artistic Entm’t, Inc. v. City of\\n\\nWarner Robins, 331 F.3d 1196, 1202 (11th Cir. 2003) (reopening discovery).\\n\\n III. The Statute and the Issues\\n\\n The Federal Food, Drug, and Cosmetic Act prohibits the introduction of\\n\\nadulterated foods into interstate commerce. 21 U.S.C. § 331(a). The FDA enforces\\n\\nthe Act. Id. § 393(b)(2)(A). The agency may bring an in rem forfeiture action in\\n\\n 3\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 4 of 27\\n\\n\\ndistrict court to condemn adulterated foods. Id. § 334(a)(1). Hi-Tech’s DMAA\\n\\nproducts were adulterated foods if they were “food additives” but not if they were\\n\\n“dietary supplements.”\\n\\n The background is this. The Dietary Supplement Health and Education Act\\n\\nof 1994, commonly referred to as “DSHEA,” amended the Federal Food, Drug,\\n\\nand Cosmetic Act to provide favorable treatment for dietary supplements. The\\n\\nstatute’s definition of “dietary supplement” includes multiple parts. 21 U.S.C.\\n\\n§ 321(ff). The only part relevant to Hi-Tech’s DMAA is this: a product that is\\n\\nintended to supplement the diet—this includes DMAA—is a dietary supplement if\\n\\nit contains “an herb or other botanical” or “a concentrate, metabolite, constituent,\\n\\nextract, or combination of” an herb or other botanical. Id. § 321(ff)(1)(C) & (F).\\n\\nThe statute describes these—as well as other substances not at issue here—as\\n\\n“dietary ingredients.”\\n\\n Under DSHEA, and subject to exceptions not relevant here, a dietary\\n\\nsupplement can be condemned as adulterated only if the FDA carries the burden of\\n\\nproving that the substance presents a “significant or unreasonable risk of illness or\\n\\ninjury” under recommended, suggested, or ordinary conditions of use. Id.\\n\\n§ 342(f)(1)(A). The FDA did not attempt to make that showing for the DMAA\\n\\nproducts it seized from Hi-Tech. A ruling that DMAA is a dietary supplement thus\\n\\nwould resolve this appeal in Hi-Tech’s favor.\\n\\n 4\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 5 of 27\\n\\n\\n On the other hand, a ruling that DMAA is a “food additive” would resolve\\n\\nthe dispute in the FDA’s favor. A substance intended for human consumption is a\\n\\nfood additive if it is not a dietary supplement and is not “generally recognized,\\n\\namong experts qualified by scientific training and experience to evaluate its safety,\\n\\nas having been adequately shown through scientific procedures . . . to be safe under\\n\\nthe conditions of its intended use.” Id. § 321(s). For a substance in common use in\\n\\nfood prior to January 1, 1958—this does not include DMAA—the adequate\\n\\nshowing of safety can be made not only by scientific procedures but also by\\n\\nexperience. There are other exceptions to this definition of “food additive,” but\\n\\nnone applies here.\\n\\n The FDA asserts that DMAA is not a dietary supplement, is not generally\\n\\nrecognized as safe, does not meet any other exception, and is therefore a food\\n\\nadditive. Hi-Tech insists that DMAA is a dietary supplement and thus is not a food\\n\\nadditive, but that even if DMAA is not a dietary supplement, DMAA is generally\\n\\nrecognized as safe and thus still is not a food additive.\\n\\n The issues thus are first, whether Hi-Tech’s DMAA products are “an herb or\\n\\nother botanical” or “a concentrate, metabolite, constituent, extract, or combination\\n\\nof” an herb or other botanical, and second, if not, whether the products are\\n\\ngenerally recognized as safe. Secondary issues are whether the FDA was entitled to\\n\\n\\n\\n\\n 5\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 6 of 27\\n\\n\\nseize and forfeit the products without engaging in rulemaking and whether the\\n\\ndistrict court should have reopened discovery.\\n\\n IV. DMAA\\n\\n The earliest known identification or use of DMAA occurred in 1944. In that\\n\\nyear Eli Lilly & Co. synthesized and patented DMAA for use as a nasal\\n\\ndecongestant. For marketing reasons, Eli Lilly asked the FDA to withdraw its\\n\\napproval of this use in 1983. At least insofar as shown by this record, DMAA was\\n\\nnot used as a dietary supplement or food additive at that time, and no health\\n\\nconcerns had been noted.\\n\\n DMAA eventually made a resurgence, this time in fitness products aimed at\\n\\nbodybuilders and other athletes. Because of DMAA’s noticeable stimulant effect,\\n\\nthe compound made its way into pre-workout energy and fat-burner products\\n\\naround the world.\\n\\n The FDA eventually adopted the position that DMAA is not a dietary\\n\\nsupplement but an unsafe food additive. The FDA issued cease-and-desist letters to\\n\\nat least some entities marketing DMAA products. Perhaps unaware of Hi-Tech’s\\n\\nmarketing of DMAA products, the FDA did not issue a cease-and-desist letter to\\n\\nHi-Tech.\\n\\n Around the same time, researchers began to find trace amounts of DMAA in\\n\\ngeraniums of the genus pelargonium. A 2013 survey concluded that overall, the\\n\\n 6\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 7 of 27\\n\\n\\nstudies showed that DMAA “is found naturally in some, but not all, geranium\\n\\nplants and extracted geranium oils.” Thomas D. Gauthier, Evidence for the\\n\\nPresence of 1,3-Dimethylamylamine (1,3-DMAA) in Geranium Plant Materials, 8\\n\\nAnalytical Chemistry Insights 29-40 (2013), available at\\n\\nhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC3682735/. Indeed, the FDA’s own\\n\\nexpert had previously participated in a study that found trace amounts of DMAA in\\n\\ngeraniums.\\n\\n Even so, this record presents a genuine factual dispute over whether trace\\n\\namounts of DMAA are naturally contained in geraniums. On the one hand, studies\\n\\nhave found trace amounts of DMAA in geraniums. On the other hand, some\\n\\nfertilizers contain DMAA that could be a source of trace amounts of DMAA in\\n\\ngeraniums, and the record includes competent testimony that there is no known\\n\\npathway by which geraniums could produce DMAA. Either way, it is clear that\\n\\nDMAA is not contained in geraniums in amounts greater than could reasonably be\\n\\ncharacterized as trace amounts. No study has found a greater amount.\\n\\n V. “Herb or Other Botanical”\\n\\n The first rule of statutory construction is to apply the plain meaning of the\\n\\nstatutory language. See, e.g., Bankston v. Then, 615 F.3d 1364, 1367 (11th Cir.\\n\\n2010). Here the meaning is not completely clear.\\n\\n\\n\\n\\n 7\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 8 of 27\\n\\n\\n Hi-Tech says DSHEA uses “botanical” to mean all plant life, nothing more\\n\\nand nothing less—that is, to mean flora, without limitation. The suggestion is\\n\\nsensible enough—“botany” is the study of plants. On the other hand, it would be\\n\\npassing strange for a writer wishing to cover the universe of plant life—to mean all\\n\\nflora—to achieve that result through the term “herb or other botanical.” Moreover,\\n\\nthe usual connotation of “botanical” when used as a noun, as recognized in\\n\\ndictionaries in use when DSHEA was enacted as well as those is use today, is a\\n\\nsubstance derived from a plant used for a limited category of purposes.\\n\\n In 1993, a year before DSHEA became law, Merriam-Webster’s defined the\\n\\nnoun “botanical” as a “a plant part or extract used esp[ecially] in skin and hair care\\n\\nproducts.” “Botanical,” Merriam-Webster’s Collegiate Dictionary 134 (10th ed.\\n\\n1993). The current edition defines the noun “botanical” as a “substance obtained or\\n\\nderived from a plant[,] such as . . . a plant part or extract used especially in skin\\n\\nand hair care products[,] a medicinal preparation derived from a plant[, or] plant\\n\\nmaterial used as a flavoring agent.” “Botanical,” Merriam-Webster’s Online\\n\\nDictionary 2019, available at https://www.merriam-\\n\\nwebster.com/dictionary/botanical. Neither definition suggests that the noun\\n\\n“botanical” includes an artificially produced substance that, entirely coincidentally,\\n\\nmay be found in trace amounts in a plant. Nor do they suggest that “botanical”\\n\\nincludes all flora.\\n\\n 8\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 9 of 27\\n\\n\\n Both the Supreme Court and the Eleventh Circuit have relied on Merriam-\\n\\nWebster’s as an aid in construing statutes. See, e.g., Octane Fitness, LLC v. ICON\\n\\nHealth & Fitness, Inc., 572 U.S. 545, 553 (2014); Burlington N. & Santa Fe Ry.\\n\\nCo. v. United States, 556 U.S. 599, 611 (2009); United States v. Zuniga-Arteaga,\\n\\n681 F.3d 1220, 1224 (11th Cir. 2012); Arriaga v. Fla. Pac. Farms, LLC, 305 F.3d\\n\\n1228, 1242 (11th Cir. 2002). This does not make these cited definitions of\\n\\n“botanical” dispositive; dictionaries are not controlling and in any event give\\n\\nexamples to convey a term’s most common uses, not necessarily to suggest limits.\\n\\nBut the narrower connotation suggested by the dictionaries is consistent with\\n\\nDSHEA’s use of the term “herb or other botanical” rather than a broader term\\n\\nplainly encompassing all plant life.\\n\\n To be sure, the difference between Hi-Tech’s broad view—all flora—and\\n\\nthe narrower dictionary definitions is not as stark as might appear at first blush.\\n\\nThat a substance derived from a plant is used in a dietary product brings it close to\\n\\nthe current dictionary definition, which includes a medicinal preparation derived\\n\\nfrom a plant.\\n\\n Still, the use of “herb or other botanical” in the statute, together with the\\n\\ndictionary definitions of a botanical as “derived from a plant,” supports a much\\n\\nnarrower construction than Hi-Tech proposes. Had Congress meant all plants and\\n\\nanything contained in them, it could have said so. It did not. At the least, the\\n\\n 9\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 10 of 27\\n\\n\\nstatutory language and dictionary definitions support a conclusion that would be\\n\\nreasonable anyway: it is unlikely that Congress used the term “herb or other\\n\\nbotanical” to mean a substance invented in a laboratory and artificially produced,\\n\\nthat can be found in a plant only in trace amounts, only coincidentally, and that has\\n\\nnever been derived from a plant for use in any medicinal, cosmetic, or dietary\\n\\nproduct.\\n\\n VI. “Constituent”\\n\\n The statutory definition of a dietary supplement extends not only to an “herb\\n\\nor other botanical” but also to “a concentrate, metabolite, constituent, extract, or\\n\\ncombination of” an herb or other botanical. 21 U.S.C. § 321(ff)(1)(F). Hi-Tech\\n\\nasserts “constituent” means anything naturally contained in. The word could be\\n\\ngiven that meaning, but the connotation is usually not so broad. Indeed, both the\\n\\n1993 edition and the current edition of Merriam-Webster’s define “constituent” as\\n\\n“an essential part.” “Constituent,” Merriam-Webster’s Collegiate Dictionary 248\\n\\n(10th ed. 1993); “Constituent,” Merriam-Webster Online Dictionary 2019,\\n\\navailable at https://www.merriam-webster.com/dictionary/constituent. This\\n\\ndefinition suggests a connotation much narrower than proposed by Hi-Tech and\\n\\ntoo narrow to include the DMAA—if there is any—contained in geraniums.\\n\\n For its part, the FDA says Hi-Tech’s proposed definition of “constituent”\\n\\nwould render superfluous the statute’s inclusion of the word “extract.” The FDA\\n\\n 10\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 11 of 27\\n\\n\\nsays the meaning of “constituent” must be informed by the other words in the\\n\\nstatutory list, under the canon noscitur a sociis. See In re Piazza, 719 F.3d 1253,\\n\\n1263 n.4 (11th Cir. 2013) (discussing this canon); Antonin Scalia & Bryan A.\\n\\nGarner, Reading Law: The Interpretation of Legal Texts (“Reading Law”) 195-98\\n\\n(2012) (same). The “most common effect of the canon is not to establish which of\\n\\ntwo totally different meanings applies but rather to limit a general term to a subset\\n\\nof all the things or actions that it covers—but only according to its ordinary\\n\\nmeaning.” Reading Law at 196.\\n\\n A concentrate or extract of a product is derived from the product in usable\\n\\nform or amount. So is a combination of the product with another substance. A\\n\\nmetabolite, too, is physically derived from a product. If, as Hi-Tech asserts,\\n\\nconstituent means anything contained in, the word is both markedly different from\\n\\nthe others in the list and awkwardly placed—the broadest term in a five-item list\\n\\nbut placed not first or last but in the center.\\n\\n None of this is dispositive. The safest conclusion is this: it is unlikely that\\n\\nCongress used the term “constituent” to mean a substance that is present in a plant\\n\\nin only trace amounts and that has never been derived from a plant for use in any\\n\\nmedicinal, cosmetic, or dietary product.\\n\\n VII. The Reason for the Statutory Presumption\\n\\n\\n\\n\\n 11\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 12 of 27\\n\\n\\n As set out above, DSHEA gives a preference to dietary supplements. The\\n\\nFDA can condemn a dietary supplement as adulterated only on a showing that it\\n\\npresents a “significant or unreasonable risk of illness or injury” under\\n\\nrecommended, suggested, or ordinary conditions of use. 21 U.S.C. § 342(f)(1)(A).\\n\\nThis is, in effect, a rebuttable presumption that the product is safe when used as\\n\\nintended.\\n\\n A principal reason for rebuttable presumptions, whether in statutes or other\\n\\nlegal constructs, is administrative convenience. When a proposition is usually true,\\n\\nit sometimes makes sense to presume it is true, subject only to rebuttal in the\\n\\noccasional instance when it is not true. Perhaps more importantly, at least in the\\n\\nregulatory context, a presumption can avoid unnecessary expense and delay—a\\n\\nperson or entity can go forward with proposed action without awaiting regulatory\\n\\napproval. This approach works best when a proposition is usually true and when\\n\\nthe rebuttable presumption is clear and easily applied—otherwise the unnecessary\\n\\nexpense and delay is not likely to be avoided.\\n\\n DSHEA well illustrates this approach. Congress thought it better to have a\\n\\nclear, administrable rule—dietary supplements are presumed safe, subject only to a\\n\\ncontrary showing—than to require a particularized inquiry in every case. See S.\\n\\nRep. No. 103-410, at 21-22 (1994). A fair inference is that herbs and other\\n\\n\\n\\n\\n 12\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 13 of 27\\n\\n\\nbotanicals and their constituents made the list of favored dietary ingredients\\n\\nbecause consuming them is ordinarily safe.\\n\\n Consuming most herbs or other botanicals, though surely not all, is safe. The\\n\\nsame is true even for most plants, and people have been consuming plants for as\\n\\nlong as there have been people. Congress reasonably could choose to treat any\\n\\nproduct derived from a plant as adulterated only on a showing that it is unsafe. A\\n\\nrebuttable presumption for anything derived from a plant would serve\\n\\nadministrative convenience and avoid delay in introducing a product to the market.\\n\\n It is a stretch, though, to apply the same reasoning to a substance invented in\\n\\na laboratory and artificially produced, that can be found in a plant, if at all, only in\\n\\ntrace amounts, only coincidentally, and that has never been derived from a plant\\n\\nfor use in any medicinal, cosmetic, or dietary product. The fact that DMAA can be\\n\\nfound in trace amounts in geraniums, if true, says absolutely nothing about whether\\n\\nconsuming the substance is safe.\\n\\n Nor does applying a rebuttable presumption to a substance of this kind serve\\n\\nadministrative convenience. It is easy enough to identify plants or substances\\n\\nactually derived from plants. But as this case illustrates, it is not always easy to\\n\\ndetermine whether a product invented in a laboratory and artificially manufactured\\n\\ncan be found in trace amounts in some plant somewhere in the world.\\n\\n\\n\\n\\n 13\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 14 of 27\\n\\n\\n There is no reason to believe that when it adopted DSHEA, Congress\\n\\nintended to put in place a rebuttable presumption that such a product is safe. We\\n\\nhold that DSHEA does not go that far.\\n\\n This does not mean that DSHEA applies only to products actually derived\\n\\nfrom plants, not those artificially manufactured. If a product is indeed a dietary\\n\\nsupplement because it contains a qualifying dietary ingredient—including, for\\n\\nexample, an herb or other botanical—a manufacturer may take the dietary\\n\\ningredient from nature or produce it artificially. But there must be a qualifying\\n\\ndietary ingredient. The ability to create a substance in a laboratory and\\n\\nmanufacture it artificially does not give a substance that status. Nor does\\n\\ncoincidentally identifying the substance in trace amounts in some plant somewhere\\n\\nin the world.\\n\\n VIII. Generally Recognized as Safe\\n\\n Hi-Tech says DMAA is generally recognized as safe—or, to quote the\\n\\nstatute’s more exacting standard, DMAA is “generally recognized, among experts\\n\\nqualified by scientific training and experience to evaluate its safety, as having been\\n\\nadequately shown through scientific procedures . . . to be safe under the conditions\\n\\nof its intended use.” 21 U.S.C. § 321(s). The FDA’s rule on this concludes that a\\n\\nsubstance meets this standard only when, based on “common knowledge\\n\\nthroughout the scientific community knowledgeable about the safety of substances\\n\\n 14\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 15 of 27\\n\\n\\ndirectly or indirectly added to food,” there is “reasonable certainty that the\\n\\nsubstance is not harmful under the conditions of its intended use.” 21 C.F.R.\\n\\n§ 170.30(a).\\n\\n As the statutory requirement for general recognition makes clear, the issue is\\n\\nnot whether, as an original matter, the factfinder in a legal proceeding would\\n\\nevaluate the evidence and conclude that a substance is safe. The issue is only\\n\\nwhether the substance is generally recognized as safe among qualified experts\\n\\nbased on adequate studies. To establish the contrary, the FDA “need only show the\\n\\nlack of the proper reputation . . . for safety of the [substance] among the\\n\\nappropriate experts, or that what reputation there is, is not based on adequate\\n\\nstudies.” United States v. Articles of Food & Drug Consisting of Coli-Trol 80,\\n\\nF4C-60 Feed Grade, Entrol-S Medicated, Entrol-P, 518 F.2d 743, 746 (5th Cir.\\n\\n1975). As a pre-Bonner decision of the Fifth Circuit, Coli-Trol remains binding in\\n\\nthis court. See Bonner v. City of Prichard, 661 F.2d 1206, 1207 (11th Cir. 1981)\\n\\n(en banc). Other circuits, too, have enforced the requirement for general\\n\\nrecognition. See United States v. Article of Food, 752 F.2d 11, 15 n.4 (1st Cir.\\n\\n1985); Premo Pharmaceutical Labs., Inc. v. United States, 629 F.2d 795, 803-05\\n\\n(2d Cir. 1980).\\n\\n The FDA made the required showing. Multiple sources, including in peer-\\n\\nreviewed publications, call into question DMAA’s safety. Among their\\n\\n 15\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 16 of 27\\n\\n\\nconclusions: DMAA may cause increases in blood pressure and hemorrhagic\\n\\nstroke; individuals with blood pressure of 120/80 mmHg or higher (much of the\\n\\nAmerican population) should avoid DMAA; use of DMAA has been associated\\n\\nwith multiple adverse events, including deaths; and DMAA may inhibit activity of\\n\\nliver enzymes and cause liver toxicity.\\n\\n After four soldiers died with DMAA in their systems, the Department of\\n\\nDefense removed all DMAA products from military exchanges and commissioned\\n\\na Safety Review Panel. The Panel issued a report finding that “deaths, hepatic\\n\\nfailure, myocardial infarction, heat stroke and rhabdomyolysis, seizure and stroke”\\n\\nwere temporally associated with service members’ “use of [DMAA-containing]\\n\\nproducts.” U.S. Dep’t of Def., Report of the Department of Defense 1,3\\n\\nDimethylamylamine (DMAA) Safety Review Panel 9 (2013). The report said this\\n\\nsuggested that some individuals “may be predisposed to severe health\\n\\nconsequences after using DMAA.” Id. The report said there appeared to be\\n\\n“significant association of DMAA use, particularly high frequency DMAA use,\\n\\nand multiple adverse events.” Id. And the report concluded that “the available\\n\\nevidence supports an elevated health risk associated with the use of DMAA-\\n\\ncontaining products.” Id. The Department continued its ban on DMAA products at\\n\\nmilitary exchanges. Id. at 10-11.\\n\\n\\n\\n\\n 16\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 17 of 27\\n\\n\\n With this track record, it is hardly surprising that the FDA’s expert in food\\n\\nchemical risk management determined that DMAA is not generally recognized as\\n\\nsafe by qualified experts.\\n\\n Hi-Tech asserts, though, that the studies and reports on which the FDA relies\\n\\ninvolve DMAA use in doses greater than Hi-Tech recommends. Hi-Tech says that\\n\\nuse of DMAA as intended does not present the same risks. Hi-Tech cites studies\\n\\nand presents expert testimony concluding that DMAA is safe at the recommended\\n\\ndoses.\\n\\n Hi-Tech’s submissions are far from conclusive. The studies use small\\n\\nsample sizes and look at short-term results. None measure the effect of DMAA in\\n\\nhigh-risk populations or on individuals with elevated blood pressure. And while\\n\\nsome but not all of the FDA’s cited studies involve high doses of DMAA, it seems\\n\\nunlikely that all the adverse events suffered by military personnel and others\\n\\nresulted from abnormal or unintended use. Correlation is not causation, but neither\\n\\nmust correlation be ignored.\\n\\n If the issue was whether DMAA is safe, Hi-Tech’s evidence would create a\\n\\ngenuine issue of fact precluding summary judgment; neither side’s evidence is\\n\\nconclusive. See Sparling v. Doyle, No. EP-13-CV-323-DCG, 2015 WL 4528759 at\\n\\n*20 (W.D. Tex. July 27, 2015) (“It is clear . . . that the scientific literature on\\n\\nDMAA presents insufficient data to conclude that DMAA is safe or that DMAA\\n\\n 17\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 18 of 27\\n\\n\\ncauses harm because the sample sizes are too small.”). But the issue is not whether\\n\\nDMAA is safe; the issue is only whether DMAA is generally recognized as safe. It\\n\\nplainly is not. On the issue of general recognition, the FDA was entitled to\\n\\nsummary judgment.\\n\\n IX. The Motion to Reopen Discovery\\n\\n The district court provided ample time for discovery—the full amount the\\n\\nparties requested. The parties submitted cross-motions for summary judgment\\n\\nwithout asking for more time or asserting that any further discovery was needed.\\n\\nBut after the court granted summary judgment for the FDA, Hi-Tech moved to\\n\\nreconsider, taking issue with the court’s legal analysis and asserting the court\\n\\nshould reopen discovery. Hi-Tech said it needed more discovery because the\\n\\ncourt’s legal analysis did not match up with the position argued by either side.\\n\\n Ours is an adversary system. When, as here, there are two sides, each side is\\n\\nafforded the opportunity to argue its position. But the court is not limited to\\n\\nchoosing one side’s position or the other’s. The court’s role is to get it right, not to\\n\\nchoose which side’s argument is better and adopt it lock, stock, and barrel. See,\\n\\ne.g., United States v. Baston, 818 F.3d 651, 663 (11th Cir. 2016) (concluding that\\n\\non a disputed legal issue, “[n]either party is correct,” and applying the correct\\n\\nstandard that neither party advocated); see also Colburn v. Odom, 911 F.3d 1110\\n\\n(11th Cir. 2018) (resolving an appeal on a ground not addressed in either side’s\\n\\n 18\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 19 of 27\\n\\n\\nbrief but essential to proper resolution of the dispute). Were it otherwise, there\\n\\nwould be no plain-error doctrine.\\n\\n Thousands of cases could be cited illustrating this principle. Indeed, the\\n\\nprinciple is so well settled that it is rarely mentioned. When a court adopts a view\\n\\nof the law that is not precisely in line with either side’s argument, the court usually\\n\\nsets out its view of the law without citing authority for the proposition that it may\\n\\ndo so. The Supreme Court has explained it this way: “[w]hen an issue or claim is\\n\\nproperly before the court, the court is not limited to the particular legal theories\\n\\nadvanced by the parties, but rather retains the independent power to identify and\\n\\napply the proper construction of governing law.” U.S. Nat’l Bk. of Oregon v.\\n\\nIndependent Ins. Agents of Am., Inc., 508 U.S. 439, 446 (1993) (quoting Kamen v.\\n\\nKemper Fin. Servs., Inc., 500 U.S. 90, 99 (1991)).\\n\\n Hi-Tech is correct that the district court did not simply accept either side’s\\n\\nview of the facts and law. Nor should the court have done so; neither side had it\\n\\njust right. Similarly, on appeal, we have not simply chosen one side’s view or the\\n\\nother’s; we have considered the arguments and provided the analysis we believe is\\n\\ncorrect. One would expect nothing less.\\n\\n Hi-Tech says, though, that it was blindsided when the district court\\n\\nemphasized that DMAA has never actually been derived from geraniums for use in\\n\\nany product. Hi-Tech says it needs more discovery to fully present its position on\\n\\n 19\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 20 of 27\\n\\n\\nthis issue—to attempt to find evidence that DMAA has in fact been derived from\\n\\ngeraniums.\\n\\n The assertion misses the mark for two reasons, either of which would be\\n\\nsufficient standing alone.\\n\\n First, Hi-Tech could not have been surprised that the court considered\\n\\nwhether DMAA has actually been derived from geraniums. The question leaps off\\n\\nthe page at anyone first considering the issues in this case. Hi-Tech asserts it does\\n\\nnot matter whether DMAA has actually been derived from geraniums—a colorable\\n\\nposition—but Hi-Tech could not have missed the possibility that a court would\\n\\ndisagree.\\n\\n Second, regardless of whether Hi-Tech recognized or should have\\n\\nrecognized that a court might find actual derivation critical, Hi-Tech had every\\n\\nincentive to fully develop the facts on this during the original discovery period. An\\n\\nintensely disputed issue was whether DMAA was contained in geraniums. Hi-Tech\\n\\nsaid yes; the FDA said no. The best support Hi-Tech could have garnered for its\\n\\nposition on this issue—as Hi-Tech surely knew—was evidence that DMAA had\\n\\nactually been derived from geraniums. The reason one can’t get blood from a\\n\\nturnip is that there is no blood in a turnip. The reason one can get juice from an\\n\\norange is that oranges are full of juice. The reason Hi-Tech found no evidence\\n\\nduring the original discovery period that DMAA had actually been derived from\\n\\n 20\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 21 of 27\\n\\n\\ngeraniums was not because Hi-Tech didn’t know to look; it was because no such\\n\\nevidence existed. Or perhaps because, despite every incentive to do so, Hi-Tech\\n\\ncouldn’t find it in the ample time it requested—and the court provided—for\\n\\ndiscovery. Hi-Tech is not entitled to more time.\\n\\n The district court did not abuse its discretion when it declined to reopen\\n\\ndiscovery.\\n\\n X. The Absence of Rulemaking\\n\\n Hi-Tech faults the FDA for bringing a forfeiture action rather than\\n\\nproceeding through rulemaking. But it is “well established” that “agencies have\\n\\ndiscretion to choose whether to proceed by rulemaking or adjudication.” RTC\\n\\nTransp. Inc. v. ICC, 731 F.2d 1502, 1505 (11th Cir. 1984). Not surprisingly, then,\\n\\nwe have upheld a forfeiture judgment in favor of the FDA against a food additive\\n\\nwithout requiring rulemaking. See United States v. Articles of Food & Drug\\n\\nConsisting of Coli-Trol 80, F4C-60 Feed Grade, Entrol-S Medicated, Entrol-P,\\n\\n518 F.2d 743, 746 (5th Cir. 1975); see also United States v. Article of Food, 752\\n\\nF.2d 11, 15-16 (1st Cir. 1985). The FDA was not required to engage in rulemaking\\n\\nbut could elect instead to proceed through a forfeiture action against Hi-Tech’s\\n\\nDMAA products.\\n\\n Proceeding in this manner did not violate the Constitution. The governing\\n\\nstatute provides notice that unapproved food additives are subject to forfeiture. 21\\n\\n 21\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 22 of 27\\n\\n\\nU.S.C. § 334(a)(1). The statute is not unconstitutionally vague, and Hi-Tech\\n\\ndoesn’t claim it is. As part of the forfeiture proceeding, Hi-Tech was afforded the\\n\\nfull range of procedural due process available in a federal court. The issues were\\n\\njoined and fully adjudicated on the merits. Due process requires nothing more.\\n\\n XI. Conclusion\\n\\n DMAA is not an “herb or other botanical.” It is not a “constituent” of an\\n\\nherb or other botanical. And it is not generally recognized by qualified experts, as\\n\\nadequately shown through scientific procedures, to be safe under the conditions of\\n\\nits intended use. The district court properly so ruled. The decision is\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 22\\n\\n Case: 17-13376 Date Filed: 08/30/2019 Page: 23 of 27\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4163,"cells":{"id":{"kind":"number","value":4657833,"string":"4,657,833"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"Richard Knight v. Florida Department of Corrections"},"decision_date":{"kind":"string","value":"2019-08-30"},"docket_number":{"kind":"string","value":"18-13390"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before TJOFLAT, JORDAN, and GRANT, Circuit Judges.\", \"parties\": \"\", \"opinions\": [{\"author\": \"GRANT, Circuit Judge:\", \"type\": \"010combined\", \"text\": \"Case: 18-13390 Date Filed: 08/30/2019 Page: 1 of 28\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 18-13390\\n ________________________\\n\\n D.C. Docket No. 0:17-cv-61921-RNS\\n\\n\\n\\nRICHARD KNIGHT,\\n\\n Petitioner-Appellant,\\n\\n versus\\n\\nFLORIDA DEPARTMENT OF CORRECTIONS,\\n\\n Respondent-Appellee.\\n ________________________\\n\\n Appeal from the United States District Court\\n for the Southern District of Florida\\n ________________________\\n\\n (August 30, 2019)\\n\\nBefore TJOFLAT, JORDAN, and GRANT, Circuit Judges.\\nGRANT, Circuit Judge:\\n Richard Knight, a Florida prisoner sentenced to death for the murders of\\n\\nOdessia Stephens and her daughter, Hanessia Mullings, appeals the district court’s\\ndenial of his federal habeas corpus petition. At this stage—almost 20 years after\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 2 of 28\\n\\n\\n\\nthe crimes were committed and more than a decade after a Florida jury found\\nKnight guilty of the murders and recommended a death sentence—Knight’s claims\\n\\nhave been winnowed down to two: first, that his death sentence is invalid under\\nHurst v. Florida, 136 S. Ct. 616 (2016), and second, that he received ineffective\\nassistance of counsel at trial. Because Hurst does not apply retroactively to\\n\\nKnight, any challenge to his death sentence on that basis is beyond our reach on\\nfederal habeas review. Nor can Knight find success in his other challenge; the\\nFlorida Supreme Court’s rejection of his ineffective-assistance claim was not an\\n\\nunreasonable application of Strickland v. Washington, 466 U.S. 668 (1984). We\\ntherefore affirm.\\n I.\\n\\n A.\\n According to evidence introduced at his murder trial, Knight lived in an\\napartment with his cousin, Hans Mullings, and Hans’s girlfriend, Odessia. Hans\\nand Odessia’s four-year-old daughter, Hanessia, also lived with them in the\\napartment. Odessia was tired of supporting Knight and one evening while Hans\\nwas out she argued with Knight, insisting that he move out the next day. After the\\nargument got heated, Knight left the house to walk around. But as he later\\nconfessed to another inmate, instead of getting less angry with Odessia once he got\\nsome air, Knight became increasingly irate. He returned to the apartment and after\\nexchanging more words with Odessia, he got a knife from the kitchen. When he\\nwent back to the master bedroom, he found Odessia and her little girl in the bed.\\nHe began stabbing Odessia and continued his attack until she stopped resisting and\\n\\n 2\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 3 of 28\\n\\n\\n\\ncurled up on the bedroom floor. He then moved on to little Hanessia, stabbing her\\nuntil his knife broke and cutting his hand in the process. As he was leaving the\\n\\nbedroom, he heard “popping noises” from where Hanessia lay on the floor, and he\\nthought that the little girl was “drowning in her own blood.” Apparently not\\nconsidering his revenge complete, he retrieved a second knife from the kitchen and\\n\\nreturned to continue his attack on Odessia. In the meantime, Odessia had crawled\\nfrom the bedroom to the living room, where she had collapsed. Knight turned her\\nover, saw that she was still alive, and started stabbing her again.\\n\\n Both Odessia and Hanessia died that night. In total, Odessia had 21 stab\\nwounds, including 14 in the neck, 24 puncture or scratch wounds, bruising and\\nligature marks consistent with having been hit and strangled with a belt, defensive\\n\\nwounds, and bruises from being hit or punched in the mouth and head. Little\\nHanessia had four stab wounds in her upper body and neck, a deep defensive\\nwound on her hand, bruises on her neck consistent with manual strangulation, and\\nbruises on her arms consistent with having been grabbed.\\n Knight showered and changed after completing his brutal acts, then headed\\nto the living room with a rag to wipe off the knives. Interrupted by a knock on the\\nfront door—it was police responding to a neighbor’s 911 call—Knight ran to his\\nroom and climbed out the window.\\n Shortly after they arrived, police encountered Knight near the apartment. He\\ntold them that he lived there, but that he did not have a key. This was odd; the\\nofficers had already found that all the doors to the apartment were locked. Knight\\nwas also visibly wet—but it was not raining. Knight explained to police that he\\n\\n 3\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 4 of 28\\n\\n\\n\\nhad been jogging, a remarkable contention from a person who was wearing long\\npants and dress shoes. He did not appear to be sweating, in any event. And\\n\\nKnight’s personal appearance subsequently revealed even more clues—he had\\nblood on the back of his shirt, scratches on his chest and midsection, a scrape on\\nhis shoulder, and fresh cuts on his hand.\\n\\n Knight was arrested and indicted for two counts of first-degree murder. A\\nFlorida jury found him guilty as charged. That same jury heard evidence and\\nargument at the penalty phase and unanimously recommended two death\\n\\nsentences—one for each murder. Consistent with Florida’s then-current death\\npenalty sentencing procedure, the judge held an additional hearing, made his own\\nfindings regarding aggravating and mitigating circumstances, and sentenced\\n\\nKnight to death. The Florida Supreme Court affirmed Knight’s convictions and\\nsentences on direct appeal. Knight v. State, 76 So. 3d 879, 890 (Fla. 2011). The\\nUnited States Supreme Court denied his petition for certiorari. Knight v. Florida,\\n566 U.S. 998, 998 (2012).\\n B.\\n Knight filed motions for state collateral relief raising the two claims at issue\\nhere, as well as others that have already been resolved. Specifically, he argued that\\nthe state court should vacate his death sentence in light of Hurst v. Florida, in\\nwhich the Supreme Court held—four years after Knight’s conviction was final—\\nthat Florida’s death penalty sentencing scheme violated the Sixth Amendment.\\n136 S. Ct. at 622. The problem identified by the Supreme Court in Hurst, and\\nargued by Knight in his post-conviction pleadings, was that the jury’s role in\\n\\n 4\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 5 of 28\\n\\n\\n\\nsentencing was to make a non-binding recommendation; the judge alone made the\\nultimate findings of fact necessary to impose the death penalty. Id. at 619, 621–22.\\n\\nKnight also argued that his guilt-phase counsel was constitutionally ineffective for\\nfailing to call an available DNA expert.\\n The Florida Supreme Court rejected his postconviction claims on the merits.\\n\\nKnight v. State, 225 So. 3d 661, 668 (Fla. 2017) (per curiam). A plurality of the\\ncourt agreed with Knight that the sentencing procedure used in his case violated\\nthe Sixth Amendment under Hurst, but also concluded that the Hurst error was\\n\\nharmless. Id. at 682. The plurality explained that under the facts of Knight’s case\\nthe penalty-phase jury had necessarily made the factual findings necessary to\\nimpose the death penalty—that “sufficient aggravators existed” and that “the\\n\\naggravation outweighed the mitigation”—when it returned a unanimous vote\\nrecommending death.1 Id. at 682–83 (citation omitted). As for his ineffective-\\nassistance claim, the court held that Knight had failed to meet his burden under\\nStrickland because he had not shown that his attorney’s decision not to call his\\nDNA expert constituted deficient performance, or that there was any reasonable\\nprobability that that decision negatively affected the outcome of his trial. Id. at\\n673–74.\\n\\n\\n\\n\\n1\\n Three out of seven justices joined the opinion on Knight’s Hurst claim. Two additional justices\\nconcurred in the result only. Knight, 225 So. 3d at 684.\\n 5\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 6 of 28\\n\\n\\n\\n C.\\n Knight filed a petition for federal habeas review in the Southern District of\\n\\nFlorida, pursuant to 28 U.S.C. § 2254. The district court denied relief but granted\\na certificate of appealability on the two claims now before us.\\n II.\\n\\n A.\\n Federal courts may grant habeas corpus relief to prisoners who are being\\ndetained “in violation of the Constitution or laws or treaties of the United States.”\\n\\n28 U.S.C. §§ 2241(c)(3); 2254(a). But our authority to award this kind of relief to\\nstate prisoners is limited—by both statute and Supreme Court precedent.\\n First, the Antiterrorism and Effective Death Penalty Act (AEDPA) limits\\n\\nour authority to award habeas relief. A federal court may not grant a state\\nprisoner’s habeas petition on any issue that was decided on the merits by the state\\ncourt unless the state court’s ruling “(1) resulted in a decision that was contrary to,\\nor involved an unreasonable application of, clearly established Federal law, as\\ndetermined by the Supreme Court of the United States; or (2) resulted in a decision\\nthat was based on an unreasonable determination of the facts in light of the\\nevidence presented in the State court proceeding.” 28 U.S.C. § 2254(d). And as\\nthe Supreme Court has explained, “clearly established” federal law means “the\\nholdings, as opposed to the dicta” from its controlling precedents at the time of the\\nrelevant state court decision. Lockyer v. Andrade, 538 U.S. 63, 71 (2003) (quoting\\nWilliams v. Taylor, 529 U.S. 362, 412 (2000)).\\n\\n\\n\\n 6\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 7 of 28\\n\\n\\n\\n A decision is “contrary to” clearly established federal law “if the state court\\narrives at a conclusion opposite to that reached by [the Supreme] Court on a\\n\\nquestion of law or if the state court decides a case differently than [the Supreme]\\nCourt has on a set of materially indistinguishable facts.” Williams, 529 U.S. at\\n413. A state court decision involves an unreasonable application of federal law “if\\n\\nthe state court identifies the correct governing legal principle from [the Supreme]\\nCourt’s decisions but unreasonably applies that principle to the facts of the\\nprisoner’s case.” Id. To justify issuance of the writ under the “unreasonable\\n\\napplication” clause, the state court’s application of Supreme Court precedent must\\nbe more than just wrong in the eyes of the federal court; it “must be ‘objectively\\nunreasonable.’” Virginia v. LeBlanc, 137 S. Ct. 1726, 1728 (2017) (quoting\\n\\nWoods v. Donald, 135 S. Ct. 1372, 1376 (2015)); see also Bell v. Cone, 535 U.S.\\n685, 694 (2002) (explaining that “an unreasonable application is different from an\\nincorrect one.”).\\n Second, Supreme Court precedent demands that in any federal habeas\\nproceeding—including collateral proceedings in capital cases—where the\\npetitioner seeks the benefit of a “new” rule of constitutional law, we must first\\ndetermine whether the rule actually qualifies as new, and then whether that rule\\napplies retroactively to the case. See Teague v. Lane, 489 U.S. 288, 300–01 (1989)\\n(plurality opinion); Penry v. Lynaugh, 492 U.S. 302, 313–14 (1989) (stating that\\nthe retroactivity approach from Teague applies in capital cases), abrogated on\\nother grounds by Atkins v. Virginia, 536 U.S. 304, 312–16, 321 (2002). In most\\ncases, we cannot disturb a state conviction based on a constitutional rule\\n\\n 7\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 8 of 28\\n\\n\\n\\nannounced after a conviction became final. Teague, 489 U.S. at 310. Only two\\nnarrow exceptions pierce this general principle of nonretroactivity: new rules that\\n\\nare “substantive rather than procedural,” and “watershed rules of criminal\\nprocedure implicating the fundamental fairness and accuracy of the criminal\\nproceeding.” Schriro v. Summerlin, 542 U.S. 348, 352–53 (2004) (internal\\n\\nquotation marks and citations omitted). In all other cases the rule applies only\\nprospectively.\\n What this means in plain English is that, in the vast majority of cases,\\n\\nprisoners will not be able to secure federal habeas relief based on a new\\nconstitutional rule—even when that rule runs in their favor. “This is but a\\nrecognition that the purpose of federal habeas corpus is to ensure that state\\n\\nconvictions comply with the federal law in existence at the time the conviction\\nbecame final, and not to provide a mechanism for the continuing reexamination of\\nfinal judgments based upon later emerging legal doctrine.” Sawyer v. Smith, 497\\nU.S. 227, 234 (1990).\\n Though these two constraints—the rule of nonretroactivity set out in Teague\\nand the deference to state court decisions mandated by AEDPA—are similar in\\nsome respects, they are nonetheless “quite separate” in their operation, and a state\\nprisoner seeking federal habeas relief must clear both hurdles to succeed. Greene\\nv. Fisher, 565 U.S. 34, 39 (2011). Accordingly—and the Supreme Court has made\\nthis clear—“in addition to performing any analysis required by AEDPA, a federal\\ncourt considering a habeas petition must conduct a threshold Teague analysis when\\n\\n\\n\\n 8\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 9 of 28\\n\\n\\n\\nthe issue is properly raised by the state.” Horn v. Banks (Banks I), 536 U.S. 266,\\n272 (2002) (per curiam).\\n\\n B.\\n Before conducting that analysis here, we pause to explain why we cannot\\nsimply accept the Florida Supreme Court’s decision to apply Hurst retroactively to\\n\\nKnight and review only its harmless-error analysis, as Knight urges us to do.\\nBecause the Florida Supreme Court had already decided to give him the benefit of\\nHurst, Knight says, the Teague retroactivity analysis no longer has any bearing in\\n\\nhis case. He is wrong. While states may fashion their own retroactivity doctrines\\nas a matter of state law, those doctrines cannot displace Teague on the federal\\nstage. Our ability to consider whether Florida applied Hurst correctly depends\\n\\nentirely on whether we can apply Hurst ourselves. So far, neither the Supreme\\nCourt nor this Circuit has answered that question by analyzing Hurst’s retroactivity\\nunder Teague.2\\n Florida, on the other hand, has its own retroactivity standard—and is free to\\ngive broader retroactive effect to new constitutional rules in state court proceedings\\n\\n\\n2\\n We have noted in passing that Hurst would not apply retroactively to a petitioner whose\\nconvictions became final long before the Supreme Court decided Ring v. Arizona, 536 U.S. 584\\n(2002), on which Hurst relied, and even before Apprendi v. New Jersey, 530 U.S. 466 (2000),\\nwhich formed the basis for Ring. See Lambrix v. Sec’y, Fla. Dep’t of Corr., 851 F.3d 1158, 1165\\nn.2 (11th Cir. 2017). And we have concluded in the context of an improvidently granted\\ncertificate of appealability that the Florida Supreme Court’s decision not to apply Hurst\\nretroactively to the same petitioner as a matter of state law was not contrary to or an\\nunreasonable application of existing Supreme Court precedents. See Lambrix v. Sec’y, Dep’t of\\nCorr., 872 F.3d 1170, 1182 (11th Cir. 2017) (per curiam). But the question of Hurst’s\\nretroactivity under Teague to a petitioner like Knight—whose convictions became final after\\nRing but before Hurst—was not squarely presented in either case, so we did not conduct that\\nanalysis.\\n 9\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 10 of 28\\n\\n\\n\\nthan Teague allows in federal cases. See Danforth v. Minnesota, 552 U.S. 264,\\n282 (2008). That is because the Teague bar “was intended to limit the authority of\\n\\nfederal courts to overturn state convictions—not to limit a state court’s authority to\\ngrant relief for violations of new rules of constitutional law when reviewing its\\nown State’s convictions.” Id. at 280–81. So when states choose to apply new rules\\n\\nof constitutional procedure that are not retroactive under Teague in federal courts,\\nthey “do not do so by misconstruing the federal Teague standard. Rather, they\\nhave developed state law to govern retroactivity in state postconviction\\n\\nproceedings.” Id. at 288–89 (emphasis in original).\\n In deciding to apply Hurst retroactively to certain state habeas cases, the\\nFlorida Supreme Court did just that. See Mosley v. State, 209 So. 3d 1248, 1274–\\n\\n83 (Fla. 2016) (per curiam). Florida’s retroactivity doctrine is unique to Florida—\\nit applies to a limited class of constitutional decisions that announce changes “of\\nfundamental significance,” which for procedural rules requires consideration of\\nthree factors: “(a) the purpose to be served by the rule, (b) the extent of reliance on\\nthe prior rule, and (c) the effect that retroactive application of the new rule would\\nhave on the administration of justice.” Asay v. State, 210 So. 3d 1, 16–17 (Fla.\\n2016) (per curiam) (citation omitted). For new constitutional rules involving the\\ndeath penalty, Florida courts also consider on a case-by-case basis whether\\nfundamental fairness requires retroactive application of the rule. Mosley, 209 So.\\n3d at 1274 (citing James v. State, 615 So. 2d 668, 668 (Fla. 1993)).3 If that sounds\\n\\n3\\n The Florida Supreme Court is now considering whether it should recede from the retroactivity\\nanalysis employed in Asay, Mosley, and James. See Owen v. State, No. SC18-810, April 24,\\n\\n 10\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 11 of 28\\n\\n\\n\\nbroader than Teague, it is for good reason—the Florida Supreme Court has itself\\nacknowledged that this retroactivity standard is “more expansive” than the federal\\n\\nrule. Johnson v. State, 904 So. 2d 400, 409 (Fla. 2005) (per curiam), abrogated on\\nunrelated grounds by Asay, 210 So. 3d at 15–16.\\n All that to say, Florida may make its own choice about the retroactivity of a\\n\\ngiven case as a matter of state law. And for Hurst, it has done so. Using its own\\nstandard, the Florida Supreme Court decided that Hurst would apply retroactively\\nin state collateral review proceedings for petitioners whose convictions had not yet\\n\\nbecome final when the U.S. Supreme Court decided Ring v. Arizona; Ring held in\\n2002 that the Sixth Amendment requires a jury (rather than a judge) to find the\\naggravating factors necessary to impose the death penalty. See Mosley, 209 So. 3d\\n\\nat 1283 (citing Ring v. Arizona, 536 U.S. 584 (2002)). The Florida Supreme Court\\nreasoned that because Hurst struck down Florida’s capital sentencing scheme\\nbased on Ring, prisoners whose cases were still pending on direct appeal when\\nRing was decided “should not suffer due to the United States Supreme Court’s\\nfourteen-year delay in applying Ring to Florida.” Id. Following that rule, because\\nKnight’s conviction became final ten years after Ring was decided, the Florida\\nSupreme Court applied Hurst retroactively in his postconviction proceeding.\\nKnight, 225 So. 3d at 682.\\n\\n\\n\\n2019 order (directing parties to brief the issue). The uncertain fate of Florida’s current\\nretroactivity doctrine offers another reason that we cannot simply rely on a state retroactivity\\ndecision as a basis for federal habeas relief. If the state doctrine were to change during our\\nreview, would we then be faced with the question of whether to apply the state’s new\\nretroactivity doctrine—retroactively?\\n 11\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 12 of 28\\n\\n\\n\\n But that state-law retroactivity determination has no significance in federal\\ncourt. Unlike state courts, lower federal courts are not free to create our own rules\\n\\nof retroactivity—if the government raises the issue, a Teague analysis is\\nmandatory. Banks I, 536 U.S. at 271–72. As we have said, “States may exercise\\ntheir collateral review power without regard to the Teague doctrine. Their doing\\n\\nso has no effect on later federal review.” Glock v. Singletary, 65 F.3d 878, 890–91\\n(11th Cir. 1995) (en banc). So we are bound to follow Teague’s retroactivity\\nprinciples whether or not the state court chose to apply the new rule in its own\\n\\ncollateral proceeding. See, e.g., Banks I, 536 U.S. at 271–72.\\n Here, then, we must conduct our own retroactivity analysis, using the\\nstandards articulated in Teague. And to repeat: Teague retroactivity is a “threshold\\n\\nquestion in every habeas case.” Caspari v. Bohlen, 510 U.S. 383, 389 (1994).\\nWhen “issues of both retroactivity and application of constitutional doctrine are\\nraised,” we must decide the retroactivity issue first. Bowen v. United States, 422\\nU.S. 916, 920 (1975). Where the State raises the issue, therefore, “federal habeas\\ncorpus courts must apply Teague before considering the merits” of the petitioner’s\\nclaims. Beard v. Banks (Banks II), 542 U.S. 406, 412 (2004) (emphasis in\\noriginal) (internal quotation marks and citation omitted).\\n What’s more, if a constitutional claim is Teague-barred, we do not reach its\\nmerits. See, e.g., id. at 410 n.2. That is because the Supreme Court’s\\n“jurisprudence concerning the ‘retroactivity’ of ‘new rules’ of constitutional law is\\nprimarily concerned, not with the question whether a constitutional violation\\noccurred, but with the availability or nonavailability of remedies.” Danforth, 552\\n\\n 12\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 13 of 28\\n\\n\\n\\nU.S. at 290–91. If the holding relied on qualifies as a new rule and does not meet\\nTeague’s strict requirements for retroactivity, then the claim is not redressable\\n\\nhere—the “nonretroactivity principle prevents a federal court from granting habeas\\ncorpus relief to a state prisoner based on a rule announced after his conviction and\\nsentence became final.” Caspari, 510 U.S. at 389 (emphasis in original). And as\\n\\nwe have said before, if “the court cannot relieve the harm of which a plaintiff\\ncomplains, the court should not take the case; in the absence of an effective remedy\\nits decision can amount to nothing more than an advisory opinion.” Wymbs v.\\n\\nRepublican State Exec. Comm., 719 F.2d 1072, 1085 (11th Cir. 1983).\\n Our authority to overturn state convictions is limited, and the retroactivity\\nprinciples articulated in Teague are tailored to those limitations. See Danforth, 552\\n\\nU.S. at 277. The fact that state courts do not face the same constraints on collateral\\nreview of their own criminal proceedings as we do does not relieve us of the\\nobligation to apply federal retroactivity standards. See id. at 280–81; Banks I, 536\\nU.S. at 271. In fact, our narrow authority as federal courts to disrupt final state-\\ncourt convictions reflects our recognition of the states’ own sovereignty. So\\nFlorida may design and apply its retroactivity principles as generously as it wishes.\\nBut notwithstanding Florida’s decision to apply Hurst—or any future decision—\\nretroactively as a matter of state law, as a federal court we are required to perform\\nthe Teague analysis to determine whether prisoners can receive retroactive relief\\nunder federal law.\\n In sum, if Hurst announced a new rule of constitutional law, but one that\\ndoes not fall into one of the exceptions to Teague’s bar on retroactivity, Knight\\n\\n 13\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 14 of 28\\n\\n\\n\\ncannot obtain federal habeas relief for any Hurst error in his sentence—regardless\\nof what Florida may choose to do under state law. And if Knight cannot obtain\\n\\nfederal habeas relief for his Hurst claim in any event, we may not offer an advisory\\nopinion on whether the claim could have merit.\\n III.\\n\\n A.\\n Turning to Teague, our analysis has three steps. First, we determine the date\\nwhen the petitioner’s conviction became final. Banks II, 542 U.S. at 411. This\\n\\nhappens when the United States Supreme Court “affirms a conviction on the merits\\non direct review or denies a petition for a writ of certiorari, or when the time for\\nfiling a certiorari petition expires.” Clay v. United States, 537 U.S. 522, 527\\n\\n(2003). Here, the Court denied Knight’s petition for a writ of certiorari on May 14,\\n2012—more than three years before Hurst was decided. Knight, 566 U.S. at 998.\\n Second, if the rule that the petitioner wants to apply had not been announced\\nby that final-conviction date, we “assay the legal landscape” as it existed at the\\ntime and determine whether existing precedent compelled the rule—that is,\\nwhether the case announced a new rule or applied an old one. Banks II, 542 U.S.\\nat 413. If—and only if—the holding was “dictated by precedent existing at the\\ntime the defendant’s conviction became final,” then the rule is not new and may be\\napplied retroactively on federal habeas review (indeed, it must be). Caspari, 510\\nU.S. at 390 (emphasis in original) (quoting Teague, 489 U.S. at 301). And that is\\nnot a light test—a rule is not dictated by prior precedent “unless it would have been\\n‘apparent to all reasonable jurists.’” Chaidez v. United States, 568 U.S. 342, 347\\n\\n 14\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 15 of 28\\n\\n\\n\\n(2013) (quoting Lambrix v. Singletary, 520 U.S. 518, 528 (1997)). The fact that a\\ndecision is “within the logical compass of” or even “controlled by” prior precedent\\n\\nis not conclusive in the Teague analysis. Butler v. McKellar, 494 U.S. 407, 415\\n(1990) (internal quotation marks and citation omitted). To “‘dictate’ a result, prior\\nprecedent must be specific; it is not enough that it name the general principle from\\n\\nwhich the assertedly new rule sprang.” Glock, 65 F.3d at 884.\\n Knight argues that Ring v. Arizona dictated the rule in Hurst. In Ring, the\\nSupreme Court held that Arizona’s capital sentencing scheme, which required\\n\\njudges alone to hear penalty-phase evidence and make factual findings relevant to\\nthe imposition of the death penalty, violated the defendant’s Sixth Amendment\\nright to a jury trial. 536 U.S. at 588–89. In doing so, the Court explicitly\\n\\noverruled its precedent upholding Arizona’s death penalty sentencing scheme, “to\\nthe extent that it allows a sentencing judge, sitting without a jury, to find an\\naggravating circumstance necessary for imposition of the death penalty.” Id. at\\n609 (emphasis added) (overruling Walton v. Arizona, 497 U.S. 639, 647–49\\n(1990)).\\n Knight wants us to find that the new Hurst rule is actually the old Ring rule\\nfor an obvious reason—if the rule is not new, and instead was binding on lower\\ncourts at the time of Knight’s conviction, then he is entitled to the benefit of the\\nrule on federal habeas review. See Yates v. Aiken, 484 U.S. 211, 216–17 & n.3\\n(1988). But Ring did not dictate the Supreme Court’s later invalidation of\\nFlorida’s death penalty sentencing scheme in Hurst. In fact, the Ring Court\\nspecifically acknowledged that Florida’s capital sentencing procedure differed\\n\\n 15\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 16 of 28\\n\\n\\n\\nfrom the Arizona scheme that it rejected. See Ring, 536 U.S. at 607–08 & n.6\\n(categorizing state capital sentencing schemes according to jury involvement in\\n\\nsentencing). In Arizona, the judge alone made the factual findings necessary to\\nimpose the death penalty and imposed that penalty entirely apart from the jury.\\nSee id. at 588. Florida’s scheme, in contrast, incorporated an advisory jury that\\n\\nconsidered penalty-phase evidence and recommended a sentence of life or death to\\nthe court. Only after the jury’s recommendation did the judge impose a sentence.\\nSee id. at 608 n.6; Fla. Stat. Ann. § 921.141 (2001).\\n\\n Hurst’s conclusion that Florida’s “hybrid” death penalty sentencing system\\nviolated the Sixth Amendment was not “apparent to all reasonable jurists” when\\nKnight’s convictions became final in 2012. We venture to count ourselves and our\\n\\ncolleagues on this Court as members of that distinguished group. And in Evans v.\\nSecretary, Florida Department of Corrections, we held that Ring did not invalidate\\nFlorida’s capital sentencing scheme. 699 F.3d 1249, 1264–65 (11th Cir. 2012).\\nWe found it significant that Florida’s statute—unlike the Arizona law at issue in\\nRing—required the penalty phase jury to find that sufficient aggravating\\ncircumstances existed before it could recommend a death sentence, and directed\\nthe sentencing court to give “great weight” to the jury’s advisory verdict. Id. at\\n1261 (citation omitted). We also noted that the Supreme Court had taken obvious\\npains in Ring to distinguish hybrid systems like Florida’s from the “judge-only”\\nsentencing scheme in Arizona and concluded that such distinctions would not have\\nbeen necessary if the Court had intended to strike down both systems. Id. at 1262.\\n\\n\\n\\n 16\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 17 of 28\\n\\n\\n\\n And we were not the only ones. In fact, Justice Alito wrote along those\\nsame lines in his dissenting opinion in Hurst: “Although the Court suggests that\\n\\ntoday’s holding follows ineluctably from Ring, the Arizona sentencing scheme at\\nissue in that case was much different from the Florida procedure now before us.”\\n136 S. Ct. at 625 (Alito, J., dissenting). After describing the “critically important\\n\\nrole” of the advisory jury in Florida’s death penalty sentencing scheme, Justice\\nAlito concluded that the “decision in Ring did not decide whether this procedure\\nviolates the Sixth Amendment, and I would not extend Ring to cover the Florida\\n\\nsystem.” Id. at 626. Clearly, reasonable jurists could—and did—disagree that\\nRing compelled the outcome in Hurst. The Alito dissent and this Court’s pre-\\nHurst holding strongly indicate that Hurst announced a new constitutional rule\\n\\nrather than applying an old one. See Banks II, 542 U.S. at 415.\\n That conclusion is only strengthened by the fact that Ring did not\\nspecifically address the continued validity of the Supreme Court’s precedents\\nupholding Florida’s death-penalty sentencing system—Spaziano v. Florida, 468\\nU.S. 447 (1984), and Hildwin v. Florida, 490 U.S. 638 (1989) (per curiam). And\\nthe Court has repeatedly instructed us to follow its precedents, even if later\\ndecisions appear to undermine them, unless and until the Court itself sets them\\naside. See, e.g., Bosse v. Oklahoma, 137 S. Ct. 1, 2 (2016) (per curiam) (“Our\\ndecisions remain binding precedent until we see fit to reconsider them, regardless\\nof whether subsequent cases have raised doubts about their continuing vitality.”\\n(citation omitted)); Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S.\\n477, 484 (1989). So Spaziano and Hildwin remained good law until the Court\\n\\n 17\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 18 of 28\\n\\n\\n\\nexplicitly overruled them in Hurst. See Saffle v. Parks, 494 U.S. 484, 488 (1990)\\n(explaining that the “explicit overruling of an earlier holding no doubt creates a\\n\\nnew rule”); see also Hurst, 136 S. Ct. at 623. It is hard to imagine that the\\nSupreme Court overruled those cases in Ring but forgot to say so until Hurst.\\n Because all these factors show that Hurst was not dictated by prior\\n\\nprecedent—and in fact explicitly overruled existing precedent upholding Florida’s\\ndeath penalty sentencing scheme—we can see that the rule in Hurst, which led to a\\nconclusion that the Florida scheme was unconstitutional, was new.\\n\\n Having determined that Hurst announced a new rule of constitutional law,\\nwe proceed to the final step in the Teague analysis—whether Hurst “falls within\\neither of the two exceptions to nonretroactivity.” Banks II, 542 U.S. at 411. Those\\n\\nexceptions, again, include (1) holdings that create substantive (not procedural)\\nrules that place “certain kinds of primary, private individual conduct beyond the\\npower of the criminal law-making authority to proscribe,” and (2) holdings that\\nconstitute “watershed rules of criminal procedure.” Teague, 489 U.S. at 311–13\\n(citation omitted).\\n The Hurst rule does not fit within either exception. To begin, substantive\\nrules include decisions that change “the range of conduct or the class of persons\\nthat the law punishes.” Schriro, 542 U.S. at 353. Procedural rules, on the other\\nhand, “regulate only the manner of determining the defendant’s culpability.” Id.\\n(emphasis in original). In considering which category the Hurst rule falls into, we\\nhave a head start because the Supreme Court has already held that Ring represented\\na “prototypical procedural rule[].” Id. And that makes sense: Ring changed the\\n\\n 18\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 19 of 28\\n\\n\\n\\npermissible procedure for sentencing in a capital case when it required “that a jury\\nrather than a judge find the essential facts” necessary to impose the death penalty.\\n\\nSee id. Because Hurst’s holding—that an advisory “jury’s mere recommendation\\nis not enough” to satisfy this procedural requirement—is an extension of the rule\\nfrom Ring, we have no trouble concluding that Hurst also announced a procedural\\n\\nrule, and not a substantive one. Hurst, 136 S. Ct. at 619.\\n The second exception is for “watershed rules of criminal procedure.” Banks\\nII, 542 U.S. at 417. This exception is extremely limited in scope—it applies “only\\n\\nto a small core of rules” so fundamental to our criminal process that it is “unlikely\\nthat many such components of basic due process have yet to emerge.” Id.\\n(citations omitted). Indeed, the watershed exception remains somewhat theoretical\\n\\nat this point; in the years following Teague, the Supreme Court has never found a\\nrule that fits. See id. And in “providing guidance as to what might fall within this\\nexception,” the Court has “repeatedly referred to the rule of Gideon v. Wainwright,\\n372 U. S. 335 (1963) (right to counsel), and only to this rule.” Id. Knight does not\\ncontend that Hurst announced a new watershed rule that compares to Gideon, and\\nwe do not see how it could have either. In short, Hurst meets neither exception,\\nand therefore is not retroactive.\\n B.\\n Because the Hurst rule is not retroactive, Knight cannot receive federal\\nhabeas relief on his Hurst claim. That is as it must be—we are conscientious about\\nthe fact that “Teague’s nonretroactivity principle acts as a limitation on the power\\nof federal courts to grant habeas corpus relief.” Id. at 412 (internal quotation\\n\\n 19\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 20 of 28\\n\\n\\n\\nmarks and citation omitted). And because we would have no lawful remedy to\\noffer even if we could identify an error, we must decline to consider whether any\\n\\nHurst error exists. That means, of course, that we also do not consider whether the\\nun-found and un-remediable error could be harmless.\\n Knight disputes this path. He argues that, even conceding the lack of an\\n\\navailable remedy for any Hurst error, whether that un-remediable error was\\nharmless is itself a separate question of federal law under Brecht v. Abrahamson,\\n507 U.S. 619, 637–38 (1993). He thus urges us to review the state court’s\\n\\nharmless-error analysis, regardless of whether we reach the merits of his Hurst\\nclaim. We will not. It strains the imagination—as well as our constitutional and\\ninstitutional respect for state courts—to suppose that we cannot remedy an error,\\n\\nbut that we can somehow remedy an erroneous state-court conclusion that the error\\nwas harmless.\\n And Brecht does not stand for the proposition that Knight asserts in any\\nevent. While Brecht established the harmless-error standard for collateral review\\nof constitutional trial errors, it did not create a stand-alone federal claim severable\\nfrom the question of whether remediable error existed in the first place.4 See\\nWilliams v. Singletary, 114 F.3d 177, 180 (11th Cir. 1997) (per curiam), as\\namended on denial of reh’g (July 29, 1997). Some underlying violation of federal\\nlaw that we can address is a necessary predicate to federal habeas relief—unless\\n\\n4\\n As an aside, what Brecht really decided was that federal courts evaluating constitutional trial\\nerror on collateral review would apply a more relaxed harmless-error standard—whether the\\nerror “had substantial and injurious effect or influence in determining the jury’s verdict”—rather\\nthan the harmless-beyond-a-reasonable-doubt standard used on direct review. Brecht, 507 U.S.\\nat 638 (quoting Kotteakos v. United States, 328 U.S. 750, 776 (1946)).\\n 20\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 21 of 28\\n\\n\\n\\nwe agree that an error has occurred, it makes no difference whether the purported\\nerror was harmless. “We have consistently applied the Brecht harmless error\\n\\nstandard only after determining that there was an error.” Id.; cf. Wilson v.\\nCorcoran, 562 U.S. 1, 6 (2010) (per curiam) (“It is not enough to note that a\\nhabeas petitioner asserts the existence of a constitutional violation; unless the\\n\\nfederal court agrees with that assertion, it may not grant relief.” (emphasis in\\noriginal)). There is no free-floating federal constitutional right to infallible\\napplication of harmless-error principles.\\n\\n And again, where our retroactivity doctrine forecloses the possibility of\\nfederal habeas relief for a constitutional error, we are constrained to stop looking.\\nSee Bowen, 422 U.S. at 920. The Supreme Court “consistently has declined to\\n\\naddress unsettled questions regarding the scope of decisions establishing new\\nconstitutional doctrine in cases in which it holds those decisions nonretroactive,”\\nand has instructed us to do the same. Id. at 920–21. This directive carries\\nconstitutional weight; as an Article III court, we are “without power to decide\\nquestions that cannot affect the rights of litigants in the case” at hand. North\\nCarolina v. Rice, 404 U.S. 244, 246 (1971) (per curiam). So even if we found\\nHurst error in Knight’s sentencing, we would still be prohibited from issuing a writ\\nof habeas corpus on that ground because Hurst is not retroactively applicable to\\nKnight under Teague. After all, Teague’s nonretroactivity command is a limitation\\non our power, not a polite suggestion. Banks II, 542 U.S. at 412. So our opinion—\\nwhatever it might be—on Knight’s Hurst claim would be purely advisory. “And it\\nis quite clear that the oldest and most consistent thread in the federal law of\\n\\n 21\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 22 of 28\\n\\n\\n\\njusticiability is that the federal courts will not give advisory opinions.” Flast v.\\nCohen, 392 U.S. 83, 96 (1968) (internal quotation marks and citation omitted).\\n\\n Where, as here, Teague bars relief before we reach the preliminary question\\nof whether constitutional error occurred at all, consideration of the secondary\\nquestion of whether any such error was harmless would be a prohibited and\\n\\npointless exercise for both the petitioner and this Court. We therefore cannot grant\\nKnight relief on his Hurst claim, whether or not it is cloaked in the garb of\\nharmless error.\\n\\n IV.\\n We now turn to Knight’s other claim. The State presented extensive DNA\\nevidence against him during the guilt phase of his trial. Kevin Noppinger, a\\n\\nserologist with the Broward County Sheriff’s Office crime lab, testified that Knight\\nhad Odessia’s blood on his hand and her DNA on his shirt when he was arrested.\\nFingernail scrapings from Odessia’s body showed that she, in turn, had Knight’s\\nDNA under her fingernails. Noppinger also tested samples from the bloody\\nclothes (boxer shorts, a shirt, and jean shorts) found under the bathroom sink. He\\nfound Knight’s blood mixed with Hanessia’s blood on the boxer shorts, Odessia’s\\nand Hanessia’s blood elsewhere on the boxer shorts and on the jean shorts, and\\nOdessia’s blood on the shirt.\\n One of the State’s other experts was Kevin McElfresh of Bode Technology\\nGroup, whose DNA analysts had conducted additional testing on different samples\\nfrom the same items of clothing. In particular, McElfresh’s group analyzed DNA\\nsamples from an unstained area of the waistband of the boxer shorts in an attempt\\n\\n 22\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 23 of 28\\n\\n\\n\\nto determine who owned them. Although Bode’s initial report stated that Knight’s\\nDNA was not on the waistband sample, McElfresh testified at trial that he had\\n\\nconducted some additional analysis and determined that some of the DNA on the\\nwaistband could have been Knight’s.\\n Knight’s guilt-phase counsel consulted with DNA expert Dr. Norah Rudin,\\n\\nwho was listed as a potential trial witness by the defense. Dr. Rudin informed\\ncounsel that, although some of Noppinger’s sample labeling practices were sloppy,\\nshe generally agreed with his conclusions about the sources of the DNA samples\\n\\nhe analyzed. She was much more critical of McElfresh’s analysis: she called his\\nmethods “fundamentally incorrect and inherently biased” and his testimony\\n“incomplete and misleading.” In her opinion, the DNA test results for the\\n\\nwaistband samples were “inconclusive.” But Dr. Rudin did not stop there—she\\nalso considered McElfresh’s testimony “relatively inconsequential when viewed in\\nthe context of the biological evidence as a whole.” She told Knight’s counsel that\\nshe did not believe that she could help his case, and that she would not call herself\\nas an expert if she were in his shoes. Counsel, it appears, agreed with her\\nperspective and did not call her at trial.\\n Testimony indicates that one additional fact persuaded counsel that calling\\nDr. Rubin to quibble with McElfresh’s methods would not be good trial strategy.\\nAt the time of Knight’s trial, Florida permitted a defendant who did not put up any\\nevidence at trial to give both initial and rebuttal closing arguments—an advantage\\nordinarily offered to the prosecution. Dr. Rudin would have been Knight’s only\\nwitness, so calling her to testify would have meant giving up the opportunity to\\n\\n 23\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 24 of 28\\n\\n\\n\\nhave the first word and the last at closing. Given that fact, she would have largely\\ncorroborated Noppinger’s testimony while also costing Knight an advantage at\\n\\nclosing arguments.\\n Nonetheless, on state collateral review, Knight contended that his counsel\\nprovided constitutionally ineffective assistance by failing to call Dr. Rudin as an\\n\\nexpert. He argued that Dr. Rudin’s criticisms of McElfresh’s methods and the\\ncrime lab’s labeling practices would have cast doubt on all of the State’s DNA\\nevidence and significantly damaged the State’s case. The Florida Supreme Court\\n\\ndisagreed, finding that Knight had not met his burden on either prong of the two-\\npart test for ineffective assistance of counsel claims set out in Strickland. See\\nKnight, 225 So. 3d at 672–74. Knight, however, argues that the state court’s\\n\\ndecision “was contrary to, or involved an unreasonable application of, clearly\\nestablished Federal law, as determined by the Supreme Court of the United States.”\\n28 U.S.C. § 2254(d)(1).\\n To begin, the Florida Supreme Court correctly identified the governing\\nstandard. Strickland is the relevant “clearly established” Supreme Court precedent\\nfor purposes of an ineffective-assistance claim. See Premo v. Moore, 562 U.S.\\n115, 118 (2011). Under Strickland, to succeed on a claim of ineffective assistance\\nof counsel, a defendant must show both that his attorney’s performance was\\ndeficient—that is, “that counsel’s representation fell below an objective standard of\\nreasonableness”—and that he was prejudiced by the inadequate performance.\\nStrickland, 466 U.S. at 687–88. In applying the first prong, courts “must indulge a\\nstrong presumption that counsel’s conduct falls within the wide range of\\n\\n 24\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 25 of 28\\n\\n\\n\\nreasonable professional assistance.” Id. at 689. And to show prejudice, the\\n“defendant must show that there is a reasonable probability that, but for counsel’s\\n\\nunprofessional errors, the result of the proceeding would have been different. A\\nreasonable probability is a probability sufficient to undermine confidence in the\\noutcome.” Id. at 694.\\n\\n The Strickland standard is “highly deferential,” as is the review of a state-\\ncourt decision under AEDPA; “when the two apply in tandem, review is ‘doubly’\\nso.” Harrington v. Richter, 562 U.S. 86, 105 (2011) (citations omitted). In\\n\\nreviewing a state court’s application of Strickland, therefore, a federal habeas court\\ncannot conduct a de novo review and reverse simply because it strongly disagrees\\nwith the state court’s conclusion. Id. at 102. Instead, for Knight to succeed on his\\n\\nineffective-assistance claim, we must conclude that the Florida Supreme Court’s\\ndecision “was an objectively unreasonable application of the Strickland standard.”\\nAllen v. Sec’y, Fla. Dep’t of Corr., 611 F.3d 740, 751 (11th Cir. 2010).\\n The Florida Supreme Court determined that counsel’s decision not to call\\nDr. Rudin at trial was reasonable trial strategy. Knight, 225 So. 3d at 674.\\nWhether the decision was actually a matter of strategy is a question of fact; thus,\\nthe state court’s finding on that issue is presumed to be correct. Provenzano v.\\nSingletary, 148 F.3d 1327, 1330 (11th Cir. 1998). We must accept all factual\\nfindings made by the state court unless the petitioner rebuts the presumption of\\ncorrectness “by clear and convincing evidence.” 28 U.S.C. § 2254(e)(1); see also\\nWood v. Allen, 542 F.3d 1281, 1285 (11th Cir. 2008). Knight has not presented\\n\\n\\n\\n 25\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 26 of 28\\n\\n\\n\\nany evidence to suggest that counsel’s decision was anything other than a matter of\\nstrategy, and we accept the state court’s finding on that point.\\n\\n Whether counsel’s strategic decision not to call Dr. Rudin was reasonable is\\na question of law, which we review through the lens of AEDPA deference. See\\nFerrell v. Hall, 640 F.3d 1199, 1223–24 (11th Cir. 2011). In assessing an\\n\\nattorney’s performance under Strickland, “strategic choices made after thorough\\ninvestigation of law and facts relevant to plausible options are virtually\\nunchallengeable.” Strickland, 466 U.S. at 690. “Which witnesses, if any, to call,\\n\\nand when to call them, is the epitome of a strategic decision, and it is one that we\\nwill seldom, if ever, second guess.” Waters v. Thomas, 46 F.3d 1506, 1512 (11th\\nCir. 1995) (en banc). We have no reason to doubt the Florida Supreme Court’s\\n\\nconclusion that counsel’s decision not to call an equivocal expert, in part to\\npreserve an advantage at closing, was well within the wide range of reasonably\\ncompetent performance.\\n The Florida Supreme Court also determined that Knight had not met his\\nburden of showing prejudice under Strickland because there was no reasonable\\nprobability that Dr. Rudin’s testimony would have made a difference in the\\noutcome of the trial, given the weight of the evidence against him. See Knight, 225\\nSo. 3d at 674. That conclusion was also reasonable.\\n Dr. Rudin generally agreed with Noppinger’s conclusions regarding the\\nDNA evidence. With or without Dr. Rudin’s testimony, therefore, there was no\\ndispute that Knight had Odessia’s blood on his hand and her DNA on his shirt\\nwhen he was arrested, or that Knight’s DNA was found under Odessia’s\\n\\n 26\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 27 of 28\\n\\n\\n\\nfingernails. Knight’s blood—as well as Odessia’s blood and Hanessia’s—was on\\nthe clothes discarded at the crime scene. And even if the jury had heard Dr.\\n\\nRudin’s criticism of McElfresh’s testimony that Knight could not be ruled out as a\\nDNA contributor for the waistband of the boxer shorts, other evidence showed that\\nthe clothes were his: the boxer shorts were the same brand and size as the ones that\\n\\nKnight was wearing when he was arrested, and Knight’s cousin testified that the\\nbloody shirt was one that Knight wore often.\\n And the evidence did not stop there. Knight was known to be home with\\n\\nHanessia less than an hour before the murders, and the jury heard evidence of\\nongoing tension with Odessia. Police found him near the scene of the murders\\nshortly after they arrived, and his answers to their questions were inconsistent with\\n\\nhis appearance. All of that is in addition to the testimony of an inmate housed with\\nKnight at the Broward County jail, who testified that Knight confessed the crime to\\nhim, providing a detailed description of events and a floorplan of the apartment. In\\nshort, even if the jury had entirely discounted McElfresh’s testimony as a result of\\nDr. Rudin’s criticism of his testing methods, the remaining evidence against\\nKnight was so strong that the chance of a not-guilty verdict still would have been\\nremote, to say the least.\\n The Florida Supreme Court’s conclusion that Knight failed to make the\\nrequired showings of deficient performance and prejudice was not an unreasonable\\napplication of Strickland.\\n\\n\\n\\n\\n 27\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 28 of 28\\n\\n\\n\\n ***\\n In sum, we are prohibited from considering Knight’s Hurst claim by\\n\\nTeague’s rule of nonretroactivity, which eliminates any possibility of relief\\nregardless of whether there was an error and regardless of whether any error was\\nharmless. And our review of Knight’s ineffective-assistance claim shows that the\\n\\nstate court’s ruling on that issue was not an unreasonable application of governing\\nSupreme Court precedent. We therefore affirm the district court’s denial of\\nKnight’s federal habeas petition.\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 28\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435086/\", \"author_raw\": \"GRANT, Circuit Judge:\"}]}"},"judge1":{"kind":"string","value":"TJOFLAT"},"judge2":{"kind":"string","value":"JORDAN"},"judge3":{"kind":"string","value":"GRANT"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"TJOFLAT\": \", Circuit\", \"JORDAN\": \", Circuit\", \"GRANT\": \", Circuit\"}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4657833/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Richard Knight v. Florida Department of Corrections\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-30\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nCase: 18-13390 Date Filed: 08/30/2019 Page: 1 of 28\n\n\n [PUBLISH]\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n ________________________\n\n No. 18-13390\n ________________________\n\n D.C. Docket No. 0:17-cv-61921-RNS\n\n\n\nRICHARD KNIGHT,\n\n Petitioner-Appellant,\n\n versus\n\nFLORIDA DEPARTMENT OF CORRECTIONS,\n\n Respondent-Appellee.\n ________________________\n\n Appeal from the United States District Court\n for the Southern District of Florida\n ________________________\n\n (August 30, 2019)\n\nBefore TJOFLAT, JORDAN, and GRANT, Circuit Judges.\nGRANT, Circuit Judge:\n Richard Knight, a Florida prisoner sentenced to death for the murders of\n\nOdessia Stephens and her daughter, Hanessia Mullings, appeals the district court’s\ndenial of his federal habeas corpus petition. At this stage—almost 20 years after\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 2 of 28\n\n\n\nthe crimes were committed and more than a decade after a Florida jury found\nKnight guilty of the murders and recommended a death sentence—Knight’s claims\n\nhave been winnowed down to two: first, that his death sentence is invalid under\nHurst v. Florida, 136 S. Ct. 616 (2016), and second, that he received ineffective\nassistance of counsel at trial. Because Hurst does not apply retroactively to\n\nKnight, any challenge to his death sentence on that basis is beyond our reach on\nfederal habeas review. Nor can Knight find success in his other challenge; the\nFlorida Supreme Court’s rejection of his ineffective-assistance claim was not an\n\nunreasonable application of Strickland v. Washington, 466 U.S. 668 (1984). We\ntherefore affirm.\n I.\n\n A.\n According to evidence introduced at his murder trial, Knight lived in an\napartment with his cousin, Hans Mullings, and Hans’s girlfriend, Odessia. Hans\nand Odessia’s four-year-old daughter, Hanessia, also lived with them in the\napartment. Odessia was tired of supporting Knight and one evening while Hans\nwas out she argued with Knight, insisting that he move out the next day. After the\nargument got heated, Knight left the house to walk around. But as he later\nconfessed to another inmate, instead of getting less angry with Odessia once he got\nsome air, Knight became increasingly irate. He returned to the apartment and after\nexchanging more words with Odessia, he got a knife from the kitchen. When he\nwent back to the master bedroom, he found Odessia and her little girl in the bed.\nHe began stabbing Odessia and continued his attack until she stopped resisting and\n\n 2\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 3 of 28\n\n\n\ncurled up on the bedroom floor. He then moved on to little Hanessia, stabbing her\nuntil his knife broke and cutting his hand in the process. As he was leaving the\n\nbedroom, he heard “popping noises” from where Hanessia lay on the floor, and he\nthought that the little girl was “drowning in her own blood.” Apparently not\nconsidering his revenge complete, he retrieved a second knife from the kitchen and\n\nreturned to continue his attack on Odessia. In the meantime, Odessia had crawled\nfrom the bedroom to the living room, where she had collapsed. Knight turned her\nover, saw that she was still alive, and started stabbing her again.\n\n Both Odessia and Hanessia died that night. In total, Odessia had 21 stab\nwounds, including 14 in the neck, 24 puncture or scratch wounds, bruising and\nligature marks consistent with having been hit and strangled with a belt, defensive\n\nwounds, and bruises from being hit or punched in the mouth and head. Little\nHanessia had four stab wounds in her upper body and neck, a deep defensive\nwound on her hand, bruises on her neck consistent with manual strangulation, and\nbruises on her arms consistent with having been grabbed.\n Knight showered and changed after completing his brutal acts, then headed\nto the living room with a rag to wipe off the knives. Interrupted by a knock on the\nfront door—it was police responding to a neighbor’s 911 call—Knight ran to his\nroom and climbed out the window.\n Shortly after they arrived, police encountered Knight near the apartment. He\ntold them that he lived there, but that he did not have a key. This was odd; the\nofficers had already found that all the doors to the apartment were locked. Knight\nwas also visibly wet—but it was not raining. Knight explained to police that he\n\n 3\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 4 of 28\n\n\n\nhad been jogging, a remarkable contention from a person who was wearing long\npants and dress shoes. He did not appear to be sweating, in any event. And\n\nKnight’s personal appearance subsequently revealed even more clues—he had\nblood on the back of his shirt, scratches on his chest and midsection, a scrape on\nhis shoulder, and fresh cuts on his hand.\n\n Knight was arrested and indicted for two counts of first-degree murder. A\nFlorida jury found him guilty as charged. That same jury heard evidence and\nargument at the penalty phase and unanimously recommended two death\n\nsentences—one for each murder. Consistent with Florida’s then-current death\npenalty sentencing procedure, the judge held an additional hearing, made his own\nfindings regarding aggravating and mitigating circumstances, and sentenced\n\nKnight to death. The Florida Supreme Court affirmed Knight’s convictions and\nsentences on direct appeal. Knight v. State, 76 So. 3d 879, 890 (Fla. 2011). The\nUnited States Supreme Court denied his petition for certiorari. Knight v. Florida,\n566 U.S. 998, 998 (2012).\n B.\n Knight filed motions for state collateral relief raising the two claims at issue\nhere, as well as others that have already been resolved. Specifically, he argued that\nthe state court should vacate his death sentence in light of Hurst v. Florida, in\nwhich the Supreme Court held—four years after Knight’s conviction was final—\nthat Florida’s death penalty sentencing scheme violated the Sixth Amendment.\n136 S. Ct. at 622. The problem identified by the Supreme Court in Hurst, and\nargued by Knight in his post-conviction pleadings, was that the jury’s role in\n\n 4\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 5 of 28\n\n\n\nsentencing was to make a non-binding recommendation; the judge alone made the\nultimate findings of fact necessary to impose the death penalty. Id. at 619, 621–22.\n\nKnight also argued that his guilt-phase counsel was constitutionally ineffective for\nfailing to call an available DNA expert.\n The Florida Supreme Court rejected his postconviction claims on the merits.\n\nKnight v. State, 225 So. 3d 661, 668 (Fla. 2017) (per curiam). A plurality of the\ncourt agreed with Knight that the sentencing procedure used in his case violated\nthe Sixth Amendment under Hurst, but also concluded that the Hurst error was\n\nharmless. Id. at 682. The plurality explained that under the facts of Knight’s case\nthe penalty-phase jury had necessarily made the factual findings necessary to\nimpose the death penalty—that “sufficient aggravators existed” and that “the\n\naggravation outweighed the mitigation”—when it returned a unanimous vote\nrecommending death.1 Id. at 682–83 (citation omitted). As for his ineffective-\nassistance claim, the court held that Knight had failed to meet his burden under\nStrickland because he had not shown that his attorney’s decision not to call his\nDNA expert constituted deficient performance, or that there was any reasonable\nprobability that that decision negatively affected the outcome of his trial. Id. at\n673–74.\n\n\n\n\n1\n Three out of seven justices joined the opinion on Knight’s Hurst claim. Two additional justices\nconcurred in the result only. Knight, 225 So. 3d at 684.\n 5\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 6 of 28\n\n\n\n C.\n Knight filed a petition for federal habeas review in the Southern District of\n\nFlorida, pursuant to 28 U.S.C. § 2254. The district court denied relief but granted\na certificate of appealability on the two claims now before us.\n II.\n\n A.\n Federal courts may grant habeas corpus relief to prisoners who are being\ndetained “in violation of the Constitution or laws or treaties of the United States.”\n\n28 U.S.C. §§ 2241(c)(3); 2254(a). But our authority to award this kind of relief to\nstate prisoners is limited—by both statute and Supreme Court precedent.\n First, the Antiterrorism and Effective Death Penalty Act (AEDPA) limits\n\nour authority to award habeas relief. A federal court may not grant a state\nprisoner’s habeas petition on any issue that was decided on the merits by the state\ncourt unless the state court’s ruling “(1) resulted in a decision that was contrary to,\nor involved an unreasonable application of, clearly established Federal law, as\ndetermined by the Supreme Court of the United States; or (2) resulted in a decision\nthat was based on an unreasonable determination of the facts in light of the\nevidence presented in the State court proceeding.” 28 U.S.C. § 2254(d). And as\nthe Supreme Court has explained, “clearly established” federal law means “the\nholdings, as opposed to the dicta” from its controlling precedents at the time of the\nrelevant state court decision. Lockyer v. Andrade, 538 U.S. 63, 71 (2003) (quoting\nWilliams v. Taylor, 529 U.S. 362, 412 (2000)).\n\n\n\n 6\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 7 of 28\n\n\n\n A decision is “contrary to” clearly established federal law “if the state court\narrives at a conclusion opposite to that reached by [the Supreme] Court on a\n\nquestion of law or if the state court decides a case differently than [the Supreme]\nCourt has on a set of materially indistinguishable facts.” Williams, 529 U.S. at\n413. A state court decision involves an unreasonable application of federal law “if\n\nthe state court identifies the correct governing legal principle from [the Supreme]\nCourt’s decisions but unreasonably applies that principle to the facts of the\nprisoner’s case.” Id. To justify issuance of the writ under the “unreasonable\n\napplication” clause, the state court’s application of Supreme Court precedent must\nbe more than just wrong in the eyes of the federal court; it “must be ‘objectively\nunreasonable.’” Virginia v. LeBlanc, 137 S. Ct. 1726, 1728 (2017) (quoting\n\nWoods v. Donald, 135 S. Ct. 1372, 1376 (2015)); see also Bell v. Cone, 535 U.S.\n685, 694 (2002) (explaining that “an unreasonable application is different from an\nincorrect one.”).\n Second, Supreme Court precedent demands that in any federal habeas\nproceeding—including collateral proceedings in capital cases—where the\npetitioner seeks the benefit of a “new” rule of constitutional law, we must first\ndetermine whether the rule actually qualifies as new, and then whether that rule\napplies retroactively to the case. See Teague v. Lane, 489 U.S. 288, 300–01 (1989)\n(plurality opinion); Penry v. Lynaugh, 492 U.S. 302, 313–14 (1989) (stating that\nthe retroactivity approach from Teague applies in capital cases), abrogated on\nother grounds by Atkins v. Virginia, 536 U.S. 304, 312–16, 321 (2002). In most\ncases, we cannot disturb a state conviction based on a constitutional rule\n\n 7\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 8 of 28\n\n\n\nannounced after a conviction became final. Teague, 489 U.S. at 310. Only two\nnarrow exceptions pierce this general principle of nonretroactivity: new rules that\n\nare “substantive rather than procedural,” and “watershed rules of criminal\nprocedure implicating the fundamental fairness and accuracy of the criminal\nproceeding.” Schriro v. Summerlin, 542 U.S. 348, 352–53 (2004) (internal\n\nquotation marks and citations omitted). In all other cases the rule applies only\nprospectively.\n What this means in plain English is that, in the vast majority of cases,\n\nprisoners will not be able to secure federal habeas relief based on a new\nconstitutional rule—even when that rule runs in their favor. “This is but a\nrecognition that the purpose of federal habeas corpus is to ensure that state\n\nconvictions comply with the federal law in existence at the time the conviction\nbecame final, and not to provide a mechanism for the continuing reexamination of\nfinal judgments based upon later emerging legal doctrine.” Sawyer v. Smith, 497\nU.S. 227, 234 (1990).\n Though these two constraints—the rule of nonretroactivity set out in Teague\nand the deference to state court decisions mandated by AEDPA—are similar in\nsome respects, they are nonetheless “quite separate” in their operation, and a state\nprisoner seeking federal habeas relief must clear both hurdles to succeed. Greene\nv. Fisher, 565 U.S. 34, 39 (2011). Accordingly—and the Supreme Court has made\nthis clear—“in addition to performing any analysis required by AEDPA, a federal\ncourt considering a habeas petition must conduct a threshold Teague analysis when\n\n\n\n 8\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 9 of 28\n\n\n\nthe issue is properly raised by the state.” Horn v. Banks (Banks I), 536 U.S. 266,\n272 (2002) (per curiam).\n\n B.\n Before conducting that analysis here, we pause to explain why we cannot\nsimply accept the Florida Supreme Court’s decision to apply Hurst retroactively to\n\nKnight and review only its harmless-error analysis, as Knight urges us to do.\nBecause the Florida Supreme Court had already decided to give him the benefit of\nHurst, Knight says, the Teague retroactivity analysis no longer has any bearing in\n\nhis case. He is wrong. While states may fashion their own retroactivity doctrines\nas a matter of state law, those doctrines cannot displace Teague on the federal\nstage. Our ability to consider whether Florida applied Hurst correctly depends\n\nentirely on whether we can apply Hurst ourselves. So far, neither the Supreme\nCourt nor this Circuit has answered that question by analyzing Hurst’s retroactivity\nunder Teague.2\n Florida, on the other hand, has its own retroactivity standard—and is free to\ngive broader retroactive effect to new constitutional rules in state court proceedings\n\n\n2\n We have noted in passing that Hurst would not apply retroactively to a petitioner whose\nconvictions became final long before the Supreme Court decided Ring v. Arizona, 536 U.S. 584\n(2002), on which Hurst relied, and even before Apprendi v. New Jersey, 530 U.S. 466 (2000),\nwhich formed the basis for Ring. See Lambrix v. Sec’y, Fla. Dep’t of Corr., 851 F.3d 1158, 1165\nn.2 (11th Cir. 2017). And we have concluded in the context of an improvidently granted\ncertificate of appealability that the Florida Supreme Court’s decision not to apply Hurst\nretroactively to the same petitioner as a matter of state law was not contrary to or an\nunreasonable application of existing Supreme Court precedents. See Lambrix v. Sec’y, Dep’t of\nCorr., 872 F.3d 1170, 1182 (11th Cir. 2017) (per curiam). But the question of Hurst’s\nretroactivity under Teague to a petitioner like Knight—whose convictions became final after\nRing but before Hurst—was not squarely presented in either case, so we did not conduct that\nanalysis.\n 9\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 10 of 28\n\n\n\nthan Teague allows in federal cases. See Danforth v. Minnesota, 552 U.S. 264,\n282 (2008). That is because the Teague bar “was intended to limit the authority of\n\nfederal courts to overturn state convictions—not to limit a state court’s authority to\ngrant relief for violations of new rules of constitutional law when reviewing its\nown State’s convictions.” Id. at 280–81. So when states choose to apply new rules\n\nof constitutional procedure that are not retroactive under Teague in federal courts,\nthey “do not do so by misconstruing the federal Teague standard. Rather, they\nhave developed state law to govern retroactivity in state postconviction\n\nproceedings.” Id. at 288–89 (emphasis in original).\n In deciding to apply Hurst retroactively to certain state habeas cases, the\nFlorida Supreme Court did just that. See Mosley v. State, 209 So. 3d 1248, 1274–\n\n83 (Fla. 2016) (per curiam). Florida’s retroactivity doctrine is unique to Florida—\nit applies to a limited class of constitutional decisions that announce changes “of\nfundamental significance,” which for procedural rules requires consideration of\nthree factors: “(a) the purpose to be served by the rule, (b) the extent of reliance on\nthe prior rule, and (c) the effect that retroactive application of the new rule would\nhave on the administration of justice.” Asay v. State, 210 So. 3d 1, 16–17 (Fla.\n2016) (per curiam) (citation omitted). For new constitutional rules involving the\ndeath penalty, Florida courts also consider on a case-by-case basis whether\nfundamental fairness requires retroactive application of the rule. Mosley, 209 So.\n3d at 1274 (citing James v. State, 615 So. 2d 668, 668 (Fla. 1993)).3 If that sounds\n\n3\n The Florida Supreme Court is now considering whether it should recede from the retroactivity\nanalysis employed in Asay, Mosley, and James. See Owen v. State, No. SC18-810, April 24,\n\n 10\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 11 of 28\n\n\n\nbroader than Teague, it is for good reason—the Florida Supreme Court has itself\nacknowledged that this retroactivity standard is “more expansive” than the federal\n\nrule. Johnson v. State, 904 So. 2d 400, 409 (Fla. 2005) (per curiam), abrogated on\nunrelated grounds by Asay, 210 So. 3d at 15–16.\n All that to say, Florida may make its own choice about the retroactivity of a\n\ngiven case as a matter of state law. And for Hurst, it has done so. Using its own\nstandard, the Florida Supreme Court decided that Hurst would apply retroactively\nin state collateral review proceedings for petitioners whose convictions had not yet\n\nbecome final when the U.S. Supreme Court decided Ring v. Arizona; Ring held in\n2002 that the Sixth Amendment requires a jury (rather than a judge) to find the\naggravating factors necessary to impose the death penalty. See Mosley, 209 So. 3d\n\nat 1283 (citing Ring v. Arizona, 536 U.S. 584 (2002)). The Florida Supreme Court\nreasoned that because Hurst struck down Florida’s capital sentencing scheme\nbased on Ring, prisoners whose cases were still pending on direct appeal when\nRing was decided “should not suffer due to the United States Supreme Court’s\nfourteen-year delay in applying Ring to Florida.” Id. Following that rule, because\nKnight’s conviction became final ten years after Ring was decided, the Florida\nSupreme Court applied Hurst retroactively in his postconviction proceeding.\nKnight, 225 So. 3d at 682.\n\n\n\n2019 order (directing parties to brief the issue). The uncertain fate of Florida’s current\nretroactivity doctrine offers another reason that we cannot simply rely on a state retroactivity\ndecision as a basis for federal habeas relief. If the state doctrine were to change during our\nreview, would we then be faced with the question of whether to apply the state’s new\nretroactivity doctrine—retroactively?\n 11\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 12 of 28\n\n\n\n But that state-law retroactivity determination has no significance in federal\ncourt. Unlike state courts, lower federal courts are not free to create our own rules\n\nof retroactivity—if the government raises the issue, a Teague analysis is\nmandatory. Banks I, 536 U.S. at 271–72. As we have said, “States may exercise\ntheir collateral review power without regard to the Teague doctrine. Their doing\n\nso has no effect on later federal review.” Glock v. Singletary, 65 F.3d 878, 890–91\n(11th Cir. 1995) (en banc). So we are bound to follow Teague’s retroactivity\nprinciples whether or not the state court chose to apply the new rule in its own\n\ncollateral proceeding. See, e.g., Banks I, 536 U.S. at 271–72.\n Here, then, we must conduct our own retroactivity analysis, using the\nstandards articulated in Teague. And to repeat: Teague retroactivity is a “threshold\n\nquestion in every habeas case.” Caspari v. Bohlen, 510 U.S. 383, 389 (1994).\nWhen “issues of both retroactivity and application of constitutional doctrine are\nraised,” we must decide the retroactivity issue first. Bowen v. United States, 422\nU.S. 916, 920 (1975). Where the State raises the issue, therefore, “federal habeas\ncorpus courts must apply Teague before considering the merits” of the petitioner’s\nclaims. Beard v. Banks (Banks II), 542 U.S. 406, 412 (2004) (emphasis in\noriginal) (internal quotation marks and citation omitted).\n What’s more, if a constitutional claim is Teague-barred, we do not reach its\nmerits. See, e.g., id. at 410 n.2. That is because the Supreme Court’s\n“jurisprudence concerning the ‘retroactivity’ of ‘new rules’ of constitutional law is\nprimarily concerned, not with the question whether a constitutional violation\noccurred, but with the availability or nonavailability of remedies.” Danforth, 552\n\n 12\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 13 of 28\n\n\n\nU.S. at 290–91. If the holding relied on qualifies as a new rule and does not meet\nTeague’s strict requirements for retroactivity, then the claim is not redressable\n\nhere—the “nonretroactivity principle prevents a federal court from granting habeas\ncorpus relief to a state prisoner based on a rule announced after his conviction and\nsentence became final.” Caspari, 510 U.S. at 389 (emphasis in original). And as\n\nwe have said before, if “the court cannot relieve the harm of which a plaintiff\ncomplains, the court should not take the case; in the absence of an effective remedy\nits decision can amount to nothing more than an advisory opinion.” Wymbs v.\n\nRepublican State Exec. Comm., 719 F.2d 1072, 1085 (11th Cir. 1983).\n Our authority to overturn state convictions is limited, and the retroactivity\nprinciples articulated in Teague are tailored to those limitations. See Danforth, 552\n\nU.S. at 277. The fact that state courts do not face the same constraints on collateral\nreview of their own criminal proceedings as we do does not relieve us of the\nobligation to apply federal retroactivity standards. See id. at 280–81; Banks I, 536\nU.S. at 271. In fact, our narrow authority as federal courts to disrupt final state-\ncourt convictions reflects our recognition of the states’ own sovereignty. So\nFlorida may design and apply its retroactivity principles as generously as it wishes.\nBut notwithstanding Florida’s decision to apply Hurst—or any future decision—\nretroactively as a matter of state law, as a federal court we are required to perform\nthe Teague analysis to determine whether prisoners can receive retroactive relief\nunder federal law.\n In sum, if Hurst announced a new rule of constitutional law, but one that\ndoes not fall into one of the exceptions to Teague’s bar on retroactivity, Knight\n\n 13\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 14 of 28\n\n\n\ncannot obtain federal habeas relief for any Hurst error in his sentence—regardless\nof what Florida may choose to do under state law. And if Knight cannot obtain\n\nfederal habeas relief for his Hurst claim in any event, we may not offer an advisory\nopinion on whether the claim could have merit.\n III.\n\n A.\n Turning to Teague, our analysis has three steps. First, we determine the date\nwhen the petitioner’s conviction became final. Banks II, 542 U.S. at 411. This\n\nhappens when the United States Supreme Court “affirms a conviction on the merits\non direct review or denies a petition for a writ of certiorari, or when the time for\nfiling a certiorari petition expires.” Clay v. United States, 537 U.S. 522, 527\n\n(2003). Here, the Court denied Knight’s petition for a writ of certiorari on May 14,\n2012—more than three years before Hurst was decided. Knight, 566 U.S. at 998.\n Second, if the rule that the petitioner wants to apply had not been announced\nby that final-conviction date, we “assay the legal landscape” as it existed at the\ntime and determine whether existing precedent compelled the rule—that is,\nwhether the case announced a new rule or applied an old one. Banks II, 542 U.S.\nat 413. If—and only if—the holding was “dictated by precedent existing at the\ntime the defendant’s conviction became final,” then the rule is not new and may be\napplied retroactively on federal habeas review (indeed, it must be). Caspari, 510\nU.S. at 390 (emphasis in original) (quoting Teague, 489 U.S. at 301). And that is\nnot a light test—a rule is not dictated by prior precedent “unless it would have been\n‘apparent to all reasonable jurists.’” Chaidez v. United States, 568 U.S. 342, 347\n\n 14\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 15 of 28\n\n\n\n(2013) (quoting Lambrix v. Singletary, 520 U.S. 518, 528 (1997)). The fact that a\ndecision is “within the logical compass of” or even “controlled by” prior precedent\n\nis not conclusive in the Teague analysis. Butler v. McKellar, 494 U.S. 407, 415\n(1990) (internal quotation marks and citation omitted). To “‘dictate’ a result, prior\nprecedent must be specific; it is not enough that it name the general principle from\n\nwhich the assertedly new rule sprang.” Glock, 65 F.3d at 884.\n Knight argues that Ring v. Arizona dictated the rule in Hurst. In Ring, the\nSupreme Court held that Arizona’s capital sentencing scheme, which required\n\njudges alone to hear penalty-phase evidence and make factual findings relevant to\nthe imposition of the death penalty, violated the defendant’s Sixth Amendment\nright to a jury trial. 536 U.S. at 588–89. In doing so, the Court explicitly\n\noverruled its precedent upholding Arizona’s death penalty sentencing scheme, “to\nthe extent that it allows a sentencing judge, sitting without a jury, to find an\naggravating circumstance necessary for imposition of the death penalty.” Id. at\n609 (emphasis added) (overruling Walton v. Arizona, 497 U.S. 639, 647–49\n(1990)).\n Knight wants us to find that the new Hurst rule is actually the old Ring rule\nfor an obvious reason—if the rule is not new, and instead was binding on lower\ncourts at the time of Knight’s conviction, then he is entitled to the benefit of the\nrule on federal habeas review. See Yates v. Aiken, 484 U.S. 211, 216–17 & n.3\n(1988). But Ring did not dictate the Supreme Court’s later invalidation of\nFlorida’s death penalty sentencing scheme in Hurst. In fact, the Ring Court\nspecifically acknowledged that Florida’s capital sentencing procedure differed\n\n 15\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 16 of 28\n\n\n\nfrom the Arizona scheme that it rejected. See Ring, 536 U.S. at 607–08 & n.6\n(categorizing state capital sentencing schemes according to jury involvement in\n\nsentencing). In Arizona, the judge alone made the factual findings necessary to\nimpose the death penalty and imposed that penalty entirely apart from the jury.\nSee id. at 588. Florida’s scheme, in contrast, incorporated an advisory jury that\n\nconsidered penalty-phase evidence and recommended a sentence of life or death to\nthe court. Only after the jury’s recommendation did the judge impose a sentence.\nSee id. at 608 n.6; Fla. Stat. Ann. § 921.141 (2001).\n\n Hurst’s conclusion that Florida’s “hybrid” death penalty sentencing system\nviolated the Sixth Amendment was not “apparent to all reasonable jurists” when\nKnight’s convictions became final in 2012. We venture to count ourselves and our\n\ncolleagues on this Court as members of that distinguished group. And in Evans v.\nSecretary, Florida Department of Corrections, we held that Ring did not invalidate\nFlorida’s capital sentencing scheme. 699 F.3d 1249, 1264–65 (11th Cir. 2012).\nWe found it significant that Florida’s statute—unlike the Arizona law at issue in\nRing—required the penalty phase jury to find that sufficient aggravating\ncircumstances existed before it could recommend a death sentence, and directed\nthe sentencing court to give “great weight” to the jury’s advisory verdict. Id. at\n1261 (citation omitted). We also noted that the Supreme Court had taken obvious\npains in Ring to distinguish hybrid systems like Florida’s from the “judge-only”\nsentencing scheme in Arizona and concluded that such distinctions would not have\nbeen necessary if the Court had intended to strike down both systems. Id. at 1262.\n\n\n\n 16\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 17 of 28\n\n\n\n And we were not the only ones. In fact, Justice Alito wrote along those\nsame lines in his dissenting opinion in Hurst: “Although the Court suggests that\n\ntoday’s holding follows ineluctably from Ring, the Arizona sentencing scheme at\nissue in that case was much different from the Florida procedure now before us.”\n136 S. Ct. at 625 (Alito, J., dissenting). After describing the “critically important\n\nrole” of the advisory jury in Florida’s death penalty sentencing scheme, Justice\nAlito concluded that the “decision in Ring did not decide whether this procedure\nviolates the Sixth Amendment, and I would not extend Ring to cover the Florida\n\nsystem.” Id. at 626. Clearly, reasonable jurists could—and did—disagree that\nRing compelled the outcome in Hurst. The Alito dissent and this Court’s pre-\nHurst holding strongly indicate that Hurst announced a new constitutional rule\n\nrather than applying an old one. See Banks II, 542 U.S. at 415.\n That conclusion is only strengthened by the fact that Ring did not\nspecifically address the continued validity of the Supreme Court’s precedents\nupholding Florida’s death-penalty sentencing system—Spaziano v. Florida, 468\nU.S. 447 (1984), and Hildwin v. Florida, 490 U.S. 638 (1989) (per curiam). And\nthe Court has repeatedly instructed us to follow its precedents, even if later\ndecisions appear to undermine them, unless and until the Court itself sets them\naside. See, e.g., Bosse v. Oklahoma, 137 S. Ct. 1, 2 (2016) (per curiam) (“Our\ndecisions remain binding precedent until we see fit to reconsider them, regardless\nof whether subsequent cases have raised doubts about their continuing vitality.”\n(citation omitted)); Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S.\n477, 484 (1989). So Spaziano and Hildwin remained good law until the Court\n\n 17\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 18 of 28\n\n\n\nexplicitly overruled them in Hurst. See Saffle v. Parks, 494 U.S. 484, 488 (1990)\n(explaining that the “explicit overruling of an earlier holding no doubt creates a\n\nnew rule”); see also Hurst, 136 S. Ct. at 623. It is hard to imagine that the\nSupreme Court overruled those cases in Ring but forgot to say so until Hurst.\n Because all these factors show that Hurst was not dictated by prior\n\nprecedent—and in fact explicitly overruled existing precedent upholding Florida’s\ndeath penalty sentencing scheme—we can see that the rule in Hurst, which led to a\nconclusion that the Florida scheme was unconstitutional, was new.\n\n Having determined that Hurst announced a new rule of constitutional law,\nwe proceed to the final step in the Teague analysis—whether Hurst “falls within\neither of the two exceptions to nonretroactivity.” Banks II, 542 U.S. at 411. Those\n\nexceptions, again, include (1) holdings that create substantive (not procedural)\nrules that place “certain kinds of primary, private individual conduct beyond the\npower of the criminal law-making authority to proscribe,” and (2) holdings that\nconstitute “watershed rules of criminal procedure.” Teague, 489 U.S. at 311–13\n(citation omitted).\n The Hurst rule does not fit within either exception. To begin, substantive\nrules include decisions that change “the range of conduct or the class of persons\nthat the law punishes.” Schriro, 542 U.S. at 353. Procedural rules, on the other\nhand, “regulate only the manner of determining the defendant’s culpability.” Id.\n(emphasis in original). In considering which category the Hurst rule falls into, we\nhave a head start because the Supreme Court has already held that Ring represented\na “prototypical procedural rule[].” Id. And that makes sense: Ring changed the\n\n 18\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 19 of 28\n\n\n\npermissible procedure for sentencing in a capital case when it required “that a jury\nrather than a judge find the essential facts” necessary to impose the death penalty.\n\nSee id. Because Hurst’s holding—that an advisory “jury’s mere recommendation\nis not enough” to satisfy this procedural requirement—is an extension of the rule\nfrom Ring, we have no trouble concluding that Hurst also announced a procedural\n\nrule, and not a substantive one. Hurst, 136 S. Ct. at 619.\n The second exception is for “watershed rules of criminal procedure.” Banks\nII, 542 U.S. at 417. This exception is extremely limited in scope—it applies “only\n\nto a small core of rules” so fundamental to our criminal process that it is “unlikely\nthat many such components of basic due process have yet to emerge.” Id.\n(citations omitted). Indeed, the watershed exception remains somewhat theoretical\n\nat this point; in the years following Teague, the Supreme Court has never found a\nrule that fits. See id. And in “providing guidance as to what might fall within this\nexception,” the Court has “repeatedly referred to the rule of Gideon v. Wainwright,\n372 U. S. 335 (1963) (right to counsel), and only to this rule.” Id. Knight does not\ncontend that Hurst announced a new watershed rule that compares to Gideon, and\nwe do not see how it could have either. In short, Hurst meets neither exception,\nand therefore is not retroactive.\n B.\n Because the Hurst rule is not retroactive, Knight cannot receive federal\nhabeas relief on his Hurst claim. That is as it must be—we are conscientious about\nthe fact that “Teague’s nonretroactivity principle acts as a limitation on the power\nof federal courts to grant habeas corpus relief.” Id. at 412 (internal quotation\n\n 19\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 20 of 28\n\n\n\nmarks and citation omitted). And because we would have no lawful remedy to\noffer even if we could identify an error, we must decline to consider whether any\n\nHurst error exists. That means, of course, that we also do not consider whether the\nun-found and un-remediable error could be harmless.\n Knight disputes this path. He argues that, even conceding the lack of an\n\navailable remedy for any Hurst error, whether that un-remediable error was\nharmless is itself a separate question of federal law under Brecht v. Abrahamson,\n507 U.S. 619, 637–38 (1993). He thus urges us to review the state court’s\n\nharmless-error analysis, regardless of whether we reach the merits of his Hurst\nclaim. We will not. It strains the imagination—as well as our constitutional and\ninstitutional respect for state courts—to suppose that we cannot remedy an error,\n\nbut that we can somehow remedy an erroneous state-court conclusion that the error\nwas harmless.\n And Brecht does not stand for the proposition that Knight asserts in any\nevent. While Brecht established the harmless-error standard for collateral review\nof constitutional trial errors, it did not create a stand-alone federal claim severable\nfrom the question of whether remediable error existed in the first place.4 See\nWilliams v. Singletary, 114 F.3d 177, 180 (11th Cir. 1997) (per curiam), as\namended on denial of reh’g (July 29, 1997). Some underlying violation of federal\nlaw that we can address is a necessary predicate to federal habeas relief—unless\n\n4\n As an aside, what Brecht really decided was that federal courts evaluating constitutional trial\nerror on collateral review would apply a more relaxed harmless-error standard—whether the\nerror “had substantial and injurious effect or influence in determining the jury’s verdict”—rather\nthan the harmless-beyond-a-reasonable-doubt standard used on direct review. Brecht, 507 U.S.\nat 638 (quoting Kotteakos v. United States, 328 U.S. 750, 776 (1946)).\n 20\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 21 of 28\n\n\n\nwe agree that an error has occurred, it makes no difference whether the purported\nerror was harmless. “We have consistently applied the Brecht harmless error\n\nstandard only after determining that there was an error.” Id.; cf. Wilson v.\nCorcoran, 562 U.S. 1, 6 (2010) (per curiam) (“It is not enough to note that a\nhabeas petitioner asserts the existence of a constitutional violation; unless the\n\nfederal court agrees with that assertion, it may not grant relief.” (emphasis in\noriginal)). There is no free-floating federal constitutional right to infallible\napplication of harmless-error principles.\n\n And again, where our retroactivity doctrine forecloses the possibility of\nfederal habeas relief for a constitutional error, we are constrained to stop looking.\nSee Bowen, 422 U.S. at 920. The Supreme Court “consistently has declined to\n\naddress unsettled questions regarding the scope of decisions establishing new\nconstitutional doctrine in cases in which it holds those decisions nonretroactive,”\nand has instructed us to do the same. Id. at 920–21. This directive carries\nconstitutional weight; as an Article III court, we are “without power to decide\nquestions that cannot affect the rights of litigants in the case” at hand. North\nCarolina v. Rice, 404 U.S. 244, 246 (1971) (per curiam). So even if we found\nHurst error in Knight’s sentencing, we would still be prohibited from issuing a writ\nof habeas corpus on that ground because Hurst is not retroactively applicable to\nKnight under Teague. After all, Teague’s nonretroactivity command is a limitation\non our power, not a polite suggestion. Banks II, 542 U.S. at 412. So our opinion—\nwhatever it might be—on Knight’s Hurst claim would be purely advisory. “And it\nis quite clear that the oldest and most consistent thread in the federal law of\n\n 21\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 22 of 28\n\n\n\njusticiability is that the federal courts will not give advisory opinions.” Flast v.\nCohen, 392 U.S. 83, 96 (1968) (internal quotation marks and citation omitted).\n\n Where, as here, Teague bars relief before we reach the preliminary question\nof whether constitutional error occurred at all, consideration of the secondary\nquestion of whether any such error was harmless would be a prohibited and\n\npointless exercise for both the petitioner and this Court. We therefore cannot grant\nKnight relief on his Hurst claim, whether or not it is cloaked in the garb of\nharmless error.\n\n IV.\n We now turn to Knight’s other claim. The State presented extensive DNA\nevidence against him during the guilt phase of his trial. Kevin Noppinger, a\n\nserologist with the Broward County Sheriff’s Office crime lab, testified that Knight\nhad Odessia’s blood on his hand and her DNA on his shirt when he was arrested.\nFingernail scrapings from Odessia’s body showed that she, in turn, had Knight’s\nDNA under her fingernails. Noppinger also tested samples from the bloody\nclothes (boxer shorts, a shirt, and jean shorts) found under the bathroom sink. He\nfound Knight’s blood mixed with Hanessia’s blood on the boxer shorts, Odessia’s\nand Hanessia’s blood elsewhere on the boxer shorts and on the jean shorts, and\nOdessia’s blood on the shirt.\n One of the State’s other experts was Kevin McElfresh of Bode Technology\nGroup, whose DNA analysts had conducted additional testing on different samples\nfrom the same items of clothing. In particular, McElfresh’s group analyzed DNA\nsamples from an unstained area of the waistband of the boxer shorts in an attempt\n\n 22\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 23 of 28\n\n\n\nto determine who owned them. Although Bode’s initial report stated that Knight’s\nDNA was not on the waistband sample, McElfresh testified at trial that he had\n\nconducted some additional analysis and determined that some of the DNA on the\nwaistband could have been Knight’s.\n Knight’s guilt-phase counsel consulted with DNA expert Dr. Norah Rudin,\n\nwho was listed as a potential trial witness by the defense. Dr. Rudin informed\ncounsel that, although some of Noppinger’s sample labeling practices were sloppy,\nshe generally agreed with his conclusions about the sources of the DNA samples\n\nhe analyzed. She was much more critical of McElfresh’s analysis: she called his\nmethods “fundamentally incorrect and inherently biased” and his testimony\n“incomplete and misleading.” In her opinion, the DNA test results for the\n\nwaistband samples were “inconclusive.” But Dr. Rudin did not stop there—she\nalso considered McElfresh’s testimony “relatively inconsequential when viewed in\nthe context of the biological evidence as a whole.” She told Knight’s counsel that\nshe did not believe that she could help his case, and that she would not call herself\nas an expert if she were in his shoes. Counsel, it appears, agreed with her\nperspective and did not call her at trial.\n Testimony indicates that one additional fact persuaded counsel that calling\nDr. Rubin to quibble with McElfresh’s methods would not be good trial strategy.\nAt the time of Knight’s trial, Florida permitted a defendant who did not put up any\nevidence at trial to give both initial and rebuttal closing arguments—an advantage\nordinarily offered to the prosecution. Dr. Rudin would have been Knight’s only\nwitness, so calling her to testify would have meant giving up the opportunity to\n\n 23\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 24 of 28\n\n\n\nhave the first word and the last at closing. Given that fact, she would have largely\ncorroborated Noppinger’s testimony while also costing Knight an advantage at\n\nclosing arguments.\n Nonetheless, on state collateral review, Knight contended that his counsel\nprovided constitutionally ineffective assistance by failing to call Dr. Rudin as an\n\nexpert. He argued that Dr. Rudin’s criticisms of McElfresh’s methods and the\ncrime lab’s labeling practices would have cast doubt on all of the State’s DNA\nevidence and significantly damaged the State’s case. The Florida Supreme Court\n\ndisagreed, finding that Knight had not met his burden on either prong of the two-\npart test for ineffective assistance of counsel claims set out in Strickland. See\nKnight, 225 So. 3d at 672–74. Knight, however, argues that the state court’s\n\ndecision “was contrary to, or involved an unreasonable application of, clearly\nestablished Federal law, as determined by the Supreme Court of the United States.”\n28 U.S.C. § 2254(d)(1).\n To begin, the Florida Supreme Court correctly identified the governing\nstandard. Strickland is the relevant “clearly established” Supreme Court precedent\nfor purposes of an ineffective-assistance claim. See Premo v. Moore, 562 U.S.\n115, 118 (2011). Under Strickland, to succeed on a claim of ineffective assistance\nof counsel, a defendant must show both that his attorney’s performance was\ndeficient—that is, “that counsel’s representation fell below an objective standard of\nreasonableness”—and that he was prejudiced by the inadequate performance.\nStrickland, 466 U.S. at 687–88. In applying the first prong, courts “must indulge a\nstrong presumption that counsel’s conduct falls within the wide range of\n\n 24\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 25 of 28\n\n\n\nreasonable professional assistance.” Id. at 689. And to show prejudice, the\n“defendant must show that there is a reasonable probability that, but for counsel’s\n\nunprofessional errors, the result of the proceeding would have been different. A\nreasonable probability is a probability sufficient to undermine confidence in the\noutcome.” Id. at 694.\n\n The Strickland standard is “highly deferential,” as is the review of a state-\ncourt decision under AEDPA; “when the two apply in tandem, review is ‘doubly’\nso.” Harrington v. Richter, 562 U.S. 86, 105 (2011) (citations omitted). In\n\nreviewing a state court’s application of Strickland, therefore, a federal habeas court\ncannot conduct a de novo review and reverse simply because it strongly disagrees\nwith the state court’s conclusion. Id. at 102. Instead, for Knight to succeed on his\n\nineffective-assistance claim, we must conclude that the Florida Supreme Court’s\ndecision “was an objectively unreasonable application of the Strickland standard.”\nAllen v. Sec’y, Fla. Dep’t of Corr., 611 F.3d 740, 751 (11th Cir. 2010).\n The Florida Supreme Court determined that counsel’s decision not to call\nDr. Rudin at trial was reasonable trial strategy. Knight, 225 So. 3d at 674.\nWhether the decision was actually a matter of strategy is a question of fact; thus,\nthe state court’s finding on that issue is presumed to be correct. Provenzano v.\nSingletary, 148 F.3d 1327, 1330 (11th Cir. 1998). We must accept all factual\nfindings made by the state court unless the petitioner rebuts the presumption of\ncorrectness “by clear and convincing evidence.” 28 U.S.C. § 2254(e)(1); see also\nWood v. Allen, 542 F.3d 1281, 1285 (11th Cir. 2008). Knight has not presented\n\n\n\n 25\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 26 of 28\n\n\n\nany evidence to suggest that counsel’s decision was anything other than a matter of\nstrategy, and we accept the state court’s finding on that point.\n\n Whether counsel’s strategic decision not to call Dr. Rudin was reasonable is\na question of law, which we review through the lens of AEDPA deference. See\nFerrell v. Hall, 640 F.3d 1199, 1223–24 (11th Cir. 2011). In assessing an\n\nattorney’s performance under Strickland, “strategic choices made after thorough\ninvestigation of law and facts relevant to plausible options are virtually\nunchallengeable.” Strickland, 466 U.S. at 690. “Which witnesses, if any, to call,\n\nand when to call them, is the epitome of a strategic decision, and it is one that we\nwill seldom, if ever, second guess.” Waters v. Thomas, 46 F.3d 1506, 1512 (11th\nCir. 1995) (en banc). We have no reason to doubt the Florida Supreme Court’s\n\nconclusion that counsel’s decision not to call an equivocal expert, in part to\npreserve an advantage at closing, was well within the wide range of reasonably\ncompetent performance.\n The Florida Supreme Court also determined that Knight had not met his\nburden of showing prejudice under Strickland because there was no reasonable\nprobability that Dr. Rudin’s testimony would have made a difference in the\noutcome of the trial, given the weight of the evidence against him. See Knight, 225\nSo. 3d at 674. That conclusion was also reasonable.\n Dr. Rudin generally agreed with Noppinger’s conclusions regarding the\nDNA evidence. With or without Dr. Rudin’s testimony, therefore, there was no\ndispute that Knight had Odessia’s blood on his hand and her DNA on his shirt\nwhen he was arrested, or that Knight’s DNA was found under Odessia’s\n\n 26\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 27 of 28\n\n\n\nfingernails. Knight’s blood—as well as Odessia’s blood and Hanessia’s—was on\nthe clothes discarded at the crime scene. And even if the jury had heard Dr.\n\nRudin’s criticism of McElfresh’s testimony that Knight could not be ruled out as a\nDNA contributor for the waistband of the boxer shorts, other evidence showed that\nthe clothes were his: the boxer shorts were the same brand and size as the ones that\n\nKnight was wearing when he was arrested, and Knight’s cousin testified that the\nbloody shirt was one that Knight wore often.\n And the evidence did not stop there. Knight was known to be home with\n\nHanessia less than an hour before the murders, and the jury heard evidence of\nongoing tension with Odessia. Police found him near the scene of the murders\nshortly after they arrived, and his answers to their questions were inconsistent with\n\nhis appearance. All of that is in addition to the testimony of an inmate housed with\nKnight at the Broward County jail, who testified that Knight confessed the crime to\nhim, providing a detailed description of events and a floorplan of the apartment. In\nshort, even if the jury had entirely discounted McElfresh’s testimony as a result of\nDr. Rudin’s criticism of his testing methods, the remaining evidence against\nKnight was so strong that the chance of a not-guilty verdict still would have been\nremote, to say the least.\n The Florida Supreme Court’s conclusion that Knight failed to make the\nrequired showings of deficient performance and prejudice was not an unreasonable\napplication of Strickland.\n\n\n\n\n 27\n\f Case: 18-13390 Date Filed: 08/30/2019 Page: 28 of 28\n\n\n\n ***\n In sum, we are prohibited from considering Knight’s Hurst claim by\n\nTeague’s rule of nonretroactivity, which eliminates any possibility of relief\nregardless of whether there was an error and regardless of whether any error was\nharmless. And our review of Knight’s ineffective-assistance claim shows that the\n\nstate court’s ruling on that issue was not an unreasonable application of governing\nSupreme Court precedent. We therefore affirm the district court’s denial of\nKnight’s federal habeas petition.\n\n AFFIRMED.\n\n\n\n\n 28\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Richard Knight v. Florida Department of Corrections\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-30\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 18-13390 Date Filed: 08/30/2019 Page: 1 of 28\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 18-13390\\n ________________________\\n\\n D.C. Docket No. 0:17-cv-61921-RNS\\n\\n\\n\\nRICHARD KNIGHT,\\n\\n Petitioner-Appellant,\\n\\n versus\\n\\nFLORIDA DEPARTMENT OF CORRECTIONS,\\n\\n Respondent-Appellee.\\n ________________________\\n\\n Appeal from the United States District Court\\n for the Southern District of Florida\\n ________________________\\n\\n (August 30, 2019)\\n\\nBefore TJOFLAT, JORDAN, and GRANT, Circuit Judges.\\nGRANT, Circuit Judge:\\n Richard Knight, a Florida prisoner sentenced to death for the murders of\\n\\nOdessia Stephens and her daughter, Hanessia Mullings, appeals the district court’s\\ndenial of his federal habeas corpus petition. At this stage—almost 20 years after\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 2 of 28\\n\\n\\n\\nthe crimes were committed and more than a decade after a Florida jury found\\nKnight guilty of the murders and recommended a death sentence—Knight’s claims\\n\\nhave been winnowed down to two: first, that his death sentence is invalid under\\nHurst v. Florida, 136 S. Ct. 616 (2016), and second, that he received ineffective\\nassistance of counsel at trial. Because Hurst does not apply retroactively to\\n\\nKnight, any challenge to his death sentence on that basis is beyond our reach on\\nfederal habeas review. Nor can Knight find success in his other challenge; the\\nFlorida Supreme Court’s rejection of his ineffective-assistance claim was not an\\n\\nunreasonable application of Strickland v. Washington, 466 U.S. 668 (1984). We\\ntherefore affirm.\\n I.\\n\\n A.\\n According to evidence introduced at his murder trial, Knight lived in an\\napartment with his cousin, Hans Mullings, and Hans’s girlfriend, Odessia. Hans\\nand Odessia’s four-year-old daughter, Hanessia, also lived with them in the\\napartment. Odessia was tired of supporting Knight and one evening while Hans\\nwas out she argued with Knight, insisting that he move out the next day. After the\\nargument got heated, Knight left the house to walk around. But as he later\\nconfessed to another inmate, instead of getting less angry with Odessia once he got\\nsome air, Knight became increasingly irate. He returned to the apartment and after\\nexchanging more words with Odessia, he got a knife from the kitchen. When he\\nwent back to the master bedroom, he found Odessia and her little girl in the bed.\\nHe began stabbing Odessia and continued his attack until she stopped resisting and\\n\\n 2\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 3 of 28\\n\\n\\n\\ncurled up on the bedroom floor. He then moved on to little Hanessia, stabbing her\\nuntil his knife broke and cutting his hand in the process. As he was leaving the\\n\\nbedroom, he heard “popping noises” from where Hanessia lay on the floor, and he\\nthought that the little girl was “drowning in her own blood.” Apparently not\\nconsidering his revenge complete, he retrieved a second knife from the kitchen and\\n\\nreturned to continue his attack on Odessia. In the meantime, Odessia had crawled\\nfrom the bedroom to the living room, where she had collapsed. Knight turned her\\nover, saw that she was still alive, and started stabbing her again.\\n\\n Both Odessia and Hanessia died that night. In total, Odessia had 21 stab\\nwounds, including 14 in the neck, 24 puncture or scratch wounds, bruising and\\nligature marks consistent with having been hit and strangled with a belt, defensive\\n\\nwounds, and bruises from being hit or punched in the mouth and head. Little\\nHanessia had four stab wounds in her upper body and neck, a deep defensive\\nwound on her hand, bruises on her neck consistent with manual strangulation, and\\nbruises on her arms consistent with having been grabbed.\\n Knight showered and changed after completing his brutal acts, then headed\\nto the living room with a rag to wipe off the knives. Interrupted by a knock on the\\nfront door—it was police responding to a neighbor’s 911 call—Knight ran to his\\nroom and climbed out the window.\\n Shortly after they arrived, police encountered Knight near the apartment. He\\ntold them that he lived there, but that he did not have a key. This was odd; the\\nofficers had already found that all the doors to the apartment were locked. Knight\\nwas also visibly wet—but it was not raining. Knight explained to police that he\\n\\n 3\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 4 of 28\\n\\n\\n\\nhad been jogging, a remarkable contention from a person who was wearing long\\npants and dress shoes. He did not appear to be sweating, in any event. And\\n\\nKnight’s personal appearance subsequently revealed even more clues—he had\\nblood on the back of his shirt, scratches on his chest and midsection, a scrape on\\nhis shoulder, and fresh cuts on his hand.\\n\\n Knight was arrested and indicted for two counts of first-degree murder. A\\nFlorida jury found him guilty as charged. That same jury heard evidence and\\nargument at the penalty phase and unanimously recommended two death\\n\\nsentences—one for each murder. Consistent with Florida’s then-current death\\npenalty sentencing procedure, the judge held an additional hearing, made his own\\nfindings regarding aggravating and mitigating circumstances, and sentenced\\n\\nKnight to death. The Florida Supreme Court affirmed Knight’s convictions and\\nsentences on direct appeal. Knight v. State, 76 So. 3d 879, 890 (Fla. 2011). The\\nUnited States Supreme Court denied his petition for certiorari. Knight v. Florida,\\n566 U.S. 998, 998 (2012).\\n B.\\n Knight filed motions for state collateral relief raising the two claims at issue\\nhere, as well as others that have already been resolved. Specifically, he argued that\\nthe state court should vacate his death sentence in light of Hurst v. Florida, in\\nwhich the Supreme Court held—four years after Knight’s conviction was final—\\nthat Florida’s death penalty sentencing scheme violated the Sixth Amendment.\\n136 S. Ct. at 622. The problem identified by the Supreme Court in Hurst, and\\nargued by Knight in his post-conviction pleadings, was that the jury’s role in\\n\\n 4\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 5 of 28\\n\\n\\n\\nsentencing was to make a non-binding recommendation; the judge alone made the\\nultimate findings of fact necessary to impose the death penalty. Id. at 619, 621–22.\\n\\nKnight also argued that his guilt-phase counsel was constitutionally ineffective for\\nfailing to call an available DNA expert.\\n The Florida Supreme Court rejected his postconviction claims on the merits.\\n\\nKnight v. State, 225 So. 3d 661, 668 (Fla. 2017) (per curiam). A plurality of the\\ncourt agreed with Knight that the sentencing procedure used in his case violated\\nthe Sixth Amendment under Hurst, but also concluded that the Hurst error was\\n\\nharmless. Id. at 682. The plurality explained that under the facts of Knight’s case\\nthe penalty-phase jury had necessarily made the factual findings necessary to\\nimpose the death penalty—that “sufficient aggravators existed” and that “the\\n\\naggravation outweighed the mitigation”—when it returned a unanimous vote\\nrecommending death.1 Id. at 682–83 (citation omitted). As for his ineffective-\\nassistance claim, the court held that Knight had failed to meet his burden under\\nStrickland because he had not shown that his attorney’s decision not to call his\\nDNA expert constituted deficient performance, or that there was any reasonable\\nprobability that that decision negatively affected the outcome of his trial. Id. at\\n673–74.\\n\\n\\n\\n\\n1\\n Three out of seven justices joined the opinion on Knight’s Hurst claim. Two additional justices\\nconcurred in the result only. Knight, 225 So. 3d at 684.\\n 5\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 6 of 28\\n\\n\\n\\n C.\\n Knight filed a petition for federal habeas review in the Southern District of\\n\\nFlorida, pursuant to 28 U.S.C. § 2254. The district court denied relief but granted\\na certificate of appealability on the two claims now before us.\\n II.\\n\\n A.\\n Federal courts may grant habeas corpus relief to prisoners who are being\\ndetained “in violation of the Constitution or laws or treaties of the United States.”\\n\\n28 U.S.C. §§ 2241(c)(3); 2254(a). But our authority to award this kind of relief to\\nstate prisoners is limited—by both statute and Supreme Court precedent.\\n First, the Antiterrorism and Effective Death Penalty Act (AEDPA) limits\\n\\nour authority to award habeas relief. A federal court may not grant a state\\nprisoner’s habeas petition on any issue that was decided on the merits by the state\\ncourt unless the state court’s ruling “(1) resulted in a decision that was contrary to,\\nor involved an unreasonable application of, clearly established Federal law, as\\ndetermined by the Supreme Court of the United States; or (2) resulted in a decision\\nthat was based on an unreasonable determination of the facts in light of the\\nevidence presented in the State court proceeding.” 28 U.S.C. § 2254(d). And as\\nthe Supreme Court has explained, “clearly established” federal law means “the\\nholdings, as opposed to the dicta” from its controlling precedents at the time of the\\nrelevant state court decision. Lockyer v. Andrade, 538 U.S. 63, 71 (2003) (quoting\\nWilliams v. Taylor, 529 U.S. 362, 412 (2000)).\\n\\n\\n\\n 6\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 7 of 28\\n\\n\\n\\n A decision is “contrary to” clearly established federal law “if the state court\\narrives at a conclusion opposite to that reached by [the Supreme] Court on a\\n\\nquestion of law or if the state court decides a case differently than [the Supreme]\\nCourt has on a set of materially indistinguishable facts.” Williams, 529 U.S. at\\n413. A state court decision involves an unreasonable application of federal law “if\\n\\nthe state court identifies the correct governing legal principle from [the Supreme]\\nCourt’s decisions but unreasonably applies that principle to the facts of the\\nprisoner’s case.” Id. To justify issuance of the writ under the “unreasonable\\n\\napplication” clause, the state court’s application of Supreme Court precedent must\\nbe more than just wrong in the eyes of the federal court; it “must be ‘objectively\\nunreasonable.’” Virginia v. LeBlanc, 137 S. Ct. 1726, 1728 (2017) (quoting\\n\\nWoods v. Donald, 135 S. Ct. 1372, 1376 (2015)); see also Bell v. Cone, 535 U.S.\\n685, 694 (2002) (explaining that “an unreasonable application is different from an\\nincorrect one.”).\\n Second, Supreme Court precedent demands that in any federal habeas\\nproceeding—including collateral proceedings in capital cases—where the\\npetitioner seeks the benefit of a “new” rule of constitutional law, we must first\\ndetermine whether the rule actually qualifies as new, and then whether that rule\\napplies retroactively to the case. See Teague v. Lane, 489 U.S. 288, 300–01 (1989)\\n(plurality opinion); Penry v. Lynaugh, 492 U.S. 302, 313–14 (1989) (stating that\\nthe retroactivity approach from Teague applies in capital cases), abrogated on\\nother grounds by Atkins v. Virginia, 536 U.S. 304, 312–16, 321 (2002). In most\\ncases, we cannot disturb a state conviction based on a constitutional rule\\n\\n 7\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 8 of 28\\n\\n\\n\\nannounced after a conviction became final. Teague, 489 U.S. at 310. Only two\\nnarrow exceptions pierce this general principle of nonretroactivity: new rules that\\n\\nare “substantive rather than procedural,” and “watershed rules of criminal\\nprocedure implicating the fundamental fairness and accuracy of the criminal\\nproceeding.” Schriro v. Summerlin, 542 U.S. 348, 352–53 (2004) (internal\\n\\nquotation marks and citations omitted). In all other cases the rule applies only\\nprospectively.\\n What this means in plain English is that, in the vast majority of cases,\\n\\nprisoners will not be able to secure federal habeas relief based on a new\\nconstitutional rule—even when that rule runs in their favor. “This is but a\\nrecognition that the purpose of federal habeas corpus is to ensure that state\\n\\nconvictions comply with the federal law in existence at the time the conviction\\nbecame final, and not to provide a mechanism for the continuing reexamination of\\nfinal judgments based upon later emerging legal doctrine.” Sawyer v. Smith, 497\\nU.S. 227, 234 (1990).\\n Though these two constraints—the rule of nonretroactivity set out in Teague\\nand the deference to state court decisions mandated by AEDPA—are similar in\\nsome respects, they are nonetheless “quite separate” in their operation, and a state\\nprisoner seeking federal habeas relief must clear both hurdles to succeed. Greene\\nv. Fisher, 565 U.S. 34, 39 (2011). Accordingly—and the Supreme Court has made\\nthis clear—“in addition to performing any analysis required by AEDPA, a federal\\ncourt considering a habeas petition must conduct a threshold Teague analysis when\\n\\n\\n\\n 8\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 9 of 28\\n\\n\\n\\nthe issue is properly raised by the state.” Horn v. Banks (Banks I), 536 U.S. 266,\\n272 (2002) (per curiam).\\n\\n B.\\n Before conducting that analysis here, we pause to explain why we cannot\\nsimply accept the Florida Supreme Court’s decision to apply Hurst retroactively to\\n\\nKnight and review only its harmless-error analysis, as Knight urges us to do.\\nBecause the Florida Supreme Court had already decided to give him the benefit of\\nHurst, Knight says, the Teague retroactivity analysis no longer has any bearing in\\n\\nhis case. He is wrong. While states may fashion their own retroactivity doctrines\\nas a matter of state law, those doctrines cannot displace Teague on the federal\\nstage. Our ability to consider whether Florida applied Hurst correctly depends\\n\\nentirely on whether we can apply Hurst ourselves. So far, neither the Supreme\\nCourt nor this Circuit has answered that question by analyzing Hurst’s retroactivity\\nunder Teague.2\\n Florida, on the other hand, has its own retroactivity standard—and is free to\\ngive broader retroactive effect to new constitutional rules in state court proceedings\\n\\n\\n2\\n We have noted in passing that Hurst would not apply retroactively to a petitioner whose\\nconvictions became final long before the Supreme Court decided Ring v. Arizona, 536 U.S. 584\\n(2002), on which Hurst relied, and even before Apprendi v. New Jersey, 530 U.S. 466 (2000),\\nwhich formed the basis for Ring. See Lambrix v. Sec’y, Fla. Dep’t of Corr., 851 F.3d 1158, 1165\\nn.2 (11th Cir. 2017). And we have concluded in the context of an improvidently granted\\ncertificate of appealability that the Florida Supreme Court’s decision not to apply Hurst\\nretroactively to the same petitioner as a matter of state law was not contrary to or an\\nunreasonable application of existing Supreme Court precedents. See Lambrix v. Sec’y, Dep’t of\\nCorr., 872 F.3d 1170, 1182 (11th Cir. 2017) (per curiam). But the question of Hurst’s\\nretroactivity under Teague to a petitioner like Knight—whose convictions became final after\\nRing but before Hurst—was not squarely presented in either case, so we did not conduct that\\nanalysis.\\n 9\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 10 of 28\\n\\n\\n\\nthan Teague allows in federal cases. See Danforth v. Minnesota, 552 U.S. 264,\\n282 (2008). That is because the Teague bar “was intended to limit the authority of\\n\\nfederal courts to overturn state convictions—not to limit a state court’s authority to\\ngrant relief for violations of new rules of constitutional law when reviewing its\\nown State’s convictions.” Id. at 280–81. So when states choose to apply new rules\\n\\nof constitutional procedure that are not retroactive under Teague in federal courts,\\nthey “do not do so by misconstruing the federal Teague standard. Rather, they\\nhave developed state law to govern retroactivity in state postconviction\\n\\nproceedings.” Id. at 288–89 (emphasis in original).\\n In deciding to apply Hurst retroactively to certain state habeas cases, the\\nFlorida Supreme Court did just that. See Mosley v. State, 209 So. 3d 1248, 1274–\\n\\n83 (Fla. 2016) (per curiam). Florida’s retroactivity doctrine is unique to Florida—\\nit applies to a limited class of constitutional decisions that announce changes “of\\nfundamental significance,” which for procedural rules requires consideration of\\nthree factors: “(a) the purpose to be served by the rule, (b) the extent of reliance on\\nthe prior rule, and (c) the effect that retroactive application of the new rule would\\nhave on the administration of justice.” Asay v. State, 210 So. 3d 1, 16–17 (Fla.\\n2016) (per curiam) (citation omitted). For new constitutional rules involving the\\ndeath penalty, Florida courts also consider on a case-by-case basis whether\\nfundamental fairness requires retroactive application of the rule. Mosley, 209 So.\\n3d at 1274 (citing James v. State, 615 So. 2d 668, 668 (Fla. 1993)).3 If that sounds\\n\\n3\\n The Florida Supreme Court is now considering whether it should recede from the retroactivity\\nanalysis employed in Asay, Mosley, and James. See Owen v. State, No. SC18-810, April 24,\\n\\n 10\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 11 of 28\\n\\n\\n\\nbroader than Teague, it is for good reason—the Florida Supreme Court has itself\\nacknowledged that this retroactivity standard is “more expansive” than the federal\\n\\nrule. Johnson v. State, 904 So. 2d 400, 409 (Fla. 2005) (per curiam), abrogated on\\nunrelated grounds by Asay, 210 So. 3d at 15–16.\\n All that to say, Florida may make its own choice about the retroactivity of a\\n\\ngiven case as a matter of state law. And for Hurst, it has done so. Using its own\\nstandard, the Florida Supreme Court decided that Hurst would apply retroactively\\nin state collateral review proceedings for petitioners whose convictions had not yet\\n\\nbecome final when the U.S. Supreme Court decided Ring v. Arizona; Ring held in\\n2002 that the Sixth Amendment requires a jury (rather than a judge) to find the\\naggravating factors necessary to impose the death penalty. See Mosley, 209 So. 3d\\n\\nat 1283 (citing Ring v. Arizona, 536 U.S. 584 (2002)). The Florida Supreme Court\\nreasoned that because Hurst struck down Florida’s capital sentencing scheme\\nbased on Ring, prisoners whose cases were still pending on direct appeal when\\nRing was decided “should not suffer due to the United States Supreme Court’s\\nfourteen-year delay in applying Ring to Florida.” Id. Following that rule, because\\nKnight’s conviction became final ten years after Ring was decided, the Florida\\nSupreme Court applied Hurst retroactively in his postconviction proceeding.\\nKnight, 225 So. 3d at 682.\\n\\n\\n\\n2019 order (directing parties to brief the issue). The uncertain fate of Florida’s current\\nretroactivity doctrine offers another reason that we cannot simply rely on a state retroactivity\\ndecision as a basis for federal habeas relief. If the state doctrine were to change during our\\nreview, would we then be faced with the question of whether to apply the state’s new\\nretroactivity doctrine—retroactively?\\n 11\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 12 of 28\\n\\n\\n\\n But that state-law retroactivity determination has no significance in federal\\ncourt. Unlike state courts, lower federal courts are not free to create our own rules\\n\\nof retroactivity—if the government raises the issue, a Teague analysis is\\nmandatory. Banks I, 536 U.S. at 271–72. As we have said, “States may exercise\\ntheir collateral review power without regard to the Teague doctrine. Their doing\\n\\nso has no effect on later federal review.” Glock v. Singletary, 65 F.3d 878, 890–91\\n(11th Cir. 1995) (en banc). So we are bound to follow Teague’s retroactivity\\nprinciples whether or not the state court chose to apply the new rule in its own\\n\\ncollateral proceeding. See, e.g., Banks I, 536 U.S. at 271–72.\\n Here, then, we must conduct our own retroactivity analysis, using the\\nstandards articulated in Teague. And to repeat: Teague retroactivity is a “threshold\\n\\nquestion in every habeas case.” Caspari v. Bohlen, 510 U.S. 383, 389 (1994).\\nWhen “issues of both retroactivity and application of constitutional doctrine are\\nraised,” we must decide the retroactivity issue first. Bowen v. United States, 422\\nU.S. 916, 920 (1975). Where the State raises the issue, therefore, “federal habeas\\ncorpus courts must apply Teague before considering the merits” of the petitioner’s\\nclaims. Beard v. Banks (Banks II), 542 U.S. 406, 412 (2004) (emphasis in\\noriginal) (internal quotation marks and citation omitted).\\n What’s more, if a constitutional claim is Teague-barred, we do not reach its\\nmerits. See, e.g., id. at 410 n.2. That is because the Supreme Court’s\\n“jurisprudence concerning the ‘retroactivity’ of ‘new rules’ of constitutional law is\\nprimarily concerned, not with the question whether a constitutional violation\\noccurred, but with the availability or nonavailability of remedies.” Danforth, 552\\n\\n 12\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 13 of 28\\n\\n\\n\\nU.S. at 290–91. If the holding relied on qualifies as a new rule and does not meet\\nTeague’s strict requirements for retroactivity, then the claim is not redressable\\n\\nhere—the “nonretroactivity principle prevents a federal court from granting habeas\\ncorpus relief to a state prisoner based on a rule announced after his conviction and\\nsentence became final.” Caspari, 510 U.S. at 389 (emphasis in original). And as\\n\\nwe have said before, if “the court cannot relieve the harm of which a plaintiff\\ncomplains, the court should not take the case; in the absence of an effective remedy\\nits decision can amount to nothing more than an advisory opinion.” Wymbs v.\\n\\nRepublican State Exec. Comm., 719 F.2d 1072, 1085 (11th Cir. 1983).\\n Our authority to overturn state convictions is limited, and the retroactivity\\nprinciples articulated in Teague are tailored to those limitations. See Danforth, 552\\n\\nU.S. at 277. The fact that state courts do not face the same constraints on collateral\\nreview of their own criminal proceedings as we do does not relieve us of the\\nobligation to apply federal retroactivity standards. See id. at 280–81; Banks I, 536\\nU.S. at 271. In fact, our narrow authority as federal courts to disrupt final state-\\ncourt convictions reflects our recognition of the states’ own sovereignty. So\\nFlorida may design and apply its retroactivity principles as generously as it wishes.\\nBut notwithstanding Florida’s decision to apply Hurst—or any future decision—\\nretroactively as a matter of state law, as a federal court we are required to perform\\nthe Teague analysis to determine whether prisoners can receive retroactive relief\\nunder federal law.\\n In sum, if Hurst announced a new rule of constitutional law, but one that\\ndoes not fall into one of the exceptions to Teague’s bar on retroactivity, Knight\\n\\n 13\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 14 of 28\\n\\n\\n\\ncannot obtain federal habeas relief for any Hurst error in his sentence—regardless\\nof what Florida may choose to do under state law. And if Knight cannot obtain\\n\\nfederal habeas relief for his Hurst claim in any event, we may not offer an advisory\\nopinion on whether the claim could have merit.\\n III.\\n\\n A.\\n Turning to Teague, our analysis has three steps. First, we determine the date\\nwhen the petitioner’s conviction became final. Banks II, 542 U.S. at 411. This\\n\\nhappens when the United States Supreme Court “affirms a conviction on the merits\\non direct review or denies a petition for a writ of certiorari, or when the time for\\nfiling a certiorari petition expires.” Clay v. United States, 537 U.S. 522, 527\\n\\n(2003). Here, the Court denied Knight’s petition for a writ of certiorari on May 14,\\n2012—more than three years before Hurst was decided. Knight, 566 U.S. at 998.\\n Second, if the rule that the petitioner wants to apply had not been announced\\nby that final-conviction date, we “assay the legal landscape” as it existed at the\\ntime and determine whether existing precedent compelled the rule—that is,\\nwhether the case announced a new rule or applied an old one. Banks II, 542 U.S.\\nat 413. If—and only if—the holding was “dictated by precedent existing at the\\ntime the defendant’s conviction became final,” then the rule is not new and may be\\napplied retroactively on federal habeas review (indeed, it must be). Caspari, 510\\nU.S. at 390 (emphasis in original) (quoting Teague, 489 U.S. at 301). And that is\\nnot a light test—a rule is not dictated by prior precedent “unless it would have been\\n‘apparent to all reasonable jurists.’” Chaidez v. United States, 568 U.S. 342, 347\\n\\n 14\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 15 of 28\\n\\n\\n\\n(2013) (quoting Lambrix v. Singletary, 520 U.S. 518, 528 (1997)). The fact that a\\ndecision is “within the logical compass of” or even “controlled by” prior precedent\\n\\nis not conclusive in the Teague analysis. Butler v. McKellar, 494 U.S. 407, 415\\n(1990) (internal quotation marks and citation omitted). To “‘dictate’ a result, prior\\nprecedent must be specific; it is not enough that it name the general principle from\\n\\nwhich the assertedly new rule sprang.” Glock, 65 F.3d at 884.\\n Knight argues that Ring v. Arizona dictated the rule in Hurst. In Ring, the\\nSupreme Court held that Arizona’s capital sentencing scheme, which required\\n\\njudges alone to hear penalty-phase evidence and make factual findings relevant to\\nthe imposition of the death penalty, violated the defendant’s Sixth Amendment\\nright to a jury trial. 536 U.S. at 588–89. In doing so, the Court explicitly\\n\\noverruled its precedent upholding Arizona’s death penalty sentencing scheme, “to\\nthe extent that it allows a sentencing judge, sitting without a jury, to find an\\naggravating circumstance necessary for imposition of the death penalty.” Id. at\\n609 (emphasis added) (overruling Walton v. Arizona, 497 U.S. 639, 647–49\\n(1990)).\\n Knight wants us to find that the new Hurst rule is actually the old Ring rule\\nfor an obvious reason—if the rule is not new, and instead was binding on lower\\ncourts at the time of Knight’s conviction, then he is entitled to the benefit of the\\nrule on federal habeas review. See Yates v. Aiken, 484 U.S. 211, 216–17 & n.3\\n(1988). But Ring did not dictate the Supreme Court’s later invalidation of\\nFlorida’s death penalty sentencing scheme in Hurst. In fact, the Ring Court\\nspecifically acknowledged that Florida’s capital sentencing procedure differed\\n\\n 15\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 16 of 28\\n\\n\\n\\nfrom the Arizona scheme that it rejected. See Ring, 536 U.S. at 607–08 & n.6\\n(categorizing state capital sentencing schemes according to jury involvement in\\n\\nsentencing). In Arizona, the judge alone made the factual findings necessary to\\nimpose the death penalty and imposed that penalty entirely apart from the jury.\\nSee id. at 588. Florida’s scheme, in contrast, incorporated an advisory jury that\\n\\nconsidered penalty-phase evidence and recommended a sentence of life or death to\\nthe court. Only after the jury’s recommendation did the judge impose a sentence.\\nSee id. at 608 n.6; Fla. Stat. Ann. § 921.141 (2001).\\n\\n Hurst’s conclusion that Florida’s “hybrid” death penalty sentencing system\\nviolated the Sixth Amendment was not “apparent to all reasonable jurists” when\\nKnight’s convictions became final in 2012. We venture to count ourselves and our\\n\\ncolleagues on this Court as members of that distinguished group. And in Evans v.\\nSecretary, Florida Department of Corrections, we held that Ring did not invalidate\\nFlorida’s capital sentencing scheme. 699 F.3d 1249, 1264–65 (11th Cir. 2012).\\nWe found it significant that Florida’s statute—unlike the Arizona law at issue in\\nRing—required the penalty phase jury to find that sufficient aggravating\\ncircumstances existed before it could recommend a death sentence, and directed\\nthe sentencing court to give “great weight” to the jury’s advisory verdict. Id. at\\n1261 (citation omitted). We also noted that the Supreme Court had taken obvious\\npains in Ring to distinguish hybrid systems like Florida’s from the “judge-only”\\nsentencing scheme in Arizona and concluded that such distinctions would not have\\nbeen necessary if the Court had intended to strike down both systems. Id. at 1262.\\n\\n\\n\\n 16\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 17 of 28\\n\\n\\n\\n And we were not the only ones. In fact, Justice Alito wrote along those\\nsame lines in his dissenting opinion in Hurst: “Although the Court suggests that\\n\\ntoday’s holding follows ineluctably from Ring, the Arizona sentencing scheme at\\nissue in that case was much different from the Florida procedure now before us.”\\n136 S. Ct. at 625 (Alito, J., dissenting). After describing the “critically important\\n\\nrole” of the advisory jury in Florida’s death penalty sentencing scheme, Justice\\nAlito concluded that the “decision in Ring did not decide whether this procedure\\nviolates the Sixth Amendment, and I would not extend Ring to cover the Florida\\n\\nsystem.” Id. at 626. Clearly, reasonable jurists could—and did—disagree that\\nRing compelled the outcome in Hurst. The Alito dissent and this Court’s pre-\\nHurst holding strongly indicate that Hurst announced a new constitutional rule\\n\\nrather than applying an old one. See Banks II, 542 U.S. at 415.\\n That conclusion is only strengthened by the fact that Ring did not\\nspecifically address the continued validity of the Supreme Court’s precedents\\nupholding Florida’s death-penalty sentencing system—Spaziano v. Florida, 468\\nU.S. 447 (1984), and Hildwin v. Florida, 490 U.S. 638 (1989) (per curiam). And\\nthe Court has repeatedly instructed us to follow its precedents, even if later\\ndecisions appear to undermine them, unless and until the Court itself sets them\\naside. See, e.g., Bosse v. Oklahoma, 137 S. Ct. 1, 2 (2016) (per curiam) (“Our\\ndecisions remain binding precedent until we see fit to reconsider them, regardless\\nof whether subsequent cases have raised doubts about their continuing vitality.”\\n(citation omitted)); Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S.\\n477, 484 (1989). So Spaziano and Hildwin remained good law until the Court\\n\\n 17\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 18 of 28\\n\\n\\n\\nexplicitly overruled them in Hurst. See Saffle v. Parks, 494 U.S. 484, 488 (1990)\\n(explaining that the “explicit overruling of an earlier holding no doubt creates a\\n\\nnew rule”); see also Hurst, 136 S. Ct. at 623. It is hard to imagine that the\\nSupreme Court overruled those cases in Ring but forgot to say so until Hurst.\\n Because all these factors show that Hurst was not dictated by prior\\n\\nprecedent—and in fact explicitly overruled existing precedent upholding Florida’s\\ndeath penalty sentencing scheme—we can see that the rule in Hurst, which led to a\\nconclusion that the Florida scheme was unconstitutional, was new.\\n\\n Having determined that Hurst announced a new rule of constitutional law,\\nwe proceed to the final step in the Teague analysis—whether Hurst “falls within\\neither of the two exceptions to nonretroactivity.” Banks II, 542 U.S. at 411. Those\\n\\nexceptions, again, include (1) holdings that create substantive (not procedural)\\nrules that place “certain kinds of primary, private individual conduct beyond the\\npower of the criminal law-making authority to proscribe,” and (2) holdings that\\nconstitute “watershed rules of criminal procedure.” Teague, 489 U.S. at 311–13\\n(citation omitted).\\n The Hurst rule does not fit within either exception. To begin, substantive\\nrules include decisions that change “the range of conduct or the class of persons\\nthat the law punishes.” Schriro, 542 U.S. at 353. Procedural rules, on the other\\nhand, “regulate only the manner of determining the defendant’s culpability.” Id.\\n(emphasis in original). In considering which category the Hurst rule falls into, we\\nhave a head start because the Supreme Court has already held that Ring represented\\na “prototypical procedural rule[].” Id. And that makes sense: Ring changed the\\n\\n 18\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 19 of 28\\n\\n\\n\\npermissible procedure for sentencing in a capital case when it required “that a jury\\nrather than a judge find the essential facts” necessary to impose the death penalty.\\n\\nSee id. Because Hurst’s holding—that an advisory “jury’s mere recommendation\\nis not enough” to satisfy this procedural requirement—is an extension of the rule\\nfrom Ring, we have no trouble concluding that Hurst also announced a procedural\\n\\nrule, and not a substantive one. Hurst, 136 S. Ct. at 619.\\n The second exception is for “watershed rules of criminal procedure.” Banks\\nII, 542 U.S. at 417. This exception is extremely limited in scope—it applies “only\\n\\nto a small core of rules” so fundamental to our criminal process that it is “unlikely\\nthat many such components of basic due process have yet to emerge.” Id.\\n(citations omitted). Indeed, the watershed exception remains somewhat theoretical\\n\\nat this point; in the years following Teague, the Supreme Court has never found a\\nrule that fits. See id. And in “providing guidance as to what might fall within this\\nexception,” the Court has “repeatedly referred to the rule of Gideon v. Wainwright,\\n372 U. S. 335 (1963) (right to counsel), and only to this rule.” Id. Knight does not\\ncontend that Hurst announced a new watershed rule that compares to Gideon, and\\nwe do not see how it could have either. In short, Hurst meets neither exception,\\nand therefore is not retroactive.\\n B.\\n Because the Hurst rule is not retroactive, Knight cannot receive federal\\nhabeas relief on his Hurst claim. That is as it must be—we are conscientious about\\nthe fact that “Teague’s nonretroactivity principle acts as a limitation on the power\\nof federal courts to grant habeas corpus relief.” Id. at 412 (internal quotation\\n\\n 19\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 20 of 28\\n\\n\\n\\nmarks and citation omitted). And because we would have no lawful remedy to\\noffer even if we could identify an error, we must decline to consider whether any\\n\\nHurst error exists. That means, of course, that we also do not consider whether the\\nun-found and un-remediable error could be harmless.\\n Knight disputes this path. He argues that, even conceding the lack of an\\n\\navailable remedy for any Hurst error, whether that un-remediable error was\\nharmless is itself a separate question of federal law under Brecht v. Abrahamson,\\n507 U.S. 619, 637–38 (1993). He thus urges us to review the state court’s\\n\\nharmless-error analysis, regardless of whether we reach the merits of his Hurst\\nclaim. We will not. It strains the imagination—as well as our constitutional and\\ninstitutional respect for state courts—to suppose that we cannot remedy an error,\\n\\nbut that we can somehow remedy an erroneous state-court conclusion that the error\\nwas harmless.\\n And Brecht does not stand for the proposition that Knight asserts in any\\nevent. While Brecht established the harmless-error standard for collateral review\\nof constitutional trial errors, it did not create a stand-alone federal claim severable\\nfrom the question of whether remediable error existed in the first place.4 See\\nWilliams v. Singletary, 114 F.3d 177, 180 (11th Cir. 1997) (per curiam), as\\namended on denial of reh’g (July 29, 1997). Some underlying violation of federal\\nlaw that we can address is a necessary predicate to federal habeas relief—unless\\n\\n4\\n As an aside, what Brecht really decided was that federal courts evaluating constitutional trial\\nerror on collateral review would apply a more relaxed harmless-error standard—whether the\\nerror “had substantial and injurious effect or influence in determining the jury’s verdict”—rather\\nthan the harmless-beyond-a-reasonable-doubt standard used on direct review. Brecht, 507 U.S.\\nat 638 (quoting Kotteakos v. United States, 328 U.S. 750, 776 (1946)).\\n 20\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 21 of 28\\n\\n\\n\\nwe agree that an error has occurred, it makes no difference whether the purported\\nerror was harmless. “We have consistently applied the Brecht harmless error\\n\\nstandard only after determining that there was an error.” Id.; cf. Wilson v.\\nCorcoran, 562 U.S. 1, 6 (2010) (per curiam) (“It is not enough to note that a\\nhabeas petitioner asserts the existence of a constitutional violation; unless the\\n\\nfederal court agrees with that assertion, it may not grant relief.” (emphasis in\\noriginal)). There is no free-floating federal constitutional right to infallible\\napplication of harmless-error principles.\\n\\n And again, where our retroactivity doctrine forecloses the possibility of\\nfederal habeas relief for a constitutional error, we are constrained to stop looking.\\nSee Bowen, 422 U.S. at 920. The Supreme Court “consistently has declined to\\n\\naddress unsettled questions regarding the scope of decisions establishing new\\nconstitutional doctrine in cases in which it holds those decisions nonretroactive,”\\nand has instructed us to do the same. Id. at 920–21. This directive carries\\nconstitutional weight; as an Article III court, we are “without power to decide\\nquestions that cannot affect the rights of litigants in the case” at hand. North\\nCarolina v. Rice, 404 U.S. 244, 246 (1971) (per curiam). So even if we found\\nHurst error in Knight’s sentencing, we would still be prohibited from issuing a writ\\nof habeas corpus on that ground because Hurst is not retroactively applicable to\\nKnight under Teague. After all, Teague’s nonretroactivity command is a limitation\\non our power, not a polite suggestion. Banks II, 542 U.S. at 412. So our opinion—\\nwhatever it might be—on Knight’s Hurst claim would be purely advisory. “And it\\nis quite clear that the oldest and most consistent thread in the federal law of\\n\\n 21\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 22 of 28\\n\\n\\n\\njusticiability is that the federal courts will not give advisory opinions.” Flast v.\\nCohen, 392 U.S. 83, 96 (1968) (internal quotation marks and citation omitted).\\n\\n Where, as here, Teague bars relief before we reach the preliminary question\\nof whether constitutional error occurred at all, consideration of the secondary\\nquestion of whether any such error was harmless would be a prohibited and\\n\\npointless exercise for both the petitioner and this Court. We therefore cannot grant\\nKnight relief on his Hurst claim, whether or not it is cloaked in the garb of\\nharmless error.\\n\\n IV.\\n We now turn to Knight’s other claim. The State presented extensive DNA\\nevidence against him during the guilt phase of his trial. Kevin Noppinger, a\\n\\nserologist with the Broward County Sheriff’s Office crime lab, testified that Knight\\nhad Odessia’s blood on his hand and her DNA on his shirt when he was arrested.\\nFingernail scrapings from Odessia’s body showed that she, in turn, had Knight’s\\nDNA under her fingernails. Noppinger also tested samples from the bloody\\nclothes (boxer shorts, a shirt, and jean shorts) found under the bathroom sink. He\\nfound Knight’s blood mixed with Hanessia’s blood on the boxer shorts, Odessia’s\\nand Hanessia’s blood elsewhere on the boxer shorts and on the jean shorts, and\\nOdessia’s blood on the shirt.\\n One of the State’s other experts was Kevin McElfresh of Bode Technology\\nGroup, whose DNA analysts had conducted additional testing on different samples\\nfrom the same items of clothing. In particular, McElfresh’s group analyzed DNA\\nsamples from an unstained area of the waistband of the boxer shorts in an attempt\\n\\n 22\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 23 of 28\\n\\n\\n\\nto determine who owned them. Although Bode’s initial report stated that Knight’s\\nDNA was not on the waistband sample, McElfresh testified at trial that he had\\n\\nconducted some additional analysis and determined that some of the DNA on the\\nwaistband could have been Knight’s.\\n Knight’s guilt-phase counsel consulted with DNA expert Dr. Norah Rudin,\\n\\nwho was listed as a potential trial witness by the defense. Dr. Rudin informed\\ncounsel that, although some of Noppinger’s sample labeling practices were sloppy,\\nshe generally agreed with his conclusions about the sources of the DNA samples\\n\\nhe analyzed. She was much more critical of McElfresh’s analysis: she called his\\nmethods “fundamentally incorrect and inherently biased” and his testimony\\n“incomplete and misleading.” In her opinion, the DNA test results for the\\n\\nwaistband samples were “inconclusive.” But Dr. Rudin did not stop there—she\\nalso considered McElfresh’s testimony “relatively inconsequential when viewed in\\nthe context of the biological evidence as a whole.” She told Knight’s counsel that\\nshe did not believe that she could help his case, and that she would not call herself\\nas an expert if she were in his shoes. Counsel, it appears, agreed with her\\nperspective and did not call her at trial.\\n Testimony indicates that one additional fact persuaded counsel that calling\\nDr. Rubin to quibble with McElfresh’s methods would not be good trial strategy.\\nAt the time of Knight’s trial, Florida permitted a defendant who did not put up any\\nevidence at trial to give both initial and rebuttal closing arguments—an advantage\\nordinarily offered to the prosecution. Dr. Rudin would have been Knight’s only\\nwitness, so calling her to testify would have meant giving up the opportunity to\\n\\n 23\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 24 of 28\\n\\n\\n\\nhave the first word and the last at closing. Given that fact, she would have largely\\ncorroborated Noppinger’s testimony while also costing Knight an advantage at\\n\\nclosing arguments.\\n Nonetheless, on state collateral review, Knight contended that his counsel\\nprovided constitutionally ineffective assistance by failing to call Dr. Rudin as an\\n\\nexpert. He argued that Dr. Rudin’s criticisms of McElfresh’s methods and the\\ncrime lab’s labeling practices would have cast doubt on all of the State’s DNA\\nevidence and significantly damaged the State’s case. The Florida Supreme Court\\n\\ndisagreed, finding that Knight had not met his burden on either prong of the two-\\npart test for ineffective assistance of counsel claims set out in Strickland. See\\nKnight, 225 So. 3d at 672–74. Knight, however, argues that the state court’s\\n\\ndecision “was contrary to, or involved an unreasonable application of, clearly\\nestablished Federal law, as determined by the Supreme Court of the United States.”\\n28 U.S.C. § 2254(d)(1).\\n To begin, the Florida Supreme Court correctly identified the governing\\nstandard. Strickland is the relevant “clearly established” Supreme Court precedent\\nfor purposes of an ineffective-assistance claim. See Premo v. Moore, 562 U.S.\\n115, 118 (2011). Under Strickland, to succeed on a claim of ineffective assistance\\nof counsel, a defendant must show both that his attorney’s performance was\\ndeficient—that is, “that counsel’s representation fell below an objective standard of\\nreasonableness”—and that he was prejudiced by the inadequate performance.\\nStrickland, 466 U.S. at 687–88. In applying the first prong, courts “must indulge a\\nstrong presumption that counsel’s conduct falls within the wide range of\\n\\n 24\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 25 of 28\\n\\n\\n\\nreasonable professional assistance.” Id. at 689. And to show prejudice, the\\n“defendant must show that there is a reasonable probability that, but for counsel’s\\n\\nunprofessional errors, the result of the proceeding would have been different. A\\nreasonable probability is a probability sufficient to undermine confidence in the\\noutcome.” Id. at 694.\\n\\n The Strickland standard is “highly deferential,” as is the review of a state-\\ncourt decision under AEDPA; “when the two apply in tandem, review is ‘doubly’\\nso.” Harrington v. Richter, 562 U.S. 86, 105 (2011) (citations omitted). In\\n\\nreviewing a state court’s application of Strickland, therefore, a federal habeas court\\ncannot conduct a de novo review and reverse simply because it strongly disagrees\\nwith the state court’s conclusion. Id. at 102. Instead, for Knight to succeed on his\\n\\nineffective-assistance claim, we must conclude that the Florida Supreme Court’s\\ndecision “was an objectively unreasonable application of the Strickland standard.”\\nAllen v. Sec’y, Fla. Dep’t of Corr., 611 F.3d 740, 751 (11th Cir. 2010).\\n The Florida Supreme Court determined that counsel’s decision not to call\\nDr. Rudin at trial was reasonable trial strategy. Knight, 225 So. 3d at 674.\\nWhether the decision was actually a matter of strategy is a question of fact; thus,\\nthe state court’s finding on that issue is presumed to be correct. Provenzano v.\\nSingletary, 148 F.3d 1327, 1330 (11th Cir. 1998). We must accept all factual\\nfindings made by the state court unless the petitioner rebuts the presumption of\\ncorrectness “by clear and convincing evidence.” 28 U.S.C. § 2254(e)(1); see also\\nWood v. Allen, 542 F.3d 1281, 1285 (11th Cir. 2008). Knight has not presented\\n\\n\\n\\n 25\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 26 of 28\\n\\n\\n\\nany evidence to suggest that counsel’s decision was anything other than a matter of\\nstrategy, and we accept the state court’s finding on that point.\\n\\n Whether counsel’s strategic decision not to call Dr. Rudin was reasonable is\\na question of law, which we review through the lens of AEDPA deference. See\\nFerrell v. Hall, 640 F.3d 1199, 1223–24 (11th Cir. 2011). In assessing an\\n\\nattorney’s performance under Strickland, “strategic choices made after thorough\\ninvestigation of law and facts relevant to plausible options are virtually\\nunchallengeable.” Strickland, 466 U.S. at 690. “Which witnesses, if any, to call,\\n\\nand when to call them, is the epitome of a strategic decision, and it is one that we\\nwill seldom, if ever, second guess.” Waters v. Thomas, 46 F.3d 1506, 1512 (11th\\nCir. 1995) (en banc). We have no reason to doubt the Florida Supreme Court’s\\n\\nconclusion that counsel’s decision not to call an equivocal expert, in part to\\npreserve an advantage at closing, was well within the wide range of reasonably\\ncompetent performance.\\n The Florida Supreme Court also determined that Knight had not met his\\nburden of showing prejudice under Strickland because there was no reasonable\\nprobability that Dr. Rudin’s testimony would have made a difference in the\\noutcome of the trial, given the weight of the evidence against him. See Knight, 225\\nSo. 3d at 674. That conclusion was also reasonable.\\n Dr. Rudin generally agreed with Noppinger’s conclusions regarding the\\nDNA evidence. With or without Dr. Rudin’s testimony, therefore, there was no\\ndispute that Knight had Odessia’s blood on his hand and her DNA on his shirt\\nwhen he was arrested, or that Knight’s DNA was found under Odessia’s\\n\\n 26\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 27 of 28\\n\\n\\n\\nfingernails. Knight’s blood—as well as Odessia’s blood and Hanessia’s—was on\\nthe clothes discarded at the crime scene. And even if the jury had heard Dr.\\n\\nRudin’s criticism of McElfresh’s testimony that Knight could not be ruled out as a\\nDNA contributor for the waistband of the boxer shorts, other evidence showed that\\nthe clothes were his: the boxer shorts were the same brand and size as the ones that\\n\\nKnight was wearing when he was arrested, and Knight’s cousin testified that the\\nbloody shirt was one that Knight wore often.\\n And the evidence did not stop there. Knight was known to be home with\\n\\nHanessia less than an hour before the murders, and the jury heard evidence of\\nongoing tension with Odessia. Police found him near the scene of the murders\\nshortly after they arrived, and his answers to their questions were inconsistent with\\n\\nhis appearance. All of that is in addition to the testimony of an inmate housed with\\nKnight at the Broward County jail, who testified that Knight confessed the crime to\\nhim, providing a detailed description of events and a floorplan of the apartment. In\\nshort, even if the jury had entirely discounted McElfresh’s testimony as a result of\\nDr. Rudin’s criticism of his testing methods, the remaining evidence against\\nKnight was so strong that the chance of a not-guilty verdict still would have been\\nremote, to say the least.\\n The Florida Supreme Court’s conclusion that Knight failed to make the\\nrequired showings of deficient performance and prejudice was not an unreasonable\\napplication of Strickland.\\n\\n\\n\\n\\n 27\\n\\f Case: 18-13390 Date Filed: 08/30/2019 Page: 28 of 28\\n\\n\\n\\n ***\\n In sum, we are prohibited from considering Knight’s Hurst claim by\\n\\nTeague’s rule of nonretroactivity, which eliminates any possibility of relief\\nregardless of whether there was an error and regardless of whether any error was\\nharmless. And our review of Knight’s ineffective-assistance claim shows that the\\n\\nstate court’s ruling on that issue was not an unreasonable application of governing\\nSupreme Court precedent. We therefore affirm the district court’s denial of\\nKnight’s federal habeas petition.\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 28\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4164,"cells":{"id":{"kind":"number","value":4657834,"string":"4,657,834"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"Mark Jenkins v. Commissioner, Alabama Department of Corrections"},"decision_date":{"kind":"string","value":"2019-08-30"},"docket_number":{"kind":"string","value":"17-12524"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before TJOFLAT, WILSON, and BRANCH, Circuit Judges.\", \"parties\": \"\", \"opinions\": [{\"author\": \"BRANCH, Circuit Judge:\", \"type\": \"010combined\", \"text\": \"Case: 17-12524 Date Filed: 08/30/2019 Page: 1 of 73\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n\\n\\n No. 17-12524\\n\\n\\n\\n D.C. Docket No. 4:08-cv-00869-VEH\\n\\nMARK ALLEN JENKINS,\\n\\n Petitioner - Appellant,\\n\\n versus\\n\\n\\n\\nCOMMISSIONER, ALABAMA DEPARTMENT OF CORRECTIONS,\\n\\n Respondent - Appellee.\\n\\n\\n\\n Appeal from the United States District Court\\n for the Northern District of Alabama\\n\\n\\n (August 30, 2019)\\n\\nBefore TJOFLAT, WILSON, and BRANCH, Circuit Judges.\\n\\nBRANCH, Circuit Judge:\\n\\n Mark Allen Jenkins, an Alabama prisoner sentenced to death for the 1989\\n\\nmurder of Tammy Ruth Hogeland, appeals the district court’s denial of his petition\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 2 of 73\\n\\n\\nfor a writ of habeas corpus, 28 U.S.C. § 2254. Before us are Jenkins’s arguments\\n\\nthat he received ineffective assistance of counsel during the penalty phase of his\\n\\ntrial and that he is intellectually disabled and therefore ineligible for the death\\n\\npenalty. After careful consideration, and with the benefit of oral argument, we\\n\\naffirm the denial of Jenkins’s habeas petition on both grounds.\\n\\n I. BACKGROUND\\n\\n A. The Crime and Arrest\\n\\n We summarize the following background narrative from opinions of the\\n\\nAlabama Court of Criminal Appeals and our own review of the record. See Jenkins\\n\\nv. State, 627 So. 2d 1034, 1037–40 (Ala. Ct. Crim. App. 1992); Jenkins v. State,\\n\\n972 So. 2d 111, 119–20 (Ala. Ct. Crim. App. 2004). The events leading up to the\\n\\nmurder began the evening of April 17, 1989. Jenkins was at the home of his\\n\\nacquaintance Christine Nicholas. He had met her several months earlier at the\\n\\nOmelet Shoppe restaurant where she worked. Jenkins was very intoxicated and\\n\\ntried to seduce her. When she resisted, Jenkins got “real mad” and asked her\\n\\nseveral times what she would do if someone came up from behind her and grabbed\\n\\nher.\\n\\n At around 1 a.m., Jenkins and Nicholas drove to the Riverchase Omelet\\n\\nShoppe. Jenkins went inside and talked with one of the waitresses, Frieda Vines.\\n\\nThe manager, Douglas Thrash, recognized Jenkins as a regular customer who knew\\n\\n\\n 2\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 3 of 73\\n\\n\\nall of the waitresses. Thrash overheard part of Jenkins’s conversation with Vines\\n\\nand heard mention of the Omelet Shoppe location near the Birmingham Airport.\\n\\nJenkins and Nicholas then returned to Nicholas’s home. Around 2 a.m., Nicholas’s\\n\\nmother asked Jenkins to leave. He did so, falling down some steps and ramming\\n\\nhis car into another vehicle in the process.\\n\\n Meanwhile, 23-year-old Tammy Hogeland was making her way to work at\\n\\nthe Omelet Shoppe. Her sister Wendy, along with Tammy’s young son, picked\\n\\nTammy up from her college classes and drove her to the Riverchase Omelet\\n\\nShoppe a little after 9:30 p.m. Sometime after 10:00, Tammy was sent to work at\\n\\nthe airport Omelet Shoppe on Tenth Avenue, since that location was unexpectedly\\n\\nshorthanded. Tammy was wearing her watch, a necklace with the words “special\\n\\nsister,” her class ring, and her diamond engagement ring. Tammy was working as a\\n\\ncook that night and was wearing a blue apron, black pants, a white shirt, and white\\n\\nshoes. She and Sarah Harris were the only employees working at the airport\\n\\nOmelet Shoppe that night.\\n\\n At about 2 a.m. on April 18, Harris saw Jenkins—whom she did not know—\\n\\ndrive up in a red sports car. His arrival was memorable to Harris because the car\\n\\nnearly jumped the curb and crashed through the restaurant’s glass wall. Jenkins got\\n\\nout of the car and came into the restaurant, appearing to be intoxicated. He walked\\n\\nover to Hogeland and began talking to her. Harris later saw Jenkins and Hogeland\\n\\n\\n 3\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 4 of 73\\n\\n\\ndrive away together in the red sports car. That was the last time anyone who knew\\n\\nHogeland ever saw her. Hogeland left behind her cigarettes, lighter, purse, and\\n\\npaycheck. She left without telling anyone, which she had never done before.\\n\\nAlthough Harris saw them drive off, she could not later say whether Hogeland left\\n\\nwith Jenkins willingly or was instead abducted.\\n\\n At around 5 a.m. that morning, Bobby and Geraldine Coe had stopped to\\n\\nbuy gas at a Chevron station on I-59 northeast of Birmingham when they saw\\n\\nJenkins drive up in the red sports car. The Coes noticed a female who appeared to\\n\\nbe “passed out” in the front passenger seat, but they could not say whether she was\\n\\nalive or dead. While Bobby Coe was pumping gas, Jenkins asked him for some\\n\\ncigarettes and said, “Looks like it’s been a long night and it looks like it’s going to\\n\\nbe a long day.” Jenkins then said “God bless you” before asking directions to I-\\n\\n459. Coe gave him directions and got back in his car. As Coe pulled out onto the\\n\\ninterstate, he saw Jenkins follow him in the red sports car. Coe then saw the car\\n\\nflash its lights, slow down, and pull to the side of the road between mile markers\\n\\n151 and 152.\\n\\n At 8 a.m., Wendy Hogeland learned that Tammy was not at the restaurant,\\n\\nand their mother called the police. Meanwhile, Jenkins went to the home of Steve\\n\\nMusser, who noticed that he was wearing the same clothes as the day before.\\n\\nJenkins told Musser that his car had been stolen the night before and asked him if\\n\\n\\n 4\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 5 of 73\\n\\n\\nhe would say that he had been with him all night. Musser refused. Christine\\n\\nNicholas also saw Jenkins in a grocery store that morning. He was looking at a\\n\\nnewspaper, making a phone call, and attempting to sell his Buick. Nicholas loaned\\n\\nhim $4 for gas. At around 10 o’clock that morning, Jenkins sold his car to Michael\\n\\nBrooks, a mechanic at a local Chevron station. Jenkins had explained that he\\n\\nneeded the money so he could visit his sick mother in California. Another\\n\\nmechanic at the service station drove Jenkins to the Greyhound bus station later\\n\\nthat morning.\\n\\n Jenkins awoke the next day on the Greyhound bus in Houston and was\\n\\nejected from the bus because his fare was used up. He then hitchhiked from\\n\\nHouston to Phoenix, then to San Diego and Los Angeles. Jenkins was first\\n\\nidentified as a suspect in Hogeland’s disappearance by Officer Mike Weems of the\\n\\nHoover Police Department on April 19. Weems ate dinner at the Riverchase\\n\\nOmelet Shoppe just about every day and had talked with Hogeland just as often.\\n\\nHe also knew Jenkins from the restaurant. Weems learned of Hogeland’s\\n\\ndisappearance from the other waitresses. Remembering how Jenkins often talked to\\n\\nHogeland and passed her notes,1 he gave Jenkins’s name to the missing persons\\n\\ninvestigator. The investigator also learned that a red Mazda RX-7 sports car, which\\n\\n\\n\\n1\\n The jury was not permitted to hear, and the record does not reveal, what exactly Jenkins had\\nsaid to Hogeland that made Officer Weems suspect him in her disappearance.\\n 5\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 6 of 73\\n\\n\\nhad been reported stolen from the service station where Jenkins worked, had been\\n\\nrecovered on I-459.\\n\\n The afternoon of April 21, a truck driver who had happened to stop on I-59\\n\\nnear mile marker 151 discovered Hogeland’s body off the side of the road. She was\\n\\nnaked, wearing only a watch, and was so badly decomposed that the body had to\\n\\nbe identified by dental records. From the fractured hyoid bone, it was determined\\n\\nthat she was manually strangled to death. Also found at the scene were Hogeland’s\\n\\napron, shoes, bra, panties, pants, and hair net, as well as some beer cans, a Mazda\\n\\nRX-7 owner’s manual, and other items later determined to have come from the red\\n\\nMazda. Her necklace and rings were never found.\\n\\n Alabama authorities issued a warrant for Jenkins’s arrest and also obtained a\\n\\nfederal fugitive warrant. An FBI agent arrested Jenkins on May 1 in front of his\\n\\nuncle’s house in Wilmington, California. Jenkins’s uncle later gave the police three\\n\\nbags of Jenkins’s personal effects. Fibers recovered from Hogeland’s body and\\n\\nclothing matched those of the Mazda, and hairs matching Hogeland’s were found\\n\\nin the Mazda. Similarly, fibers from Jenkins’s clothing placed him in the Mazda,\\n\\nand fibers of his jeans were found on Hogeland’s apron. A bootprint near\\n\\nHogeland’s body matched a combat boot from among the belongings Jenkins’s\\n\\nuncle gave to the police, and Jenkins’s roommate said the boots looked just like\\n\\nthose that Jenkins wore.\\n\\n\\n 6\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 7 of 73\\n\\n\\n A business card like those that had been in the Mazda’s glove box,\\n\\nbelonging to an owner of the Mazda, was recovered from Jenkins’s wallet after his\\n\\narrest in California. Later, Jenkins’s cellmate in the Alabama jail reported that\\n\\nJenkins told him “he had done the crime” and was worried that the couple from the\\n\\ngas station would identify him or that the police would find his fingerprint on a\\n\\nbeer can at the scene. A St. Clair County grand jury indicted Jenkins for capital\\n\\nmurder.\\n\\n B. The Trial and Sentencing\\n\\n Jenkins was represented at trial in 1991 by attorneys Douglas Scofield and\\n\\nStan Downey. Scofield, the lead attorney, had been practicing criminal defense in\\n\\nBirmingham since 1984 but had never before tried a capital murder case. Jenkins’s\\n\\nlandlord and his grandmother had arranged the referral and representation before\\n\\nthe court appointed him. Scofield agreed to the appointment on the condition that a\\n\\nSt. Clair County lawyer also be appointed. That lawyer was Downey, who had\\n\\nmore experience than Scofield but whose capital experience was limited to one\\n\\ncase that had not gone to trial. The two attorneys had agreed that Downey would\\n\\nbe responsible for jury selection, but Scofield eventually took over during the voir\\n\\ndire. Per their agreement, Scofield was responsible for the guilt-phase trial,\\n\\nalthough Downey participated by interviewing Jenkins several times. Scofield also\\n\\nmet with Jenkins around a dozen times. Jenkins had told him about his background\\n\\n\\n 7\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 8 of 73\\n\\n\\nand childhood, including the fact that he had had a difficult childhood and had\\n\\nlived on the streets since age 11.\\n\\n Throughout the weeklong trial, Scofield pursued an outright acquittal. In\\n\\nparticular, he challenged Harris’s identification of Jenkins as the man who drove\\n\\naway with Hogeland, and he challenged the medical examiner’s opinions about the\\n\\ntime and cause of Hogeland’s death. In his closing argument, Scofield took a wide-\\n\\nranging approach to undermining the State’s wholly circumstantial case against\\n\\nJenkins. He attacked the State’s timeline of the crime, questioning whether Jenkins\\n\\nin his intoxicated state would have had the time and the ability to leave Nicholas’s\\n\\nhouse, steal the Mazda, and abduct and kill Hogeland. He reminded the jury of the\\n\\nweakness of Harris’s identification of Jenkins and the Coes’ identification of the\\n\\nwoman in the car. He pointed out that no physical evidence definitively linked\\n\\nJenkins to the Mazda, to Hogeland’s body, or to her jewelry, which was never\\n\\nrecovered. And he argued that the evidence provided no plausible motive for\\n\\nJenkins to have kidnapped, robbed, or killed Hogeland. In sum, he urged the jury\\n\\nto find reasonable doubt and acquit Jenkins. At the very least, he argued, they\\n\\nshould find that the State failed to prove Jenkins’s intent to rob and kidnap\\n\\nHogeland, which were necessary to convict Jenkins of capital murder. He further\\n\\nargued that, if the jury thought Jenkins killed Hogeland as a crime of passion, they\\n\\ncould convict him only of manslaughter. The jury deliberated for three hours and\\n\\n\\n 8\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 9 of 73\\n\\n\\nfifteen minutes before convicting Jenkins of capital murder, Ala. Code § 13A-5-\\n\\n40(a)(1) & (2), i.e., murder committed in the course of kidnapping and robbery.\\n\\n The penalty phase took place that same afternoon. A week before trial,\\n\\nScofield and Downey had discussed the penalty phase, and Scofield understood\\n\\nthat Downey would be responsible for handling the penalty phase. Nonetheless, the\\n\\nwitness who was present to testify on Jenkins’s behalf during the penalty phase\\n\\nwas better acquainted with Scofield, so Downey suggested that afternoon that\\n\\nScofield conduct the direct examination. Scofield urged Downey to interview the\\n\\nwitness quickly and prepare to examine him, and Downey did so.\\n\\n Downey handled the penalty presentation to the jury. The State did not\\n\\npresent any additional aggravation evidence, resting on the evidence presented\\n\\nduring the guilt phase. One witness, Lonnie Seal, testified on Jenkins’s behalf in\\n\\norder, in Downey’s words, to “reveal another side of Mark Jenkins to you that you\\n\\ndon’t know anything about.” Seal had met Jenkins in Fontana, California, in 1988\\n\\nwhen they both worked at the same garage. Seal and his wife became friends with\\n\\nJenkins, and he visited in their home. When the Seals decided to move to Alabama\\n\\na few months later, Jenkins agreed to help with their move and to live with the\\n\\nSeals in Alabama for a time. It took Lonnie Seal several weeks to find a job, but\\n\\nJenkins found a job in two days and contributed toward the family’s rent and\\n\\n\\n\\n\\n 9\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 10 of 73\\n\\n\\ngroceries. Jenkins continued to visit the Seals after he moved into his own home,\\n\\nand he would always bring a small gift for and play with the Seals’ infant son.\\n\\n Downey’s closing argument to the jury requested a sentence of life without\\n\\nparole. Downey reminded the jurors that they had agreed during voir dire that\\n\\ndeath was not always the appropriate punishment for homicide. Quoting the Bible,2\\n\\nhe reminded the jury that the Old Testament permitted capital punishment only\\n\\nwhen supported by the testimony of two or three eyewitnesses, whereas in this case\\n\\nthere had been only circumstantial evidence. He asserted from Seal’s testimony\\n\\nthat “Mark had a side of him that was generous, that was giving, and that was\\n\\nkind,” and argued that the deterrent and retributive purposes of capital punishment\\n\\nwould not be served in this case.\\n\\n The jury deliberated for 50 minutes before returning its nonbinding\\n\\nrecommendation of death by a vote of 10 to 2. See Ala. Code § 13A-5-46(e) (1981)\\n\\n(amended 2017 to make jury’s verdict binding). After a separate hearing, the court\\n\\nfound the two statutory aggravating factors for which the State had argued: that the\\n\\nmurder was committed during the commission of a robbery, and that it was\\n\\ncommitted during the commission of a kidnapping. See Ala. Code § 13A-5-49(4).\\n\\n\\n\\n\\n2\\n “If anyone kills a person, the murderer shall be put to death on the evidence of witnesses. But\\nno person shall be put to death on the testimony of one witness.” Numbers 35:30 (ESV). Downey\\nalso paraphrased Deuteronomy 17:6 as “On the evidence of two witnesses or three witnesses, he\\nis [sic] shall be put to death. He shall not be put to death on the evidence of one witness.”\\n 10\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 11 of 73\\n\\n\\nThe court also found two statutory mitigating factors: that Jenkins had no\\n\\nsignificant history of prior criminal activity, and his age at the time of the crime\\n\\n(21). See id. § 13A-5-51(1), (7). The court considered but rejected the statutory\\n\\nmitigation that, due to his consumption of alcohol, Jenkins’s capacity to appreciate\\n\\nthe criminality of his conduct or to conform his conduct to the requirements of law\\n\\nwas substantially impaired. See id. § 13A-5-51(6). The court found that the\\n\\naggravating circumstances outweighed the mitigating circumstances and, in\\n\\naccordance with the jury’s recommendation, imposed a sentence of death.\\n\\n C. The State and Federal Post-Conviction Proceedings\\n\\n On direct appeal, the Alabama Court of Criminal Appeals affirmed Jenkins’s\\n\\nconviction and death sentence. Jenkins v. State, 627 So. 2d 1034 (Ala. Ct. Crim.\\n\\nApp. 1992), aff’d, 627 So. 2d 1054 (Ala. 1993), cert. denied, 511 U.S. 1012\\n\\n(1994). In 1995, Jenkins filed his first petition for state postconviction relief, Ala.\\n\\nR. Crim. P. 32,3 alleging, among many other things, that he received ineffective\\n\\nassistance of counsel during his penalty phase. The Rule 32 trial court conducted a\\n\\nthree-day hearing at which evidence was presented about the performance of\\n\\nJenkins’s attorneys.\\n\\n\\n\\n\\n3\\n “Subject to the limitations of Rule 32.2, any defendant who has been convicted of a criminal\\noffense may institute a proceeding in the court of original conviction to secure appropriate relief\\non the ground that . . . [t]he constitution of the United States or of the State of Alabama requires\\na new trial, a new sentence proceeding, or other relief. . . .” Ala. R. Crim. P. 32.1.\\n 11\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 12 of 73\\n\\n\\n Scofield testified about the representation that he and Downey provided. In\\n\\nrelevant part, Scofield asserted that he and Downey had “decided that [developing\\n\\nmitigation for the penalty phase] would be something that Stan [Downey] would\\n\\nhandle.” Downey later told Scofield that he hadn’t done any such investigation, but\\n\\nScofield admitted that, “[a]part from what he has told me, I don’t know what he\\n\\nhas done.” Scofield testified that Downey “didn’t appear prepared” for the penalty\\n\\nphase, based on Downey’s last-minute suggestion that Scofield examine Seal since\\n\\nhe had talked with him before.\\n\\n In preparation for the penalty phase, Scofield himself did not interview any\\n\\nmember of Jenkins’s family, but he did not know whether Downey interviewed\\n\\nany family members. Scofield did interview Jenkins’s landlord and some other\\n\\nacquaintances but did not do so specifically for the purpose of developing\\n\\nmitigation. He had met with Jenkins around a dozen times, and Jenkins had told\\n\\nhim his life story including his experience of abuse as a child. In hindsight,\\n\\nScofield said he had a “lack of appreciation of mitigating circumstances.” He\\n\\nrepeatedly denied in his Rule 32 testimony that his omissions during the penalty\\n\\nphase were strategic: “I didn’t understand what I was doing. . . . I just didn’t\\n\\nunderstand what I needed to do.”\\n\\n For reasons that remain unclear, Downey—who is since deceased—did not\\n\\ntestify or submit an affidavit in the Rule 32 proceeding. Scofield answered,\\n\\n\\n 12\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 13 of 73\\n\\n\\n“Obviously not,” when asked whether he had personal knowledge of everything\\n\\nDowney may or may not have done to prepare for the penalty phase. The record of\\n\\nthe extent of Downey’s pretrial preparation is thus limited to Downey’s detailed\\n\\nfee declaration, which he submitted to the State of Alabama in documentation of\\n\\nthe compensation due him as appointed counsel. See Ala. Code § 15-12-21. In it,\\n\\nDowney reported that he spent 71 hours4 on out-of-court preparation for trial. The\\n\\nreport details 35 hours of interviews with Jenkins in 18 separate visits to the jail\\n\\nover the course of 17 months, as well as one substantive conversation with\\n\\nJenkins’s grandmother.5\\n\\n The Rule 32 trial court also heard testimony from four family members and\\n\\none friend of Jenkins’s. The family members presented for the first time a stark\\n\\ndepiction of Jenkins as an unloved and abused child in California. Jenkins’s older\\n\\nbrother Michael Jenkins testified that all of his siblings have different fathers and\\n\\nthat his parents often used drugs and fought. Michael testified that Jenkins was\\n\\nbeaten frequently as a child by his stepfather, if not every day then every other day\\n\\n\\n\\n4\\n In Jenkins’s state postconviction appeal, the Alabama Court of Criminal Appeals erroneously\\nstated that Downey’s fee declaration reports 171 hours of pretrial preparation. See Jenkins, 972\\nSo. 2d at 137. We discuss the effect of this discrepancy infra note 7.\\n5\\n The state appellate court also noted that Downey had moved for a continuance in October 1989\\nso that he could conduct “[f]urther discovery and investigation (including a possible trip to\\nCalifornia).” Id. We note that Scofield had orally suggested to the trial court in September 1989\\nthat more time might be needed with respect to “evidence about the defendant” and witnesses in\\nCalifornia, as well as possible psychological evaluation.\\n 13\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 14 of 73\\n\\n\\nor every three days. Jenkins was beaten all over his body with a variety of\\n\\nimplements. Michael was beaten too. Jenkins’s stepfather would lock Jenkins in\\n\\nhis bedroom for a couple of hours at a time, sometimes without dinner. Some of\\n\\nthe beatings were prompted by Jenkins’s bedwetting or soiling. Michael testified\\n\\nthat Jenkins’s stepfather sometimes made Jenkins wear his soiled underwear on his\\n\\nhead and even forced him to eat his own feces a few times. Michael testified that\\n\\nJenkins’s bladder and bowel control problems began after a camping trip he took\\n\\nwith his stepfather and stepgrandfather.\\n\\n Michael testified that, as children, he and Jenkins were made to work on car\\n\\nengines late into the night and to shovel horse manure daily. The family moved\\n\\nfrequently and the children attended many different schools as a result. Michael\\n\\ntestified that he was Jenkins’s only friend as a child and that Jenkins did poorly in\\n\\nschool. Jenkins began running away from home as a young teenager and lived with\\n\\nvarious other family members or on the streets. Michael said that Jenkins was a\\n\\ngood brother to him and had asked him to testify at his trial, but he never heard\\n\\nfrom his attorneys or found out when the trial was. Michael admitted to having\\n\\ncognitive and memory problems due to a head injury years earlier.\\n\\n Jenkins’s cousin Tammy Pitts testified that, as an infant, Jenkins’s mother\\n\\nbeat and neglected him, leaving him dirty and in soiled diapers. His stepfather also\\n\\nbeat Jenkins daily until he left home at age 13, sometimes bruising him so badly\\n\\n\\n 14\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 15 of 73\\n\\n\\nthat he was laid up in bed for days or weeks. Jenkins had problems with\\n\\nbedwetting and soiling, and his stepfather made him wear soiled clothes to school.\\n\\nJenkins would be locked in his filthy bedroom around the clock except to attend\\n\\nschool and do chores. He was usually not allowed to eat dinner with the rest of the\\n\\nfamily and subsisted on scraps and dog food. The entire family belittled Jenkins\\n\\nbecause his biological father was Hispanic and his complexion was therefore\\n\\ndarker. Jenkins’s parents drank a lot and used drugs every day. Pitts knew these\\n\\nthings because she had daily contact with the family and lived with them off and\\n\\non, and Jenkins lived with her for a time after he ran away from home. No one had\\n\\ncontacted her about testifying at Jenkins’s trial.\\n\\n Jenkins’s second cousin Betty DeLavega testified that her relationship with\\n\\nJenkins was like that of an aunt. She lived with the Jenkins family for five months\\n\\nwhen Jenkins was 10 or 11. She saw Jenkins’s stepfather slap Jenkins “[q]uite a\\n\\nfew times,” and once, he made Jenkins eat his own feces. DeLavega testified that\\n\\nJenkins’s parents used drugs nearly every day and withheld affection from Jenkins\\n\\nand Michael. No one contacted her about Jenkins’s trial; Jenkins’s grandmother\\n\\nhad asked her to testify in the Rule 32 hearing about any abuse she had witnessed.\\n\\n Jenkins’s grandmother Doris Wagoner testified that while Jenkins’s mother\\n\\nwas pregnant with him, her husband was in prison and she drank and took drugs.\\n\\nWhen Jenkins was born, his mother initially put him up for adoption, then changed\\n\\n\\n 15\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 16 of 73\\n\\n\\nher mind weeks later. Jenkins was a lethargic infant and his mother cared more\\n\\nabout partying than about her children. Wagoner testified that she heard of\\n\\nJenkins’s mistreatment as a child but didn’t see it. She could tell from the way that\\n\\nJenkins feared his stepfather that he was being abused, and she knew he slept in a\\n\\nfilthy bed. After Jenkins was arrested for murder, Wagoner spoke with Scofield,\\n\\nwho asked for money relating to his representation of Jenkins. She did not\\n\\nremember whether she talked with Scofield about Jenkins’s background. She could\\n\\nnot remember why she didn’t testify at the trial; she might have been ill, but she\\n\\nwas probably busy.\\n\\n Sherry Seal, the wife of Lonnie Seal, who had testified in the penalty phase\\n\\nof the trial, testified that Jenkins was respectful and generous and that she trusted\\n\\nhim with her children. She related how he supported her family after moving to\\n\\nAlabama, giving them “pretty much” his entire paycheck. She said she would still\\n\\nfeel safe around Jenkins. Two jailers also testified that Jenkins was a respectful,\\n\\ncourteous, model inmate who never complained or caused any problems.\\n\\n Jenkins also introduced medical, school, and juvenile court records from his\\n\\nchildhood. The medical records document his premature birth and his mother’s\\n\\nintention to give him up for adoption. The school records reflect a struggling\\n\\nstudent who was repeatedly promoted to the next grade because of his age, despite\\n\\nserious academic difficulties, chronic truancy, and frequent school changes. The\\n\\n\\n 16\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 17 of 73\\n\\n\\njuvenile court records document Jenkins’s placement in juvenile hall starting at age\\n\\n14, following repeated acts of grand theft auto and running away.\\n\\n The Rule 32 hearing concluded with the testimony of two psychologists\\n\\nabout Jenkins’s history of psychological trauma stemming from his childhood\\n\\nabuse. The defense expert, Dr. David Lisak, had met with Jenkins three times and\\n\\ninterviewed a dozen of his family members, family friends, and associates. Dr.\\n\\nLisak testified that the pervasive abuse and neglect in Jenkins’s childhood impaired\\n\\nhis ability to cope with trauma as an adult. He opined that Jenkins’s childhood\\n\\nabuse adversely affected his cognitive and emotional development and other\\n\\nabilities. He described how Jenkins had been neglected as an infant and sexually\\n\\nabused by his grandfather at age four, which led to a lifetime of bladder and bowel\\n\\ndysfunction. He also discussed the hypothesis that children who are abused are at\\n\\ngreater risk for perpetrating violence later in life, but he acknowledged there would\\n\\nhave been other causative factors in Jenkins’s case. Dr. Lisak described Jenkins’s\\n\\nintellectual capacity as “somewhat borderline,” though he did not perform any\\n\\ntesting of his own.\\n\\n The State’s expert, Dr. Karl Kirkland, had met with Jenkins for several hours\\n\\nand administered several tests. On the Wechsler Adult Intelligence Scale, Jenkins\\n\\n“scored in the range of borderline intellectual functioning which is between mild\\n\\nmental retardation and low average intellectual functioning” with an overall IQ of\\n\\n\\n 17\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 18 of 73\\n\\n\\n76. Dr. Kirkland agreed that Jenkins had been “reared in a chaotic dysfunctional\\n\\nfamily” and generally agreed with Dr. Lisak’s assessment of the negative effects of\\n\\nchildhood trauma.\\n\\n The court considered all of this evidence before denying Rule 32 relief in\\n\\nfull. As relevant here, it found that some of the evidence about the severity of\\n\\nabuse that Jenkins suffered as a child was not credible. It further found that Jenkins\\n\\nsuffered no prejudice from any deficient performance by his penalty-phase counsel\\n\\nbecause the aggravating circumstances outweighed any additional mitigating\\n\\nevidence Downey might have adduced.\\n\\n The Alabama Court of Criminal Appeals affirmed the denial of Rule 32\\n\\nrelief. Jenkins v. State, 972 So. 2d 111 (Ala. Ct. Crim. App. 2004). In relevant part,\\n\\nthe appeals court rejected Jenkins’s ineffective assistance claim, finding that\\n\\ncounsel reasonably chose to pursue a penalty-phase strategy of residual doubt and\\n\\ngood character. Id. at 147. It explained that the new evidence was self-\\n\\ncontradictory in that it asserted that childhood abuse would have made Jenkins a\\n\\nviolent adult yet also asserted that Jenkins was meek and mild. Id. The appeals\\n\\ncourt also agreed with the lower court’s conclusion that Jenkins suffered no\\n\\nprejudice from any error by his counsel because the aggravating circumstances still\\n\\noutweighed any new mitigating ones. Id. at 148. The court also noted that evidence\\n\\n\\n\\n\\n 18\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 19 of 73\\n\\n\\nabout Jenkins’s good behavior in jail awaiting trial was “minimally mitigating” and\\n\\nwould not have affected the sentence. Id. at 149.\\n\\n Because the Supreme Court’s decision in Atkins v. Virginia6 had issued\\n\\nduring the pendency of the Rule 32 appeal, the Alabama appeals court ordered\\n\\nsupplemental briefing on the possible impact of Atkins upon Jenkins’s sentence.\\n\\nJenkins asked the court to stay the proceedings and conduct an evidentiary hearing\\n\\nafter the Alabama legislature enacted a statute about executing the mentally\\n\\nretarded or, in the alternative, to vacate his sentence and remand after the Alabama\\n\\nlegislature acted. He later asserted that he is mentally retarded and therefore\\n\\nineligible for the death penalty. The appeals court proceeded to reject Jenkins’s\\n\\nAtkins claim on the merits because the record established that his IQ was 76 and\\n\\nthat he maintained relationships and employment. Id. at 155. The Alabama\\n\\nSupreme Court summarily affirmed, in relevant part. Ex parte Jenkins, 972 So. 2d\\n\\n159, 165 (Ala. 2005), remanded to 972 So. 2d 165 (Ala. Ct. Crim. App. 2005)\\n\\n(juror misconduct claim), cert. denied, 552 U.S. 1167 (2008).\\n\\n In 2008, Jenkins filed his first petition for a writ of habeas corpus in federal\\n\\ncourt pursuant to 28 U.S.C. § 2254. The federal proceedings were stayed for\\n\\nseveral years while Jenkins pursued additional, unrelated Rule 32 relief in the state\\n\\n\\n6\\n 536 U.S. 304, 321 (2002) (holding that the execution of mentally retarded offenders violates the\\nEighth Amendment). The rule announced in Atkins is retroactive to cases on collateral review. In\\nre Holladay, 331 F.3d 1169, 1172–73 (11th Cir. 2003).\\n 19\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 20 of 73\\n\\n\\ncourts, ultimately unsuccessfully. In the district court, Jenkins moved for an\\n\\nevidentiary hearing on his Atkins claim. After reviewing the state court record,\\n\\nwith a focus on the Rule 32 testimony of Drs. Lisak and Kirkland, the district court\\n\\ndenied the motion in 2015. Noting that our Circuit has accepted the State of\\n\\nAlabama’s definition of mental retardation following Atkins, the court found that\\n\\nJenkins is not intellectually disabled under that standard. In a 347-page order, the\\n\\ndistrict court also denied relief on all of Jenkins’s claims including ineffective\\n\\nassistance of counsel during the penalty phase and affirmed as reasonable the state\\n\\ncourt’s conclusion that Jenkins failed to establish prejudice from any error by his\\n\\ncounsel. The district court later denied reconsideration and a certificate of\\n\\nappealability.\\n\\n Jenkins now appeals. Our Court granted Jenkins a certificate of\\n\\nappealability, 28 U.S.C. § 2253(c), on two issues:\\n\\n 1) Whether the district court erred in denying Appellant’s claim\\n that his trial counsel rendered ineffective assistance of counsel in the\\n investigation and presentation of mitigating evidence during the\\n penalty phase of Appellant’s 1991 trial; and\\n\\n 2) Whether the district court erred in denying Appellant’s claim\\n that he is intellectually disabled and ineligible for the death penalty\\n under Atkins v. Virginia, 536 U.S. 304 (2002).\\n\\n II. STANDARDS OF REVIEW\\n\\n We review the denial of a petition for a writ of habeas corpus de novo.\\n\\nMorrow v. Warden, Ga. Diagnostic Prison, 886 F.3d 1138, 1146 (11th Cir. 2018).\\n\\n 20\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 21 of 73\\n\\n\\nThe Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) instructs that\\n\\na writ of habeas corpus shall not be granted to a state prisoner unless the state\\n\\nadjudication of a claim “resulted in a decision that was contrary to, or involved an\\n\\nunreasonable application of, clearly established Federal law, as determined by the\\n\\nSupreme Court of the United States.” 28 U.S.C. § 2254(d)(1). “[C]learly\\n\\nestablished Federal law” means “the holdings, as opposed to the dicta, of [the\\n\\nSupreme] Court’s decisions as of the time of the relevant state-court decision.”\\n\\nWilliams v. Taylor, 529 U.S. 362, 412 (2000). “[W]hen the last state court to\\n\\ndecide a prisoner’s federal claim explains its decision on the merits in a reasoned\\n\\nopinion . . . a federal habeas court simply reviews the specific reasons given by the\\n\\nstate court and defers to those reasons if they are reasonable.” Wilson v. Sellers,\\n\\n138 S. Ct. 1188, 1192 (2018). A state court’s decision is reasonable “so long as\\n\\n‘fairminded jurists could disagree’ on the correctness of the state court’s decision.”\\n\\nHarrington v. Richter, 562 U.S. 86, 101 (2011) (quoting Yarborough v. Alvarado,\\n\\n541 U.S. 652, 664 (2004)). “[E]ven a strong case for relief does not mean the state\\n\\ncourt’s contrary conclusion was unreasonable.” Id. at 102. Rather, a prisoner must\\n\\nshow that the state court’s ruling “was so lacking in justification that there was an\\n\\nerror well understood and comprehended in existing law beyond any possibility for\\n\\nfairminded disagreement.” Id. at 103. That ruling must have been “‘objectively\\n\\nunreasonable,’ not merely wrong; even ‘clear error’ will not suffice.” White v.\\n\\n\\n 21\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 22 of 73\\n\\n\\nWoodall, 572 U.S. 415, 419 (2014) (quoting Lockyer v. Andrade, 538 U.S. 63, 75\\n\\n(2003)).\\n\\n In addition, “a determination of a factual issue made by a State court shall be\\n\\npresumed to be correct,” and the prisoner bears “the burden of rebutting the\\n\\npresumption of correctness by clear and convincing evidence.” 28 U.S.C.\\n\\n§ 2254(e)(1). AEDPA further instructs that a writ of habeas corpus may not be\\n\\ngranted unless the state court adjudication “resulted in a decision that was based on\\n\\nan unreasonable determination of the facts in light of the evidence presented in the\\n\\nState court proceeding.” Id. § 2254(d)(2). When a federal habeas court makes a\\n\\nfactual finding as part of its habeas determination, we review for clear error, which\\n\\nis a “highly deferential” standard of review. Thomas v. Allen, 607 F.3d 749, 752\\n\\n(11th Cir. 2010) (quoting Holton v. City of Thomasville Sch. Dist., 425 F.3d 1325,\\n\\n1350 (11th Cir. 2005)). “A finding is ‘clearly erroneous’ when although there is\\n\\nevidence to support it, the reviewing court on the entire evidence is left with the\\n\\ndefinite and firm conviction that a mistake has been committed.” United States v.\\n\\nU.S. Gypsum Co., 333 U.S. 364, 395 (1948).\\n\\n III. DISCUSSION\\n\\n A. Ineffective Assistance of Counsel\\n\\n A prisoner alleging that he received ineffective assistance of counsel in\\n\\nviolation of the Sixth Amendment must establish two elements. Strickland v.\\n\\n\\n 22\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 23 of 73\\n\\n\\nWashington, 466 U.S. 668, 687 (1984). “First, the defendant must show that\\n\\ncounsel’s performance was deficient.” Id. In other words, he must show “that\\n\\ncounsel made errors so serious that counsel was not functioning as the ‘counsel’\\n\\nguaranteed the defendant by the Sixth Amendment.” Id. “Second, the defendant\\n\\nmust show that the deficient performance prejudiced the defense.” Id. We discuss\\n\\neach of these elements in turn.\\n\\n 1. Deficient Performance\\n\\n A “doubly” deferential lens sits before our eyes as we review the\\n\\nperformance of Jenkins’s counsel during the penalty phase. Richter, 562 U.S. at\\n\\n105 (quoting Knowles v. Mirzayance, 556 U.S. 111, 123 (2009)). This redoubled\\n\\ndeference arises out of the tandem application of AEDPA’s “highly deferential”\\n\\nreview of a state court’s decision with Strickland’s “highly deferential” review of\\n\\nan attorney’s performance. Id. (quoting Lindh v. Murphy, 521 U.S. 320, 333 n.7\\n\\n(1997), and Strickland, 466 U.S. at 689). Under the Strickland standard, “counsel\\n\\nis strongly presumed to have rendered adequate assistance and made all significant\\n\\ndecisions in the exercise of reasonable professional judgment.” Strickland, 466\\n\\nU.S. at 690. And under AEPDA, we must deny habeas relief if “there is any\\n\\nreasonable argument that counsel satisfied Strickland’s deferential standard.”\\n\\nRichter, 562 U.S. at 105.\\n\\n\\n\\n\\n 23\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 24 of 73\\n\\n\\n Jenkins argues that his counsel performed deficiently by failing to\\n\\n“investigate compelling mitigating evidence or develop a penalty-phase theory.”\\n\\nThe state court disagreed. It quoted decisions of our Court that hold that “there is\\n\\nno per se rule that evidence of a criminal defendant’s troubled childhood must\\n\\nalways be presented as mitigating evidence.” Jenkins, 972 So. 2d at 144 (quoting\\n\\nMarek v. Singletary, 62 F.3d 1295, 1300 (11th Cir. 1995)). Noting that “Downey\\n\\nwas in charge of the penalty phase” but “we do not have the benefit of Downey’s\\n\\ntestimony as to what occurred and why,” it reviewed the trial record to find that\\n\\n“counsel argued residual doubt and Jenkins’s good character at the penalty phase.”\\n\\nId. at 145–46. It concluded that “Downey’s decision to concentrate on reasonable\\n\\ndoubt and to portray Jenkins as a good person was reasonable under the\\n\\ncircumstances.” Id. at 147.\\n\\n Jenkins insists, however, that counsel’s failure to develop mitigating\\n\\nevidence about his childhood abuse was “not the product of strategy, but rather\\n\\narose from ingnorance and inexeperiance [sic].” The problem for Jenkins is that,\\n\\nlike the state courts, this Court also does not know what Downey did or why. The\\n\\nrecord is silent as to his thoughts and intentions as he prepared for the penalty\\n\\nphase. Given that Jenkins bore the burden of overcoming the strong presumption of\\n\\ncompetence and of proving Downey’s deficient performance, we are perplexed\\n\\nwhy Jenkins’s new counsel did not have Downey testify or submit an affidavit in\\n\\n\\n 24\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 25 of 73\\n\\n\\nthe Rule 32 proceeding. See Strickland, 466 U.S. at 687 (“The defendant must\\n\\nshow that counsel’s performance was deficient.”). Nor did Jenkins testify to the\\n\\nsubstance of his conversations with his counsel. “The reasonableness of a trial\\n\\ncounsel’s acts, including lack of investigation or excluding character witnesses\\n\\nfrom the sentencing phase, depends ‘critically’ upon what information the client\\n\\ncommunicated to counsel.” Chandler v. United States, 218 F.3d 1305, 1324 (11th\\n\\nCir. 2000) (en banc) (quoting Strickland, 466 U.S. at 691). There is simply no\\n\\nrecord evidence to support Jenkins’s central assertion that Downey made no\\n\\ntactical choices in his preparation for the penalty phase.\\n\\n Jenkins instead relies on Scofield’s Rule 32 testimony about how he and\\n\\nDowney prepared for the trial and sentencing. We acknowledge that Scofield\\n\\nrepeatedly testified that he did not really know what he was doing and that he\\n\\nfailed to make strategic choices with respect to mitigating circumstances and\\n\\nsentencing. But Scofield was equally clear that Downey—the more experienced\\n\\nlawyer—was in fact responsible for the penalty phase. And Scofield also\\n\\nrepeatedly testified that he did not know what Downey may or may not have done\\n\\nto prepare. Furthermore, his lack of knowledge about Downey’s preparation was\\n\\nunderstandable. He and Downey were not colleagues who shared an office and\\n\\nmight reasonably be expected to know what the other was doing much of the time.\\n\\nScofield and Downey did not even work in the same city.\\n\\n\\n 25\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 26 of 73\\n\\n\\n This case is unlike those in which we have found deficient performance\\n\\nwhere the record showed that counsel investigated insufficiently because he\\n\\n“waited until the eleventh hour” to begin preparing for the penalty phase while\\n\\nfaced with “overwhelming evidence of guilt.” See, e.g., Johnson v. Sec’y, D.O.C.,\\n\\n643 F.3d 907, 932 (11th Cir. 2011). The limited record that Jenkins developed\\n\\nestablishes that Downey spent at least 71 hours preparing for trial, spread out over\\n\\n17 months. The state court interpreted this sparse record as reflecting a penalty-\\n\\nphase strategy of residual doubt, the logical extension of Downey’s guilt-phase\\n\\ninvestigation of the evidence and Scofield’s pursuit at trial of an acquittal based on\\n\\nreasonable doubt. See Jenkins, 972 So. 2d at 147. In such a situation, “[a] lawyer’s\\n\\ntime and effort in preparing to defend his client in the guilt phase of a capital case\\n\\ncontinues to count at the sentencing phase.” Tarver v. Hopper, 169 F.3d 710, 715\\n\\n(11th Cir. 1999).\\n\\n We decline to dictate how many hours or how many sessions an attorney\\n\\nmust spend interviewing his counsel in order to provide effective assistance. “No\\n\\nabsolute rules dictate what is reasonable performance for lawyers.” Chandler, 218\\n\\nF.3d at 1317 (citing Strickland, 466 U.S. at 688–89). But even this limited record\\n\\nundermines Jenkins’s assertion that Downey prepared inadequately. Downey’s 71\\n\\nhours of preparation included 35 hours of jailhouse conversations with Jenkins\\n\\n\\n\\n\\n 26\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 27 of 73\\n\\n\\nduring 18 separate meetings.7 According to Downey’s notes, Scofield was present\\n\\nfor only five of those meetings (12 hours’ worth), leaving 23 hours over 13\\n\\nmeetings between Jenkins and Downey alone. Downey’s notes reveal little about\\n\\nthe content of those conversations beyond broad notations such as “evidence,\\n\\nwitnesses, general status,” and discussions with Jenkins of the evidence and\\n\\ntestimony against him were surely relevant to Downey’s penalty-phase strategy of\\n\\nresidual doubt.\\n\\n But the record does not reveal that Downey knew or should have known that\\n\\nJenkins’s childhood merited investigation beyond that which he may have done.\\n\\nThe record shows that Downey developed enough of a rapport with Jenkins that,\\n\\naccording to his fee declaration, he “g[ave a] Christmas gift [and] discuss[ed] how\\n\\nmuch to tell his family.” But how much Downey knew about Jenkins’s family\\n\\nbackground can only be guessed from Scofield’s Rule 32 hearing testimony.\\n\\nScofield testified that he discussed Jenkins’s background during his dozen or so\\n\\njailhouse meetings with him, and that Downey was present for some of those\\n\\nmeetings. Scofield—and possibly Downey—learned that Jenkins experienced\\n\\nabuse as a child, that his stepfather was brutal, and that he lived on the streets and\\n\\n\\n\\n\\n7\\n We note that the state court clearly erroneously stated that Downey’s fee declaration\\ndocumented 171 hours of pretrial preparation. See Jenkins, 972 So. 2d at 137. However, the state\\ncourt accurately stated that “he spent over 25 hours talking with Jenkins in more than 10 visits to\\nthe jail.” Id. That statement was not an unreasonable determination of the facts.\\n 27\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 28 of 73\\n\\n\\nin juvenile institutions because of his brutal home life. It is against this backdrop\\n\\nthat Downey’s presumptive “decision not to investigate must be directly assessed\\n\\nfor reasonableness in all the circumstances, applying a heavy measure of deference\\n\\nto counsel’s judgments.” Strickland, 466 U.S. at 691.\\n\\n That assessment depends a great deal on what Jenkins told his counsel about\\n\\nhis background. See id. But we do not know whether Downey was present when\\n\\nScofield learned about Jenkins’s childhood abuse, nor what else Downey may have\\n\\nlearned during his 13 solo discussions with Jenkins. Therefore, “given the absence\\n\\nof evidence in the record, we must assume counsel carried out his professional\\n\\nresponsibility and discussed mitigation with his client.” Chandler, 218 F.3d at\\n\\n1324. We will not “turn that presumption on its head by giving [Jenkins] the\\n\\nbenefit of the doubt where it is unclear what [Downey] did or did not do.” Williams\\n\\nv. Head, 185 F.3d 1223, 1235 (11th Cir. 1999). We will not assume that Jenkins\\n\\ntold Downey more than he told Scofield, and Scofield denied remembering that\\n\\nJenkins told him about the severity or frequency of his stepfather’s physical abuse.\\n\\nHe denied remembering Jenkins telling him about being beaten daily, being locked\\n\\nin his room 24/7, or being forced to eat his feces. And there is no suggestion in the\\n\\nrecord that counsel knew or had reason to investigate the allegation of sexual abuse\\n\\nby his stepgrandfather. The record simply does not reflect that Jenkins told either\\n\\nof his lawyers about the severe abuse that some of the Rule 32 witnesses later\\n\\n\\n 28\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 29 of 73\\n\\n\\ndescribed. We will not fault those lawyers for failing to present as mitigation this\\n\\nlimited evidence that they may have reasonably believed to be unremarkable. See\\n\\nStrickland, 466 U.S. at 691.\\n\\n To be sure, counsel’s failure “to discover all reasonably available mitigating\\n\\nevidence” may constitute deficient performance in some cases. Wiggins v. Smith,\\n\\n539 U.S. 510, 524 (2003) (quoting ABA Guidelines for the Appointment and\\n\\nPerformance of Counsel in Death Penalty Cases § 11.4.1(C) (1989)). Our Court\\n\\nhas identified deficient performance in several instances where counsel had good\\n\\nreason to investigate the defendant’s background but failed to do so. See Morrow,\\n\\n886 F.3d at 1147 (collecting cases). But we assess “a particular decision not to\\n\\ninvestigate . . . for reasonableness in all the circumstances, applying a heavy\\n\\nmeasure of deference to counsel’s judgments.” Strickland, 466 U.S. at 691. Thus,\\n\\nwe have rejected a hard rule that counsel’s “strategic decisions can be considered\\n\\nreasonable only if they are preceded by a ‘thorough investigation.’” Williams, 185\\n\\nF.3d at 1237 (quoting Rogers v. Zant, 13 F.3d 384, 387 (11th Cir. 1994)). Because\\n\\nwe simply do not know why Downey chose not to investigate Jenkins’s childhood\\n\\nmore thoroughly, Downey is entitled to the presumption that this strategic decision\\n\\nwas reasonable.\\n\\n And though we have also rejected “a per se rule of ineffective assistance\\n\\nwhere counsel does not consult family members,” Holladay v. Haley, 209 F.3d\\n\\n\\n 29\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 30 of 73\\n\\n\\n1243, 1252 (11th Cir. 2000), we note that Downey did not fail altogether to consult\\n\\nfamily members. Downey had at least one substantive conversation with Jenkins’s\\n\\ngrandmother, and we know from her Rule 32 hearing testimony that she never\\n\\nwitnessed Jenkins’s childhood abuse. Given the silence of the record and the strong\\n\\npresumption of competence to which Downey is entitled, we decline to fault him\\n\\nfor not doing more when Jenkins and his grandmother did not and could not,\\n\\nrespectively, provide a basis for doing so. Cf. United States v. Pease, 240 F.3d 938,\\n\\n941–92 (11th Cir. 2001) (“we cannot say as a matter of law that reliance on a\\n\\nclient’s statements is per se deficient performance”). Nor did Jenkins’s counsel\\n\\nhave an absolute duty to put on abusive-childhood evidence, even if they knew of\\n\\nit. See Chandler, 218 F.3d at 1319 (“Counsel is not required to present every\\n\\nnonfrivolous defense; nor is counsel required to present all mitigation evidence,\\n\\neven if the additional mitigation evidence would not have been incompatible with\\n\\ncounsel’s strategy.”); Waters v. Thomas, 46 F.3d 1506, 1511 (11th Cir. 1995) (en\\n\\nbanc) (“Our decisions are inconsistent with any notion that counsel must present\\n\\nall available mitigating circumstance evidence.”).\\n\\n Instead, we affirm as a reasonable application of Strickland the state court’s\\n\\nconclusion that Downey’s strategic decision to emphasize residual doubt was\\n\\nreasonable. See Jenkins, 972 So. 2d at 147. As we have noted, “[a] lawyer’s time\\n\\nand effort in preparing to defend his client in the guilt phase of a capital case\\n\\n\\n 30\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 31 of 73\\n\\n\\ncontinues to count at the sentencing phase.” Tarver, 169 F.3d at 715. This\\n\\nreasoning is especially applicable where, as here, counsel pursued the reasonable\\n\\nstrategy of arguing for acquittal at the guilt phase and for life at the sentencing\\n\\nphase by highlighting doubt about the evidence. “[R]esidual doubt is perhaps the\\n\\nmost effective strategy to employ at sentencing.” Chandler, 218 F.3d at 1320 n.28.\\n\\n“[E]specially when—as in this case—the evidence of guilt was not overwhelming,\\n\\nwe expect that petitioners can rarely (if ever) prove a lawyer to be ineffective for\\n\\nrelying on this seemingly reasonable strategy to defend his client.” Id. at 1320\\n\\n(footnote omitted); see also Tarver, 169 F.3d at 715–16 (collecting law review\\n\\narticles on the merits of residual-doubt strategy). Not only can such a strategy be\\n\\nconstitutionally adequate, in some cases it may be “the most effective performance\\n\\nin defense to the death penalty.” Tarver, 169 F.3d at 716.\\n\\n We cannot say that Downey’s manifest strategy of residual doubt was\\n\\nobjectively unreasonable under the circumstances. Jenkins’s case is precisely the\\n\\nsort where reminding the jury of the misgivings it may have held about the State’s\\n\\ncase could have been effective. Scofield’s guilt-phase closing argument from that\\n\\nvery morning, which explicitly emphasized doubt and the weakness of the State’s\\n\\nevidence, would have been fresh in the minds of the jury.8 Although we are not\\n\\n\\n8\\n During the penalty phase of a capital case, “[e]vidence presented at the trial of the case may be\\nconsidered insofar as it is relevant to the aggravating and mitigating circumstances without the\\nnecessity of re-introducing that evidence at the sentence hearing.” Ala. Code § 13A-5-45(c).\\n 31\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 32 of 73\\n\\n\\nquestioning the jury’s guilty verdict or the sufficiency of the evidence, which are\\n\\nbeyond the scope of our habeas review, we acknowledge that the State’s case\\n\\nagainst Jenkins was—as Downey reminded the jury during the penalty phase—\\n\\nentirely circumstantial. Indeed, Scofield maintained post-conviction that the case\\n\\nfor reasonable doubt was “[v]ery strong.” No one saw Jenkins commit either the\\n\\nmurder or the robbery and kidnapping that elevated Hogeland’s murder to a capital\\n\\ncrime. Scofield’s guilt-phase closing argument challenged the inferential leaps the\\n\\nState was asking the jury to make in finding that Jenkins committed these capital-\\n\\npredicate offenses and that he formed the intent to kill. This situation was not one\\n\\nwhere a guilt-phase strategy of acquittal—and therefore a penalty-phase strategy of\\n\\nresidual doubt—was unreasonable in light of the evidence. Cf. Johnson, 643 F.3d\\n\\nat 932–33 (deficient sentencing investigation where defendant had confessed to\\n\\ntwo murders).\\n\\n Thus, although Downey’s penalty-phase closing argument did not\\n\\nspecifically invoke lingering doubt, it clearly reminded the jury of the entirely\\n\\ncircumstantial nature of the State’s guilt-phase case. See Chandler, 218 F.3d at\\n\\n1320 & n.6 (a strategy of “focus[ing] on obtaining an acquittal and then, at\\n\\nsentencing, on lingering doubt” is reasonable, even if counsel does not use the\\n\\nwords “lingering doubt” or “residual doubt”). Downey closed this reminder by\\n\\nexplaining to the jury how a foundational ancient legal system proscribed capital\\n\\n\\n 32\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 33 of 73\\n\\n\\npunishment in the absence of eyewitness testimony—the exact situation here. The\\n\\nstate court found that Downey’s penalty-phase closing argument reflected a\\n\\nstrategic “decision to concentrate on reasonable doubt,” see Jenkins, 972 So. 2d at\\n\\n147, a tactical choice entitled to a “strong presumption of correctness,” Marek, 62\\n\\nF.3d at 1300 & n.3.\\n\\n Jenkins suggests other theories of mitigation that Downey might have more\\n\\neffectively pursued,9 but our deferential review under Strickland does not ask\\n\\nwhether counsel could possibly or ideally have been more effective. “The test for\\n\\nineffectiveness is not whether counsel could have done more.” Waters, 46 F.3d at\\n\\n1518. We do not ask whether an attorney’s representation “deviated from best\\n\\npractices or common custom,” and we should resist the temptation to second-guess\\n\\nan attorney with the benefit of our hindsight. Richter, 562 U.S. at 105. We ask only\\n\\n“whether, in light of all the circumstances, the identified acts or omissions were\\n\\n\\n\\n\\n9\\n In particular, Jenkins argues that his counsel should have presented “model prisoner” evidence.\\nThe state court noted, “Good conduct during pretrial incarceration is not necessarily a mitigating\\ncircumstance,” before finding that Jenkins had not met his burden of establishing that his counsel\\neven had information about his conduct in jail. Jenkins, 972 So. 2d at 149. We agree, and we\\ndecline to impose a requirement of sua sponte investigation of this kind in every capital case. Cf.\\nWilliams, 529 U.S. at 396 (finding deficient performance based in part on not returning a call\\nfrom a prison ministry volunteer who offered to vouch for defendant’s good behavior in prison).\\n“Surely, counsel is not required to call a witness to testify to facts such as lack of violent nature\\nwhen the jury has rejected such an approach and has found that the defendant is guilty of\\nmurder.” Griffin v. Wainwright, 760 F.2d 1505, 1512 (11th Cir. 1985), vacated on other grounds,\\n476 U.S. 1112 (1986), reaff’d, 874 F.2d 1397 (11th Cir. 1989).\\n 33\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 34 of 73\\n\\n\\noutside the range of professionally competent assistance.” Strickland, 466 U.S. at\\n\\n690.\\n\\n We conclude that the state court did not unreasonably determine that Jenkins\\n\\nfailed to establish objectively incompetent performance by his counsel during the\\n\\npenalty phase of his trial. Our inquiry is not “what the best lawyers would have\\n\\ndone. Nor is the test even what most good lawyers would have done. We ask only\\n\\nwhether some reasonable lawyer at the trial could have acted, in the circumstances,\\n\\nas defense counsel acted at trial. . . . We are not interested in grading lawyers’\\n\\nperformances.” White v. Singletary, 972 F.2d 1218, 1220–21 (11th Cir. 1992). And\\n\\n“[i]f fairminded jurists could reasonably disagree” about the reasonableness of\\n\\ncounsel’s performance, “then habeas relief is due to be denied.” Johnson, 643 F.3d\\n\\nat 932.\\n\\n 2. Prejudice\\n\\n Even if Jenkins had demonstrated that his counsel performed as no\\n\\nreasonable lawyer could have, he must also demonstrate a “reasonable probability\\n\\nthat, but for counsel’s unprofessional errors, the result of the proceeding would\\n\\nhave been different.” Strickland, 466 U.S. at 694. Given that the jury here\\n\\nrecommended a sentence of death by the narrowest possible vote, 10 to 2, Jenkins\\n\\nneed establish only “a reasonable probability that at least one juror would have\\n\\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537. “A\\n\\n\\n 34\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 35 of 73\\n\\n\\nreasonable probability is a probability sufficient to undermine confidence in the\\n\\noutcome.” Strickland, 466 U.S. at 694. In measuring that probability here, we will\\n\\n“evaluate the totality of the available mitigation evidence—both that adduced at\\n\\ntrial, and the evidence adduced in the habeas proceeding in reweighing it against\\n\\nthe evidence in aggravation.” Williams, 529 U.S. at 397–98.\\n\\n We begin our analysis by clarifying what exactly Jenkins is arguing about\\n\\nprejudice. We do not know precisely which aggravating and mitigating factors the\\n\\njury may have implicitly found before making its recommendation of death. The\\n\\ntrial court found at sentencing two statutory aggravating factors: the robbery and\\n\\nthe kidnapping of the victim. See Ala. Code § 13A-5-49(4). Because none of the\\n\\nmitigation evidence offered in the Rule 32 hearing bears on those factors, we see\\n\\nno reasonable probability that the aggravating factors would have been determined\\n\\ndifferently by the jury or the court. The court also found at sentencing two\\n\\nstatutory mitigating factors: no significant history of prior criminal activity and\\n\\nage. See id. § 13A-5-51(1), (7). There was also strong evidence of Jenkins’s\\n\\nintoxication, which was a key part of Jenkins’s guilt-phase defense and could have\\n\\nsupported a third mitigating factor, impaired capacity. See id. § 13A-5-51(6).10\\n\\n\\n10\\n The jury had heard evidence about all three of these statutory mitigators during the guilt phase\\nof the trial. Downey’s penalty-phase opening statement reviewed for the jury all seven of the\\npossible statutory mitigating factors, and the court also recited them in its jury instructions,\\nreminding the jury that they could consider guilt-phase evidence as well as non-statutory\\nmitigation. Jenkins has not established that there is a substantial probability that the jury’s\\nrecommendation would have been different had it been more explicitly reminded of Jenkins’s\\n 35\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 36 of 73\\n\\n\\nJenkins does not argue that the jury or the court would have found other statutory\\n\\nmitigators in light of the Rule 32 evidence. Rather, he contends that the jury might\\n\\nhave found non-statutory mitigation based on Jenkins’s abusive childhood, low\\n\\nintelligence, or mental health problems that would have outweighed the\\n\\naggravating factors.11\\n\\n The Rule 32 trial court found that Jenkins had not demonstrated prejudice\\n\\neven assuming that counsel had performed deficiently. See Jenkins, 972 So. 2d at\\n\\n138. It concluded that he had not shown a reasonable probability that the sentencer\\n\\n“would have concluded that a weighing of the aggravating and mitigating\\n\\ncircumstances did not warrant death” upon a reweighing of the evidence. See id. It\\n\\nemphasized the depravity of the crime and the aggravating circumstances and\\n\\nfound that the mitigating witnesses were not credible for various reasons. See id.\\n\\nThe state appellate court affirmed, declining to disturb the credibility\\n\\ndeterminations of the trial court. Id. at 142–44. Independently reweighing the\\n\\n\\nage, criminal history, and intoxication, given that the trial court, the Rule 32 trial court, and the\\nstate appellate court all explicitly considered those mitigators and found them outweighed by the\\naggravators.\\n11\\n Alabama law lists seven statutory mitigating circumstances, but states that mitigation “shall\\n. . . not be limited to” those seven circumstances. Ala. Code § 13A-5-51. Further, “mitigating\\ncircumstances shall include any aspect of a defendant’s character or record and any of the\\ncircumstances of the offense that the defendant offers as a basis for a sentence of life\\nimprisonment without parole instead of death, and any other relevant mitigating circumstances\\nwhich the defendant offers as a basis for a sentence of life imprisonment without parole instead\\nof death.” Id. § 13A-5-52. A jury may recommend a sentence of death only if it finds that “one or\\nmore aggravating circumstances . . . exist and that they outweigh the mitigating circumstances, if\\nany.” Id. § 13A-5-46(e)(3).\\n 36\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 37 of 73\\n\\n\\nevidence, it concluded that, “[g]iven the aggravating circumstances that were\\n\\nproven by the State and the facts surrounding Hogeland’s murder, we, like the\\n\\ncircuit court, are confident that death was the appropriate punishment for Jenkins’s\\n\\nactions.” Id. at 148.\\n\\n Our review of this evidence does not show that it is likely that the jury\\n\\nwould have concluded that death was not the appropriate punishment if it had\\n\\nheard the Rule 32 evidence. Much of the proffered mitigation evidence is, in our\\n\\nview, much less mitigating than Jenkins asserts. In particular, the testimony of\\n\\nJenkins’s older brother Michael was likely to have been a double-edged sword. See\\n\\nEvans v. Sec’y, Dep’t of Corr., 703 F.3d 1316, 1327 (11th Cir. 2013) (en banc) (“it\\n\\nis reasonable to conclude that a defendant was not prejudiced when his mitigation\\n\\nevidence was a two-edged sword” (quoting Ponticelli v. Sec’y, Fla. Dep’t of Corr.,\\n\\n690 F.2d 1271, 1296 (11th Cir. 2012)). Michael was the closest relative to testify at\\n\\nthe Rule 32 hearing and his testimony about the abuse Jenkins experienced should\\n\\nhave been the most reliable, since he experienced much of the same physical abuse\\n\\nhimself. The brothers shared the same filthy bedroom, received beatings from the\\n\\nsame paddle, and were frequently absent from school together. But Michael’s\\n\\ntestimony contained some discrepancies, and he admitted to cognitive and memory\\n\\nproblems. Perhaps Michael’s testimony would have made Jenkins appear more\\n\\nsympathetic in the eyes of the jury—but perhaps not. While Jenkins ran away from\\n\\n\\n 37\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 38 of 73\\n\\n\\nhome and eventually committed murder, Michael responded to the same\\n\\nbackground by running away from home, graduating from high school, and\\n\\nproviding health care for both the terminally ill and people with substance use\\n\\ndisorders. It is possible that the jury could even have found Jenkins to be more\\n\\nculpable for not overcoming his difficult upbringing and to pose more of a danger\\n\\nof future violence.\\n\\n We also find reasonable the state court’s identification of other problems\\n\\nwith the mitigating evidence Jenkins argues should have been presented. Hearing\\n\\nthat evidence firsthand in the Rule 32 proceeding, the trial court found the\\n\\ntestimony of the cousins less credible in light of the lack of physical abuse\\n\\ndocumented in the school health records. See Jenkins, 972 So. 2d at 139 (noting\\n\\nthat school records observed a rash and gingivitis). It faulted Pitts in particular for\\n\\npurporting to have witnessed ongoing, severe abuse but taking only minimal,\\n\\nineffective steps to obtain outside help. See id. It further noted that DeLavega and\\n\\nWagoner actually testified that they never saw Jenkins beaten, and it noted all of\\n\\nthe witnesses’ interest in getting Jenkins off death row. All of these factors would\\n\\nlikely have caused a reasonable jury to discount their testimony. See id. at 139–42.\\n\\nWe must accept the trial court’s views of the Rule 32 testimony. “In the absence of\\n\\nclear and convincing evidence, we have no power on federal habeas review to\\n\\nrevisit the state court’s credibility determinations.” Bishop v. Warden, 726 F.3d\\n\\n\\n 38\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 39 of 73\\n\\n\\n1243, 1259 (11th Cir. 2013) (citing Marshall v. Lonberger, 459 U.S. 422, 434\\n\\n(1983)).\\n\\n The state court also noted that Wagoner testified that she was not available\\n\\nto testify at the trial, and it found that Seal’s testimony would have been\\n\\ncumulative to that of her husband. Thus, counsel’s failure to call those two\\n\\nwitnesses could not have been prejudicial to Jenkins. See id. at 141, 142. Neither of\\n\\nthese conclusions involves an unreasonable determination of the facts. We agree\\n\\nthat the Rule 32 testimony of Sherry Seal was substantially similar to the testimony\\n\\nof her husband Lonnie Seal that the jury heard during the penalty phase. Both\\n\\ntestified about meeting Jenkins in California, Jenkins’s assistance with their move\\n\\nto Alabama, and his residence with them in their Alabama home. Sherry added\\n\\nonly that Jenkins sometimes wet his bed. It was not unreasonable for the state court\\n\\nto conclude that Sherry’s testimony was cumulative of her husband’s, and “a\\n\\npetitioner cannot satisfy the prejudice prong of the Strickland test with evidence\\n\\nthat is merely cumulative of evidence already presented.” Rose v. McNeil, 634 F.3d\\n\\n1224, 1243 (11th Cir. 2011).\\n\\n Further, the school, medical, and juvenile court records that Jenkins\\n\\nintroduced in the Rule 32 proceeding have their own problems. As noted, they do\\n\\nnot document or even corroborate the severe level of abuse alleged by some of the\\n\\nwitnesses. Though these documents tell a sad story, they do not create a substantial\\n\\n\\n 39\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 40 of 73\\n\\n\\nlikelihood that the jury would have recommended life. Although they portray\\n\\nJenkins as a victim of truly unfortunate circumstances, they also show his unsavory\\n\\ndelinquent behavior. That delinquency included a history of stealing cars, and\\n\\ngiven that the jury likely inferred that Jenkins stole the red Mazda, new\\n\\ninformation about his history of theft may not have been especially mitigating.\\n\\nSimilarly, we agree with the state court that evidence of Jenkins’s compliant\\n\\nconduct in pretrial detention is “minimally mitigating” where there had been no\\n\\nevidence of jail misconduct as an aggravator. See Jenkins, 972 So. 2d at 149; cf.\\n\\nSkipper v. South Carolina, 476 U.S. 1, 5 & n.1 (1986) (good jail behavior is\\n\\n“potentially mitigating” especially “[w]here the prosecution specifically relies on a\\n\\nprediction of future dangerousness in asking for the death penalty”).\\n\\n Jenkins argues that allowing the jury to see his experience of mental,\\n\\nemotional, and cognitive difficulties, compounded by years of abuse and neglect\\n\\nand considered alongside his mild character and good jail behavior as an adult,\\n\\nwould have convinced them that Hogeland’s murder was an “aberration” that\\n\\ndeserved a punishment less severe than death. That interpretation of Jenkins’s life\\n\\nstory may be conceivable. But more than mere conceivability is required to\\n\\nestablish prejudice: “The likelihood of a different result must be substantial, not\\n\\njust conceivable.” Richter, 562 U.S. at 112 (emphasis added). We do not find the\\n\\nstate court’s decision that a different result was not substantially likely to be an\\n\\n\\n 40\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 41 of 73\\n\\n\\nunreasonable application of Strickland. As we have noted, we think it equally\\n\\nlikely that the jury could have rejected this alternate narrative out of hand in light\\n\\nof having just convicted him of the violent murder of a young mother. We also\\n\\nthink it possible that the jury could have viewed the murder as just one step in a\\n\\nlogical progression of degeneracy that began with behavioral problems at school,\\n\\ncontinued through the juvenile court system, and escalated past grand theft auto\\n\\ninto kidnapping and murder. Taken separately as well as together, the proffered\\n\\nevidence does not compel us to conclude that a verdict of life would have been\\n\\nsubstantially likely. Because fairminded jurists could disagree about the state\\n\\ncourt’s conclusion about Strickland prejudice, we affirm the denial of habeas relief\\n\\non this ground.\\n\\n B. Intellectual Disability\\n\\n Jenkins argues that the district court erred when it denied his claim that he is\\n\\nintellectually disabled and ineligible for the death penalty under Atkins v. Virginia,\\n\\n536 U.S. 304 (2002). The Supreme Court held in Atkins that executing an offender\\n\\nwho has mental retardation12 violates the Eighth Amendment’s ban on excessive\\n\\n\\n\\n\\n12\\n In 2002, the Atkins Court followed prevailing convention in referring to “mentally retarded”\\noffenders. We will use that now-disfavored term only when quoting or discussing older sources.\\nSince 2014, the Supreme Court has exclusively used the terms “intellectually disabled person”\\nand “person with intellectual disability.” See, e.g., Hall v. Florida, 572 U.S. 701, 704 (2014);\\nMoore v. Texas, 139 S. Ct. 666, 672 (2019).\\n 41\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 42 of 73\\n\\n\\npunishments. Id. at 321. It defined mental retardation by quoting the American\\n\\nAssociation on Mental Retardation:\\n\\n Mental retardation refers to substantial limitations in present\\n functioning. It is characterized by significantly subaverage intellectual\\n functioning, existing concurrently with related limitations in two or\\n more of the following applicable adaptive skill areas: communication,\\n self-care, home living, social skills, community use, self-direction,\\n health and safety, functional academics, leisure, and work. Mental\\n retardation manifests before age 18.\\n\\nId. at 308 n.3. Alabama courts have followed that tripartite definition, adopting\\n\\nwhat they consider the “broadest” definition of mental retardation for Atkins\\n\\npurposes:\\n\\n [A] defendant, to be considered mentally retarded, must have\\n significantly subaverage intellectual functioning (an IQ of 70 or\\n below), and significant or substantial deficits in adaptive behavior.\\n Additionally, these problems must have manifested themselves during\\n the developmental period (i.e., before the defendant reached age 18).\\n\\nEx parte Perkins, 851 So. 2d 453, 456 (Ala. 2002).\\n\\n The clinical evidence adduced during the Rule 32 hearing with respect to\\n\\nJenkins’s intellectual ability was that his IQ was 76, reflecting “borderline\\n\\nintellectual functioning.” The record also included the report of the Lunacy\\n\\nCommission13 which found that Jenkins was competent to stand trial, as well as\\n\\n\\n\\n13\\n Alabama law requires a capital defendant to be “committed to the Department of Mental\\nHealth and Mental Retardation for examination” if there is a question about his capacity to stand\\ntrial or if there will be a mental-disease-or-defect defense. Ala. Code § 15-16-22(a). “The\\nassigned mental health professional(s) shall examine the defendant with respect to determining\\nthe presence of any mental disease or defect which, if determined to be present, would affect the\\n 42\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 43 of 73\\n\\n\\nJenkins’s school and juvenile court records. The court also considered all the\\n\\ntestimony it had heard about Jenkins and reviewed the transcript of his murder trial\\n\\nbefore rejecting Jenkins’s claim of mental retardation.\\n\\n On appeal, after reciting the Perkins standard, the Alabama Court of\\n\\nCriminal Appeals explained that Jenkins had not established mental retardation:\\n\\n Dr. Kirkland testified that he performed psychological tests on Jenkins\\n and that Jenkins’s IQ was 76. There was evidence presented at\\n Jenkins’s trial indicating that Jenkins maintained relationships with\\n other individuals and that he had been employed by P.S. Edwards\\n Landscaping Company, Cotton Lowe 76 Service Station, and\\n Paramount Painting Company. The record fails to show that Jenkins\\n meets the most liberal view of mental retardation adopted by the\\n Alabama Supreme Court in Perkins.\\n\\nJenkins, 972 So. 2d at 155. In this habeas appeal, Jenkins raises arguments about\\n\\nall three components of the intellectual disability showing under Atkins and\\n\\nPerkins, which we address in turn. He also appeals the denial of an evidentiary\\n\\nhearing on this issue.\\n\\n 1. Intellectual Function\\n\\n The first component of intellectual disability under Perkins is “significantly\\n\\nsubaverage intellectual functioning (an IQ of 70 or below).” Perkins, 851 So. 2d at\\n\\n456. Jenkins argues that the use of a strict IQ cutoff of 70 was contrary to or an\\n\\nunreasonable application of Atkins.\\n\\n\\ncapacity of the defendant to proceed or continue to trial or which would affect the defendant’s\\ncriminal responsibility at the time of the commission of the crime.” Id.\\n 43\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 44 of 73\\n\\n\\n We disagree. The state court’s decision that Jenkins failed to establish\\n\\nmental retardation in this respect was a reasonable application of the record and of\\n\\nAtkins. As a threshold matter, we note that the state court did not say it was\\n\\napplying a strict cutoff of 70; it merely stated that Dr. Kirkland had measured\\n\\nJenkins’s IQ as 76. Jenkins, 972 So. 2d at 155. Under our deferential review, the\\n\\nquestion is whether the decision that Jenkins had not established the intellectual\\n\\nfunctioning aspect of intellectual disability was an unreasonable application of\\n\\nAtkins. The Supreme Court explained in Atkins that it was leaving “to the State[s]\\n\\nthe task of developing appropriate ways to enforce the constitutional restriction\\n\\nupon [their] execution of sentences.” Atkins, 536 U.S. at 317 (quoting Ford v.\\n\\nWainwright, 477 U.S. 399, 416–17 (1986)). In line with that invitation, our Circuit\\n\\nhas accepted Alabama’s judicial definition of intellectual disability as laid out in\\n\\nPerkins.14 See Thomas, 607 F.3d at 752–53. Jenkins objects, however, that the\\n\\nPerkins standard, which he characterizes as articulating a “strict cutoff” of 70,\\n\\nconflicts with more recent Supreme Court precedent that instead requires a\\n\\n“clinical” definition that goes beyond a single IQ score.\\n\\n In particular, Jenkins points to Hall v. Florida, in which the Supreme Court\\n\\nheld that when an offender’s IQ is near but greater than 70, courts must take into\\n\\n\\n\\n14\\n Notwithstanding the urging of the Alabama Supreme Court, see Perkins, 851 So. 2d at 455\\nn.1, the Alabama legislature has not enacted a statutory definition of intellectual disability.\\n 44\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 45 of 73\\n\\n\\naccount the IQ test’s standard error of measurement as well as non-IQ evidence of\\n\\nintellectual function. See Hall v. Florida, 572 U.S. 701, 723 (2014) (“This Court\\n\\nagrees with the medical experts that when a defendant’s IQ test score falls within\\n\\nthe test’s acknowledged and inherent margin of error, the defendant must be able\\n\\nto present additional evidence of intellectual disability, including testimony\\n\\nregarding adaptive deficits.”). But, as Jenkins concedes, our Circuit has\\n\\nspecifically held that Hall is not retroactive to cases on collateral review. See In re\\n\\nHenry, 757 F.3d 1151, 1161 (11th Cir. 2014) (explaining that Hall does not supply\\n\\na new substantive rule but rather “merely provides new procedures for ensuring\\n\\nthat States do not execute members of an already protected group”). We also note\\n\\nthat Alabama courts have found the Perkins standard to comport with Hall in that\\n\\nit does not prevent an offender from presenting the standard error of measurement\\n\\nor non-IQ evidence of intellectual disability. See Reeves v. State, 226 So. 3d 711,\\n\\n728–29 (Ala. Ct. Crim. App. 2016). Here, the state court did not unreasonably\\n\\napply Atkins when it analyzed the intellectual functioning component of\\n\\nintellectual disability.\\n\\n Jenkins also argues that the state court’s decision on this component in his\\n\\nRule 32 proceeding resulted from an unreasonable determination of the facts. He\\n\\nasserts that his IQ score of 76, when combined with the test’s standard error of\\n\\nmeasurement and considered alongside the Flynn effect (which we discuss below),\\n\\n\\n 45\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 46 of 73\\n\\n\\ndoes show that he has significantly subaverage intellectual function. For the\\n\\nreasons that follow, we disagree.\\n\\n Most fundamentally, neither of the clinicians who testified about Jenkins’s\\n\\nintellectual ability in the Rule 32 hearing opined that he was mentally retarded.\\n\\nThe closest thing Jenkins can point to is the testimony of Dr. Kirkland that his IQ\\n\\nis two standard deviations below the mean. But his entire testimony, in context, did\\n\\nnot establish intellectual disability. To the contrary: Dr. Kirkland testified that\\n\\nJenkins “scored in the range of borderline intellectual functioning which is\\n\\nbetween mild mental retardation and low average intellectual functioning.”\\n\\nFurthermore, Jenkins’s own expert Dr. Lisak fully reviewed Dr. Kirkland’s clinical\\n\\nfindings. He agreed with those findings, concluding that the IQ results show\\n\\n“borderline intelligence.” It is tremendously significant that no clinical assessment\\n\\nin the entire record of these proceedings has found that Jenkins has mental\\n\\nretardation or intellectual disability.\\n\\n Jenkins nonetheless argues that his IQ score of 76 establishes the intellectual\\n\\ncomponent of intellectual ability when the standard error of measurement and the\\n\\nFlynn effect are considered alongside that score. It was not unreasonable for the\\n\\nstate court to disagree. As we have discussed, Jenkins is not retroactively entitled\\n\\nto have the state courts take into account the margin of error of the IQ test. That\\n\\nstatistical margin of error—5 points in this case—takes into account characteristics\\n\\n\\n 46\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 47 of 73\\n\\n\\nof the test and variations due to chance in order to state how confident the tester is\\n\\nthat the measured result accurately states Jenkins’s true, unknowable IQ. A result\\n\\nof 76 with a standard error of measurement of 5 simply means that the tester is\\n\\n68% confident that Jenkins’s true IQ falls between 71 and 81, and 95% confident\\n\\nthat his true IQ falls between 66 and 86. See David H. Kaye & David A. Freedman,\\n\\nReference Guide on Statistics, in Reference Manual on Scientific Evidence 211,\\n\\n243–44 (3d ed. 2011); see also Hall, 572 U.S. at 738–39 (Alito, J., dissenting). Of\\n\\ncourse, Jenkins would like us to consider only the possibility that the test result\\n\\noverstated his IQ—but the standard error of measurement is a two-way street.\\n\\n“[T]he standard error of measurement is a bi-directional concept that does not\\n\\ncarry with it a presumption that an individual’s IQ falls to the bottom of his IQ\\n\\nrange.” Ledford v. Warden, Ga. Diag. & Clas. Prison, 818 F.3d 600, 641 (11th Cir.\\n\\n2016).\\n\\n Even if Jenkins were somehow entitled to subtract the 5-point margin of\\n\\nerror from his score and thereby lower his “true” IQ to 71, he still would not satisfy\\n\\nthe clinical intellectual component of intellectual disability. Thus, he sought in the\\n\\ndistrict court to introduce evidence about the Flynn effect in order to bring his IQ\\n\\ndown to 70 or below. As our Court has explained, the Flynn effect “recognizes the\\n\\nfact that IQ test scores have been increasing over time”:\\n\\n The Flynn effect acknowledges that as an intelligence test ages, or\\n moves farther from the date on which it was standardized, or normed,\\n 47\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 48 of 73\\n\\n\\n the mean score of the population as a whole on that assessment\\n instrument increases, thereby artificially inflating the IQ scores of\\n individual test subjects. Therefore, the IQ test scores must be\\n recalibrated to keep all test subjects on a level playing field.\\n\\nThomas, 607 F.3d at 753. We have said that “[a]n evaluator may also consider the\\n\\n‘Flynn effect’” in assessing an offender’s possible intellectual disability, id.\\n\\n(emphasis added), but neither we nor the Alabama Supreme Court nor the U.S.\\n\\nSupreme Court have said that a court must consider it in order to reasonably apply\\n\\nAtkins. See, e.g., Reeves, 226 So. 3d at 739 (“This Court has repeatedly held that a\\n\\ncircuit court is not required to accept, consider, or apply the ‘Flynn Effect’ in\\n\\ndetermining intellectual disability.”).\\n\\n More problematically, Jenkins did not raise the Flynn effect issue in the\\n\\nAlabama state courts. Jenkins asserts that he implicitly raised it as part and parcel\\n\\nof his Atkins claim during his Rule 32 appeal, but we do not agree that the\\n\\nAlabama courts had a full and fair opportunity to consider this issue, because it\\n\\nwas not explicitly presented to them. See Kelley v. Sec’y for the Dep’t of Corr., 377\\n\\nF.3d 1317, 1349–50 (11th Cir. 2004) (holding that one ineffective-assistance\\n\\nargument was unexhausted despite the petitioner’s general presentation of a\\n\\nStrickland claim in the state courts). Thus, under AEDPA, Jenkins is procedurally\\n\\nbarred from raising this unexhausted claim in his federal habeas petition. See 28\\n\\nU.S.C. § 2254(b)(1)(A), (c); Mason v. Allen, 605 F.3d 1114, 1119 (11th Cir. 2010).\\n\\nWe will allow a prisoner to overcome a procedural bar if he establishes cause for\\n 48\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 49 of 73\\n\\n\\nhis default and prejudice therefrom. See, e.g., Smith v. Jones, 256 F.3d 1135, 1138\\n\\n(11th Cir. 2001). But Jenkins cannot establish prejudice from being barred from\\n\\nasserting this claim now because the Supreme Court has not “clearly established”\\n\\nthat a court must consider the Flynn effect in assessing intellectual disability. Thus,\\n\\nunder AEDPA, it could not form the basis for a federal habeas claim. See 28\\n\\nU.S.C. § 2254(d)(1). We therefore conclude that the state court did not\\n\\nunreasonably determine the facts or unreasonably apply Atkins with respect to the\\n\\nintellectual component of intellectual disability.\\n\\n 2. Adaptive Function\\n\\n The second component of intellectual disability is “significant or substantial\\n\\ndeficits in adaptive behavior.” Perkins, 851 So. 2d at 456. Jenkins argues that the\\n\\nstate court, which mentioned only Jenkins’s successes in employment and social\\n\\nskills, see Jenkins, 972 So. 2d at 155, failed to consider his deficits. He adds that\\n\\neven the mention of employment success does not tell the whole story. Jenkins\\n\\ninsists that his jobs were menial, his employers took advantage of him, and he held\\n\\nthe jobs for only a few months each.\\n\\n But the state court’s terse discussion of Jenkins’s adaptive behavior did not\\n\\ninvolve an unreasonable determination of the facts or an unreasonable application\\n\\nof Atkins. To the contrary, its decision on this issue was reasonable in light of both\\n\\nthe record and Atkins.\\n\\n\\n 49\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 50 of 73\\n\\n\\n Atkins elaborated that the adaptive-skill component of intellectual disability\\n\\nrequires substantial present limitation in at least two of the following areas:\\n\\n“communication, self-care, home living, social skills, community use, self-\\n\\ndirection, health and safety, functional academics, leisure, and work.” Atkins, 536\\n\\nU.S. at 308 n.3 (quoting definition of American Association on Mental\\n\\nRetardation). The record supports the state court’s judgment that Jenkins does not\\n\\nhave substantial deficits in at least two areas. To be sure, we will assume that\\n\\nJenkins may be substantially limited in functional academics; Dr. Kirkland\\n\\nassessed his reading, spelling, and arithmetic skills at a third-grade level. But the\\n\\nrecord does not compel a finding of substantial present deficits in the areas of\\n\\ncommunication, self-care, home living, and work. Although Jenkins emphasizes\\n\\nhis transient employment, the record shows he was able to find work quickly, to\\n\\nwork hard to support the Seals’ household until Lonnie Seal found a job, and, later,\\n\\nto move into his own home. That record evidence is inconsistent with a substantial\\n\\ndeficit in the area of work.\\n\\n As the State observes, the facts of the crime also fail to show significant\\n\\ndeficits in the areas of communication, self-care, community use, and self-\\n\\ndirection. Jenkins was able to communicate well enough to solicit an alibi and to\\n\\nsell his car, including writing out a bill of sale. He then was able to arrange for his\\n\\nflight from Alabama, including purchasing a bus ticket and hitchhiking across the\\n\\n\\n 50\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 51 of 73\\n\\n\\ncountry. Indeed, throughout his trial and postconviction proceedings, the people\\n\\nwho knew Jenkins as an adult consistently described a man who did not have\\n\\nserious difficulties in communicating, forming relationships, working, and caring\\n\\nfor himself. No witness mentioned these kinds of difficulties in his or her\\n\\ntestimony about Jenkins. There was some evidence that Jenkins was “not\\n\\narticulate” and that his house was messy, but that testimony does not rise to the\\n\\nlevel of overwhelming evidence of substantial functional deficits that would\\n\\nsupport a conclusion that the state court unreasonably determined the facts.\\n\\n Of course, the record also contains evidence of Jenkins’s childhood\\n\\nacademic and social deficits, attributable at least in part to his parents’ neglect, but\\n\\nJenkins’s childhood is not directly relevant to our consideration of his present\\n\\nlimitations. Although we acknowledge that the Supreme Court has noted that such\\n\\n“risk factors” may support further exploration of the possibility of intellectual\\n\\ndisability, see Moore v. Texas, 137 S. Ct. 1039, 1051 (2017), the state court record\\n\\ncontains the results of that exploration. Overall, that record supports the state\\n\\ncourt’s conclusion that Jenkins does not have substantial deficits in adaptive\\n\\nbehaviors.\\n\\n 3. Juvenile Onset\\n\\n The third and final component of intellectual disability is manifestation of\\n\\nthe first and second components “before the defendant reached age 18.” Perkins,\\n\\n\\n 51\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 52 of 73\\n\\n\\n851 So. 2d at 456. Jenkins asserts, the State assumes, and we agree that the state\\n\\ncourt did not rule upon this issue, so there is nothing to which we must defer under\\n\\nAEDPA. See, e.g., Brumfield v. Cain, 135 S. Ct. 2269, 2282 (2015). Thus, we will\\n\\nreview the district court’s legal conclusions on this issue de novo, and its factual\\n\\nfindings for clear error. See, e.g., Grossman v. McDonough, 466 F.3d 1325, 1335\\n\\n(11th Cir. 2006). Jenkins argues that he has satisfied the age-of-onset component\\n\\nby showing “a wide range of evidence of deficits in both intellectual and adaptive\\n\\nfunctioning throughout Jenkins’s childhood.”\\n\\n In fact, the ample record of Jenkins’s childhood, as developed in the Rule 32\\n\\nproceeding, does not point to intellectual disability before age 18. School\\n\\nassessments at age 12 reported an IQ score of 83, an adaptive behavior rating of\\n\\n“low,” and noted overall an intellectual capacity of “average,” a diagnosis of\\n\\ndyslexia, and “poor functional skills.” The psychologist attributed Jenkins’s\\n\\ndifficulties to a combination of “[f]requent moves, changes in schools and family\\n\\ninstability” with “[l]ow ability coupled with a specific learning disability” and\\n\\n“[l]ack of responsibility for self control and poor use of time.” Although Jenkins\\n\\nfailed several grades of school, teachers additionally attributed his difficulties to\\n\\nhis behavioral problems and his high number of absences from school.\\n\\n Similarly, IQ tests ordered by the juvenile court at age 14 gave scores of 81\\n\\nand 86 on different tests, in the “[d]ull–[n]ormal” range and reflecting “a definite\\n\\n\\n 52\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 53 of 73\\n\\n\\nlearning disability.” These records all show a child with serious academic deficits\\n\\nand some intellectual and adaptive deficits, but they do not clearly show an\\n\\nintellectually disabled child. In particular, Jenkins’s childhood IQ scores of 83, 81,\\n\\nand 86 fail to point to intellectual disability or mental retardation. We are not left\\n\\nwith the definite and firm conviction that the district court erred when it found that\\n\\n“Jenkins’s [childhood] intelligence scores and adaptive functioning skills were\\n\\nattributable to his learning disability and other circumstances in his life, rather than\\n\\nmental retardation.”\\n\\n In sum, we agree with the district court that the decision of the Alabama\\n\\nCourt of Criminal Appeals that Jenkins does not have intellectual disability was\\n\\nnot contrary to or an unreasonable application of Atkins v. Virginia.\\n\\n 4. Evidentiary Hearing\\n\\n Finally, Jenkins argues that the district court erred when it denied his request\\n\\nfor an evidentiary hearing on his Atkins claim. If Jenkins diligently attempted to\\n\\npresent this claim in the state court, we review for abuse of discretion. Burgess v.\\n\\nComm’r, Ala. Dep’t of Corr., 723 F.3d 1308, 1320 (11th Cir. 2013). “The district\\n\\ncourt abuses its discretion where ‘such a hearing could enable an applicant to prove\\n\\nthe petition’s factual allegations, which, if true, would entitle the applicant to\\n\\nfederal habeas relief.’” Id. (quoting Schriro v. Landrigan, 550 U.S. 465, 474\\n\\n(2007)). If Jenkins did not diligently attempt to develop the factual basis for this\\n\\n\\n 53\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 54 of 73\\n\\n\\nclaim in the state court, absent certain circumstances, he is not entitled to a hearing\\n\\nand we will affirm the district court’s denial. 28 U.S.C. § 2254(e)(2).\\n\\n The parties strenuously dispute whether Jenkins diligently attempted to\\n\\npresent this issue in the state court. On the one hand, the State argues that Jenkins\\n\\nfailed to allege that he was mentally retarded even after the U.S. Supreme Court\\n\\ndecided Atkins in 2002. On the other hand, Jenkins asserts that he raised the\\n\\npossibility of mental retardation as early as his original Rule 32 petition in 1995.\\n\\n As we have noted, shortly after the Supreme Court decided Atkins, the\\n\\nAlabama Court of Criminal Appeals ordered supplemental briefing on the possible\\n\\nimpact of Atkins on Jenkins. Jenkins’s Rule 32 appeal was pending at that point,\\n\\nhaving been fully briefed and orally argued in the state appellate court. Having\\n\\nreviewed the state court records, we agree with the State that Jenkins did not\\n\\ndiligently attempt to present his Atkins claim in the state courts.\\n\\n “Diligence will require in the usual case that the prisoner, at a minimum,\\n\\nseek an evidentiary hearing in state court.” Williams, 529 U.S. at 437. The record\\n\\nclearly shows that Jenkins did not do so. In the initial supplemental brief on Atkins\\n\\nhe filed in 2002, Jenkins merely requested a future hearing after the Alabama\\n\\nlegislature’s enactment of a statute on executing the mentally retarded. Seventeen\\n\\nyears later, that contingency is still in the future. Only in his reply supplemental\\n\\nbrief did Jenkins then assert, for the first time ever, that he is mentally retarded.\\n\\n\\n 54\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 55 of 73\\n\\n\\n We acknowledge that Jenkins’s initial Rule 32 petition included in its\\n\\nallegations of ineffective assistance of counsel a list of unpresented mitigating\\n\\nevidence including his “developmental[] impair[ment]” and “learning disabilities,\\n\\nlow intelligence, poor comprehension and retarded socialization skills.” We\\n\\nconclude that these tangential Strickland references to his mental and social\\n\\nabilities do not amount to diligence. Thus, because Jenkins did not diligently\\n\\nattempt to develop the factual basis for his Atkins claim in the state courts, we\\n\\naffirm the district court’s denial of his request for an evidentiary hearing.\\n\\n Even if Jenkins had diligently attempted to present this claim in the state\\n\\ncourt, however, we conclude in the alternative that the district court did not abuse\\n\\nits discretion when it denied a hearing. We agree, in light of the ample evidence\\n\\nabout intellectual ability that Jenkins adduced in support of his ineffective-\\n\\nassistance claim, that such a hearing would not have enabled Jenkins to prove that\\n\\nhe is intellectually disabled. See Burgess, 723 F.3d at 1320.\\n\\n IV. CONCLUSION\\n\\n The denial of Jenkins’s petition for a writ of habeas corpus is\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 55\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 56 of 73\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435087/\", \"author_raw\": \"BRANCH, Circuit Judge:\"}, {\"author\": \"WILSON, Circuit Judge, dissenting\", \"type\": \"dissent\", \"text\": \"WILSON, Circuit Judge, dissenting:\\n\\n Mark Jenkins must prove that the Alabama state courts acted unreasonably\\n\\nin denying him post-conviction collateral relief. 28 U.S.C. § 2254(d)(1). Jenkins\\n\\nhas satisfied that burden as to his ineffective assistance of counsel claim. He has\\n\\nalso demonstrated that he is entitled to an evidentiary hearing regarding his Atkins\\n\\nv. Virginia claim. Because the majority came to the opposite conclusions, I\\n\\ndissent.\\n\\n I. Ineffective Assistance of Counsel at Penalty Phase\\n\\n Jenkins first asserts that his trial counsel’s failure to investigate Jenkins’\\n\\nabusive upbringing and to present any mitigating evidence at the penalty phase of\\n\\nhis trial constituted ineffective assistance of counsel. The majority rejects this\\n\\ncontention, concluding that the Alabama state courts did not act unreasonably in\\n\\ndenying Jenkins’ ineffective assistance of counsel claim on collateral review. I\\n\\ndisagree and maintain that Jenkins established both deficient performance and\\n\\nprejudice, entitling him to habeas relief. See Strickland v. Washington, 466 U.S.\\n\\n668, 687 (1984).\\n\\n A. Deficient Performance\\n\\n The Alabama state courts concluded that Jenkins’ counsel, Stan Downey,\\n\\ndid not perform deficiently at the penalty phase. The majority finds that\\n\\nconclusion reasonable primarily because Downey (1) was not obligated to\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 57 of 73\\n\\n\\ninvestigate Jenkins’ abusive childhood and (2) reasonably pursued a residual doubt\\n\\nstrategy at the penalty phase. But further review of the record requires me to part\\n\\ncompany with the majority’s reasoning and, ultimately, persuades me that the\\n\\nAlabama courts’ conclusion was indeed unreasonable. See 28 U.S.C. § 2254(d)(1)\\n\\n(providing that a writ of habeas corpus should not be granted to a state prisoner\\n\\nunless the state court adjudication involved “a decision that was contrary to, or\\n\\ninvolved an unreasonable application of, clearly established Federal law, as\\n\\ndetermined by the Supreme Court of the United States”).\\n\\n Jenkins offers numerous examples of how his counsel performed deficiently\\n\\nat the penalty phase of trial. First, counsel failed to thoroughly investigate Jenkins’\\n\\nabusive childhood—an error that constitutes deficient performance. In Williams v.\\n\\nAllen, we held that counsel’s failure to investigate mitigating evidence at the\\n\\npenalty stage constituted ineffective assistance of counsel because counsel drew\\n\\nupon only three sources of information—a psychologist’s report, a presentence\\n\\ninvestigation report, and an interview with the appellant’s mother. 542 F.3d 1326,\\n\\n1340 (11th Cir. 2008); see also Debruce v. Comm’r, Ala. Dep’t of Corr., 758 F.3d\\n\\n1263, 1270, 1273–74 (11th Cir. 2014) (finding counsel’s performance deficient\\n\\nwhere counsel interviewed the petitioner and his mother, but failed to contact\\n\\nexpert witnesses or other family members). In that case, we concluded that\\n\\n\\n\\n\\n 57\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 58 of 73\\n\\n\\ncounsel’s lack of investigation resulted in “an incomplete and misleading\\n\\nunderstanding of Williams’s life history.” Williams, 542 F.3d at 1340.\\n\\n Jenkins’ counsel did even less than the attorneys in Williams and Debruce.\\n\\nThat is, Jenkins’ counsel conducted no investigation. Downey did not consider\\n\\nJenkins’ poor school records, mention Jenkins’ juvenile history, produce\\n\\ndocumentary evidence that Jenkins suffered from severe depression, or pursue\\n\\nrecords demonstrating Jenkins’ significant deficits in cognitive functions. We have\\n\\npreviously found that a counsel’s failure to investigate these rudimentary examples\\n\\nof mitigating evidence constitutes ineffective assistance of counsel. See Daniel v.\\n\\nComm’r, Ala. Dep’t of Corr., 822 F.3d 1248 (2016) (reversing and remanding\\n\\ndistrict court’s dismissal of the ineffective assistance of counsel claim and noting\\n\\nhow counsel should have more thoroughly investigated Daniel’s history of\\n\\nphysical and sexual abuse, his poor school records, and evidence of a potential\\n\\nintellectual disability).\\n\\n Given the lack of preparation, it is not surprising that Downey failed to call\\n\\nany witnesses to testify at the penalty stage about Jenkins’ childhood, thus failing\\n\\nto give a complete depiction of Jenkins’ “excruciating life history.” Wiggins v.\\n\\nSmith, 539 U.S. 510, 537 (2003) (finding deficient performance where counsel did\\n\\nnot adequately investigate petitioner’s life history for mitigating evidence); see\\n\\nalso Debruce, 758 F.3d at 1270, 1273–74; Johnson v. Sec’y, Dep’t of Corr., 643\\n\\n\\n 58\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 59 of 73\\n\\n\\nF.3d 907, 932 (11th Cir. 2011) (“No reasonable attorney, after being told by his\\n\\nclient that he had an abusive upbringing, would fail to interview members of his\\n\\nfamily who were readily available and could corroborate or refute the allegations\\n\\nof abuse.”).1 If Downey had effectively investigated—or investigated at all—he\\n\\nwould have been aware of several available and informed witnesses. Ultimately,\\n\\nonly one witness testified at the penalty phase: Lonnie Seal.2 Downey met with\\n\\nSeal for only fifteen minutes before the penalty phase began. That fifteen minutes\\n\\nrepresents the extent of Downey’s preparation for the penalty phase—a wholly\\n\\ninadequate amount of time under Supreme Court precedent. See Williams v.\\n\\nTaylor, 529 U.S. 362, 395 (2000) (finding mitigation investigation deficient where\\n\\n“counsel did not begin to prepare for that phase of the proceeding until a week\\n\\nbefore the trial”).\\n\\n Despite ample evidence of deficient performance, the Alabama Court of\\n\\nCriminal Appeals (CCA) concluded that Downey performed reasonably. The\\n\\nCCA’s decision rested on three explanations: (1) because Downey did not testify,\\n\\ncourts must presume he acted reasonably; (2) based on the record, Downey played\\n\\n\\n1\\n The majority goes to great pains to maintain that Downey did not know about Jenkins’ abusive\\nchildhood. This assertion is directly refuted by the record. Douglas Scofield, Jenkins’ lead\\nattorney, testified that Jenkins described his traumatic upbringing at meetings at which Downey\\nwas present.\\n2\\n It is unclear whether Downey contacted Seal and requested that he testify during the penalty\\nphase of trial. Given that Downey urged Scofield to conduct Seal’s direct examination because\\nScofield was better acquainted with Seal, it is likely that Scofield made the arrangements.\\n 59\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 60 of 73\\n\\n\\nan active role investigating and preparing for the penalty phase of trial; and (3)\\n\\nDowney’s decision to pursue a residual doubt strategy was reasonable.3 Because\\n\\neach of these findings constituted an unreasonable determination of the facts or an\\n\\nunreasonable application of clearly established federal law, Jenkins is entitled to\\n\\nhabeas relief. See 28 U.S.C. § 2254(d)(1).\\n\\n The CCA’s first rationale—presuming Downey acted reasonably because\\n\\nDowney did not testify at the Rule 32 hearing—was based on both an unreasonable\\n\\ndetermination of facts and an unreasonable application of clearly established law.\\n\\nThe CCA stated that “the record is virtually silent as to what actions were or were\\n\\nnot taken or what was or was not done by Mr. Downey at trial and why” and\\n\\ntherefore assumed that Downey’s actions “could have been reasonable and\\n\\nstrategic under the circumstances.” The majority likewise assumes Downey acted\\n\\nreasonably. This conclusion is based, in large part, on Downey’s failure to testify\\n\\nat the Rule 32 hearing. According to the majority, “[t]he problem for Jenkins is\\n\\nthat, like the state courts, this Court also does not know what Downey did or why.”\\n\\n But the record clearly illustrates Downey’s failure to investigate and overall\\n\\nlack of preparation. First, Downey’s fee declarations—which the majority\\n\\nconcedes were “detailed”—showed no investigation of mitigating circumstances,\\n\\nand no preparation for the penalty phase. Importantly, Downey’s fee declarations\\n\\n\\n3\\n The majority opinion endorsed the CCA’s first and third explanations.\\n 60\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 61 of 73\\n\\n\\nwere exceptionally detailed, recounting his work in five-minute increments. The\\n\\nsingle reference to the “mitigating” phase of trial in the declarations concerned a\\n\\nmeeting with Scofield that occurred only a week before trial. Other than that\\n\\nvague entry, no notation indicates that Downey did anything to prepare for the\\n\\npenalty phase of trial during the nearly two years he represented Jenkins.\\n\\n Second, and more persuasively, Douglas Scofield, Jenkins’ lead attorney,\\n\\nexplicitly testified that Downey told him that Downey had not investigated\\n\\npotential mitigating evidence. Both the Alabama courts and the majority ignore\\n\\nthat fact and instead focus on Scofield’s testimony that, “[a]part from what he has\\n\\ntold me, I don’t know what he has done.” This statement, to which the majority\\n\\nassigns great weight, is hardly illuminating. Of course Scofield knows only what\\n\\nDowney told him. And what Downey told him was this: Downey did not\\n\\ninvestigate potentially mitigating evidence before the penalty phase.\\n\\n In the face of this direct evidence of deficient performance, the Alabama\\n\\ncourts and the majority pivot to Downey’s failure to testify at the Rule 32 hearing.\\n\\nWithout his testimony, the majority reasons, we must assume that Downey’s\\n\\ndecision not to investigate was strategic and therefore his representation\\n\\nreasonable. I disagree, and I have deep reservations about the practical effects of\\n\\nthe majority’s understanding of the law. The majority suggests (if not holds) that\\n\\nthe absence of testimony from an allegedly deficient attorney per se means that the\\n\\n\\n 61\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 62 of 73\\n\\n\\nattorney’s actions were reasonable. It is true that “counsel is strongly presumed to\\n\\nhave rendered adequate assistance,” Strickland, 466 U.S. at 690. But a petitioner\\n\\nmay overcome that burden with evidence of inadequate assistance. And that\\n\\nevidence may take various forms. To satisfy the burden, it is not necessary for the\\n\\nallegedly deficient counsel to testify and admit wrongdoing. Nor is it necessary for\\n\\nthe record to reflect counsel’s “thoughts and intentions as he prepared for the\\n\\npenalty phase,” as the majority seems to require. When evaluating whether a\\n\\ncounsel’s performance was deficient, we ask whether his performance “fell below\\n\\nan objective standard of reasonableness.” Strickland, 466 U.S. 687–88. We do not\\n\\ncontemplate what counsel thought or intended. We consider whether his\\n\\nperformance was objectively unreasonable. See id. Because the Alabama courts\\n\\ndid not apply the proper standard, their conclusion was an unreasonable application\\n\\nof clearly established law. See id.\\n\\n The CCA’s second reason for denying Jenkins relief was also based on an\\n\\nunreasonable application of clearly established federal law. The CCA concluded\\n\\nthat Downey had adequately investigated and prepared for the penalty phase of\\n\\ntrial, emphasizing that Downey met with Jenkins on several occasions and spoke\\n\\nwith Jenkins’ grandmother. Based on these findings, the CCA determined that\\n\\nDowney provided adequate assistance of counsel during the penalty phase. This is\\n\\nan unreasonable application of clearly established law that “[a]n attorney has a\\n\\n\\n 62\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 63 of 73\\n\\n\\nduty to conduct a reasonable investigation, including an investigation of the\\n\\ndefendant’s background, for possible mitigating evidence.” Porter v. Singletary,\\n\\n14 F.3d 554, 557 (11th Cir. 1994).\\n\\n As previously discussed, it is established in the record that Downey knew of\\n\\nJenkins’ horrible upbringing yet still failed to investigate potentially mitigating\\n\\nevidence. “A reasonable investigation . . . should have included, at a minimum,\\n\\ninterviewing other family members who could corroborate the evidence of abuse\\n\\nand speak to the resulting impact on [Jenkins].” Williams, 542 F.3d at 1340. But\\n\\nDowney did not conduct a single interview—of a family member or otherwise—in\\n\\npreparation of the penalty phase.4 Because Downey did not conduct an\\n\\ninvestigation, let alone a reasonable investigation, the CCA’s conclusion that\\n\\nDowney adequately investigated potentially mitigating evidence was an\\n\\nunreasonable application of clearly established law.\\n\\n The CCA’s third rationale for its decision—that Jenkins’ counsel reasonably\\n\\nand strategically decided to focus on residual doubt—was both an unreasonable\\n\\napplication of clearly established law and an unreasonable determination of the\\n\\nfacts. In Williams v. Taylor, the Supreme Court held that otherwise reasonable\\n\\ntrial strategies cannot be justified unless counsel first conducted a constitutionally\\n\\n\\n4\\n Downey’s fee declaration reflects one telephone call he had with Jenkins’ grandmother, Doris\\nWagoner. But that call occurred 18 months before trial and, according to Wagoner, concerned a\\nrequest for money.\\n 63\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 64 of 73\\n\\n\\nadequate investigation. 529 U.S. at 395–96. And here, Downey conducted no\\n\\nsuch investigation. It cannot be said, therefore, that Downey’s decision—if one\\n\\ncan even call it that—to pursue a residual doubt strategy was strategically\\n\\nreasonable. See id. The CCA’s conclusion to the contrary was an unreasonable\\n\\napplication of clearly established law.\\n\\n The conclusion was likewise an unreasonable determination of the facts. On\\n\\ndirect appeal, the Alabama CCA noted that “[t]he evidence overwhelmingly\\n\\npointed to the guilt of the appellant.” Jenkins v. State, 627 So. 2d 1054 (Ala. Crim.\\n\\nApp. Feb. 27, 1992). Then, on appeal from Jenkins’ Rule 32 petition, the CCA\\n\\nagain acknowledged how “[t]he evidence establishing Jenkins’s guilt was\\n\\noverwhelming.” Jenkins v. State, 972 So. 2d 111, 157 (Ala. Crim. App. Feb. 27,\\n\\n2004); see also id. at 156 (“The amount of evidence incriminating Jenkins . . . was\\n\\noverwhelming.”). Despite the “overwhelming” evidence of Jenkins’ guilt, the\\n\\nCCA concluded that Jenkins’ counsel acted reasonably in presenting a residual\\n\\ndoubt defense during the penalty phase of trial. But no reasonable attorney would\\n\\nemploy a residual doubt strategy in the face of such overwhelming evidence. See\\n\\nJohnson, 643 F.3d at 932–33. The Alabama courts’ contrary conclusion, therefore,\\n\\nwas an unreasonable determination of the facts.\\n\\n\\n\\n\\n 64\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 65 of 73\\n\\n\\n B. Prejudice\\n\\n Jenkins’ Rule 32 petition also pleaded sufficient facts to show that his\\n\\ncounsel’s deficient performance prejudiced his defense, satisfying the second\\n\\nStrickland prong. To establish prejudice, a petitioner must show a “reasonable\\n\\nprobability that, but for counsel’s unprofessional errors, the result of the\\n\\nproceeding would have been different.” Strickland, 466 U.S. at 694. Here, the\\n\\njury voted in favor of the death penalty by the slimmest margin permitted under\\n\\nAlabama law: ten votes in favor and two votes against. Jenkins need only\\n\\nestablish, then, “a reasonable probability that at least one juror would have struck\\n\\na different balance.” Wiggins, 539 U.S. at 537 (emphasis added). In measuring\\n\\nthat probability, we “evaluate the totality of the available mitigation evidence—\\n\\nboth that adduced at trial, and the evidence adduced in the habeas proceeding in\\n\\nreweighing it against the evidence in aggravation.” Williams, 529 U.S. at 397.\\n\\n The testimony elicited at the Rule 32 hearing was horrendous. Four family\\n\\nmembers offered detailed, consistent testimony about Jenkins’ traumatic\\n\\nchildhood. The majority describes that testimony, highlighting some of the most\\n\\nappalling facts elicited at the Rule 32 hearing. For example, Jenkins’ stepfather\\n\\nwould beat Jenkins on a semi-daily basis, would lock him in a room for hours\\n\\nwithout food, and even made Jenkins eat his own feces. In addition to the\\n\\nmitigating facts detailed by the majority, the following came to light during the\\n\\n\\n 65\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 66 of 73\\n\\n\\nRule 32 hearing. Jenkins’ mother had an affair while her husband, Jenkins’\\n\\nstepfather, was in prison. When Jenkins’ stepfather was released, he took out his\\n\\nfrustrations about the affair on Jenkins. Because Jenkins’ biological father was\\n\\nJenkins’ stepfather referred to Jenkins in demeaning names like “Mexican trash,”\\n\\nand “Puerto Rican puke.” The stepfather also called him “piece of shit” and\\n\\n“bastard.”\\n\\n The Rule 32 testimony also detailed the alcohol substance abuse, drug\\n\\naddiction, and incessant domestic violence that surrounded Jenkins’ childhood.\\n\\nAnd, most shockingly, the testimony described how Jenkins’ grandfather sexually\\n\\nabused Jenkins when Jenkins was just four years old. As a result of that sexual\\n\\nabuse, Jenkins, suffered bladder and bowel movement issues. Jenkins’\\n\\nincontinence became a trigger for his stepfather: when Jenkins had an accident, his\\n\\nstepfather would beat him or subject him to severe degradation. For example,\\n\\nJenkins’ stepfather would hang Jenkins’ soiled sheets in the front yard for all the\\n\\nneighbors to see, make Jenkins wear his soiled clothes on his head, and, most\\n\\ndisturbingly, make Jenkins eat his own feces. The only way to escape the abuse\\n\\nwas by running away. So, by age twelve, Jenkins lived on the streets.5\\n\\n\\n\\n5\\n The CCA found parts of the family members’ testimony not credible, though it did not\\nexplicitly discredit testimony describing some of the worst abuse. Given the grotesque nature of\\nthe family members’ testimony, and the fact that their testimony was consistent, there is\\nreasonable probability that at least one more juror would have been compelled to vote against a\\ndeath sentence.\\n 66\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 67 of 73\\n\\n\\n Jenkins also introduced medical, school, and juvenile court records from his\\n\\nchildhood, which shed further light on his difficult childhood and intellectual\\n\\ndeficiencies. Juvenile records show Jenkins as “an emotionally disturbed child”\\n\\nwith “psychological problems” and expressed “grave concerns for the emotional\\n\\nwill being of this adolescent”. Moreover, two psychologists testified about\\n\\nJenkins’ history of psychological trauma stemming from his childhood abuse—\\n\\nweighty evidence that is not even considered in the majority’s discussion. Finally,\\n\\ntwo prison guards testified that Jenkins was the ideal prisoner during pretrial\\n\\ndetention. One of the guards even described Jenkins as the best inmate he had\\n\\nevery supervised.\\n\\n Given this compelling mitigating evidence, it is not difficult to conclude that\\n\\nJenkins established “a reasonable probability that at least one juror would have\\n\\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537; see\\n\\nalso Daniel, 822 F.3d at 1276 (listing several factors that support a prejudice\\n\\nfinding, including (1) evidence of impaired intellectual functioning; (2) evidence of\\n\\nchildhood sexual abuse; and (3) the fact that the jury voted only 10–2 in\\n\\nrecommending a death sentence).\\n\\n I. Evidentiary Hearing\\n\\n Jenkins also argues that he is entitled to habeas relief under Atkins v.\\n\\nVirginia, 563 U.S. 304 (2002). He contends that he satisfies the three requirements\\n\\n\\n 67\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 68 of 73\\n\\n\\nnecessary to establish intellectual disability—(1) subaverage intellectual\\n\\nfunctioning; (2) deficits in adaptive behavior; and (3) onset before the age of 18,\\n\\nsee Ex parte Perkins, 851 So. 2d 453, 456 (Ala. 2002)—thus, exempting him from\\n\\nimplementation of the death penalty. Alternatively, Jenkins argues that we should\\n\\nremand for an evidentiary hearing on the Atkins issue.\\n\\n The majority dismisses each argument. First, it concludes that the CCA did\\n\\nnot act unreasonably in concluding that Jenkins failed to satisfy any of the Perkins\\n\\nprongs. Second, it determines that Jenkins is not entitled to an evidentiary hearing\\n\\nbecause he “did not diligently attempt to develop the factual basis for this claim in\\n\\nthe state court.” I disagree.\\n\\n First, it is not possible to undergo a substantive Atkins analysis based on the\\n\\nrecord before us. For one, the CCA’s evaluation of Jenkins’ Atkins claim was a\\n\\nmere three sentences. And, more significantly, Jenkins never had an opportunity to\\n\\npresent evidence on his Atkins claim. To deny Jenkins’ claim, therefore, the\\n\\nmajority relies on evidence from unrelated parts of the record. Such an approach is\\n\\nimproper because, “[p]rior to Atkins, [the petitioner] could not have been expected\\n\\nto necessarily present evidence sufficient to support an Atkins claim because such\\n\\nevidence constituted a double edged sword.” Burgess v. Comm’r, Ala. Dep’t of\\n\\nCorr., 723 F.3d 1308, 1320 (11th Cir. 2013) (internal citation omitted). The\\n\\nevidence was a double-edged sword, we explained, because prior to Atkins,\\n\\n\\n 68\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 69 of 73\\n\\n\\nevidence of intellectual disability could suggest that a defendant would be\\n\\ndangerous in the future; therefore, “a defendant could reasonably decide not to\\n\\nhighlight his mental retardation.” Id. at 1318. For that reason, courts should not\\n\\nrely on pre-Atkins evidence—evidence that was presented in a wholly distinct\\n\\ncontext—to support a finding that a petitioner is not intellectually disabled. So,\\n\\nhere, instead of attempting to make a medical diagnosis based on an insufficient\\n\\nrecord, we should remand for an evidentiary hearing. The majority refuses to do\\n\\nso, reasoning that “Jenkins did not diligently attempt to present his Atkins claim in\\n\\nthe state courts.” I disagree.\\n\\n “[I]f the petitioner was diligent in developing the record in the state habeas\\n\\nproceedings, ‘a federal court may grant an evidentiary hearing without further\\n\\nregard for the provisions of § 2254(3)(2).” Burgess, 723 F.3d at 1320 (citation\\n\\nomitted). “‘Diligence [for purposes of § 2254(d)] depends on whether the prisoner\\n\\nmade a reasonable attempt, in light of the information available at the time, to\\n\\ninvestigate and pursue claims in state court,’ and ‘will require in the usual case that\\n\\nprisoner, at a minimum, seek an evidentiary hearing in state court.’” Id. (quoting\\n\\nWilliams, 529 U.S. at 435 (emphasis added)).\\n\\n The majority concludes that Jenkins failed to diligently pursue his Atkins\\n\\nclaim in state court. But again, based on binding precedent and the record, I must\\n\\npart company with the majority. In conducting its analysis, the majority does not\\n\\n\\n 69\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 70 of 73\\n\\n\\nadequately consider whether Jenkins “made a reasonable attempt, in light of the\\n\\ncircumstances available at the time, to investigate and pursue claims in state\\n\\ncourt.” Id. (emphasis added) (quoting Williams, 529 U.S. at 435). Specifically, the\\n\\nmajority did not consider the unusual procedural posture of Jenkins’ claim.\\n\\n The Supreme Court issued Atkins v. Virginia while Jenkins’ petition for\\n\\npost-conviction collateral relief was pending before the CCA. The CCA, therefore,\\n\\nrequested supplemental briefing on the effect, if any, Atkins had on Jenkins’\\n\\npetition. At that point, neither the Alabama Supreme Court nor the Alabama\\n\\nlegislature had interpreted the groundbreaking case, making Jenkins’ counsel’s\\n\\nnext step unclear. Jenkins’ counsel nonetheless made a reasonable attempt to\\n\\npursue Jenkins’ Atkins claim—he requested that the CCA “vacate his death\\n\\nsentence and remand the case back to the lower court for further proceedings.”\\n\\nThe request appears even more reasonable when considered alongside procedurally\\n\\nanalogous cases decided around the same time.\\n\\n At the time Jenkins’ counsel was preparing the supplemental briefing, there\\n\\nhad been three other post-conviction petitions for relief that shared Jenkins’\\n\\nunusual procedural posture. See Clemons v. State, 55 So. 3d 314 (Ala. Crim. App.\\n\\nAug. 29, 2003); Wood v. State, 891 So. 2d 398 (Ala. Crim. App. Apr. 25, 2003);\\n\\nTarver v. State, 940 So. 2d 312 (Ala. Crim. App. Feb. 27, 2004). These three cases\\n\\nare similar to Jenkins’ petition in that (1) the petitioners were sentenced to death,\\n\\n\\n 70\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 71 of 73\\n\\n\\n(2) the petitioners filed Rule 32 petitions, (3) their Rule 32 petitions were\\n\\ndismissed at the trial court level, (4) while their appeals before the CCA were\\n\\npending, Atkins was decided, and (5) given the Atkins decision, each petitioner\\n\\nsubmitted supplemental briefing. In their supplemental briefs, Clemons, Wood,\\n\\nand Tarver each argued that he was intellectually disabled under Atkins and thus it\\n\\nwould be unconstitutional to execute him. None of the petitioners asked for an\\n\\nevidentiary hearing on their Atkins claim or requested a remand for further\\n\\nproceedings. Despite the lack of request for a hearing, the CCA remanded for an\\n\\nevidentiary hearing in each case.\\n\\n With these cases in mind, it would have been reasonable for Jenkins to think\\n\\nthat broadly stating his Atkins claim would constitute “diligent” pursuit of his\\n\\nclaim. Nonetheless, Jenkins did more than these other petitioners: he specifically\\n\\nrequested that the CCA “vacate his death sentence and remand the case back to the\\n\\nlower court for further proceedings.” Given “the information available at the\\n\\ntime,” Jenkins’ made a “reasonable attempt” to pursue his claims.\\n\\n I also cannot come to terms with the majority’s ultimate conclusion. The\\n\\nmajority determines that Jenkins did not act diligently in pursuing his Atkins claim\\n\\nbecause “[t]he record clearly shows that Jenkins did not [seek an evidentiary\\n\\nhearing].” But Jenkins did ask the CCA to “remand his case for further\\n\\nproceedings.” And, in Alabama, post-conviction relief petitioners with factually\\n\\n\\n 71\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 72 of 73\\n\\n\\nsufficient claims are entitled to an evidentiary hearing. Ala. R. Crim. P. 32.9. So,\\n\\nif the CCA had granted Jenkins’ request to remand “for further proceedings,”\\n\\nJenkins would have received an evidentiary hearing. Jenkins’ request that the\\n\\nCCA “remand for further proceedings,” then, was effectively equivalent to a\\n\\nrequest for an evidentiary hearing. The majority’s superficial conclusion to the\\n\\ncontrary misses the point.\\n\\n II. Conclusion\\n\\n Jenkins’ Rule 32 evidentiary hearing spanned three days. Four family\\n\\nmembers testified about his horrific upbringing. Two psychiatrists testified about\\n\\nhow that upbringing affected Jenkins as an adult. Two prison guards testified\\n\\nabout Jenkins’ exemplary behavior while in pretrial detention. And one friend\\n\\ntestified about Jenkins’ character. Evidence introduced also documented Jenkins’\\n\\nacademic failings and his juvenile record. In short, Jenkins presented an\\n\\nabundance of compelling mitigating evidence. But the jury never heard any of that\\n\\nevidence. Downey failed to introduce it at the penalty phase. In fact, Downey\\n\\nfailed to investigate any mitigating evidence whatsoever. Downey’s failures\\n\\ncannot be attributed to strategic decisions. Nor can they be called reasonable. His\\n\\nperformance can only be described as deficient. And that deficient performance\\n\\nwas surely prejudicial—there is more than a reasonable probability that the\\n\\nhorrifying and detailed Rule 32 testimony would have persuaded one juror to vote\\n\\n\\n 72\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 73 of 73\\n\\n\\nagainst sentencing Jenkins to death. Jenkins is therefore entitled to habeas relief\\n\\non his ineffective assistance of counsel claim.\\n\\n Jenkins is also entitled to a hearing on his Atkins claim. Though Jenkins\\n\\npresented ample mitigating evidence at the Rule 32 hearing, he never had an\\n\\nopportunity to present evidence concerning (1) his intellectual functioning, (2) his\\n\\nadaptive deficits, or (3) whether his potential intellectual disability manifested\\n\\nbefore Jenkins reached 18—the three requirements necessary to establish\\n\\nintellectual disability and thus exempt an individual from a death sentence. The\\n\\nCCA nonetheless decided to evaluate Jenkins’ Atkins claim, apparently believing\\n\\nthat the record, which included no direct evidence concerning whether Jenkins was\\n\\nintellectually disabled, was sufficient to resolve the complex medical issue. In just\\n\\nthree sentences, the CCA arrived at a diagnosis: not intellectually disabled.\\n\\n Despite the insufficient factual record, the majority declines to remand with\\n\\ninstructions to conduct an evidentiary hearing—what I view as the proper\\n\\nrecourse—because Jenkins did not use the phrase “evidentiary hearing” when\\n\\nrequesting further proceedings on his Atkins claim in the Alabama courts. Such a\\n\\nrigid understanding of the standard should not bar a death row petitioner from\\n\\nbringing a constitutional, and potentially lifesaving, claim. I therefore dissent.\\n\\n\\n\\n\\n 73\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435087/\", \"author_raw\": \"WILSON, Circuit Judge, dissenting\"}]}"},"judge1":{"kind":"string","value":"TJOFLAT"},"judge2":{"kind":"string","value":"WILSON"},"judge3":{"kind":"string","value":"BRANCH"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"TJOFLAT\": \", Circuit\", \"WILSON\": \", Circuit\", \"BRANCH\": \", Circuit\"}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":1,"string":"1"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4657834/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Mark Jenkins v. Commissioner, Alabama Department of Corrections\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-30\nOpinion type: majority\nHas dissenting opinion: Yes\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nCase: 17-12524 Date Filed: 08/30/2019 Page: 1 of 73\n\n\n [PUBLISH]\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n\n\n No. 17-12524\n\n\n\n D.C. Docket No. 4:08-cv-00869-VEH\n\nMARK ALLEN JENKINS,\n\n Petitioner - Appellant,\n\n versus\n\n\n\nCOMMISSIONER, ALABAMA DEPARTMENT OF CORRECTIONS,\n\n Respondent - Appellee.\n\n\n\n Appeal from the United States District Court\n for the Northern District of Alabama\n\n\n (August 30, 2019)\n\nBefore TJOFLAT, WILSON, and BRANCH, Circuit Judges.\n\nBRANCH, Circuit Judge:\n\n Mark Allen Jenkins, an Alabama prisoner sentenced to death for the 1989\n\nmurder of Tammy Ruth Hogeland, appeals the district court’s denial of his petition\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 2 of 73\n\n\nfor a writ of habeas corpus, 28 U.S.C. § 2254. Before us are Jenkins’s arguments\n\nthat he received ineffective assistance of counsel during the penalty phase of his\n\ntrial and that he is intellectually disabled and therefore ineligible for the death\n\npenalty. After careful consideration, and with the benefit of oral argument, we\n\naffirm the denial of Jenkins’s habeas petition on both grounds.\n\n I. BACKGROUND\n\n A. The Crime and Arrest\n\n We summarize the following background narrative from opinions of the\n\nAlabama Court of Criminal Appeals and our own review of the record. See Jenkins\n\nv. State, 627 So. 2d 1034, 1037–40 (Ala. Ct. Crim. App. 1992); Jenkins v. State,\n\n972 So. 2d 111, 119–20 (Ala. Ct. Crim. App. 2004). The events leading up to the\n\nmurder began the evening of April 17, 1989. Jenkins was at the home of his\n\nacquaintance Christine Nicholas. He had met her several months earlier at the\n\nOmelet Shoppe restaurant where she worked. Jenkins was very intoxicated and\n\ntried to seduce her. When she resisted, Jenkins got “real mad” and asked her\n\nseveral times what she would do if someone came up from behind her and grabbed\n\nher.\n\n At around 1 a.m., Jenkins and Nicholas drove to the Riverchase Omelet\n\nShoppe. Jenkins went inside and talked with one of the waitresses, Frieda Vines.\n\nThe manager, Douglas Thrash, recognized Jenkins as a regular customer who knew\n\n\n 2\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 3 of 73\n\n\nall of the waitresses. Thrash overheard part of Jenkins’s conversation with Vines\n\nand heard mention of the Omelet Shoppe location near the Birmingham Airport.\n\nJenkins and Nicholas then returned to Nicholas’s home. Around 2 a.m., Nicholas’s\n\nmother asked Jenkins to leave. He did so, falling down some steps and ramming\n\nhis car into another vehicle in the process.\n\n Meanwhile, 23-year-old Tammy Hogeland was making her way to work at\n\nthe Omelet Shoppe. Her sister Wendy, along with Tammy’s young son, picked\n\nTammy up from her college classes and drove her to the Riverchase Omelet\n\nShoppe a little after 9:30 p.m. Sometime after 10:00, Tammy was sent to work at\n\nthe airport Omelet Shoppe on Tenth Avenue, since that location was unexpectedly\n\nshorthanded. Tammy was wearing her watch, a necklace with the words “special\n\nsister,” her class ring, and her diamond engagement ring. Tammy was working as a\n\ncook that night and was wearing a blue apron, black pants, a white shirt, and white\n\nshoes. She and Sarah Harris were the only employees working at the airport\n\nOmelet Shoppe that night.\n\n At about 2 a.m. on April 18, Harris saw Jenkins—whom she did not know—\n\ndrive up in a red sports car. His arrival was memorable to Harris because the car\n\nnearly jumped the curb and crashed through the restaurant’s glass wall. Jenkins got\n\nout of the car and came into the restaurant, appearing to be intoxicated. He walked\n\nover to Hogeland and began talking to her. Harris later saw Jenkins and Hogeland\n\n\n 3\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 4 of 73\n\n\ndrive away together in the red sports car. That was the last time anyone who knew\n\nHogeland ever saw her. Hogeland left behind her cigarettes, lighter, purse, and\n\npaycheck. She left without telling anyone, which she had never done before.\n\nAlthough Harris saw them drive off, she could not later say whether Hogeland left\n\nwith Jenkins willingly or was instead abducted.\n\n At around 5 a.m. that morning, Bobby and Geraldine Coe had stopped to\n\nbuy gas at a Chevron station on I-59 northeast of Birmingham when they saw\n\nJenkins drive up in the red sports car. The Coes noticed a female who appeared to\n\nbe “passed out” in the front passenger seat, but they could not say whether she was\n\nalive or dead. While Bobby Coe was pumping gas, Jenkins asked him for some\n\ncigarettes and said, “Looks like it’s been a long night and it looks like it’s going to\n\nbe a long day.” Jenkins then said “God bless you” before asking directions to I-\n\n459. Coe gave him directions and got back in his car. As Coe pulled out onto the\n\ninterstate, he saw Jenkins follow him in the red sports car. Coe then saw the car\n\nflash its lights, slow down, and pull to the side of the road between mile markers\n\n151 and 152.\n\n At 8 a.m., Wendy Hogeland learned that Tammy was not at the restaurant,\n\nand their mother called the police. Meanwhile, Jenkins went to the home of Steve\n\nMusser, who noticed that he was wearing the same clothes as the day before.\n\nJenkins told Musser that his car had been stolen the night before and asked him if\n\n\n 4\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 5 of 73\n\n\nhe would say that he had been with him all night. Musser refused. Christine\n\nNicholas also saw Jenkins in a grocery store that morning. He was looking at a\n\nnewspaper, making a phone call, and attempting to sell his Buick. Nicholas loaned\n\nhim $4 for gas. At around 10 o’clock that morning, Jenkins sold his car to Michael\n\nBrooks, a mechanic at a local Chevron station. Jenkins had explained that he\n\nneeded the money so he could visit his sick mother in California. Another\n\nmechanic at the service station drove Jenkins to the Greyhound bus station later\n\nthat morning.\n\n Jenkins awoke the next day on the Greyhound bus in Houston and was\n\nejected from the bus because his fare was used up. He then hitchhiked from\n\nHouston to Phoenix, then to San Diego and Los Angeles. Jenkins was first\n\nidentified as a suspect in Hogeland’s disappearance by Officer Mike Weems of the\n\nHoover Police Department on April 19. Weems ate dinner at the Riverchase\n\nOmelet Shoppe just about every day and had talked with Hogeland just as often.\n\nHe also knew Jenkins from the restaurant. Weems learned of Hogeland’s\n\ndisappearance from the other waitresses. Remembering how Jenkins often talked to\n\nHogeland and passed her notes,1 he gave Jenkins’s name to the missing persons\n\ninvestigator. The investigator also learned that a red Mazda RX-7 sports car, which\n\n\n\n1\n The jury was not permitted to hear, and the record does not reveal, what exactly Jenkins had\nsaid to Hogeland that made Officer Weems suspect him in her disappearance.\n 5\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 6 of 73\n\n\nhad been reported stolen from the service station where Jenkins worked, had been\n\nrecovered on I-459.\n\n The afternoon of April 21, a truck driver who had happened to stop on I-59\n\nnear mile marker 151 discovered Hogeland’s body off the side of the road. She was\n\nnaked, wearing only a watch, and was so badly decomposed that the body had to\n\nbe identified by dental records. From the fractured hyoid bone, it was determined\n\nthat she was manually strangled to death. Also found at the scene were Hogeland’s\n\napron, shoes, bra, panties, pants, and hair net, as well as some beer cans, a Mazda\n\nRX-7 owner’s manual, and other items later determined to have come from the red\n\nMazda. Her necklace and rings were never found.\n\n Alabama authorities issued a warrant for Jenkins’s arrest and also obtained a\n\nfederal fugitive warrant. An FBI agent arrested Jenkins on May 1 in front of his\n\nuncle’s house in Wilmington, California. Jenkins’s uncle later gave the police three\n\nbags of Jenkins’s personal effects. Fibers recovered from Hogeland’s body and\n\nclothing matched those of the Mazda, and hairs matching Hogeland’s were found\n\nin the Mazda. Similarly, fibers from Jenkins’s clothing placed him in the Mazda,\n\nand fibers of his jeans were found on Hogeland’s apron. A bootprint near\n\nHogeland’s body matched a combat boot from among the belongings Jenkins’s\n\nuncle gave to the police, and Jenkins’s roommate said the boots looked just like\n\nthose that Jenkins wore.\n\n\n 6\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 7 of 73\n\n\n A business card like those that had been in the Mazda’s glove box,\n\nbelonging to an owner of the Mazda, was recovered from Jenkins’s wallet after his\n\narrest in California. Later, Jenkins’s cellmate in the Alabama jail reported that\n\nJenkins told him “he had done the crime” and was worried that the couple from the\n\ngas station would identify him or that the police would find his fingerprint on a\n\nbeer can at the scene. A St. Clair County grand jury indicted Jenkins for capital\n\nmurder.\n\n B. The Trial and Sentencing\n\n Jenkins was represented at trial in 1991 by attorneys Douglas Scofield and\n\nStan Downey. Scofield, the lead attorney, had been practicing criminal defense in\n\nBirmingham since 1984 but had never before tried a capital murder case. Jenkins’s\n\nlandlord and his grandmother had arranged the referral and representation before\n\nthe court appointed him. Scofield agreed to the appointment on the condition that a\n\nSt. Clair County lawyer also be appointed. That lawyer was Downey, who had\n\nmore experience than Scofield but whose capital experience was limited to one\n\ncase that had not gone to trial. The two attorneys had agreed that Downey would\n\nbe responsible for jury selection, but Scofield eventually took over during the voir\n\ndire. Per their agreement, Scofield was responsible for the guilt-phase trial,\n\nalthough Downey participated by interviewing Jenkins several times. Scofield also\n\nmet with Jenkins around a dozen times. Jenkins had told him about his background\n\n\n 7\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 8 of 73\n\n\nand childhood, including the fact that he had had a difficult childhood and had\n\nlived on the streets since age 11.\n\n Throughout the weeklong trial, Scofield pursued an outright acquittal. In\n\nparticular, he challenged Harris’s identification of Jenkins as the man who drove\n\naway with Hogeland, and he challenged the medical examiner’s opinions about the\n\ntime and cause of Hogeland’s death. In his closing argument, Scofield took a wide-\n\nranging approach to undermining the State’s wholly circumstantial case against\n\nJenkins. He attacked the State’s timeline of the crime, questioning whether Jenkins\n\nin his intoxicated state would have had the time and the ability to leave Nicholas’s\n\nhouse, steal the Mazda, and abduct and kill Hogeland. He reminded the jury of the\n\nweakness of Harris’s identification of Jenkins and the Coes’ identification of the\n\nwoman in the car. He pointed out that no physical evidence definitively linked\n\nJenkins to the Mazda, to Hogeland’s body, or to her jewelry, which was never\n\nrecovered. And he argued that the evidence provided no plausible motive for\n\nJenkins to have kidnapped, robbed, or killed Hogeland. In sum, he urged the jury\n\nto find reasonable doubt and acquit Jenkins. At the very least, he argued, they\n\nshould find that the State failed to prove Jenkins’s intent to rob and kidnap\n\nHogeland, which were necessary to convict Jenkins of capital murder. He further\n\nargued that, if the jury thought Jenkins killed Hogeland as a crime of passion, they\n\ncould convict him only of manslaughter. The jury deliberated for three hours and\n\n\n 8\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 9 of 73\n\n\nfifteen minutes before convicting Jenkins of capital murder, Ala. Code § 13A-5-\n\n40(a)(1) & (2), i.e., murder committed in the course of kidnapping and robbery.\n\n The penalty phase took place that same afternoon. A week before trial,\n\nScofield and Downey had discussed the penalty phase, and Scofield understood\n\nthat Downey would be responsible for handling the penalty phase. Nonetheless, the\n\nwitness who was present to testify on Jenkins’s behalf during the penalty phase\n\nwas better acquainted with Scofield, so Downey suggested that afternoon that\n\nScofield conduct the direct examination. Scofield urged Downey to interview the\n\nwitness quickly and prepare to examine him, and Downey did so.\n\n Downey handled the penalty presentation to the jury. The State did not\n\npresent any additional aggravation evidence, resting on the evidence presented\n\nduring the guilt phase. One witness, Lonnie Seal, testified on Jenkins’s behalf in\n\norder, in Downey’s words, to “reveal another side of Mark Jenkins to you that you\n\ndon’t know anything about.” Seal had met Jenkins in Fontana, California, in 1988\n\nwhen they both worked at the same garage. Seal and his wife became friends with\n\nJenkins, and he visited in their home. When the Seals decided to move to Alabama\n\na few months later, Jenkins agreed to help with their move and to live with the\n\nSeals in Alabama for a time. It took Lonnie Seal several weeks to find a job, but\n\nJenkins found a job in two days and contributed toward the family’s rent and\n\n\n\n\n 9\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 10 of 73\n\n\ngroceries. Jenkins continued to visit the Seals after he moved into his own home,\n\nand he would always bring a small gift for and play with the Seals’ infant son.\n\n Downey’s closing argument to the jury requested a sentence of life without\n\nparole. Downey reminded the jurors that they had agreed during voir dire that\n\ndeath was not always the appropriate punishment for homicide. Quoting the Bible,2\n\nhe reminded the jury that the Old Testament permitted capital punishment only\n\nwhen supported by the testimony of two or three eyewitnesses, whereas in this case\n\nthere had been only circumstantial evidence. He asserted from Seal’s testimony\n\nthat “Mark had a side of him that was generous, that was giving, and that was\n\nkind,” and argued that the deterrent and retributive purposes of capital punishment\n\nwould not be served in this case.\n\n The jury deliberated for 50 minutes before returning its nonbinding\n\nrecommendation of death by a vote of 10 to 2. See Ala. Code § 13A-5-46(e) (1981)\n\n(amended 2017 to make jury’s verdict binding). After a separate hearing, the court\n\nfound the two statutory aggravating factors for which the State had argued: that the\n\nmurder was committed during the commission of a robbery, and that it was\n\ncommitted during the commission of a kidnapping. See Ala. Code § 13A-5-49(4).\n\n\n\n\n2\n “If anyone kills a person, the murderer shall be put to death on the evidence of witnesses. But\nno person shall be put to death on the testimony of one witness.” Numbers 35:30 (ESV). Downey\nalso paraphrased Deuteronomy 17:6 as “On the evidence of two witnesses or three witnesses, he\nis [sic] shall be put to death. He shall not be put to death on the evidence of one witness.”\n 10\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 11 of 73\n\n\nThe court also found two statutory mitigating factors: that Jenkins had no\n\nsignificant history of prior criminal activity, and his age at the time of the crime\n\n(21). See id. § 13A-5-51(1), (7). The court considered but rejected the statutory\n\nmitigation that, due to his consumption of alcohol, Jenkins’s capacity to appreciate\n\nthe criminality of his conduct or to conform his conduct to the requirements of law\n\nwas substantially impaired. See id. § 13A-5-51(6). The court found that the\n\naggravating circumstances outweighed the mitigating circumstances and, in\n\naccordance with the jury’s recommendation, imposed a sentence of death.\n\n C. The State and Federal Post-Conviction Proceedings\n\n On direct appeal, the Alabama Court of Criminal Appeals affirmed Jenkins’s\n\nconviction and death sentence. Jenkins v. State, 627 So. 2d 1034 (Ala. Ct. Crim.\n\nApp. 1992), aff’d, 627 So. 2d 1054 (Ala. 1993), cert. denied, 511 U.S. 1012\n\n(1994). In 1995, Jenkins filed his first petition for state postconviction relief, Ala.\n\nR. Crim. P. 32,3 alleging, among many other things, that he received ineffective\n\nassistance of counsel during his penalty phase. The Rule 32 trial court conducted a\n\nthree-day hearing at which evidence was presented about the performance of\n\nJenkins’s attorneys.\n\n\n\n\n3\n “Subject to the limitations of Rule 32.2, any defendant who has been convicted of a criminal\noffense may institute a proceeding in the court of original conviction to secure appropriate relief\non the ground that . . . [t]he constitution of the United States or of the State of Alabama requires\na new trial, a new sentence proceeding, or other relief. . . .” Ala. R. Crim. P. 32.1.\n 11\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 12 of 73\n\n\n Scofield testified about the representation that he and Downey provided. In\n\nrelevant part, Scofield asserted that he and Downey had “decided that [developing\n\nmitigation for the penalty phase] would be something that Stan [Downey] would\n\nhandle.” Downey later told Scofield that he hadn’t done any such investigation, but\n\nScofield admitted that, “[a]part from what he has told me, I don’t know what he\n\nhas done.” Scofield testified that Downey “didn’t appear prepared” for the penalty\n\nphase, based on Downey’s last-minute suggestion that Scofield examine Seal since\n\nhe had talked with him before.\n\n In preparation for the penalty phase, Scofield himself did not interview any\n\nmember of Jenkins’s family, but he did not know whether Downey interviewed\n\nany family members. Scofield did interview Jenkins’s landlord and some other\n\nacquaintances but did not do so specifically for the purpose of developing\n\nmitigation. He had met with Jenkins around a dozen times, and Jenkins had told\n\nhim his life story including his experience of abuse as a child. In hindsight,\n\nScofield said he had a “lack of appreciation of mitigating circumstances.” He\n\nrepeatedly denied in his Rule 32 testimony that his omissions during the penalty\n\nphase were strategic: “I didn’t understand what I was doing. . . . I just didn’t\n\nunderstand what I needed to do.”\n\n For reasons that remain unclear, Downey—who is since deceased—did not\n\ntestify or submit an affidavit in the Rule 32 proceeding. Scofield answered,\n\n\n 12\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 13 of 73\n\n\n“Obviously not,” when asked whether he had personal knowledge of everything\n\nDowney may or may not have done to prepare for the penalty phase. The record of\n\nthe extent of Downey’s pretrial preparation is thus limited to Downey’s detailed\n\nfee declaration, which he submitted to the State of Alabama in documentation of\n\nthe compensation due him as appointed counsel. See Ala. Code § 15-12-21. In it,\n\nDowney reported that he spent 71 hours4 on out-of-court preparation for trial. The\n\nreport details 35 hours of interviews with Jenkins in 18 separate visits to the jail\n\nover the course of 17 months, as well as one substantive conversation with\n\nJenkins’s grandmother.5\n\n The Rule 32 trial court also heard testimony from four family members and\n\none friend of Jenkins’s. The family members presented for the first time a stark\n\ndepiction of Jenkins as an unloved and abused child in California. Jenkins’s older\n\nbrother Michael Jenkins testified that all of his siblings have different fathers and\n\nthat his parents often used drugs and fought. Michael testified that Jenkins was\n\nbeaten frequently as a child by his stepfather, if not every day then every other day\n\n\n\n4\n In Jenkins’s state postconviction appeal, the Alabama Court of Criminal Appeals erroneously\nstated that Downey’s fee declaration reports 171 hours of pretrial preparation. See Jenkins, 972\nSo. 2d at 137. We discuss the effect of this discrepancy infra note 7.\n5\n The state appellate court also noted that Downey had moved for a continuance in October 1989\nso that he could conduct “[f]urther discovery and investigation (including a possible trip to\nCalifornia).” Id. We note that Scofield had orally suggested to the trial court in September 1989\nthat more time might be needed with respect to “evidence about the defendant” and witnesses in\nCalifornia, as well as possible psychological evaluation.\n 13\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 14 of 73\n\n\nor every three days. Jenkins was beaten all over his body with a variety of\n\nimplements. Michael was beaten too. Jenkins’s stepfather would lock Jenkins in\n\nhis bedroom for a couple of hours at a time, sometimes without dinner. Some of\n\nthe beatings were prompted by Jenkins’s bedwetting or soiling. Michael testified\n\nthat Jenkins’s stepfather sometimes made Jenkins wear his soiled underwear on his\n\nhead and even forced him to eat his own feces a few times. Michael testified that\n\nJenkins’s bladder and bowel control problems began after a camping trip he took\n\nwith his stepfather and stepgrandfather.\n\n Michael testified that, as children, he and Jenkins were made to work on car\n\nengines late into the night and to shovel horse manure daily. The family moved\n\nfrequently and the children attended many different schools as a result. Michael\n\ntestified that he was Jenkins’s only friend as a child and that Jenkins did poorly in\n\nschool. Jenkins began running away from home as a young teenager and lived with\n\nvarious other family members or on the streets. Michael said that Jenkins was a\n\ngood brother to him and had asked him to testify at his trial, but he never heard\n\nfrom his attorneys or found out when the trial was. Michael admitted to having\n\ncognitive and memory problems due to a head injury years earlier.\n\n Jenkins’s cousin Tammy Pitts testified that, as an infant, Jenkins’s mother\n\nbeat and neglected him, leaving him dirty and in soiled diapers. His stepfather also\n\nbeat Jenkins daily until he left home at age 13, sometimes bruising him so badly\n\n\n 14\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 15 of 73\n\n\nthat he was laid up in bed for days or weeks. Jenkins had problems with\n\nbedwetting and soiling, and his stepfather made him wear soiled clothes to school.\n\nJenkins would be locked in his filthy bedroom around the clock except to attend\n\nschool and do chores. He was usually not allowed to eat dinner with the rest of the\n\nfamily and subsisted on scraps and dog food. The entire family belittled Jenkins\n\nbecause his biological father was Hispanic and his complexion was therefore\n\ndarker. Jenkins’s parents drank a lot and used drugs every day. Pitts knew these\n\nthings because she had daily contact with the family and lived with them off and\n\non, and Jenkins lived with her for a time after he ran away from home. No one had\n\ncontacted her about testifying at Jenkins’s trial.\n\n Jenkins’s second cousin Betty DeLavega testified that her relationship with\n\nJenkins was like that of an aunt. She lived with the Jenkins family for five months\n\nwhen Jenkins was 10 or 11. She saw Jenkins’s stepfather slap Jenkins “[q]uite a\n\nfew times,” and once, he made Jenkins eat his own feces. DeLavega testified that\n\nJenkins’s parents used drugs nearly every day and withheld affection from Jenkins\n\nand Michael. No one contacted her about Jenkins’s trial; Jenkins’s grandmother\n\nhad asked her to testify in the Rule 32 hearing about any abuse she had witnessed.\n\n Jenkins’s grandmother Doris Wagoner testified that while Jenkins’s mother\n\nwas pregnant with him, her husband was in prison and she drank and took drugs.\n\nWhen Jenkins was born, his mother initially put him up for adoption, then changed\n\n\n 15\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 16 of 73\n\n\nher mind weeks later. Jenkins was a lethargic infant and his mother cared more\n\nabout partying than about her children. Wagoner testified that she heard of\n\nJenkins’s mistreatment as a child but didn’t see it. She could tell from the way that\n\nJenkins feared his stepfather that he was being abused, and she knew he slept in a\n\nfilthy bed. After Jenkins was arrested for murder, Wagoner spoke with Scofield,\n\nwho asked for money relating to his representation of Jenkins. She did not\n\nremember whether she talked with Scofield about Jenkins’s background. She could\n\nnot remember why she didn’t testify at the trial; she might have been ill, but she\n\nwas probably busy.\n\n Sherry Seal, the wife of Lonnie Seal, who had testified in the penalty phase\n\nof the trial, testified that Jenkins was respectful and generous and that she trusted\n\nhim with her children. She related how he supported her family after moving to\n\nAlabama, giving them “pretty much” his entire paycheck. She said she would still\n\nfeel safe around Jenkins. Two jailers also testified that Jenkins was a respectful,\n\ncourteous, model inmate who never complained or caused any problems.\n\n Jenkins also introduced medical, school, and juvenile court records from his\n\nchildhood. The medical records document his premature birth and his mother’s\n\nintention to give him up for adoption. The school records reflect a struggling\n\nstudent who was repeatedly promoted to the next grade because of his age, despite\n\nserious academic difficulties, chronic truancy, and frequent school changes. The\n\n\n 16\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 17 of 73\n\n\njuvenile court records document Jenkins’s placement in juvenile hall starting at age\n\n14, following repeated acts of grand theft auto and running away.\n\n The Rule 32 hearing concluded with the testimony of two psychologists\n\nabout Jenkins’s history of psychological trauma stemming from his childhood\n\nabuse. The defense expert, Dr. David Lisak, had met with Jenkins three times and\n\ninterviewed a dozen of his family members, family friends, and associates. Dr.\n\nLisak testified that the pervasive abuse and neglect in Jenkins’s childhood impaired\n\nhis ability to cope with trauma as an adult. He opined that Jenkins’s childhood\n\nabuse adversely affected his cognitive and emotional development and other\n\nabilities. He described how Jenkins had been neglected as an infant and sexually\n\nabused by his grandfather at age four, which led to a lifetime of bladder and bowel\n\ndysfunction. He also discussed the hypothesis that children who are abused are at\n\ngreater risk for perpetrating violence later in life, but he acknowledged there would\n\nhave been other causative factors in Jenkins’s case. Dr. Lisak described Jenkins’s\n\nintellectual capacity as “somewhat borderline,” though he did not perform any\n\ntesting of his own.\n\n The State’s expert, Dr. Karl Kirkland, had met with Jenkins for several hours\n\nand administered several tests. On the Wechsler Adult Intelligence Scale, Jenkins\n\n“scored in the range of borderline intellectual functioning which is between mild\n\nmental retardation and low average intellectual functioning” with an overall IQ of\n\n\n 17\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 18 of 73\n\n\n76. Dr. Kirkland agreed that Jenkins had been “reared in a chaotic dysfunctional\n\nfamily” and generally agreed with Dr. Lisak’s assessment of the negative effects of\n\nchildhood trauma.\n\n The court considered all of this evidence before denying Rule 32 relief in\n\nfull. As relevant here, it found that some of the evidence about the severity of\n\nabuse that Jenkins suffered as a child was not credible. It further found that Jenkins\n\nsuffered no prejudice from any deficient performance by his penalty-phase counsel\n\nbecause the aggravating circumstances outweighed any additional mitigating\n\nevidence Downey might have adduced.\n\n The Alabama Court of Criminal Appeals affirmed the denial of Rule 32\n\nrelief. Jenkins v. State, 972 So. 2d 111 (Ala. Ct. Crim. App. 2004). In relevant part,\n\nthe appeals court rejected Jenkins’s ineffective assistance claim, finding that\n\ncounsel reasonably chose to pursue a penalty-phase strategy of residual doubt and\n\ngood character. Id. at 147. It explained that the new evidence was self-\n\ncontradictory in that it asserted that childhood abuse would have made Jenkins a\n\nviolent adult yet also asserted that Jenkins was meek and mild. Id. The appeals\n\ncourt also agreed with the lower court’s conclusion that Jenkins suffered no\n\nprejudice from any error by his counsel because the aggravating circumstances still\n\noutweighed any new mitigating ones. Id. at 148. The court also noted that evidence\n\n\n\n\n 18\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 19 of 73\n\n\nabout Jenkins’s good behavior in jail awaiting trial was “minimally mitigating” and\n\nwould not have affected the sentence. Id. at 149.\n\n Because the Supreme Court’s decision in Atkins v. Virginia6 had issued\n\nduring the pendency of the Rule 32 appeal, the Alabama appeals court ordered\n\nsupplemental briefing on the possible impact of Atkins upon Jenkins’s sentence.\n\nJenkins asked the court to stay the proceedings and conduct an evidentiary hearing\n\nafter the Alabama legislature enacted a statute about executing the mentally\n\nretarded or, in the alternative, to vacate his sentence and remand after the Alabama\n\nlegislature acted. He later asserted that he is mentally retarded and therefore\n\nineligible for the death penalty. The appeals court proceeded to reject Jenkins’s\n\nAtkins claim on the merits because the record established that his IQ was 76 and\n\nthat he maintained relationships and employment. Id. at 155. The Alabama\n\nSupreme Court summarily affirmed, in relevant part. Ex parte Jenkins, 972 So. 2d\n\n159, 165 (Ala. 2005), remanded to 972 So. 2d 165 (Ala. Ct. Crim. App. 2005)\n\n(juror misconduct claim), cert. denied, 552 U.S. 1167 (2008).\n\n In 2008, Jenkins filed his first petition for a writ of habeas corpus in federal\n\ncourt pursuant to 28 U.S.C. § 2254. The federal proceedings were stayed for\n\nseveral years while Jenkins pursued additional, unrelated Rule 32 relief in the state\n\n\n6\n 536 U.S. 304, 321 (2002) (holding that the execution of mentally retarded offenders violates the\nEighth Amendment). The rule announced in Atkins is retroactive to cases on collateral review. In\nre Holladay, 331 F.3d 1169, 1172–73 (11th Cir. 2003).\n 19\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 20 of 73\n\n\ncourts, ultimately unsuccessfully. In the district court, Jenkins moved for an\n\nevidentiary hearing on his Atkins claim. After reviewing the state court record,\n\nwith a focus on the Rule 32 testimony of Drs. Lisak and Kirkland, the district court\n\ndenied the motion in 2015. Noting that our Circuit has accepted the State of\n\nAlabama’s definition of mental retardation following Atkins, the court found that\n\nJenkins is not intellectually disabled under that standard. In a 347-page order, the\n\ndistrict court also denied relief on all of Jenkins’s claims including ineffective\n\nassistance of counsel during the penalty phase and affirmed as reasonable the state\n\ncourt’s conclusion that Jenkins failed to establish prejudice from any error by his\n\ncounsel. The district court later denied reconsideration and a certificate of\n\nappealability.\n\n Jenkins now appeals. Our Court granted Jenkins a certificate of\n\nappealability, 28 U.S.C. § 2253(c), on two issues:\n\n 1) Whether the district court erred in denying Appellant’s claim\n that his trial counsel rendered ineffective assistance of counsel in the\n investigation and presentation of mitigating evidence during the\n penalty phase of Appellant’s 1991 trial; and\n\n 2) Whether the district court erred in denying Appellant’s claim\n that he is intellectually disabled and ineligible for the death penalty\n under Atkins v. Virginia, 536 U.S. 304 (2002).\n\n II. STANDARDS OF REVIEW\n\n We review the denial of a petition for a writ of habeas corpus de novo.\n\nMorrow v. Warden, Ga. Diagnostic Prison, 886 F.3d 1138, 1146 (11th Cir. 2018).\n\n 20\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 21 of 73\n\n\nThe Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) instructs that\n\na writ of habeas corpus shall not be granted to a state prisoner unless the state\n\nadjudication of a claim “resulted in a decision that was contrary to, or involved an\n\nunreasonable application of, clearly established Federal law, as determined by the\n\nSupreme Court of the United States.” 28 U.S.C. § 2254(d)(1). “[C]learly\n\nestablished Federal law” means “the holdings, as opposed to the dicta, of [the\n\nSupreme] Court’s decisions as of the time of the relevant state-court decision.”\n\nWilliams v. Taylor, 529 U.S. 362, 412 (2000). “[W]hen the last state court to\n\ndecide a prisoner’s federal claim explains its decision on the merits in a reasoned\n\nopinion . . . a federal habeas court simply reviews the specific reasons given by the\n\nstate court and defers to those reasons if they are reasonable.” Wilson v. Sellers,\n\n138 S. Ct. 1188, 1192 (2018). A state court’s decision is reasonable “so long as\n\n‘fairminded jurists could disagree’ on the correctness of the state court’s decision.”\n\nHarrington v. Richter, 562 U.S. 86, 101 (2011) (quoting Yarborough v. Alvarado,\n\n541 U.S. 652, 664 (2004)). “[E]ven a strong case for relief does not mean the state\n\ncourt’s contrary conclusion was unreasonable.” Id. at 102. Rather, a prisoner must\n\nshow that the state court’s ruling “was so lacking in justification that there was an\n\nerror well understood and comprehended in existing law beyond any possibility for\n\nfairminded disagreement.” Id. at 103. That ruling must have been “‘objectively\n\nunreasonable,’ not merely wrong; even ‘clear error’ will not suffice.” White v.\n\n\n 21\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 22 of 73\n\n\nWoodall, 572 U.S. 415, 419 (2014) (quoting Lockyer v. Andrade, 538 U.S. 63, 75\n\n(2003)).\n\n In addition, “a determination of a factual issue made by a State court shall be\n\npresumed to be correct,” and the prisoner bears “the burden of rebutting the\n\npresumption of correctness by clear and convincing evidence.” 28 U.S.C.\n\n§ 2254(e)(1). AEDPA further instructs that a writ of habeas corpus may not be\n\ngranted unless the state court adjudication “resulted in a decision that was based on\n\nan unreasonable determination of the facts in light of the evidence presented in the\n\nState court proceeding.” Id. § 2254(d)(2). When a federal habeas court makes a\n\nfactual finding as part of its habeas determination, we review for clear error, which\n\nis a “highly deferential” standard of review. Thomas v. Allen, 607 F.3d 749, 752\n\n(11th Cir. 2010) (quoting Holton v. City of Thomasville Sch. Dist., 425 F.3d 1325,\n\n1350 (11th Cir. 2005)). “A finding is ‘clearly erroneous’ when although there is\n\nevidence to support it, the reviewing court on the entire evidence is left with the\n\ndefinite and firm conviction that a mistake has been committed.” United States v.\n\nU.S. Gypsum Co., 333 U.S. 364, 395 (1948).\n\n III. DISCUSSION\n\n A. Ineffective Assistance of Counsel\n\n A prisoner alleging that he received ineffective assistance of counsel in\n\nviolation of the Sixth Amendment must establish two elements. Strickland v.\n\n\n 22\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 23 of 73\n\n\nWashington, 466 U.S. 668, 687 (1984). “First, the defendant must show that\n\ncounsel’s performance was deficient.” Id. In other words, he must show “that\n\ncounsel made errors so serious that counsel was not functioning as the ‘counsel’\n\nguaranteed the defendant by the Sixth Amendment.” Id. “Second, the defendant\n\nmust show that the deficient performance prejudiced the defense.” Id. We discuss\n\neach of these elements in turn.\n\n 1. Deficient Performance\n\n A “doubly” deferential lens sits before our eyes as we review the\n\nperformance of Jenkins’s counsel during the penalty phase. Richter, 562 U.S. at\n\n105 (quoting Knowles v. Mirzayance, 556 U.S. 111, 123 (2009)). This redoubled\n\ndeference arises out of the tandem application of AEDPA’s “highly deferential”\n\nreview of a state court’s decision with Strickland’s “highly deferential” review of\n\nan attorney’s performance. Id. (quoting Lindh v. Murphy, 521 U.S. 320, 333 n.7\n\n(1997), and Strickland, 466 U.S. at 689). Under the Strickland standard, “counsel\n\nis strongly presumed to have rendered adequate assistance and made all significant\n\ndecisions in the exercise of reasonable professional judgment.” Strickland, 466\n\nU.S. at 690. And under AEPDA, we must deny habeas relief if “there is any\n\nreasonable argument that counsel satisfied Strickland’s deferential standard.”\n\nRichter, 562 U.S. at 105.\n\n\n\n\n 23\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 24 of 73\n\n\n Jenkins argues that his counsel performed deficiently by failing to\n\n“investigate compelling mitigating evidence or develop a penalty-phase theory.”\n\nThe state court disagreed. It quoted decisions of our Court that hold that “there is\n\nno per se rule that evidence of a criminal defendant’s troubled childhood must\n\nalways be presented as mitigating evidence.” Jenkins, 972 So. 2d at 144 (quoting\n\nMarek v. Singletary, 62 F.3d 1295, 1300 (11th Cir. 1995)). Noting that “Downey\n\nwas in charge of the penalty phase” but “we do not have the benefit of Downey’s\n\ntestimony as to what occurred and why,” it reviewed the trial record to find that\n\n“counsel argued residual doubt and Jenkins’s good character at the penalty phase.”\n\nId. at 145–46. It concluded that “Downey’s decision to concentrate on reasonable\n\ndoubt and to portray Jenkins as a good person was reasonable under the\n\ncircumstances.” Id. at 147.\n\n Jenkins insists, however, that counsel’s failure to develop mitigating\n\nevidence about his childhood abuse was “not the product of strategy, but rather\n\narose from ingnorance and inexeperiance [sic].” The problem for Jenkins is that,\n\nlike the state courts, this Court also does not know what Downey did or why. The\n\nrecord is silent as to his thoughts and intentions as he prepared for the penalty\n\nphase. Given that Jenkins bore the burden of overcoming the strong presumption of\n\ncompetence and of proving Downey’s deficient performance, we are perplexed\n\nwhy Jenkins’s new counsel did not have Downey testify or submit an affidavit in\n\n\n 24\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 25 of 73\n\n\nthe Rule 32 proceeding. See Strickland, 466 U.S. at 687 (“The defendant must\n\nshow that counsel’s performance was deficient.”). Nor did Jenkins testify to the\n\nsubstance of his conversations with his counsel. “The reasonableness of a trial\n\ncounsel’s acts, including lack of investigation or excluding character witnesses\n\nfrom the sentencing phase, depends ‘critically’ upon what information the client\n\ncommunicated to counsel.” Chandler v. United States, 218 F.3d 1305, 1324 (11th\n\nCir. 2000) (en banc) (quoting Strickland, 466 U.S. at 691). There is simply no\n\nrecord evidence to support Jenkins’s central assertion that Downey made no\n\ntactical choices in his preparation for the penalty phase.\n\n Jenkins instead relies on Scofield’s Rule 32 testimony about how he and\n\nDowney prepared for the trial and sentencing. We acknowledge that Scofield\n\nrepeatedly testified that he did not really know what he was doing and that he\n\nfailed to make strategic choices with respect to mitigating circumstances and\n\nsentencing. But Scofield was equally clear that Downey—the more experienced\n\nlawyer—was in fact responsible for the penalty phase. And Scofield also\n\nrepeatedly testified that he did not know what Downey may or may not have done\n\nto prepare. Furthermore, his lack of knowledge about Downey’s preparation was\n\nunderstandable. He and Downey were not colleagues who shared an office and\n\nmight reasonably be expected to know what the other was doing much of the time.\n\nScofield and Downey did not even work in the same city.\n\n\n 25\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 26 of 73\n\n\n This case is unlike those in which we have found deficient performance\n\nwhere the record showed that counsel investigated insufficiently because he\n\n“waited until the eleventh hour” to begin preparing for the penalty phase while\n\nfaced with “overwhelming evidence of guilt.” See, e.g., Johnson v. Sec’y, D.O.C.,\n\n643 F.3d 907, 932 (11th Cir. 2011). The limited record that Jenkins developed\n\nestablishes that Downey spent at least 71 hours preparing for trial, spread out over\n\n17 months. The state court interpreted this sparse record as reflecting a penalty-\n\nphase strategy of residual doubt, the logical extension of Downey’s guilt-phase\n\ninvestigation of the evidence and Scofield’s pursuit at trial of an acquittal based on\n\nreasonable doubt. See Jenkins, 972 So. 2d at 147. In such a situation, “[a] lawyer’s\n\ntime and effort in preparing to defend his client in the guilt phase of a capital case\n\ncontinues to count at the sentencing phase.” Tarver v. Hopper, 169 F.3d 710, 715\n\n(11th Cir. 1999).\n\n We decline to dictate how many hours or how many sessions an attorney\n\nmust spend interviewing his counsel in order to provide effective assistance. “No\n\nabsolute rules dictate what is reasonable performance for lawyers.” Chandler, 218\n\nF.3d at 1317 (citing Strickland, 466 U.S. at 688–89). But even this limited record\n\nundermines Jenkins’s assertion that Downey prepared inadequately. Downey’s 71\n\nhours of preparation included 35 hours of jailhouse conversations with Jenkins\n\n\n\n\n 26\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 27 of 73\n\n\nduring 18 separate meetings.7 According to Downey’s notes, Scofield was present\n\nfor only five of those meetings (12 hours’ worth), leaving 23 hours over 13\n\nmeetings between Jenkins and Downey alone. Downey’s notes reveal little about\n\nthe content of those conversations beyond broad notations such as “evidence,\n\nwitnesses, general status,” and discussions with Jenkins of the evidence and\n\ntestimony against him were surely relevant to Downey’s penalty-phase strategy of\n\nresidual doubt.\n\n But the record does not reveal that Downey knew or should have known that\n\nJenkins’s childhood merited investigation beyond that which he may have done.\n\nThe record shows that Downey developed enough of a rapport with Jenkins that,\n\naccording to his fee declaration, he “g[ave a] Christmas gift [and] discuss[ed] how\n\nmuch to tell his family.” But how much Downey knew about Jenkins’s family\n\nbackground can only be guessed from Scofield’s Rule 32 hearing testimony.\n\nScofield testified that he discussed Jenkins’s background during his dozen or so\n\njailhouse meetings with him, and that Downey was present for some of those\n\nmeetings. Scofield—and possibly Downey—learned that Jenkins experienced\n\nabuse as a child, that his stepfather was brutal, and that he lived on the streets and\n\n\n\n\n7\n We note that the state court clearly erroneously stated that Downey’s fee declaration\ndocumented 171 hours of pretrial preparation. See Jenkins, 972 So. 2d at 137. However, the state\ncourt accurately stated that “he spent over 25 hours talking with Jenkins in more than 10 visits to\nthe jail.” Id. That statement was not an unreasonable determination of the facts.\n 27\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 28 of 73\n\n\nin juvenile institutions because of his brutal home life. It is against this backdrop\n\nthat Downey’s presumptive “decision not to investigate must be directly assessed\n\nfor reasonableness in all the circumstances, applying a heavy measure of deference\n\nto counsel’s judgments.” Strickland, 466 U.S. at 691.\n\n That assessment depends a great deal on what Jenkins told his counsel about\n\nhis background. See id. But we do not know whether Downey was present when\n\nScofield learned about Jenkins’s childhood abuse, nor what else Downey may have\n\nlearned during his 13 solo discussions with Jenkins. Therefore, “given the absence\n\nof evidence in the record, we must assume counsel carried out his professional\n\nresponsibility and discussed mitigation with his client.” Chandler, 218 F.3d at\n\n1324. We will not “turn that presumption on its head by giving [Jenkins] the\n\nbenefit of the doubt where it is unclear what [Downey] did or did not do.” Williams\n\nv. Head, 185 F.3d 1223, 1235 (11th Cir. 1999). We will not assume that Jenkins\n\ntold Downey more than he told Scofield, and Scofield denied remembering that\n\nJenkins told him about the severity or frequency of his stepfather’s physical abuse.\n\nHe denied remembering Jenkins telling him about being beaten daily, being locked\n\nin his room 24/7, or being forced to eat his feces. And there is no suggestion in the\n\nrecord that counsel knew or had reason to investigate the allegation of sexual abuse\n\nby his stepgrandfather. The record simply does not reflect that Jenkins told either\n\nof his lawyers about the severe abuse that some of the Rule 32 witnesses later\n\n\n 28\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 29 of 73\n\n\ndescribed. We will not fault those lawyers for failing to present as mitigation this\n\nlimited evidence that they may have reasonably believed to be unremarkable. See\n\nStrickland, 466 U.S. at 691.\n\n To be sure, counsel’s failure “to discover all reasonably available mitigating\n\nevidence” may constitute deficient performance in some cases. Wiggins v. Smith,\n\n539 U.S. 510, 524 (2003) (quoting ABA Guidelines for the Appointment and\n\nPerformance of Counsel in Death Penalty Cases § 11.4.1(C) (1989)). Our Court\n\nhas identified deficient performance in several instances where counsel had good\n\nreason to investigate the defendant’s background but failed to do so. See Morrow,\n\n886 F.3d at 1147 (collecting cases). But we assess “a particular decision not to\n\ninvestigate . . . for reasonableness in all the circumstances, applying a heavy\n\nmeasure of deference to counsel’s judgments.” Strickland, 466 U.S. at 691. Thus,\n\nwe have rejected a hard rule that counsel’s “strategic decisions can be considered\n\nreasonable only if they are preceded by a ‘thorough investigation.’” Williams, 185\n\nF.3d at 1237 (quoting Rogers v. Zant, 13 F.3d 384, 387 (11th Cir. 1994)). Because\n\nwe simply do not know why Downey chose not to investigate Jenkins’s childhood\n\nmore thoroughly, Downey is entitled to the presumption that this strategic decision\n\nwas reasonable.\n\n And though we have also rejected “a per se rule of ineffective assistance\n\nwhere counsel does not consult family members,” Holladay v. Haley, 209 F.3d\n\n\n 29\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 30 of 73\n\n\n1243, 1252 (11th Cir. 2000), we note that Downey did not fail altogether to consult\n\nfamily members. Downey had at least one substantive conversation with Jenkins’s\n\ngrandmother, and we know from her Rule 32 hearing testimony that she never\n\nwitnessed Jenkins’s childhood abuse. Given the silence of the record and the strong\n\npresumption of competence to which Downey is entitled, we decline to fault him\n\nfor not doing more when Jenkins and his grandmother did not and could not,\n\nrespectively, provide a basis for doing so. Cf. United States v. Pease, 240 F.3d 938,\n\n941–92 (11th Cir. 2001) (“we cannot say as a matter of law that reliance on a\n\nclient’s statements is per se deficient performance”). Nor did Jenkins’s counsel\n\nhave an absolute duty to put on abusive-childhood evidence, even if they knew of\n\nit. See Chandler, 218 F.3d at 1319 (“Counsel is not required to present every\n\nnonfrivolous defense; nor is counsel required to present all mitigation evidence,\n\neven if the additional mitigation evidence would not have been incompatible with\n\ncounsel’s strategy.”); Waters v. Thomas, 46 F.3d 1506, 1511 (11th Cir. 1995) (en\n\nbanc) (“Our decisions are inconsistent with any notion that counsel must present\n\nall available mitigating circumstance evidence.”).\n\n Instead, we affirm as a reasonable application of Strickland the state court’s\n\nconclusion that Downey’s strategic decision to emphasize residual doubt was\n\nreasonable. See Jenkins, 972 So. 2d at 147. As we have noted, “[a] lawyer’s time\n\nand effort in preparing to defend his client in the guilt phase of a capital case\n\n\n 30\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 31 of 73\n\n\ncontinues to count at the sentencing phase.” Tarver, 169 F.3d at 715. This\n\nreasoning is especially applicable where, as here, counsel pursued the reasonable\n\nstrategy of arguing for acquittal at the guilt phase and for life at the sentencing\n\nphase by highlighting doubt about the evidence. “[R]esidual doubt is perhaps the\n\nmost effective strategy to employ at sentencing.” Chandler, 218 F.3d at 1320 n.28.\n\n“[E]specially when—as in this case—the evidence of guilt was not overwhelming,\n\nwe expect that petitioners can rarely (if ever) prove a lawyer to be ineffective for\n\nrelying on this seemingly reasonable strategy to defend his client.” Id. at 1320\n\n(footnote omitted); see also Tarver, 169 F.3d at 715–16 (collecting law review\n\narticles on the merits of residual-doubt strategy). Not only can such a strategy be\n\nconstitutionally adequate, in some cases it may be “the most effective performance\n\nin defense to the death penalty.” Tarver, 169 F.3d at 716.\n\n We cannot say that Downey’s manifest strategy of residual doubt was\n\nobjectively unreasonable under the circumstances. Jenkins’s case is precisely the\n\nsort where reminding the jury of the misgivings it may have held about the State’s\n\ncase could have been effective. Scofield’s guilt-phase closing argument from that\n\nvery morning, which explicitly emphasized doubt and the weakness of the State’s\n\nevidence, would have been fresh in the minds of the jury.8 Although we are not\n\n\n8\n During the penalty phase of a capital case, “[e]vidence presented at the trial of the case may be\nconsidered insofar as it is relevant to the aggravating and mitigating circumstances without the\nnecessity of re-introducing that evidence at the sentence hearing.” Ala. Code § 13A-5-45(c).\n 31\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 32 of 73\n\n\nquestioning the jury’s guilty verdict or the sufficiency of the evidence, which are\n\nbeyond the scope of our habeas review, we acknowledge that the State’s case\n\nagainst Jenkins was—as Downey reminded the jury during the penalty phase—\n\nentirely circumstantial. Indeed, Scofield maintained post-conviction that the case\n\nfor reasonable doubt was “[v]ery strong.” No one saw Jenkins commit either the\n\nmurder or the robbery and kidnapping that elevated Hogeland’s murder to a capital\n\ncrime. Scofield’s guilt-phase closing argument challenged the inferential leaps the\n\nState was asking the jury to make in finding that Jenkins committed these capital-\n\npredicate offenses and that he formed the intent to kill. This situation was not one\n\nwhere a guilt-phase strategy of acquittal—and therefore a penalty-phase strategy of\n\nresidual doubt—was unreasonable in light of the evidence. Cf. Johnson, 643 F.3d\n\nat 932–33 (deficient sentencing investigation where defendant had confessed to\n\ntwo murders).\n\n Thus, although Downey’s penalty-phase closing argument did not\n\nspecifically invoke lingering doubt, it clearly reminded the jury of the entirely\n\ncircumstantial nature of the State’s guilt-phase case. See Chandler, 218 F.3d at\n\n1320 & n.6 (a strategy of “focus[ing] on obtaining an acquittal and then, at\n\nsentencing, on lingering doubt” is reasonable, even if counsel does not use the\n\nwords “lingering doubt” or “residual doubt”). Downey closed this reminder by\n\nexplaining to the jury how a foundational ancient legal system proscribed capital\n\n\n 32\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 33 of 73\n\n\npunishment in the absence of eyewitness testimony—the exact situation here. The\n\nstate court found that Downey’s penalty-phase closing argument reflected a\n\nstrategic “decision to concentrate on reasonable doubt,” see Jenkins, 972 So. 2d at\n\n147, a tactical choice entitled to a “strong presumption of correctness,” Marek, 62\n\nF.3d at 1300 & n.3.\n\n Jenkins suggests other theories of mitigation that Downey might have more\n\neffectively pursued,9 but our deferential review under Strickland does not ask\n\nwhether counsel could possibly or ideally have been more effective. “The test for\n\nineffectiveness is not whether counsel could have done more.” Waters, 46 F.3d at\n\n1518. We do not ask whether an attorney’s representation “deviated from best\n\npractices or common custom,” and we should resist the temptation to second-guess\n\nan attorney with the benefit of our hindsight. Richter, 562 U.S. at 105. We ask only\n\n“whether, in light of all the circumstances, the identified acts or omissions were\n\n\n\n\n9\n In particular, Jenkins argues that his counsel should have presented “model prisoner” evidence.\nThe state court noted, “Good conduct during pretrial incarceration is not necessarily a mitigating\ncircumstance,” before finding that Jenkins had not met his burden of establishing that his counsel\neven had information about his conduct in jail. Jenkins, 972 So. 2d at 149. We agree, and we\ndecline to impose a requirement of sua sponte investigation of this kind in every capital case. Cf.\nWilliams, 529 U.S. at 396 (finding deficient performance based in part on not returning a call\nfrom a prison ministry volunteer who offered to vouch for defendant’s good behavior in prison).\n“Surely, counsel is not required to call a witness to testify to facts such as lack of violent nature\nwhen the jury has rejected such an approach and has found that the defendant is guilty of\nmurder.” Griffin v. Wainwright, 760 F.2d 1505, 1512 (11th Cir. 1985), vacated on other grounds,\n476 U.S. 1112 (1986), reaff’d, 874 F.2d 1397 (11th Cir. 1989).\n 33\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 34 of 73\n\n\noutside the range of professionally competent assistance.” Strickland, 466 U.S. at\n\n690.\n\n We conclude that the state court did not unreasonably determine that Jenkins\n\nfailed to establish objectively incompetent performance by his counsel during the\n\npenalty phase of his trial. Our inquiry is not “what the best lawyers would have\n\ndone. Nor is the test even what most good lawyers would have done. We ask only\n\nwhether some reasonable lawyer at the trial could have acted, in the circumstances,\n\nas defense counsel acted at trial. . . . We are not interested in grading lawyers’\n\nperformances.” White v. Singletary, 972 F.2d 1218, 1220–21 (11th Cir. 1992). And\n\n“[i]f fairminded jurists could reasonably disagree” about the reasonableness of\n\ncounsel’s performance, “then habeas relief is due to be denied.” Johnson, 643 F.3d\n\nat 932.\n\n 2. Prejudice\n\n Even if Jenkins had demonstrated that his counsel performed as no\n\nreasonable lawyer could have, he must also demonstrate a “reasonable probability\n\nthat, but for counsel’s unprofessional errors, the result of the proceeding would\n\nhave been different.” Strickland, 466 U.S. at 694. Given that the jury here\n\nrecommended a sentence of death by the narrowest possible vote, 10 to 2, Jenkins\n\nneed establish only “a reasonable probability that at least one juror would have\n\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537. “A\n\n\n 34\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 35 of 73\n\n\nreasonable probability is a probability sufficient to undermine confidence in the\n\noutcome.” Strickland, 466 U.S. at 694. In measuring that probability here, we will\n\n“evaluate the totality of the available mitigation evidence—both that adduced at\n\ntrial, and the evidence adduced in the habeas proceeding in reweighing it against\n\nthe evidence in aggravation.” Williams, 529 U.S. at 397–98.\n\n We begin our analysis by clarifying what exactly Jenkins is arguing about\n\nprejudice. We do not know precisely which aggravating and mitigating factors the\n\njury may have implicitly found before making its recommendation of death. The\n\ntrial court found at sentencing two statutory aggravating factors: the robbery and\n\nthe kidnapping of the victim. See Ala. Code § 13A-5-49(4). Because none of the\n\nmitigation evidence offered in the Rule 32 hearing bears on those factors, we see\n\nno reasonable probability that the aggravating factors would have been determined\n\ndifferently by the jury or the court. The court also found at sentencing two\n\nstatutory mitigating factors: no significant history of prior criminal activity and\n\nage. See id. § 13A-5-51(1), (7). There was also strong evidence of Jenkins’s\n\nintoxication, which was a key part of Jenkins’s guilt-phase defense and could have\n\nsupported a third mitigating factor, impaired capacity. See id. § 13A-5-51(6).10\n\n\n10\n The jury had heard evidence about all three of these statutory mitigators during the guilt phase\nof the trial. Downey’s penalty-phase opening statement reviewed for the jury all seven of the\npossible statutory mitigating factors, and the court also recited them in its jury instructions,\nreminding the jury that they could consider guilt-phase evidence as well as non-statutory\nmitigation. Jenkins has not established that there is a substantial probability that the jury’s\nrecommendation would have been different had it been more explicitly reminded of Jenkins’s\n 35\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 36 of 73\n\n\nJenkins does not argue that the jury or the court would have found other statutory\n\nmitigators in light of the Rule 32 evidence. Rather, he contends that the jury might\n\nhave found non-statutory mitigation based on Jenkins’s abusive childhood, low\n\nintelligence, or mental health problems that would have outweighed the\n\naggravating factors.11\n\n The Rule 32 trial court found that Jenkins had not demonstrated prejudice\n\neven assuming that counsel had performed deficiently. See Jenkins, 972 So. 2d at\n\n138. It concluded that he had not shown a reasonable probability that the sentencer\n\n“would have concluded that a weighing of the aggravating and mitigating\n\ncircumstances did not warrant death” upon a reweighing of the evidence. See id. It\n\nemphasized the depravity of the crime and the aggravating circumstances and\n\nfound that the mitigating witnesses were not credible for various reasons. See id.\n\nThe state appellate court affirmed, declining to disturb the credibility\n\ndeterminations of the trial court. Id. at 142–44. Independently reweighing the\n\n\nage, criminal history, and intoxication, given that the trial court, the Rule 32 trial court, and the\nstate appellate court all explicitly considered those mitigators and found them outweighed by the\naggravators.\n11\n Alabama law lists seven statutory mitigating circumstances, but states that mitigation “shall\n. . . not be limited to” those seven circumstances. Ala. Code § 13A-5-51. Further, “mitigating\ncircumstances shall include any aspect of a defendant’s character or record and any of the\ncircumstances of the offense that the defendant offers as a basis for a sentence of life\nimprisonment without parole instead of death, and any other relevant mitigating circumstances\nwhich the defendant offers as a basis for a sentence of life imprisonment without parole instead\nof death.” Id. § 13A-5-52. A jury may recommend a sentence of death only if it finds that “one or\nmore aggravating circumstances . . . exist and that they outweigh the mitigating circumstances, if\nany.” Id. § 13A-5-46(e)(3).\n 36\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 37 of 73\n\n\nevidence, it concluded that, “[g]iven the aggravating circumstances that were\n\nproven by the State and the facts surrounding Hogeland’s murder, we, like the\n\ncircuit court, are confident that death was the appropriate punishment for Jenkins’s\n\nactions.” Id. at 148.\n\n Our review of this evidence does not show that it is likely that the jury\n\nwould have concluded that death was not the appropriate punishment if it had\n\nheard the Rule 32 evidence. Much of the proffered mitigation evidence is, in our\n\nview, much less mitigating than Jenkins asserts. In particular, the testimony of\n\nJenkins’s older brother Michael was likely to have been a double-edged sword. See\n\nEvans v. Sec’y, Dep’t of Corr., 703 F.3d 1316, 1327 (11th Cir. 2013) (en banc) (“it\n\nis reasonable to conclude that a defendant was not prejudiced when his mitigation\n\nevidence was a two-edged sword” (quoting Ponticelli v. Sec’y, Fla. Dep’t of Corr.,\n\n690 F.2d 1271, 1296 (11th Cir. 2012)). Michael was the closest relative to testify at\n\nthe Rule 32 hearing and his testimony about the abuse Jenkins experienced should\n\nhave been the most reliable, since he experienced much of the same physical abuse\n\nhimself. The brothers shared the same filthy bedroom, received beatings from the\n\nsame paddle, and were frequently absent from school together. But Michael’s\n\ntestimony contained some discrepancies, and he admitted to cognitive and memory\n\nproblems. Perhaps Michael’s testimony would have made Jenkins appear more\n\nsympathetic in the eyes of the jury—but perhaps not. While Jenkins ran away from\n\n\n 37\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 38 of 73\n\n\nhome and eventually committed murder, Michael responded to the same\n\nbackground by running away from home, graduating from high school, and\n\nproviding health care for both the terminally ill and people with substance use\n\ndisorders. It is possible that the jury could even have found Jenkins to be more\n\nculpable for not overcoming his difficult upbringing and to pose more of a danger\n\nof future violence.\n\n We also find reasonable the state court’s identification of other problems\n\nwith the mitigating evidence Jenkins argues should have been presented. Hearing\n\nthat evidence firsthand in the Rule 32 proceeding, the trial court found the\n\ntestimony of the cousins less credible in light of the lack of physical abuse\n\ndocumented in the school health records. See Jenkins, 972 So. 2d at 139 (noting\n\nthat school records observed a rash and gingivitis). It faulted Pitts in particular for\n\npurporting to have witnessed ongoing, severe abuse but taking only minimal,\n\nineffective steps to obtain outside help. See id. It further noted that DeLavega and\n\nWagoner actually testified that they never saw Jenkins beaten, and it noted all of\n\nthe witnesses’ interest in getting Jenkins off death row. All of these factors would\n\nlikely have caused a reasonable jury to discount their testimony. See id. at 139–42.\n\nWe must accept the trial court’s views of the Rule 32 testimony. “In the absence of\n\nclear and convincing evidence, we have no power on federal habeas review to\n\nrevisit the state court’s credibility determinations.” Bishop v. Warden, 726 F.3d\n\n\n 38\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 39 of 73\n\n\n1243, 1259 (11th Cir. 2013) (citing Marshall v. Lonberger, 459 U.S. 422, 434\n\n(1983)).\n\n The state court also noted that Wagoner testified that she was not available\n\nto testify at the trial, and it found that Seal’s testimony would have been\n\ncumulative to that of her husband. Thus, counsel’s failure to call those two\n\nwitnesses could not have been prejudicial to Jenkins. See id. at 141, 142. Neither of\n\nthese conclusions involves an unreasonable determination of the facts. We agree\n\nthat the Rule 32 testimony of Sherry Seal was substantially similar to the testimony\n\nof her husband Lonnie Seal that the jury heard during the penalty phase. Both\n\ntestified about meeting Jenkins in California, Jenkins’s assistance with their move\n\nto Alabama, and his residence with them in their Alabama home. Sherry added\n\nonly that Jenkins sometimes wet his bed. It was not unreasonable for the state court\n\nto conclude that Sherry’s testimony was cumulative of her husband’s, and “a\n\npetitioner cannot satisfy the prejudice prong of the Strickland test with evidence\n\nthat is merely cumulative of evidence already presented.” Rose v. McNeil, 634 F.3d\n\n1224, 1243 (11th Cir. 2011).\n\n Further, the school, medical, and juvenile court records that Jenkins\n\nintroduced in the Rule 32 proceeding have their own problems. As noted, they do\n\nnot document or even corroborate the severe level of abuse alleged by some of the\n\nwitnesses. Though these documents tell a sad story, they do not create a substantial\n\n\n 39\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 40 of 73\n\n\nlikelihood that the jury would have recommended life. Although they portray\n\nJenkins as a victim of truly unfortunate circumstances, they also show his unsavory\n\ndelinquent behavior. That delinquency included a history of stealing cars, and\n\ngiven that the jury likely inferred that Jenkins stole the red Mazda, new\n\ninformation about his history of theft may not have been especially mitigating.\n\nSimilarly, we agree with the state court that evidence of Jenkins’s compliant\n\nconduct in pretrial detention is “minimally mitigating” where there had been no\n\nevidence of jail misconduct as an aggravator. See Jenkins, 972 So. 2d at 149; cf.\n\nSkipper v. South Carolina, 476 U.S. 1, 5 & n.1 (1986) (good jail behavior is\n\n“potentially mitigating” especially “[w]here the prosecution specifically relies on a\n\nprediction of future dangerousness in asking for the death penalty”).\n\n Jenkins argues that allowing the jury to see his experience of mental,\n\nemotional, and cognitive difficulties, compounded by years of abuse and neglect\n\nand considered alongside his mild character and good jail behavior as an adult,\n\nwould have convinced them that Hogeland’s murder was an “aberration” that\n\ndeserved a punishment less severe than death. That interpretation of Jenkins’s life\n\nstory may be conceivable. But more than mere conceivability is required to\n\nestablish prejudice: “The likelihood of a different result must be substantial, not\n\njust conceivable.” Richter, 562 U.S. at 112 (emphasis added). We do not find the\n\nstate court’s decision that a different result was not substantially likely to be an\n\n\n 40\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 41 of 73\n\n\nunreasonable application of Strickland. As we have noted, we think it equally\n\nlikely that the jury could have rejected this alternate narrative out of hand in light\n\nof having just convicted him of the violent murder of a young mother. We also\n\nthink it possible that the jury could have viewed the murder as just one step in a\n\nlogical progression of degeneracy that began with behavioral problems at school,\n\ncontinued through the juvenile court system, and escalated past grand theft auto\n\ninto kidnapping and murder. Taken separately as well as together, the proffered\n\nevidence does not compel us to conclude that a verdict of life would have been\n\nsubstantially likely. Because fairminded jurists could disagree about the state\n\ncourt’s conclusion about Strickland prejudice, we affirm the denial of habeas relief\n\non this ground.\n\n B. Intellectual Disability\n\n Jenkins argues that the district court erred when it denied his claim that he is\n\nintellectually disabled and ineligible for the death penalty under Atkins v. Virginia,\n\n536 U.S. 304 (2002). The Supreme Court held in Atkins that executing an offender\n\nwho has mental retardation12 violates the Eighth Amendment’s ban on excessive\n\n\n\n\n12\n In 2002, the Atkins Court followed prevailing convention in referring to “mentally retarded”\noffenders. We will use that now-disfavored term only when quoting or discussing older sources.\nSince 2014, the Supreme Court has exclusively used the terms “intellectually disabled person”\nand “person with intellectual disability.” See, e.g., Hall v. Florida, 572 U.S. 701, 704 (2014);\nMoore v. Texas, 139 S. Ct. 666, 672 (2019).\n 41\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 42 of 73\n\n\npunishments. Id. at 321. It defined mental retardation by quoting the American\n\nAssociation on Mental Retardation:\n\n Mental retardation refers to substantial limitations in present\n functioning. It is characterized by significantly subaverage intellectual\n functioning, existing concurrently with related limitations in two or\n more of the following applicable adaptive skill areas: communication,\n self-care, home living, social skills, community use, self-direction,\n health and safety, functional academics, leisure, and work. Mental\n retardation manifests before age 18.\n\nId. at 308 n.3. Alabama courts have followed that tripartite definition, adopting\n\nwhat they consider the “broadest” definition of mental retardation for Atkins\n\npurposes:\n\n [A] defendant, to be considered mentally retarded, must have\n significantly subaverage intellectual functioning (an IQ of 70 or\n below), and significant or substantial deficits in adaptive behavior.\n Additionally, these problems must have manifested themselves during\n the developmental period (i.e., before the defendant reached age 18).\n\nEx parte Perkins, 851 So. 2d 453, 456 (Ala. 2002).\n\n The clinical evidence adduced during the Rule 32 hearing with respect to\n\nJenkins’s intellectual ability was that his IQ was 76, reflecting “borderline\n\nintellectual functioning.” The record also included the report of the Lunacy\n\nCommission13 which found that Jenkins was competent to stand trial, as well as\n\n\n\n13\n Alabama law requires a capital defendant to be “committed to the Department of Mental\nHealth and Mental Retardation for examination” if there is a question about his capacity to stand\ntrial or if there will be a mental-disease-or-defect defense. Ala. Code § 15-16-22(a). “The\nassigned mental health professional(s) shall examine the defendant with respect to determining\nthe presence of any mental disease or defect which, if determined to be present, would affect the\n 42\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 43 of 73\n\n\nJenkins’s school and juvenile court records. The court also considered all the\n\ntestimony it had heard about Jenkins and reviewed the transcript of his murder trial\n\nbefore rejecting Jenkins’s claim of mental retardation.\n\n On appeal, after reciting the Perkins standard, the Alabama Court of\n\nCriminal Appeals explained that Jenkins had not established mental retardation:\n\n Dr. Kirkland testified that he performed psychological tests on Jenkins\n and that Jenkins’s IQ was 76. There was evidence presented at\n Jenkins’s trial indicating that Jenkins maintained relationships with\n other individuals and that he had been employed by P.S. Edwards\n Landscaping Company, Cotton Lowe 76 Service Station, and\n Paramount Painting Company. The record fails to show that Jenkins\n meets the most liberal view of mental retardation adopted by the\n Alabama Supreme Court in Perkins.\n\nJenkins, 972 So. 2d at 155. In this habeas appeal, Jenkins raises arguments about\n\nall three components of the intellectual disability showing under Atkins and\n\nPerkins, which we address in turn. He also appeals the denial of an evidentiary\n\nhearing on this issue.\n\n 1. Intellectual Function\n\n The first component of intellectual disability under Perkins is “significantly\n\nsubaverage intellectual functioning (an IQ of 70 or below).” Perkins, 851 So. 2d at\n\n456. Jenkins argues that the use of a strict IQ cutoff of 70 was contrary to or an\n\nunreasonable application of Atkins.\n\n\ncapacity of the defendant to proceed or continue to trial or which would affect the defendant’s\ncriminal responsibility at the time of the commission of the crime.” Id.\n 43\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 44 of 73\n\n\n We disagree. The state court’s decision that Jenkins failed to establish\n\nmental retardation in this respect was a reasonable application of the record and of\n\nAtkins. As a threshold matter, we note that the state court did not say it was\n\napplying a strict cutoff of 70; it merely stated that Dr. Kirkland had measured\n\nJenkins’s IQ as 76. Jenkins, 972 So. 2d at 155. Under our deferential review, the\n\nquestion is whether the decision that Jenkins had not established the intellectual\n\nfunctioning aspect of intellectual disability was an unreasonable application of\n\nAtkins. The Supreme Court explained in Atkins that it was leaving “to the State[s]\n\nthe task of developing appropriate ways to enforce the constitutional restriction\n\nupon [their] execution of sentences.” Atkins, 536 U.S. at 317 (quoting Ford v.\n\nWainwright, 477 U.S. 399, 416–17 (1986)). In line with that invitation, our Circuit\n\nhas accepted Alabama’s judicial definition of intellectual disability as laid out in\n\nPerkins.14 See Thomas, 607 F.3d at 752–53. Jenkins objects, however, that the\n\nPerkins standard, which he characterizes as articulating a “strict cutoff” of 70,\n\nconflicts with more recent Supreme Court precedent that instead requires a\n\n“clinical” definition that goes beyond a single IQ score.\n\n In particular, Jenkins points to Hall v. Florida, in which the Supreme Court\n\nheld that when an offender’s IQ is near but greater than 70, courts must take into\n\n\n\n14\n Notwithstanding the urging of the Alabama Supreme Court, see Perkins, 851 So. 2d at 455\nn.1, the Alabama legislature has not enacted a statutory definition of intellectual disability.\n 44\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 45 of 73\n\n\naccount the IQ test’s standard error of measurement as well as non-IQ evidence of\n\nintellectual function. See Hall v. Florida, 572 U.S. 701, 723 (2014) (“This Court\n\nagrees with the medical experts that when a defendant’s IQ test score falls within\n\nthe test’s acknowledged and inherent margin of error, the defendant must be able\n\nto present additional evidence of intellectual disability, including testimony\n\nregarding adaptive deficits.”). But, as Jenkins concedes, our Circuit has\n\nspecifically held that Hall is not retroactive to cases on collateral review. See In re\n\nHenry, 757 F.3d 1151, 1161 (11th Cir. 2014) (explaining that Hall does not supply\n\na new substantive rule but rather “merely provides new procedures for ensuring\n\nthat States do not execute members of an already protected group”). We also note\n\nthat Alabama courts have found the Perkins standard to comport with Hall in that\n\nit does not prevent an offender from presenting the standard error of measurement\n\nor non-IQ evidence of intellectual disability. See Reeves v. State, 226 So. 3d 711,\n\n728–29 (Ala. Ct. Crim. App. 2016). Here, the state court did not unreasonably\n\napply Atkins when it analyzed the intellectual functioning component of\n\nintellectual disability.\n\n Jenkins also argues that the state court’s decision on this component in his\n\nRule 32 proceeding resulted from an unreasonable determination of the facts. He\n\nasserts that his IQ score of 76, when combined with the test’s standard error of\n\nmeasurement and considered alongside the Flynn effect (which we discuss below),\n\n\n 45\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 46 of 73\n\n\ndoes show that he has significantly subaverage intellectual function. For the\n\nreasons that follow, we disagree.\n\n Most fundamentally, neither of the clinicians who testified about Jenkins’s\n\nintellectual ability in the Rule 32 hearing opined that he was mentally retarded.\n\nThe closest thing Jenkins can point to is the testimony of Dr. Kirkland that his IQ\n\nis two standard deviations below the mean. But his entire testimony, in context, did\n\nnot establish intellectual disability. To the contrary: Dr. Kirkland testified that\n\nJenkins “scored in the range of borderline intellectual functioning which is\n\nbetween mild mental retardation and low average intellectual functioning.”\n\nFurthermore, Jenkins’s own expert Dr. Lisak fully reviewed Dr. Kirkland’s clinical\n\nfindings. He agreed with those findings, concluding that the IQ results show\n\n“borderline intelligence.” It is tremendously significant that no clinical assessment\n\nin the entire record of these proceedings has found that Jenkins has mental\n\nretardation or intellectual disability.\n\n Jenkins nonetheless argues that his IQ score of 76 establishes the intellectual\n\ncomponent of intellectual ability when the standard error of measurement and the\n\nFlynn effect are considered alongside that score. It was not unreasonable for the\n\nstate court to disagree. As we have discussed, Jenkins is not retroactively entitled\n\nto have the state courts take into account the margin of error of the IQ test. That\n\nstatistical margin of error—5 points in this case—takes into account characteristics\n\n\n 46\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 47 of 73\n\n\nof the test and variations due to chance in order to state how confident the tester is\n\nthat the measured result accurately states Jenkins’s true, unknowable IQ. A result\n\nof 76 with a standard error of measurement of 5 simply means that the tester is\n\n68% confident that Jenkins’s true IQ falls between 71 and 81, and 95% confident\n\nthat his true IQ falls between 66 and 86. See David H. Kaye & David A. Freedman,\n\nReference Guide on Statistics, in Reference Manual on Scientific Evidence 211,\n\n243–44 (3d ed. 2011); see also Hall, 572 U.S. at 738–39 (Alito, J., dissenting). Of\n\ncourse, Jenkins would like us to consider only the possibility that the test result\n\noverstated his IQ—but the standard error of measurement is a two-way street.\n\n“[T]he standard error of measurement is a bi-directional concept that does not\n\ncarry with it a presumption that an individual’s IQ falls to the bottom of his IQ\n\nrange.” Ledford v. Warden, Ga. Diag. & Clas. Prison, 818 F.3d 600, 641 (11th Cir.\n\n2016).\n\n Even if Jenkins were somehow entitled to subtract the 5-point margin of\n\nerror from his score and thereby lower his “true” IQ to 71, he still would not satisfy\n\nthe clinical intellectual component of intellectual disability. Thus, he sought in the\n\ndistrict court to introduce evidence about the Flynn effect in order to bring his IQ\n\ndown to 70 or below. As our Court has explained, the Flynn effect “recognizes the\n\nfact that IQ test scores have been increasing over time”:\n\n The Flynn effect acknowledges that as an intelligence test ages, or\n moves farther from the date on which it was standardized, or normed,\n 47\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 48 of 73\n\n\n the mean score of the population as a whole on that assessment\n instrument increases, thereby artificially inflating the IQ scores of\n individual test subjects. Therefore, the IQ test scores must be\n recalibrated to keep all test subjects on a level playing field.\n\nThomas, 607 F.3d at 753. We have said that “[a]n evaluator may also consider the\n\n‘Flynn effect’” in assessing an offender’s possible intellectual disability, id.\n\n(emphasis added), but neither we nor the Alabama Supreme Court nor the U.S.\n\nSupreme Court have said that a court must consider it in order to reasonably apply\n\nAtkins. See, e.g., Reeves, 226 So. 3d at 739 (“This Court has repeatedly held that a\n\ncircuit court is not required to accept, consider, or apply the ‘Flynn Effect’ in\n\ndetermining intellectual disability.”).\n\n More problematically, Jenkins did not raise the Flynn effect issue in the\n\nAlabama state courts. Jenkins asserts that he implicitly raised it as part and parcel\n\nof his Atkins claim during his Rule 32 appeal, but we do not agree that the\n\nAlabama courts had a full and fair opportunity to consider this issue, because it\n\nwas not explicitly presented to them. See Kelley v. Sec’y for the Dep’t of Corr., 377\n\nF.3d 1317, 1349–50 (11th Cir. 2004) (holding that one ineffective-assistance\n\nargument was unexhausted despite the petitioner’s general presentation of a\n\nStrickland claim in the state courts). Thus, under AEDPA, Jenkins is procedurally\n\nbarred from raising this unexhausted claim in his federal habeas petition. See 28\n\nU.S.C. § 2254(b)(1)(A), (c); Mason v. Allen, 605 F.3d 1114, 1119 (11th Cir. 2010).\n\nWe will allow a prisoner to overcome a procedural bar if he establishes cause for\n 48\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 49 of 73\n\n\nhis default and prejudice therefrom. See, e.g., Smith v. Jones, 256 F.3d 1135, 1138\n\n(11th Cir. 2001). But Jenkins cannot establish prejudice from being barred from\n\nasserting this claim now because the Supreme Court has not “clearly established”\n\nthat a court must consider the Flynn effect in assessing intellectual disability. Thus,\n\nunder AEDPA, it could not form the basis for a federal habeas claim. See 28\n\nU.S.C. § 2254(d)(1). We therefore conclude that the state court did not\n\nunreasonably determine the facts or unreasonably apply Atkins with respect to the\n\nintellectual component of intellectual disability.\n\n 2. Adaptive Function\n\n The second component of intellectual disability is “significant or substantial\n\ndeficits in adaptive behavior.” Perkins, 851 So. 2d at 456. Jenkins argues that the\n\nstate court, which mentioned only Jenkins’s successes in employment and social\n\nskills, see Jenkins, 972 So. 2d at 155, failed to consider his deficits. He adds that\n\neven the mention of employment success does not tell the whole story. Jenkins\n\ninsists that his jobs were menial, his employers took advantage of him, and he held\n\nthe jobs for only a few months each.\n\n But the state court’s terse discussion of Jenkins’s adaptive behavior did not\n\ninvolve an unreasonable determination of the facts or an unreasonable application\n\nof Atkins. To the contrary, its decision on this issue was reasonable in light of both\n\nthe record and Atkins.\n\n\n 49\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 50 of 73\n\n\n Atkins elaborated that the adaptive-skill component of intellectual disability\n\nrequires substantial present limitation in at least two of the following areas:\n\n“communication, self-care, home living, social skills, community use, self-\n\ndirection, health and safety, functional academics, leisure, and work.” Atkins, 536\n\nU.S. at 308 n.3 (quoting definition of American Association on Mental\n\nRetardation). The record supports the state court’s judgment that Jenkins does not\n\nhave substantial deficits in at least two areas. To be sure, we will assume that\n\nJenkins may be substantially limited in functional academics; Dr. Kirkland\n\nassessed his reading, spelling, and arithmetic skills at a third-grade level. But the\n\nrecord does not compel a finding of substantial present deficits in the areas of\n\ncommunication, self-care, home living, and work. Although Jenkins emphasizes\n\nhis transient employment, the record shows he was able to find work quickly, to\n\nwork hard to support the Seals’ household until Lonnie Seal found a job, and, later,\n\nto move into his own home. That record evidence is inconsistent with a substantial\n\ndeficit in the area of work.\n\n As the State observes, the facts of the crime also fail to show significant\n\ndeficits in the areas of communication, self-care, community use, and self-\n\ndirection. Jenkins was able to communicate well enough to solicit an alibi and to\n\nsell his car, including writing out a bill of sale. He then was able to arrange for his\n\nflight from Alabama, including purchasing a bus ticket and hitchhiking across the\n\n\n 50\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 51 of 73\n\n\ncountry. Indeed, throughout his trial and postconviction proceedings, the people\n\nwho knew Jenkins as an adult consistently described a man who did not have\n\nserious difficulties in communicating, forming relationships, working, and caring\n\nfor himself. No witness mentioned these kinds of difficulties in his or her\n\ntestimony about Jenkins. There was some evidence that Jenkins was “not\n\narticulate” and that his house was messy, but that testimony does not rise to the\n\nlevel of overwhelming evidence of substantial functional deficits that would\n\nsupport a conclusion that the state court unreasonably determined the facts.\n\n Of course, the record also contains evidence of Jenkins’s childhood\n\nacademic and social deficits, attributable at least in part to his parents’ neglect, but\n\nJenkins’s childhood is not directly relevant to our consideration of his present\n\nlimitations. Although we acknowledge that the Supreme Court has noted that such\n\n“risk factors” may support further exploration of the possibility of intellectual\n\ndisability, see Moore v. Texas, 137 S. Ct. 1039, 1051 (2017), the state court record\n\ncontains the results of that exploration. Overall, that record supports the state\n\ncourt’s conclusion that Jenkins does not have substantial deficits in adaptive\n\nbehaviors.\n\n 3. Juvenile Onset\n\n The third and final component of intellectual disability is manifestation of\n\nthe first and second components “before the defendant reached age 18.” Perkins,\n\n\n 51\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 52 of 73\n\n\n851 So. 2d at 456. Jenkins asserts, the State assumes, and we agree that the state\n\ncourt did not rule upon this issue, so there is nothing to which we must defer under\n\nAEDPA. See, e.g., Brumfield v. Cain, 135 S. Ct. 2269, 2282 (2015). Thus, we will\n\nreview the district court’s legal conclusions on this issue de novo, and its factual\n\nfindings for clear error. See, e.g., Grossman v. McDonough, 466 F.3d 1325, 1335\n\n(11th Cir. 2006). Jenkins argues that he has satisfied the age-of-onset component\n\nby showing “a wide range of evidence of deficits in both intellectual and adaptive\n\nfunctioning throughout Jenkins’s childhood.”\n\n In fact, the ample record of Jenkins’s childhood, as developed in the Rule 32\n\nproceeding, does not point to intellectual disability before age 18. School\n\nassessments at age 12 reported an IQ score of 83, an adaptive behavior rating of\n\n“low,” and noted overall an intellectual capacity of “average,” a diagnosis of\n\ndyslexia, and “poor functional skills.” The psychologist attributed Jenkins’s\n\ndifficulties to a combination of “[f]requent moves, changes in schools and family\n\ninstability” with “[l]ow ability coupled with a specific learning disability” and\n\n“[l]ack of responsibility for self control and poor use of time.” Although Jenkins\n\nfailed several grades of school, teachers additionally attributed his difficulties to\n\nhis behavioral problems and his high number of absences from school.\n\n Similarly, IQ tests ordered by the juvenile court at age 14 gave scores of 81\n\nand 86 on different tests, in the “[d]ull–[n]ormal” range and reflecting “a definite\n\n\n 52\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 53 of 73\n\n\nlearning disability.” These records all show a child with serious academic deficits\n\nand some intellectual and adaptive deficits, but they do not clearly show an\n\nintellectually disabled child. In particular, Jenkins’s childhood IQ scores of 83, 81,\n\nand 86 fail to point to intellectual disability or mental retardation. We are not left\n\nwith the definite and firm conviction that the district court erred when it found that\n\n“Jenkins’s [childhood] intelligence scores and adaptive functioning skills were\n\nattributable to his learning disability and other circumstances in his life, rather than\n\nmental retardation.”\n\n In sum, we agree with the district court that the decision of the Alabama\n\nCourt of Criminal Appeals that Jenkins does not have intellectual disability was\n\nnot contrary to or an unreasonable application of Atkins v. Virginia.\n\n 4. Evidentiary Hearing\n\n Finally, Jenkins argues that the district court erred when it denied his request\n\nfor an evidentiary hearing on his Atkins claim. If Jenkins diligently attempted to\n\npresent this claim in the state court, we review for abuse of discretion. Burgess v.\n\nComm’r, Ala. Dep’t of Corr., 723 F.3d 1308, 1320 (11th Cir. 2013). “The district\n\ncourt abuses its discretion where ‘such a hearing could enable an applicant to prove\n\nthe petition’s factual allegations, which, if true, would entitle the applicant to\n\nfederal habeas relief.’” Id. (quoting Schriro v. Landrigan, 550 U.S. 465, 474\n\n(2007)). If Jenkins did not diligently attempt to develop the factual basis for this\n\n\n 53\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 54 of 73\n\n\nclaim in the state court, absent certain circumstances, he is not entitled to a hearing\n\nand we will affirm the district court’s denial. 28 U.S.C. § 2254(e)(2).\n\n The parties strenuously dispute whether Jenkins diligently attempted to\n\npresent this issue in the state court. On the one hand, the State argues that Jenkins\n\nfailed to allege that he was mentally retarded even after the U.S. Supreme Court\n\ndecided Atkins in 2002. On the other hand, Jenkins asserts that he raised the\n\npossibility of mental retardation as early as his original Rule 32 petition in 1995.\n\n As we have noted, shortly after the Supreme Court decided Atkins, the\n\nAlabama Court of Criminal Appeals ordered supplemental briefing on the possible\n\nimpact of Atkins on Jenkins. Jenkins’s Rule 32 appeal was pending at that point,\n\nhaving been fully briefed and orally argued in the state appellate court. Having\n\nreviewed the state court records, we agree with the State that Jenkins did not\n\ndiligently attempt to present his Atkins claim in the state courts.\n\n “Diligence will require in the usual case that the prisoner, at a minimum,\n\nseek an evidentiary hearing in state court.” Williams, 529 U.S. at 437. The record\n\nclearly shows that Jenkins did not do so. In the initial supplemental brief on Atkins\n\nhe filed in 2002, Jenkins merely requested a future hearing after the Alabama\n\nlegislature’s enactment of a statute on executing the mentally retarded. Seventeen\n\nyears later, that contingency is still in the future. Only in his reply supplemental\n\nbrief did Jenkins then assert, for the first time ever, that he is mentally retarded.\n\n\n 54\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 55 of 73\n\n\n We acknowledge that Jenkins’s initial Rule 32 petition included in its\n\nallegations of ineffective assistance of counsel a list of unpresented mitigating\n\nevidence including his “developmental[] impair[ment]” and “learning disabilities,\n\nlow intelligence, poor comprehension and retarded socialization skills.” We\n\nconclude that these tangential Strickland references to his mental and social\n\nabilities do not amount to diligence. Thus, because Jenkins did not diligently\n\nattempt to develop the factual basis for his Atkins claim in the state courts, we\n\naffirm the district court’s denial of his request for an evidentiary hearing.\n\n Even if Jenkins had diligently attempted to present this claim in the state\n\ncourt, however, we conclude in the alternative that the district court did not abuse\n\nits discretion when it denied a hearing. We agree, in light of the ample evidence\n\nabout intellectual ability that Jenkins adduced in support of his ineffective-\n\nassistance claim, that such a hearing would not have enabled Jenkins to prove that\n\nhe is intellectually disabled. See Burgess, 723 F.3d at 1320.\n\n IV. CONCLUSION\n\n The denial of Jenkins’s petition for a writ of habeas corpus is\n\n AFFIRMED.\n\n\n\n\n 55\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 56 of 73\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Mark Jenkins v. Commissioner, Alabama Department of Corrections\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-30\\nOpinion type: majority\\nHas dissenting opinion: Yes\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 17-12524 Date Filed: 08/30/2019 Page: 1 of 73\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n\\n\\n No. 17-12524\\n\\n\\n\\n D.C. Docket No. 4:08-cv-00869-VEH\\n\\nMARK ALLEN JENKINS,\\n\\n Petitioner - Appellant,\\n\\n versus\\n\\n\\n\\nCOMMISSIONER, ALABAMA DEPARTMENT OF CORRECTIONS,\\n\\n Respondent - Appellee.\\n\\n\\n\\n Appeal from the United States District Court\\n for the Northern District of Alabama\\n\\n\\n (August 30, 2019)\\n\\nBefore TJOFLAT, WILSON, and BRANCH, Circuit Judges.\\n\\nBRANCH, Circuit Judge:\\n\\n Mark Allen Jenkins, an Alabama prisoner sentenced to death for the 1989\\n\\nmurder of Tammy Ruth Hogeland, appeals the district court’s denial of his petition\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 2 of 73\\n\\n\\nfor a writ of habeas corpus, 28 U.S.C. § 2254. Before us are Jenkins’s arguments\\n\\nthat he received ineffective assistance of counsel during the penalty phase of his\\n\\ntrial and that he is intellectually disabled and therefore ineligible for the death\\n\\npenalty. After careful consideration, and with the benefit of oral argument, we\\n\\naffirm the denial of Jenkins’s habeas petition on both grounds.\\n\\n I. BACKGROUND\\n\\n A. The Crime and Arrest\\n\\n We summarize the following background narrative from opinions of the\\n\\nAlabama Court of Criminal Appeals and our own review of the record. See Jenkins\\n\\nv. State, 627 So. 2d 1034, 1037–40 (Ala. Ct. Crim. App. 1992); Jenkins v. State,\\n\\n972 So. 2d 111, 119–20 (Ala. Ct. Crim. App. 2004). The events leading up to the\\n\\nmurder began the evening of April 17, 1989. Jenkins was at the home of his\\n\\nacquaintance Christine Nicholas. He had met her several months earlier at the\\n\\nOmelet Shoppe restaurant where she worked. Jenkins was very intoxicated and\\n\\ntried to seduce her. When she resisted, Jenkins got “real mad” and asked her\\n\\nseveral times what she would do if someone came up from behind her and grabbed\\n\\nher.\\n\\n At around 1 a.m., Jenkins and Nicholas drove to the Riverchase Omelet\\n\\nShoppe. Jenkins went inside and talked with one of the waitresses, Frieda Vines.\\n\\nThe manager, Douglas Thrash, recognized Jenkins as a regular customer who knew\\n\\n\\n 2\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 3 of 73\\n\\n\\nall of the waitresses. Thrash overheard part of Jenkins’s conversation with Vines\\n\\nand heard mention of the Omelet Shoppe location near the Birmingham Airport.\\n\\nJenkins and Nicholas then returned to Nicholas’s home. Around 2 a.m., Nicholas’s\\n\\nmother asked Jenkins to leave. He did so, falling down some steps and ramming\\n\\nhis car into another vehicle in the process.\\n\\n Meanwhile, 23-year-old Tammy Hogeland was making her way to work at\\n\\nthe Omelet Shoppe. Her sister Wendy, along with Tammy’s young son, picked\\n\\nTammy up from her college classes and drove her to the Riverchase Omelet\\n\\nShoppe a little after 9:30 p.m. Sometime after 10:00, Tammy was sent to work at\\n\\nthe airport Omelet Shoppe on Tenth Avenue, since that location was unexpectedly\\n\\nshorthanded. Tammy was wearing her watch, a necklace with the words “special\\n\\nsister,” her class ring, and her diamond engagement ring. Tammy was working as a\\n\\ncook that night and was wearing a blue apron, black pants, a white shirt, and white\\n\\nshoes. She and Sarah Harris were the only employees working at the airport\\n\\nOmelet Shoppe that night.\\n\\n At about 2 a.m. on April 18, Harris saw Jenkins—whom she did not know—\\n\\ndrive up in a red sports car. His arrival was memorable to Harris because the car\\n\\nnearly jumped the curb and crashed through the restaurant’s glass wall. Jenkins got\\n\\nout of the car and came into the restaurant, appearing to be intoxicated. He walked\\n\\nover to Hogeland and began talking to her. Harris later saw Jenkins and Hogeland\\n\\n\\n 3\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 4 of 73\\n\\n\\ndrive away together in the red sports car. That was the last time anyone who knew\\n\\nHogeland ever saw her. Hogeland left behind her cigarettes, lighter, purse, and\\n\\npaycheck. She left without telling anyone, which she had never done before.\\n\\nAlthough Harris saw them drive off, she could not later say whether Hogeland left\\n\\nwith Jenkins willingly or was instead abducted.\\n\\n At around 5 a.m. that morning, Bobby and Geraldine Coe had stopped to\\n\\nbuy gas at a Chevron station on I-59 northeast of Birmingham when they saw\\n\\nJenkins drive up in the red sports car. The Coes noticed a female who appeared to\\n\\nbe “passed out” in the front passenger seat, but they could not say whether she was\\n\\nalive or dead. While Bobby Coe was pumping gas, Jenkins asked him for some\\n\\ncigarettes and said, “Looks like it’s been a long night and it looks like it’s going to\\n\\nbe a long day.” Jenkins then said “God bless you” before asking directions to I-\\n\\n459. Coe gave him directions and got back in his car. As Coe pulled out onto the\\n\\ninterstate, he saw Jenkins follow him in the red sports car. Coe then saw the car\\n\\nflash its lights, slow down, and pull to the side of the road between mile markers\\n\\n151 and 152.\\n\\n At 8 a.m., Wendy Hogeland learned that Tammy was not at the restaurant,\\n\\nand their mother called the police. Meanwhile, Jenkins went to the home of Steve\\n\\nMusser, who noticed that he was wearing the same clothes as the day before.\\n\\nJenkins told Musser that his car had been stolen the night before and asked him if\\n\\n\\n 4\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 5 of 73\\n\\n\\nhe would say that he had been with him all night. Musser refused. Christine\\n\\nNicholas also saw Jenkins in a grocery store that morning. He was looking at a\\n\\nnewspaper, making a phone call, and attempting to sell his Buick. Nicholas loaned\\n\\nhim $4 for gas. At around 10 o’clock that morning, Jenkins sold his car to Michael\\n\\nBrooks, a mechanic at a local Chevron station. Jenkins had explained that he\\n\\nneeded the money so he could visit his sick mother in California. Another\\n\\nmechanic at the service station drove Jenkins to the Greyhound bus station later\\n\\nthat morning.\\n\\n Jenkins awoke the next day on the Greyhound bus in Houston and was\\n\\nejected from the bus because his fare was used up. He then hitchhiked from\\n\\nHouston to Phoenix, then to San Diego and Los Angeles. Jenkins was first\\n\\nidentified as a suspect in Hogeland’s disappearance by Officer Mike Weems of the\\n\\nHoover Police Department on April 19. Weems ate dinner at the Riverchase\\n\\nOmelet Shoppe just about every day and had talked with Hogeland just as often.\\n\\nHe also knew Jenkins from the restaurant. Weems learned of Hogeland’s\\n\\ndisappearance from the other waitresses. Remembering how Jenkins often talked to\\n\\nHogeland and passed her notes,1 he gave Jenkins’s name to the missing persons\\n\\ninvestigator. The investigator also learned that a red Mazda RX-7 sports car, which\\n\\n\\n\\n1\\n The jury was not permitted to hear, and the record does not reveal, what exactly Jenkins had\\nsaid to Hogeland that made Officer Weems suspect him in her disappearance.\\n 5\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 6 of 73\\n\\n\\nhad been reported stolen from the service station where Jenkins worked, had been\\n\\nrecovered on I-459.\\n\\n The afternoon of April 21, a truck driver who had happened to stop on I-59\\n\\nnear mile marker 151 discovered Hogeland’s body off the side of the road. She was\\n\\nnaked, wearing only a watch, and was so badly decomposed that the body had to\\n\\nbe identified by dental records. From the fractured hyoid bone, it was determined\\n\\nthat she was manually strangled to death. Also found at the scene were Hogeland’s\\n\\napron, shoes, bra, panties, pants, and hair net, as well as some beer cans, a Mazda\\n\\nRX-7 owner’s manual, and other items later determined to have come from the red\\n\\nMazda. Her necklace and rings were never found.\\n\\n Alabama authorities issued a warrant for Jenkins’s arrest and also obtained a\\n\\nfederal fugitive warrant. An FBI agent arrested Jenkins on May 1 in front of his\\n\\nuncle’s house in Wilmington, California. Jenkins’s uncle later gave the police three\\n\\nbags of Jenkins’s personal effects. Fibers recovered from Hogeland’s body and\\n\\nclothing matched those of the Mazda, and hairs matching Hogeland’s were found\\n\\nin the Mazda. Similarly, fibers from Jenkins’s clothing placed him in the Mazda,\\n\\nand fibers of his jeans were found on Hogeland’s apron. A bootprint near\\n\\nHogeland’s body matched a combat boot from among the belongings Jenkins’s\\n\\nuncle gave to the police, and Jenkins’s roommate said the boots looked just like\\n\\nthose that Jenkins wore.\\n\\n\\n 6\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 7 of 73\\n\\n\\n A business card like those that had been in the Mazda’s glove box,\\n\\nbelonging to an owner of the Mazda, was recovered from Jenkins’s wallet after his\\n\\narrest in California. Later, Jenkins’s cellmate in the Alabama jail reported that\\n\\nJenkins told him “he had done the crime” and was worried that the couple from the\\n\\ngas station would identify him or that the police would find his fingerprint on a\\n\\nbeer can at the scene. A St. Clair County grand jury indicted Jenkins for capital\\n\\nmurder.\\n\\n B. The Trial and Sentencing\\n\\n Jenkins was represented at trial in 1991 by attorneys Douglas Scofield and\\n\\nStan Downey. Scofield, the lead attorney, had been practicing criminal defense in\\n\\nBirmingham since 1984 but had never before tried a capital murder case. Jenkins’s\\n\\nlandlord and his grandmother had arranged the referral and representation before\\n\\nthe court appointed him. Scofield agreed to the appointment on the condition that a\\n\\nSt. Clair County lawyer also be appointed. That lawyer was Downey, who had\\n\\nmore experience than Scofield but whose capital experience was limited to one\\n\\ncase that had not gone to trial. The two attorneys had agreed that Downey would\\n\\nbe responsible for jury selection, but Scofield eventually took over during the voir\\n\\ndire. Per their agreement, Scofield was responsible for the guilt-phase trial,\\n\\nalthough Downey participated by interviewing Jenkins several times. Scofield also\\n\\nmet with Jenkins around a dozen times. Jenkins had told him about his background\\n\\n\\n 7\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 8 of 73\\n\\n\\nand childhood, including the fact that he had had a difficult childhood and had\\n\\nlived on the streets since age 11.\\n\\n Throughout the weeklong trial, Scofield pursued an outright acquittal. In\\n\\nparticular, he challenged Harris’s identification of Jenkins as the man who drove\\n\\naway with Hogeland, and he challenged the medical examiner’s opinions about the\\n\\ntime and cause of Hogeland’s death. In his closing argument, Scofield took a wide-\\n\\nranging approach to undermining the State’s wholly circumstantial case against\\n\\nJenkins. He attacked the State’s timeline of the crime, questioning whether Jenkins\\n\\nin his intoxicated state would have had the time and the ability to leave Nicholas’s\\n\\nhouse, steal the Mazda, and abduct and kill Hogeland. He reminded the jury of the\\n\\nweakness of Harris’s identification of Jenkins and the Coes’ identification of the\\n\\nwoman in the car. He pointed out that no physical evidence definitively linked\\n\\nJenkins to the Mazda, to Hogeland’s body, or to her jewelry, which was never\\n\\nrecovered. And he argued that the evidence provided no plausible motive for\\n\\nJenkins to have kidnapped, robbed, or killed Hogeland. In sum, he urged the jury\\n\\nto find reasonable doubt and acquit Jenkins. At the very least, he argued, they\\n\\nshould find that the State failed to prove Jenkins’s intent to rob and kidnap\\n\\nHogeland, which were necessary to convict Jenkins of capital murder. He further\\n\\nargued that, if the jury thought Jenkins killed Hogeland as a crime of passion, they\\n\\ncould convict him only of manslaughter. The jury deliberated for three hours and\\n\\n\\n 8\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 9 of 73\\n\\n\\nfifteen minutes before convicting Jenkins of capital murder, Ala. Code § 13A-5-\\n\\n40(a)(1) & (2), i.e., murder committed in the course of kidnapping and robbery.\\n\\n The penalty phase took place that same afternoon. A week before trial,\\n\\nScofield and Downey had discussed the penalty phase, and Scofield understood\\n\\nthat Downey would be responsible for handling the penalty phase. Nonetheless, the\\n\\nwitness who was present to testify on Jenkins’s behalf during the penalty phase\\n\\nwas better acquainted with Scofield, so Downey suggested that afternoon that\\n\\nScofield conduct the direct examination. Scofield urged Downey to interview the\\n\\nwitness quickly and prepare to examine him, and Downey did so.\\n\\n Downey handled the penalty presentation to the jury. The State did not\\n\\npresent any additional aggravation evidence, resting on the evidence presented\\n\\nduring the guilt phase. One witness, Lonnie Seal, testified on Jenkins’s behalf in\\n\\norder, in Downey’s words, to “reveal another side of Mark Jenkins to you that you\\n\\ndon’t know anything about.” Seal had met Jenkins in Fontana, California, in 1988\\n\\nwhen they both worked at the same garage. Seal and his wife became friends with\\n\\nJenkins, and he visited in their home. When the Seals decided to move to Alabama\\n\\na few months later, Jenkins agreed to help with their move and to live with the\\n\\nSeals in Alabama for a time. It took Lonnie Seal several weeks to find a job, but\\n\\nJenkins found a job in two days and contributed toward the family’s rent and\\n\\n\\n\\n\\n 9\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 10 of 73\\n\\n\\ngroceries. Jenkins continued to visit the Seals after he moved into his own home,\\n\\nand he would always bring a small gift for and play with the Seals’ infant son.\\n\\n Downey’s closing argument to the jury requested a sentence of life without\\n\\nparole. Downey reminded the jurors that they had agreed during voir dire that\\n\\ndeath was not always the appropriate punishment for homicide. Quoting the Bible,2\\n\\nhe reminded the jury that the Old Testament permitted capital punishment only\\n\\nwhen supported by the testimony of two or three eyewitnesses, whereas in this case\\n\\nthere had been only circumstantial evidence. He asserted from Seal’s testimony\\n\\nthat “Mark had a side of him that was generous, that was giving, and that was\\n\\nkind,” and argued that the deterrent and retributive purposes of capital punishment\\n\\nwould not be served in this case.\\n\\n The jury deliberated for 50 minutes before returning its nonbinding\\n\\nrecommendation of death by a vote of 10 to 2. See Ala. Code § 13A-5-46(e) (1981)\\n\\n(amended 2017 to make jury’s verdict binding). After a separate hearing, the court\\n\\nfound the two statutory aggravating factors for which the State had argued: that the\\n\\nmurder was committed during the commission of a robbery, and that it was\\n\\ncommitted during the commission of a kidnapping. See Ala. Code § 13A-5-49(4).\\n\\n\\n\\n\\n2\\n “If anyone kills a person, the murderer shall be put to death on the evidence of witnesses. But\\nno person shall be put to death on the testimony of one witness.” Numbers 35:30 (ESV). Downey\\nalso paraphrased Deuteronomy 17:6 as “On the evidence of two witnesses or three witnesses, he\\nis [sic] shall be put to death. He shall not be put to death on the evidence of one witness.”\\n 10\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 11 of 73\\n\\n\\nThe court also found two statutory mitigating factors: that Jenkins had no\\n\\nsignificant history of prior criminal activity, and his age at the time of the crime\\n\\n(21). See id. § 13A-5-51(1), (7). The court considered but rejected the statutory\\n\\nmitigation that, due to his consumption of alcohol, Jenkins’s capacity to appreciate\\n\\nthe criminality of his conduct or to conform his conduct to the requirements of law\\n\\nwas substantially impaired. See id. § 13A-5-51(6). The court found that the\\n\\naggravating circumstances outweighed the mitigating circumstances and, in\\n\\naccordance with the jury’s recommendation, imposed a sentence of death.\\n\\n C. The State and Federal Post-Conviction Proceedings\\n\\n On direct appeal, the Alabama Court of Criminal Appeals affirmed Jenkins’s\\n\\nconviction and death sentence. Jenkins v. State, 627 So. 2d 1034 (Ala. Ct. Crim.\\n\\nApp. 1992), aff’d, 627 So. 2d 1054 (Ala. 1993), cert. denied, 511 U.S. 1012\\n\\n(1994). In 1995, Jenkins filed his first petition for state postconviction relief, Ala.\\n\\nR. Crim. P. 32,3 alleging, among many other things, that he received ineffective\\n\\nassistance of counsel during his penalty phase. The Rule 32 trial court conducted a\\n\\nthree-day hearing at which evidence was presented about the performance of\\n\\nJenkins’s attorneys.\\n\\n\\n\\n\\n3\\n “Subject to the limitations of Rule 32.2, any defendant who has been convicted of a criminal\\noffense may institute a proceeding in the court of original conviction to secure appropriate relief\\non the ground that . . . [t]he constitution of the United States or of the State of Alabama requires\\na new trial, a new sentence proceeding, or other relief. . . .” Ala. R. Crim. P. 32.1.\\n 11\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 12 of 73\\n\\n\\n Scofield testified about the representation that he and Downey provided. In\\n\\nrelevant part, Scofield asserted that he and Downey had “decided that [developing\\n\\nmitigation for the penalty phase] would be something that Stan [Downey] would\\n\\nhandle.” Downey later told Scofield that he hadn’t done any such investigation, but\\n\\nScofield admitted that, “[a]part from what he has told me, I don’t know what he\\n\\nhas done.” Scofield testified that Downey “didn’t appear prepared” for the penalty\\n\\nphase, based on Downey’s last-minute suggestion that Scofield examine Seal since\\n\\nhe had talked with him before.\\n\\n In preparation for the penalty phase, Scofield himself did not interview any\\n\\nmember of Jenkins’s family, but he did not know whether Downey interviewed\\n\\nany family members. Scofield did interview Jenkins’s landlord and some other\\n\\nacquaintances but did not do so specifically for the purpose of developing\\n\\nmitigation. He had met with Jenkins around a dozen times, and Jenkins had told\\n\\nhim his life story including his experience of abuse as a child. In hindsight,\\n\\nScofield said he had a “lack of appreciation of mitigating circumstances.” He\\n\\nrepeatedly denied in his Rule 32 testimony that his omissions during the penalty\\n\\nphase were strategic: “I didn’t understand what I was doing. . . . I just didn’t\\n\\nunderstand what I needed to do.”\\n\\n For reasons that remain unclear, Downey—who is since deceased—did not\\n\\ntestify or submit an affidavit in the Rule 32 proceeding. Scofield answered,\\n\\n\\n 12\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 13 of 73\\n\\n\\n“Obviously not,” when asked whether he had personal knowledge of everything\\n\\nDowney may or may not have done to prepare for the penalty phase. The record of\\n\\nthe extent of Downey’s pretrial preparation is thus limited to Downey’s detailed\\n\\nfee declaration, which he submitted to the State of Alabama in documentation of\\n\\nthe compensation due him as appointed counsel. See Ala. Code § 15-12-21. In it,\\n\\nDowney reported that he spent 71 hours4 on out-of-court preparation for trial. The\\n\\nreport details 35 hours of interviews with Jenkins in 18 separate visits to the jail\\n\\nover the course of 17 months, as well as one substantive conversation with\\n\\nJenkins’s grandmother.5\\n\\n The Rule 32 trial court also heard testimony from four family members and\\n\\none friend of Jenkins’s. The family members presented for the first time a stark\\n\\ndepiction of Jenkins as an unloved and abused child in California. Jenkins’s older\\n\\nbrother Michael Jenkins testified that all of his siblings have different fathers and\\n\\nthat his parents often used drugs and fought. Michael testified that Jenkins was\\n\\nbeaten frequently as a child by his stepfather, if not every day then every other day\\n\\n\\n\\n4\\n In Jenkins’s state postconviction appeal, the Alabama Court of Criminal Appeals erroneously\\nstated that Downey’s fee declaration reports 171 hours of pretrial preparation. See Jenkins, 972\\nSo. 2d at 137. We discuss the effect of this discrepancy infra note 7.\\n5\\n The state appellate court also noted that Downey had moved for a continuance in October 1989\\nso that he could conduct “[f]urther discovery and investigation (including a possible trip to\\nCalifornia).” Id. We note that Scofield had orally suggested to the trial court in September 1989\\nthat more time might be needed with respect to “evidence about the defendant” and witnesses in\\nCalifornia, as well as possible psychological evaluation.\\n 13\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 14 of 73\\n\\n\\nor every three days. Jenkins was beaten all over his body with a variety of\\n\\nimplements. Michael was beaten too. Jenkins’s stepfather would lock Jenkins in\\n\\nhis bedroom for a couple of hours at a time, sometimes without dinner. Some of\\n\\nthe beatings were prompted by Jenkins’s bedwetting or soiling. Michael testified\\n\\nthat Jenkins’s stepfather sometimes made Jenkins wear his soiled underwear on his\\n\\nhead and even forced him to eat his own feces a few times. Michael testified that\\n\\nJenkins’s bladder and bowel control problems began after a camping trip he took\\n\\nwith his stepfather and stepgrandfather.\\n\\n Michael testified that, as children, he and Jenkins were made to work on car\\n\\nengines late into the night and to shovel horse manure daily. The family moved\\n\\nfrequently and the children attended many different schools as a result. Michael\\n\\ntestified that he was Jenkins’s only friend as a child and that Jenkins did poorly in\\n\\nschool. Jenkins began running away from home as a young teenager and lived with\\n\\nvarious other family members or on the streets. Michael said that Jenkins was a\\n\\ngood brother to him and had asked him to testify at his trial, but he never heard\\n\\nfrom his attorneys or found out when the trial was. Michael admitted to having\\n\\ncognitive and memory problems due to a head injury years earlier.\\n\\n Jenkins’s cousin Tammy Pitts testified that, as an infant, Jenkins’s mother\\n\\nbeat and neglected him, leaving him dirty and in soiled diapers. His stepfather also\\n\\nbeat Jenkins daily until he left home at age 13, sometimes bruising him so badly\\n\\n\\n 14\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 15 of 73\\n\\n\\nthat he was laid up in bed for days or weeks. Jenkins had problems with\\n\\nbedwetting and soiling, and his stepfather made him wear soiled clothes to school.\\n\\nJenkins would be locked in his filthy bedroom around the clock except to attend\\n\\nschool and do chores. He was usually not allowed to eat dinner with the rest of the\\n\\nfamily and subsisted on scraps and dog food. The entire family belittled Jenkins\\n\\nbecause his biological father was Hispanic and his complexion was therefore\\n\\ndarker. Jenkins’s parents drank a lot and used drugs every day. Pitts knew these\\n\\nthings because she had daily contact with the family and lived with them off and\\n\\non, and Jenkins lived with her for a time after he ran away from home. No one had\\n\\ncontacted her about testifying at Jenkins’s trial.\\n\\n Jenkins’s second cousin Betty DeLavega testified that her relationship with\\n\\nJenkins was like that of an aunt. She lived with the Jenkins family for five months\\n\\nwhen Jenkins was 10 or 11. She saw Jenkins’s stepfather slap Jenkins “[q]uite a\\n\\nfew times,” and once, he made Jenkins eat his own feces. DeLavega testified that\\n\\nJenkins’s parents used drugs nearly every day and withheld affection from Jenkins\\n\\nand Michael. No one contacted her about Jenkins’s trial; Jenkins’s grandmother\\n\\nhad asked her to testify in the Rule 32 hearing about any abuse she had witnessed.\\n\\n Jenkins’s grandmother Doris Wagoner testified that while Jenkins’s mother\\n\\nwas pregnant with him, her husband was in prison and she drank and took drugs.\\n\\nWhen Jenkins was born, his mother initially put him up for adoption, then changed\\n\\n\\n 15\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 16 of 73\\n\\n\\nher mind weeks later. Jenkins was a lethargic infant and his mother cared more\\n\\nabout partying than about her children. Wagoner testified that she heard of\\n\\nJenkins’s mistreatment as a child but didn’t see it. She could tell from the way that\\n\\nJenkins feared his stepfather that he was being abused, and she knew he slept in a\\n\\nfilthy bed. After Jenkins was arrested for murder, Wagoner spoke with Scofield,\\n\\nwho asked for money relating to his representation of Jenkins. She did not\\n\\nremember whether she talked with Scofield about Jenkins’s background. She could\\n\\nnot remember why she didn’t testify at the trial; she might have been ill, but she\\n\\nwas probably busy.\\n\\n Sherry Seal, the wife of Lonnie Seal, who had testified in the penalty phase\\n\\nof the trial, testified that Jenkins was respectful and generous and that she trusted\\n\\nhim with her children. She related how he supported her family after moving to\\n\\nAlabama, giving them “pretty much” his entire paycheck. She said she would still\\n\\nfeel safe around Jenkins. Two jailers also testified that Jenkins was a respectful,\\n\\ncourteous, model inmate who never complained or caused any problems.\\n\\n Jenkins also introduced medical, school, and juvenile court records from his\\n\\nchildhood. The medical records document his premature birth and his mother’s\\n\\nintention to give him up for adoption. The school records reflect a struggling\\n\\nstudent who was repeatedly promoted to the next grade because of his age, despite\\n\\nserious academic difficulties, chronic truancy, and frequent school changes. The\\n\\n\\n 16\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 17 of 73\\n\\n\\njuvenile court records document Jenkins’s placement in juvenile hall starting at age\\n\\n14, following repeated acts of grand theft auto and running away.\\n\\n The Rule 32 hearing concluded with the testimony of two psychologists\\n\\nabout Jenkins’s history of psychological trauma stemming from his childhood\\n\\nabuse. The defense expert, Dr. David Lisak, had met with Jenkins three times and\\n\\ninterviewed a dozen of his family members, family friends, and associates. Dr.\\n\\nLisak testified that the pervasive abuse and neglect in Jenkins’s childhood impaired\\n\\nhis ability to cope with trauma as an adult. He opined that Jenkins’s childhood\\n\\nabuse adversely affected his cognitive and emotional development and other\\n\\nabilities. He described how Jenkins had been neglected as an infant and sexually\\n\\nabused by his grandfather at age four, which led to a lifetime of bladder and bowel\\n\\ndysfunction. He also discussed the hypothesis that children who are abused are at\\n\\ngreater risk for perpetrating violence later in life, but he acknowledged there would\\n\\nhave been other causative factors in Jenkins’s case. Dr. Lisak described Jenkins’s\\n\\nintellectual capacity as “somewhat borderline,” though he did not perform any\\n\\ntesting of his own.\\n\\n The State’s expert, Dr. Karl Kirkland, had met with Jenkins for several hours\\n\\nand administered several tests. On the Wechsler Adult Intelligence Scale, Jenkins\\n\\n“scored in the range of borderline intellectual functioning which is between mild\\n\\nmental retardation and low average intellectual functioning” with an overall IQ of\\n\\n\\n 17\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 18 of 73\\n\\n\\n76. Dr. Kirkland agreed that Jenkins had been “reared in a chaotic dysfunctional\\n\\nfamily” and generally agreed with Dr. Lisak’s assessment of the negative effects of\\n\\nchildhood trauma.\\n\\n The court considered all of this evidence before denying Rule 32 relief in\\n\\nfull. As relevant here, it found that some of the evidence about the severity of\\n\\nabuse that Jenkins suffered as a child was not credible. It further found that Jenkins\\n\\nsuffered no prejudice from any deficient performance by his penalty-phase counsel\\n\\nbecause the aggravating circumstances outweighed any additional mitigating\\n\\nevidence Downey might have adduced.\\n\\n The Alabama Court of Criminal Appeals affirmed the denial of Rule 32\\n\\nrelief. Jenkins v. State, 972 So. 2d 111 (Ala. Ct. Crim. App. 2004). In relevant part,\\n\\nthe appeals court rejected Jenkins’s ineffective assistance claim, finding that\\n\\ncounsel reasonably chose to pursue a penalty-phase strategy of residual doubt and\\n\\ngood character. Id. at 147. It explained that the new evidence was self-\\n\\ncontradictory in that it asserted that childhood abuse would have made Jenkins a\\n\\nviolent adult yet also asserted that Jenkins was meek and mild. Id. The appeals\\n\\ncourt also agreed with the lower court’s conclusion that Jenkins suffered no\\n\\nprejudice from any error by his counsel because the aggravating circumstances still\\n\\noutweighed any new mitigating ones. Id. at 148. The court also noted that evidence\\n\\n\\n\\n\\n 18\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 19 of 73\\n\\n\\nabout Jenkins’s good behavior in jail awaiting trial was “minimally mitigating” and\\n\\nwould not have affected the sentence. Id. at 149.\\n\\n Because the Supreme Court’s decision in Atkins v. Virginia6 had issued\\n\\nduring the pendency of the Rule 32 appeal, the Alabama appeals court ordered\\n\\nsupplemental briefing on the possible impact of Atkins upon Jenkins’s sentence.\\n\\nJenkins asked the court to stay the proceedings and conduct an evidentiary hearing\\n\\nafter the Alabama legislature enacted a statute about executing the mentally\\n\\nretarded or, in the alternative, to vacate his sentence and remand after the Alabama\\n\\nlegislature acted. He later asserted that he is mentally retarded and therefore\\n\\nineligible for the death penalty. The appeals court proceeded to reject Jenkins’s\\n\\nAtkins claim on the merits because the record established that his IQ was 76 and\\n\\nthat he maintained relationships and employment. Id. at 155. The Alabama\\n\\nSupreme Court summarily affirmed, in relevant part. Ex parte Jenkins, 972 So. 2d\\n\\n159, 165 (Ala. 2005), remanded to 972 So. 2d 165 (Ala. Ct. Crim. App. 2005)\\n\\n(juror misconduct claim), cert. denied, 552 U.S. 1167 (2008).\\n\\n In 2008, Jenkins filed his first petition for a writ of habeas corpus in federal\\n\\ncourt pursuant to 28 U.S.C. § 2254. The federal proceedings were stayed for\\n\\nseveral years while Jenkins pursued additional, unrelated Rule 32 relief in the state\\n\\n\\n6\\n 536 U.S. 304, 321 (2002) (holding that the execution of mentally retarded offenders violates the\\nEighth Amendment). The rule announced in Atkins is retroactive to cases on collateral review. In\\nre Holladay, 331 F.3d 1169, 1172–73 (11th Cir. 2003).\\n 19\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 20 of 73\\n\\n\\ncourts, ultimately unsuccessfully. In the district court, Jenkins moved for an\\n\\nevidentiary hearing on his Atkins claim. After reviewing the state court record,\\n\\nwith a focus on the Rule 32 testimony of Drs. Lisak and Kirkland, the district court\\n\\ndenied the motion in 2015. Noting that our Circuit has accepted the State of\\n\\nAlabama’s definition of mental retardation following Atkins, the court found that\\n\\nJenkins is not intellectually disabled under that standard. In a 347-page order, the\\n\\ndistrict court also denied relief on all of Jenkins’s claims including ineffective\\n\\nassistance of counsel during the penalty phase and affirmed as reasonable the state\\n\\ncourt’s conclusion that Jenkins failed to establish prejudice from any error by his\\n\\ncounsel. The district court later denied reconsideration and a certificate of\\n\\nappealability.\\n\\n Jenkins now appeals. Our Court granted Jenkins a certificate of\\n\\nappealability, 28 U.S.C. § 2253(c), on two issues:\\n\\n 1) Whether the district court erred in denying Appellant’s claim\\n that his trial counsel rendered ineffective assistance of counsel in the\\n investigation and presentation of mitigating evidence during the\\n penalty phase of Appellant’s 1991 trial; and\\n\\n 2) Whether the district court erred in denying Appellant’s claim\\n that he is intellectually disabled and ineligible for the death penalty\\n under Atkins v. Virginia, 536 U.S. 304 (2002).\\n\\n II. STANDARDS OF REVIEW\\n\\n We review the denial of a petition for a writ of habeas corpus de novo.\\n\\nMorrow v. Warden, Ga. Diagnostic Prison, 886 F.3d 1138, 1146 (11th Cir. 2018).\\n\\n 20\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 21 of 73\\n\\n\\nThe Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) instructs that\\n\\na writ of habeas corpus shall not be granted to a state prisoner unless the state\\n\\nadjudication of a claim “resulted in a decision that was contrary to, or involved an\\n\\nunreasonable application of, clearly established Federal law, as determined by the\\n\\nSupreme Court of the United States.” 28 U.S.C. § 2254(d)(1). “[C]learly\\n\\nestablished Federal law” means “the holdings, as opposed to the dicta, of [the\\n\\nSupreme] Court’s decisions as of the time of the relevant state-court decision.”\\n\\nWilliams v. Taylor, 529 U.S. 362, 412 (2000). “[W]hen the last state court to\\n\\ndecide a prisoner’s federal claim explains its decision on the merits in a reasoned\\n\\nopinion . . . a federal habeas court simply reviews the specific reasons given by the\\n\\nstate court and defers to those reasons if they are reasonable.” Wilson v. Sellers,\\n\\n138 S. Ct. 1188, 1192 (2018). A state court’s decision is reasonable “so long as\\n\\n‘fairminded jurists could disagree’ on the correctness of the state court’s decision.”\\n\\nHarrington v. Richter, 562 U.S. 86, 101 (2011) (quoting Yarborough v. Alvarado,\\n\\n541 U.S. 652, 664 (2004)). “[E]ven a strong case for relief does not mean the state\\n\\ncourt’s contrary conclusion was unreasonable.” Id. at 102. Rather, a prisoner must\\n\\nshow that the state court’s ruling “was so lacking in justification that there was an\\n\\nerror well understood and comprehended in existing law beyond any possibility for\\n\\nfairminded disagreement.” Id. at 103. That ruling must have been “‘objectively\\n\\nunreasonable,’ not merely wrong; even ‘clear error’ will not suffice.” White v.\\n\\n\\n 21\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 22 of 73\\n\\n\\nWoodall, 572 U.S. 415, 419 (2014) (quoting Lockyer v. Andrade, 538 U.S. 63, 75\\n\\n(2003)).\\n\\n In addition, “a determination of a factual issue made by a State court shall be\\n\\npresumed to be correct,” and the prisoner bears “the burden of rebutting the\\n\\npresumption of correctness by clear and convincing evidence.” 28 U.S.C.\\n\\n§ 2254(e)(1). AEDPA further instructs that a writ of habeas corpus may not be\\n\\ngranted unless the state court adjudication “resulted in a decision that was based on\\n\\nan unreasonable determination of the facts in light of the evidence presented in the\\n\\nState court proceeding.” Id. § 2254(d)(2). When a federal habeas court makes a\\n\\nfactual finding as part of its habeas determination, we review for clear error, which\\n\\nis a “highly deferential” standard of review. Thomas v. Allen, 607 F.3d 749, 752\\n\\n(11th Cir. 2010) (quoting Holton v. City of Thomasville Sch. Dist., 425 F.3d 1325,\\n\\n1350 (11th Cir. 2005)). “A finding is ‘clearly erroneous’ when although there is\\n\\nevidence to support it, the reviewing court on the entire evidence is left with the\\n\\ndefinite and firm conviction that a mistake has been committed.” United States v.\\n\\nU.S. Gypsum Co., 333 U.S. 364, 395 (1948).\\n\\n III. DISCUSSION\\n\\n A. Ineffective Assistance of Counsel\\n\\n A prisoner alleging that he received ineffective assistance of counsel in\\n\\nviolation of the Sixth Amendment must establish two elements. Strickland v.\\n\\n\\n 22\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 23 of 73\\n\\n\\nWashington, 466 U.S. 668, 687 (1984). “First, the defendant must show that\\n\\ncounsel’s performance was deficient.” Id. In other words, he must show “that\\n\\ncounsel made errors so serious that counsel was not functioning as the ‘counsel’\\n\\nguaranteed the defendant by the Sixth Amendment.” Id. “Second, the defendant\\n\\nmust show that the deficient performance prejudiced the defense.” Id. We discuss\\n\\neach of these elements in turn.\\n\\n 1. Deficient Performance\\n\\n A “doubly” deferential lens sits before our eyes as we review the\\n\\nperformance of Jenkins’s counsel during the penalty phase. Richter, 562 U.S. at\\n\\n105 (quoting Knowles v. Mirzayance, 556 U.S. 111, 123 (2009)). This redoubled\\n\\ndeference arises out of the tandem application of AEDPA’s “highly deferential”\\n\\nreview of a state court’s decision with Strickland’s “highly deferential” review of\\n\\nan attorney’s performance. Id. (quoting Lindh v. Murphy, 521 U.S. 320, 333 n.7\\n\\n(1997), and Strickland, 466 U.S. at 689). Under the Strickland standard, “counsel\\n\\nis strongly presumed to have rendered adequate assistance and made all significant\\n\\ndecisions in the exercise of reasonable professional judgment.” Strickland, 466\\n\\nU.S. at 690. And under AEPDA, we must deny habeas relief if “there is any\\n\\nreasonable argument that counsel satisfied Strickland’s deferential standard.”\\n\\nRichter, 562 U.S. at 105.\\n\\n\\n\\n\\n 23\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 24 of 73\\n\\n\\n Jenkins argues that his counsel performed deficiently by failing to\\n\\n“investigate compelling mitigating evidence or develop a penalty-phase theory.”\\n\\nThe state court disagreed. It quoted decisions of our Court that hold that “there is\\n\\nno per se rule that evidence of a criminal defendant’s troubled childhood must\\n\\nalways be presented as mitigating evidence.” Jenkins, 972 So. 2d at 144 (quoting\\n\\nMarek v. Singletary, 62 F.3d 1295, 1300 (11th Cir. 1995)). Noting that “Downey\\n\\nwas in charge of the penalty phase” but “we do not have the benefit of Downey’s\\n\\ntestimony as to what occurred and why,” it reviewed the trial record to find that\\n\\n“counsel argued residual doubt and Jenkins’s good character at the penalty phase.”\\n\\nId. at 145–46. It concluded that “Downey’s decision to concentrate on reasonable\\n\\ndoubt and to portray Jenkins as a good person was reasonable under the\\n\\ncircumstances.” Id. at 147.\\n\\n Jenkins insists, however, that counsel’s failure to develop mitigating\\n\\nevidence about his childhood abuse was “not the product of strategy, but rather\\n\\narose from ingnorance and inexeperiance [sic].” The problem for Jenkins is that,\\n\\nlike the state courts, this Court also does not know what Downey did or why. The\\n\\nrecord is silent as to his thoughts and intentions as he prepared for the penalty\\n\\nphase. Given that Jenkins bore the burden of overcoming the strong presumption of\\n\\ncompetence and of proving Downey’s deficient performance, we are perplexed\\n\\nwhy Jenkins’s new counsel did not have Downey testify or submit an affidavit in\\n\\n\\n 24\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 25 of 73\\n\\n\\nthe Rule 32 proceeding. See Strickland, 466 U.S. at 687 (“The defendant must\\n\\nshow that counsel’s performance was deficient.”). Nor did Jenkins testify to the\\n\\nsubstance of his conversations with his counsel. “The reasonableness of a trial\\n\\ncounsel’s acts, including lack of investigation or excluding character witnesses\\n\\nfrom the sentencing phase, depends ‘critically’ upon what information the client\\n\\ncommunicated to counsel.” Chandler v. United States, 218 F.3d 1305, 1324 (11th\\n\\nCir. 2000) (en banc) (quoting Strickland, 466 U.S. at 691). There is simply no\\n\\nrecord evidence to support Jenkins’s central assertion that Downey made no\\n\\ntactical choices in his preparation for the penalty phase.\\n\\n Jenkins instead relies on Scofield’s Rule 32 testimony about how he and\\n\\nDowney prepared for the trial and sentencing. We acknowledge that Scofield\\n\\nrepeatedly testified that he did not really know what he was doing and that he\\n\\nfailed to make strategic choices with respect to mitigating circumstances and\\n\\nsentencing. But Scofield was equally clear that Downey—the more experienced\\n\\nlawyer—was in fact responsible for the penalty phase. And Scofield also\\n\\nrepeatedly testified that he did not know what Downey may or may not have done\\n\\nto prepare. Furthermore, his lack of knowledge about Downey’s preparation was\\n\\nunderstandable. He and Downey were not colleagues who shared an office and\\n\\nmight reasonably be expected to know what the other was doing much of the time.\\n\\nScofield and Downey did not even work in the same city.\\n\\n\\n 25\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 26 of 73\\n\\n\\n This case is unlike those in which we have found deficient performance\\n\\nwhere the record showed that counsel investigated insufficiently because he\\n\\n“waited until the eleventh hour” to begin preparing for the penalty phase while\\n\\nfaced with “overwhelming evidence of guilt.” See, e.g., Johnson v. Sec’y, D.O.C.,\\n\\n643 F.3d 907, 932 (11th Cir. 2011). The limited record that Jenkins developed\\n\\nestablishes that Downey spent at least 71 hours preparing for trial, spread out over\\n\\n17 months. The state court interpreted this sparse record as reflecting a penalty-\\n\\nphase strategy of residual doubt, the logical extension of Downey’s guilt-phase\\n\\ninvestigation of the evidence and Scofield’s pursuit at trial of an acquittal based on\\n\\nreasonable doubt. See Jenkins, 972 So. 2d at 147. In such a situation, “[a] lawyer’s\\n\\ntime and effort in preparing to defend his client in the guilt phase of a capital case\\n\\ncontinues to count at the sentencing phase.” Tarver v. Hopper, 169 F.3d 710, 715\\n\\n(11th Cir. 1999).\\n\\n We decline to dictate how many hours or how many sessions an attorney\\n\\nmust spend interviewing his counsel in order to provide effective assistance. “No\\n\\nabsolute rules dictate what is reasonable performance for lawyers.” Chandler, 218\\n\\nF.3d at 1317 (citing Strickland, 466 U.S. at 688–89). But even this limited record\\n\\nundermines Jenkins’s assertion that Downey prepared inadequately. Downey’s 71\\n\\nhours of preparation included 35 hours of jailhouse conversations with Jenkins\\n\\n\\n\\n\\n 26\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 27 of 73\\n\\n\\nduring 18 separate meetings.7 According to Downey’s notes, Scofield was present\\n\\nfor only five of those meetings (12 hours’ worth), leaving 23 hours over 13\\n\\nmeetings between Jenkins and Downey alone. Downey’s notes reveal little about\\n\\nthe content of those conversations beyond broad notations such as “evidence,\\n\\nwitnesses, general status,” and discussions with Jenkins of the evidence and\\n\\ntestimony against him were surely relevant to Downey’s penalty-phase strategy of\\n\\nresidual doubt.\\n\\n But the record does not reveal that Downey knew or should have known that\\n\\nJenkins’s childhood merited investigation beyond that which he may have done.\\n\\nThe record shows that Downey developed enough of a rapport with Jenkins that,\\n\\naccording to his fee declaration, he “g[ave a] Christmas gift [and] discuss[ed] how\\n\\nmuch to tell his family.” But how much Downey knew about Jenkins’s family\\n\\nbackground can only be guessed from Scofield’s Rule 32 hearing testimony.\\n\\nScofield testified that he discussed Jenkins’s background during his dozen or so\\n\\njailhouse meetings with him, and that Downey was present for some of those\\n\\nmeetings. Scofield—and possibly Downey—learned that Jenkins experienced\\n\\nabuse as a child, that his stepfather was brutal, and that he lived on the streets and\\n\\n\\n\\n\\n7\\n We note that the state court clearly erroneously stated that Downey’s fee declaration\\ndocumented 171 hours of pretrial preparation. See Jenkins, 972 So. 2d at 137. However, the state\\ncourt accurately stated that “he spent over 25 hours talking with Jenkins in more than 10 visits to\\nthe jail.” Id. That statement was not an unreasonable determination of the facts.\\n 27\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 28 of 73\\n\\n\\nin juvenile institutions because of his brutal home life. It is against this backdrop\\n\\nthat Downey’s presumptive “decision not to investigate must be directly assessed\\n\\nfor reasonableness in all the circumstances, applying a heavy measure of deference\\n\\nto counsel’s judgments.” Strickland, 466 U.S. at 691.\\n\\n That assessment depends a great deal on what Jenkins told his counsel about\\n\\nhis background. See id. But we do not know whether Downey was present when\\n\\nScofield learned about Jenkins’s childhood abuse, nor what else Downey may have\\n\\nlearned during his 13 solo discussions with Jenkins. Therefore, “given the absence\\n\\nof evidence in the record, we must assume counsel carried out his professional\\n\\nresponsibility and discussed mitigation with his client.” Chandler, 218 F.3d at\\n\\n1324. We will not “turn that presumption on its head by giving [Jenkins] the\\n\\nbenefit of the doubt where it is unclear what [Downey] did or did not do.” Williams\\n\\nv. Head, 185 F.3d 1223, 1235 (11th Cir. 1999). We will not assume that Jenkins\\n\\ntold Downey more than he told Scofield, and Scofield denied remembering that\\n\\nJenkins told him about the severity or frequency of his stepfather’s physical abuse.\\n\\nHe denied remembering Jenkins telling him about being beaten daily, being locked\\n\\nin his room 24/7, or being forced to eat his feces. And there is no suggestion in the\\n\\nrecord that counsel knew or had reason to investigate the allegation of sexual abuse\\n\\nby his stepgrandfather. The record simply does not reflect that Jenkins told either\\n\\nof his lawyers about the severe abuse that some of the Rule 32 witnesses later\\n\\n\\n 28\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 29 of 73\\n\\n\\ndescribed. We will not fault those lawyers for failing to present as mitigation this\\n\\nlimited evidence that they may have reasonably believed to be unremarkable. See\\n\\nStrickland, 466 U.S. at 691.\\n\\n To be sure, counsel’s failure “to discover all reasonably available mitigating\\n\\nevidence” may constitute deficient performance in some cases. Wiggins v. Smith,\\n\\n539 U.S. 510, 524 (2003) (quoting ABA Guidelines for the Appointment and\\n\\nPerformance of Counsel in Death Penalty Cases § 11.4.1(C) (1989)). Our Court\\n\\nhas identified deficient performance in several instances where counsel had good\\n\\nreason to investigate the defendant’s background but failed to do so. See Morrow,\\n\\n886 F.3d at 1147 (collecting cases). But we assess “a particular decision not to\\n\\ninvestigate . . . for reasonableness in all the circumstances, applying a heavy\\n\\nmeasure of deference to counsel’s judgments.” Strickland, 466 U.S. at 691. Thus,\\n\\nwe have rejected a hard rule that counsel’s “strategic decisions can be considered\\n\\nreasonable only if they are preceded by a ‘thorough investigation.’” Williams, 185\\n\\nF.3d at 1237 (quoting Rogers v. Zant, 13 F.3d 384, 387 (11th Cir. 1994)). Because\\n\\nwe simply do not know why Downey chose not to investigate Jenkins’s childhood\\n\\nmore thoroughly, Downey is entitled to the presumption that this strategic decision\\n\\nwas reasonable.\\n\\n And though we have also rejected “a per se rule of ineffective assistance\\n\\nwhere counsel does not consult family members,” Holladay v. Haley, 209 F.3d\\n\\n\\n 29\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 30 of 73\\n\\n\\n1243, 1252 (11th Cir. 2000), we note that Downey did not fail altogether to consult\\n\\nfamily members. Downey had at least one substantive conversation with Jenkins’s\\n\\ngrandmother, and we know from her Rule 32 hearing testimony that she never\\n\\nwitnessed Jenkins’s childhood abuse. Given the silence of the record and the strong\\n\\npresumption of competence to which Downey is entitled, we decline to fault him\\n\\nfor not doing more when Jenkins and his grandmother did not and could not,\\n\\nrespectively, provide a basis for doing so. Cf. United States v. Pease, 240 F.3d 938,\\n\\n941–92 (11th Cir. 2001) (“we cannot say as a matter of law that reliance on a\\n\\nclient’s statements is per se deficient performance”). Nor did Jenkins’s counsel\\n\\nhave an absolute duty to put on abusive-childhood evidence, even if they knew of\\n\\nit. See Chandler, 218 F.3d at 1319 (“Counsel is not required to present every\\n\\nnonfrivolous defense; nor is counsel required to present all mitigation evidence,\\n\\neven if the additional mitigation evidence would not have been incompatible with\\n\\ncounsel’s strategy.”); Waters v. Thomas, 46 F.3d 1506, 1511 (11th Cir. 1995) (en\\n\\nbanc) (“Our decisions are inconsistent with any notion that counsel must present\\n\\nall available mitigating circumstance evidence.”).\\n\\n Instead, we affirm as a reasonable application of Strickland the state court’s\\n\\nconclusion that Downey’s strategic decision to emphasize residual doubt was\\n\\nreasonable. See Jenkins, 972 So. 2d at 147. As we have noted, “[a] lawyer’s time\\n\\nand effort in preparing to defend his client in the guilt phase of a capital case\\n\\n\\n 30\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 31 of 73\\n\\n\\ncontinues to count at the sentencing phase.” Tarver, 169 F.3d at 715. This\\n\\nreasoning is especially applicable where, as here, counsel pursued the reasonable\\n\\nstrategy of arguing for acquittal at the guilt phase and for life at the sentencing\\n\\nphase by highlighting doubt about the evidence. “[R]esidual doubt is perhaps the\\n\\nmost effective strategy to employ at sentencing.” Chandler, 218 F.3d at 1320 n.28.\\n\\n“[E]specially when—as in this case—the evidence of guilt was not overwhelming,\\n\\nwe expect that petitioners can rarely (if ever) prove a lawyer to be ineffective for\\n\\nrelying on this seemingly reasonable strategy to defend his client.” Id. at 1320\\n\\n(footnote omitted); see also Tarver, 169 F.3d at 715–16 (collecting law review\\n\\narticles on the merits of residual-doubt strategy). Not only can such a strategy be\\n\\nconstitutionally adequate, in some cases it may be “the most effective performance\\n\\nin defense to the death penalty.” Tarver, 169 F.3d at 716.\\n\\n We cannot say that Downey’s manifest strategy of residual doubt was\\n\\nobjectively unreasonable under the circumstances. Jenkins’s case is precisely the\\n\\nsort where reminding the jury of the misgivings it may have held about the State’s\\n\\ncase could have been effective. Scofield’s guilt-phase closing argument from that\\n\\nvery morning, which explicitly emphasized doubt and the weakness of the State’s\\n\\nevidence, would have been fresh in the minds of the jury.8 Although we are not\\n\\n\\n8\\n During the penalty phase of a capital case, “[e]vidence presented at the trial of the case may be\\nconsidered insofar as it is relevant to the aggravating and mitigating circumstances without the\\nnecessity of re-introducing that evidence at the sentence hearing.” Ala. Code § 13A-5-45(c).\\n 31\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 32 of 73\\n\\n\\nquestioning the jury’s guilty verdict or the sufficiency of the evidence, which are\\n\\nbeyond the scope of our habeas review, we acknowledge that the State’s case\\n\\nagainst Jenkins was—as Downey reminded the jury during the penalty phase—\\n\\nentirely circumstantial. Indeed, Scofield maintained post-conviction that the case\\n\\nfor reasonable doubt was “[v]ery strong.” No one saw Jenkins commit either the\\n\\nmurder or the robbery and kidnapping that elevated Hogeland’s murder to a capital\\n\\ncrime. Scofield’s guilt-phase closing argument challenged the inferential leaps the\\n\\nState was asking the jury to make in finding that Jenkins committed these capital-\\n\\npredicate offenses and that he formed the intent to kill. This situation was not one\\n\\nwhere a guilt-phase strategy of acquittal—and therefore a penalty-phase strategy of\\n\\nresidual doubt—was unreasonable in light of the evidence. Cf. Johnson, 643 F.3d\\n\\nat 932–33 (deficient sentencing investigation where defendant had confessed to\\n\\ntwo murders).\\n\\n Thus, although Downey’s penalty-phase closing argument did not\\n\\nspecifically invoke lingering doubt, it clearly reminded the jury of the entirely\\n\\ncircumstantial nature of the State’s guilt-phase case. See Chandler, 218 F.3d at\\n\\n1320 & n.6 (a strategy of “focus[ing] on obtaining an acquittal and then, at\\n\\nsentencing, on lingering doubt” is reasonable, even if counsel does not use the\\n\\nwords “lingering doubt” or “residual doubt”). Downey closed this reminder by\\n\\nexplaining to the jury how a foundational ancient legal system proscribed capital\\n\\n\\n 32\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 33 of 73\\n\\n\\npunishment in the absence of eyewitness testimony—the exact situation here. The\\n\\nstate court found that Downey’s penalty-phase closing argument reflected a\\n\\nstrategic “decision to concentrate on reasonable doubt,” see Jenkins, 972 So. 2d at\\n\\n147, a tactical choice entitled to a “strong presumption of correctness,” Marek, 62\\n\\nF.3d at 1300 & n.3.\\n\\n Jenkins suggests other theories of mitigation that Downey might have more\\n\\neffectively pursued,9 but our deferential review under Strickland does not ask\\n\\nwhether counsel could possibly or ideally have been more effective. “The test for\\n\\nineffectiveness is not whether counsel could have done more.” Waters, 46 F.3d at\\n\\n1518. We do not ask whether an attorney’s representation “deviated from best\\n\\npractices or common custom,” and we should resist the temptation to second-guess\\n\\nan attorney with the benefit of our hindsight. Richter, 562 U.S. at 105. We ask only\\n\\n“whether, in light of all the circumstances, the identified acts or omissions were\\n\\n\\n\\n\\n9\\n In particular, Jenkins argues that his counsel should have presented “model prisoner” evidence.\\nThe state court noted, “Good conduct during pretrial incarceration is not necessarily a mitigating\\ncircumstance,” before finding that Jenkins had not met his burden of establishing that his counsel\\neven had information about his conduct in jail. Jenkins, 972 So. 2d at 149. We agree, and we\\ndecline to impose a requirement of sua sponte investigation of this kind in every capital case. Cf.\\nWilliams, 529 U.S. at 396 (finding deficient performance based in part on not returning a call\\nfrom a prison ministry volunteer who offered to vouch for defendant’s good behavior in prison).\\n“Surely, counsel is not required to call a witness to testify to facts such as lack of violent nature\\nwhen the jury has rejected such an approach and has found that the defendant is guilty of\\nmurder.” Griffin v. Wainwright, 760 F.2d 1505, 1512 (11th Cir. 1985), vacated on other grounds,\\n476 U.S. 1112 (1986), reaff’d, 874 F.2d 1397 (11th Cir. 1989).\\n 33\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 34 of 73\\n\\n\\noutside the range of professionally competent assistance.” Strickland, 466 U.S. at\\n\\n690.\\n\\n We conclude that the state court did not unreasonably determine that Jenkins\\n\\nfailed to establish objectively incompetent performance by his counsel during the\\n\\npenalty phase of his trial. Our inquiry is not “what the best lawyers would have\\n\\ndone. Nor is the test even what most good lawyers would have done. We ask only\\n\\nwhether some reasonable lawyer at the trial could have acted, in the circumstances,\\n\\nas defense counsel acted at trial. . . . We are not interested in grading lawyers’\\n\\nperformances.” White v. Singletary, 972 F.2d 1218, 1220–21 (11th Cir. 1992). And\\n\\n“[i]f fairminded jurists could reasonably disagree” about the reasonableness of\\n\\ncounsel’s performance, “then habeas relief is due to be denied.” Johnson, 643 F.3d\\n\\nat 932.\\n\\n 2. Prejudice\\n\\n Even if Jenkins had demonstrated that his counsel performed as no\\n\\nreasonable lawyer could have, he must also demonstrate a “reasonable probability\\n\\nthat, but for counsel’s unprofessional errors, the result of the proceeding would\\n\\nhave been different.” Strickland, 466 U.S. at 694. Given that the jury here\\n\\nrecommended a sentence of death by the narrowest possible vote, 10 to 2, Jenkins\\n\\nneed establish only “a reasonable probability that at least one juror would have\\n\\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537. “A\\n\\n\\n 34\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 35 of 73\\n\\n\\nreasonable probability is a probability sufficient to undermine confidence in the\\n\\noutcome.” Strickland, 466 U.S. at 694. In measuring that probability here, we will\\n\\n“evaluate the totality of the available mitigation evidence—both that adduced at\\n\\ntrial, and the evidence adduced in the habeas proceeding in reweighing it against\\n\\nthe evidence in aggravation.” Williams, 529 U.S. at 397–98.\\n\\n We begin our analysis by clarifying what exactly Jenkins is arguing about\\n\\nprejudice. We do not know precisely which aggravating and mitigating factors the\\n\\njury may have implicitly found before making its recommendation of death. The\\n\\ntrial court found at sentencing two statutory aggravating factors: the robbery and\\n\\nthe kidnapping of the victim. See Ala. Code § 13A-5-49(4). Because none of the\\n\\nmitigation evidence offered in the Rule 32 hearing bears on those factors, we see\\n\\nno reasonable probability that the aggravating factors would have been determined\\n\\ndifferently by the jury or the court. The court also found at sentencing two\\n\\nstatutory mitigating factors: no significant history of prior criminal activity and\\n\\nage. See id. § 13A-5-51(1), (7). There was also strong evidence of Jenkins’s\\n\\nintoxication, which was a key part of Jenkins’s guilt-phase defense and could have\\n\\nsupported a third mitigating factor, impaired capacity. See id. § 13A-5-51(6).10\\n\\n\\n10\\n The jury had heard evidence about all three of these statutory mitigators during the guilt phase\\nof the trial. Downey’s penalty-phase opening statement reviewed for the jury all seven of the\\npossible statutory mitigating factors, and the court also recited them in its jury instructions,\\nreminding the jury that they could consider guilt-phase evidence as well as non-statutory\\nmitigation. Jenkins has not established that there is a substantial probability that the jury’s\\nrecommendation would have been different had it been more explicitly reminded of Jenkins’s\\n 35\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 36 of 73\\n\\n\\nJenkins does not argue that the jury or the court would have found other statutory\\n\\nmitigators in light of the Rule 32 evidence. Rather, he contends that the jury might\\n\\nhave found non-statutory mitigation based on Jenkins’s abusive childhood, low\\n\\nintelligence, or mental health problems that would have outweighed the\\n\\naggravating factors.11\\n\\n The Rule 32 trial court found that Jenkins had not demonstrated prejudice\\n\\neven assuming that counsel had performed deficiently. See Jenkins, 972 So. 2d at\\n\\n138. It concluded that he had not shown a reasonable probability that the sentencer\\n\\n“would have concluded that a weighing of the aggravating and mitigating\\n\\ncircumstances did not warrant death” upon a reweighing of the evidence. See id. It\\n\\nemphasized the depravity of the crime and the aggravating circumstances and\\n\\nfound that the mitigating witnesses were not credible for various reasons. See id.\\n\\nThe state appellate court affirmed, declining to disturb the credibility\\n\\ndeterminations of the trial court. Id. at 142–44. Independently reweighing the\\n\\n\\nage, criminal history, and intoxication, given that the trial court, the Rule 32 trial court, and the\\nstate appellate court all explicitly considered those mitigators and found them outweighed by the\\naggravators.\\n11\\n Alabama law lists seven statutory mitigating circumstances, but states that mitigation “shall\\n. . . not be limited to” those seven circumstances. Ala. Code § 13A-5-51. Further, “mitigating\\ncircumstances shall include any aspect of a defendant’s character or record and any of the\\ncircumstances of the offense that the defendant offers as a basis for a sentence of life\\nimprisonment without parole instead of death, and any other relevant mitigating circumstances\\nwhich the defendant offers as a basis for a sentence of life imprisonment without parole instead\\nof death.” Id. § 13A-5-52. A jury may recommend a sentence of death only if it finds that “one or\\nmore aggravating circumstances . . . exist and that they outweigh the mitigating circumstances, if\\nany.” Id. § 13A-5-46(e)(3).\\n 36\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 37 of 73\\n\\n\\nevidence, it concluded that, “[g]iven the aggravating circumstances that were\\n\\nproven by the State and the facts surrounding Hogeland’s murder, we, like the\\n\\ncircuit court, are confident that death was the appropriate punishment for Jenkins’s\\n\\nactions.” Id. at 148.\\n\\n Our review of this evidence does not show that it is likely that the jury\\n\\nwould have concluded that death was not the appropriate punishment if it had\\n\\nheard the Rule 32 evidence. Much of the proffered mitigation evidence is, in our\\n\\nview, much less mitigating than Jenkins asserts. In particular, the testimony of\\n\\nJenkins’s older brother Michael was likely to have been a double-edged sword. See\\n\\nEvans v. Sec’y, Dep’t of Corr., 703 F.3d 1316, 1327 (11th Cir. 2013) (en banc) (“it\\n\\nis reasonable to conclude that a defendant was not prejudiced when his mitigation\\n\\nevidence was a two-edged sword” (quoting Ponticelli v. Sec’y, Fla. Dep’t of Corr.,\\n\\n690 F.2d 1271, 1296 (11th Cir. 2012)). Michael was the closest relative to testify at\\n\\nthe Rule 32 hearing and his testimony about the abuse Jenkins experienced should\\n\\nhave been the most reliable, since he experienced much of the same physical abuse\\n\\nhimself. The brothers shared the same filthy bedroom, received beatings from the\\n\\nsame paddle, and were frequently absent from school together. But Michael’s\\n\\ntestimony contained some discrepancies, and he admitted to cognitive and memory\\n\\nproblems. Perhaps Michael’s testimony would have made Jenkins appear more\\n\\nsympathetic in the eyes of the jury—but perhaps not. While Jenkins ran away from\\n\\n\\n 37\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 38 of 73\\n\\n\\nhome and eventually committed murder, Michael responded to the same\\n\\nbackground by running away from home, graduating from high school, and\\n\\nproviding health care for both the terminally ill and people with substance use\\n\\ndisorders. It is possible that the jury could even have found Jenkins to be more\\n\\nculpable for not overcoming his difficult upbringing and to pose more of a danger\\n\\nof future violence.\\n\\n We also find reasonable the state court’s identification of other problems\\n\\nwith the mitigating evidence Jenkins argues should have been presented. Hearing\\n\\nthat evidence firsthand in the Rule 32 proceeding, the trial court found the\\n\\ntestimony of the cousins less credible in light of the lack of physical abuse\\n\\ndocumented in the school health records. See Jenkins, 972 So. 2d at 139 (noting\\n\\nthat school records observed a rash and gingivitis). It faulted Pitts in particular for\\n\\npurporting to have witnessed ongoing, severe abuse but taking only minimal,\\n\\nineffective steps to obtain outside help. See id. It further noted that DeLavega and\\n\\nWagoner actually testified that they never saw Jenkins beaten, and it noted all of\\n\\nthe witnesses’ interest in getting Jenkins off death row. All of these factors would\\n\\nlikely have caused a reasonable jury to discount their testimony. See id. at 139–42.\\n\\nWe must accept the trial court’s views of the Rule 32 testimony. “In the absence of\\n\\nclear and convincing evidence, we have no power on federal habeas review to\\n\\nrevisit the state court’s credibility determinations.” Bishop v. Warden, 726 F.3d\\n\\n\\n 38\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 39 of 73\\n\\n\\n1243, 1259 (11th Cir. 2013) (citing Marshall v. Lonberger, 459 U.S. 422, 434\\n\\n(1983)).\\n\\n The state court also noted that Wagoner testified that she was not available\\n\\nto testify at the trial, and it found that Seal’s testimony would have been\\n\\ncumulative to that of her husband. Thus, counsel’s failure to call those two\\n\\nwitnesses could not have been prejudicial to Jenkins. See id. at 141, 142. Neither of\\n\\nthese conclusions involves an unreasonable determination of the facts. We agree\\n\\nthat the Rule 32 testimony of Sherry Seal was substantially similar to the testimony\\n\\nof her husband Lonnie Seal that the jury heard during the penalty phase. Both\\n\\ntestified about meeting Jenkins in California, Jenkins’s assistance with their move\\n\\nto Alabama, and his residence with them in their Alabama home. Sherry added\\n\\nonly that Jenkins sometimes wet his bed. It was not unreasonable for the state court\\n\\nto conclude that Sherry’s testimony was cumulative of her husband’s, and “a\\n\\npetitioner cannot satisfy the prejudice prong of the Strickland test with evidence\\n\\nthat is merely cumulative of evidence already presented.” Rose v. McNeil, 634 F.3d\\n\\n1224, 1243 (11th Cir. 2011).\\n\\n Further, the school, medical, and juvenile court records that Jenkins\\n\\nintroduced in the Rule 32 proceeding have their own problems. As noted, they do\\n\\nnot document or even corroborate the severe level of abuse alleged by some of the\\n\\nwitnesses. Though these documents tell a sad story, they do not create a substantial\\n\\n\\n 39\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 40 of 73\\n\\n\\nlikelihood that the jury would have recommended life. Although they portray\\n\\nJenkins as a victim of truly unfortunate circumstances, they also show his unsavory\\n\\ndelinquent behavior. That delinquency included a history of stealing cars, and\\n\\ngiven that the jury likely inferred that Jenkins stole the red Mazda, new\\n\\ninformation about his history of theft may not have been especially mitigating.\\n\\nSimilarly, we agree with the state court that evidence of Jenkins’s compliant\\n\\nconduct in pretrial detention is “minimally mitigating” where there had been no\\n\\nevidence of jail misconduct as an aggravator. See Jenkins, 972 So. 2d at 149; cf.\\n\\nSkipper v. South Carolina, 476 U.S. 1, 5 & n.1 (1986) (good jail behavior is\\n\\n“potentially mitigating” especially “[w]here the prosecution specifically relies on a\\n\\nprediction of future dangerousness in asking for the death penalty”).\\n\\n Jenkins argues that allowing the jury to see his experience of mental,\\n\\nemotional, and cognitive difficulties, compounded by years of abuse and neglect\\n\\nand considered alongside his mild character and good jail behavior as an adult,\\n\\nwould have convinced them that Hogeland’s murder was an “aberration” that\\n\\ndeserved a punishment less severe than death. That interpretation of Jenkins’s life\\n\\nstory may be conceivable. But more than mere conceivability is required to\\n\\nestablish prejudice: “The likelihood of a different result must be substantial, not\\n\\njust conceivable.” Richter, 562 U.S. at 112 (emphasis added). We do not find the\\n\\nstate court’s decision that a different result was not substantially likely to be an\\n\\n\\n 40\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 41 of 73\\n\\n\\nunreasonable application of Strickland. As we have noted, we think it equally\\n\\nlikely that the jury could have rejected this alternate narrative out of hand in light\\n\\nof having just convicted him of the violent murder of a young mother. We also\\n\\nthink it possible that the jury could have viewed the murder as just one step in a\\n\\nlogical progression of degeneracy that began with behavioral problems at school,\\n\\ncontinued through the juvenile court system, and escalated past grand theft auto\\n\\ninto kidnapping and murder. Taken separately as well as together, the proffered\\n\\nevidence does not compel us to conclude that a verdict of life would have been\\n\\nsubstantially likely. Because fairminded jurists could disagree about the state\\n\\ncourt’s conclusion about Strickland prejudice, we affirm the denial of habeas relief\\n\\non this ground.\\n\\n B. Intellectual Disability\\n\\n Jenkins argues that the district court erred when it denied his claim that he is\\n\\nintellectually disabled and ineligible for the death penalty under Atkins v. Virginia,\\n\\n536 U.S. 304 (2002). The Supreme Court held in Atkins that executing an offender\\n\\nwho has mental retardation12 violates the Eighth Amendment’s ban on excessive\\n\\n\\n\\n\\n12\\n In 2002, the Atkins Court followed prevailing convention in referring to “mentally retarded”\\noffenders. We will use that now-disfavored term only when quoting or discussing older sources.\\nSince 2014, the Supreme Court has exclusively used the terms “intellectually disabled person”\\nand “person with intellectual disability.” See, e.g., Hall v. Florida, 572 U.S. 701, 704 (2014);\\nMoore v. Texas, 139 S. Ct. 666, 672 (2019).\\n 41\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 42 of 73\\n\\n\\npunishments. Id. at 321. It defined mental retardation by quoting the American\\n\\nAssociation on Mental Retardation:\\n\\n Mental retardation refers to substantial limitations in present\\n functioning. It is characterized by significantly subaverage intellectual\\n functioning, existing concurrently with related limitations in two or\\n more of the following applicable adaptive skill areas: communication,\\n self-care, home living, social skills, community use, self-direction,\\n health and safety, functional academics, leisure, and work. Mental\\n retardation manifests before age 18.\\n\\nId. at 308 n.3. Alabama courts have followed that tripartite definition, adopting\\n\\nwhat they consider the “broadest” definition of mental retardation for Atkins\\n\\npurposes:\\n\\n [A] defendant, to be considered mentally retarded, must have\\n significantly subaverage intellectual functioning (an IQ of 70 or\\n below), and significant or substantial deficits in adaptive behavior.\\n Additionally, these problems must have manifested themselves during\\n the developmental period (i.e., before the defendant reached age 18).\\n\\nEx parte Perkins, 851 So. 2d 453, 456 (Ala. 2002).\\n\\n The clinical evidence adduced during the Rule 32 hearing with respect to\\n\\nJenkins’s intellectual ability was that his IQ was 76, reflecting “borderline\\n\\nintellectual functioning.” The record also included the report of the Lunacy\\n\\nCommission13 which found that Jenkins was competent to stand trial, as well as\\n\\n\\n\\n13\\n Alabama law requires a capital defendant to be “committed to the Department of Mental\\nHealth and Mental Retardation for examination” if there is a question about his capacity to stand\\ntrial or if there will be a mental-disease-or-defect defense. Ala. Code § 15-16-22(a). “The\\nassigned mental health professional(s) shall examine the defendant with respect to determining\\nthe presence of any mental disease or defect which, if determined to be present, would affect the\\n 42\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 43 of 73\\n\\n\\nJenkins’s school and juvenile court records. The court also considered all the\\n\\ntestimony it had heard about Jenkins and reviewed the transcript of his murder trial\\n\\nbefore rejecting Jenkins’s claim of mental retardation.\\n\\n On appeal, after reciting the Perkins standard, the Alabama Court of\\n\\nCriminal Appeals explained that Jenkins had not established mental retardation:\\n\\n Dr. Kirkland testified that he performed psychological tests on Jenkins\\n and that Jenkins’s IQ was 76. There was evidence presented at\\n Jenkins’s trial indicating that Jenkins maintained relationships with\\n other individuals and that he had been employed by P.S. Edwards\\n Landscaping Company, Cotton Lowe 76 Service Station, and\\n Paramount Painting Company. The record fails to show that Jenkins\\n meets the most liberal view of mental retardation adopted by the\\n Alabama Supreme Court in Perkins.\\n\\nJenkins, 972 So. 2d at 155. In this habeas appeal, Jenkins raises arguments about\\n\\nall three components of the intellectual disability showing under Atkins and\\n\\nPerkins, which we address in turn. He also appeals the denial of an evidentiary\\n\\nhearing on this issue.\\n\\n 1. Intellectual Function\\n\\n The first component of intellectual disability under Perkins is “significantly\\n\\nsubaverage intellectual functioning (an IQ of 70 or below).” Perkins, 851 So. 2d at\\n\\n456. Jenkins argues that the use of a strict IQ cutoff of 70 was contrary to or an\\n\\nunreasonable application of Atkins.\\n\\n\\ncapacity of the defendant to proceed or continue to trial or which would affect the defendant’s\\ncriminal responsibility at the time of the commission of the crime.” Id.\\n 43\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 44 of 73\\n\\n\\n We disagree. The state court’s decision that Jenkins failed to establish\\n\\nmental retardation in this respect was a reasonable application of the record and of\\n\\nAtkins. As a threshold matter, we note that the state court did not say it was\\n\\napplying a strict cutoff of 70; it merely stated that Dr. Kirkland had measured\\n\\nJenkins’s IQ as 76. Jenkins, 972 So. 2d at 155. Under our deferential review, the\\n\\nquestion is whether the decision that Jenkins had not established the intellectual\\n\\nfunctioning aspect of intellectual disability was an unreasonable application of\\n\\nAtkins. The Supreme Court explained in Atkins that it was leaving “to the State[s]\\n\\nthe task of developing appropriate ways to enforce the constitutional restriction\\n\\nupon [their] execution of sentences.” Atkins, 536 U.S. at 317 (quoting Ford v.\\n\\nWainwright, 477 U.S. 399, 416–17 (1986)). In line with that invitation, our Circuit\\n\\nhas accepted Alabama’s judicial definition of intellectual disability as laid out in\\n\\nPerkins.14 See Thomas, 607 F.3d at 752–53. Jenkins objects, however, that the\\n\\nPerkins standard, which he characterizes as articulating a “strict cutoff” of 70,\\n\\nconflicts with more recent Supreme Court precedent that instead requires a\\n\\n“clinical” definition that goes beyond a single IQ score.\\n\\n In particular, Jenkins points to Hall v. Florida, in which the Supreme Court\\n\\nheld that when an offender’s IQ is near but greater than 70, courts must take into\\n\\n\\n\\n14\\n Notwithstanding the urging of the Alabama Supreme Court, see Perkins, 851 So. 2d at 455\\nn.1, the Alabama legislature has not enacted a statutory definition of intellectual disability.\\n 44\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 45 of 73\\n\\n\\naccount the IQ test’s standard error of measurement as well as non-IQ evidence of\\n\\nintellectual function. See Hall v. Florida, 572 U.S. 701, 723 (2014) (“This Court\\n\\nagrees with the medical experts that when a defendant’s IQ test score falls within\\n\\nthe test’s acknowledged and inherent margin of error, the defendant must be able\\n\\nto present additional evidence of intellectual disability, including testimony\\n\\nregarding adaptive deficits.”). But, as Jenkins concedes, our Circuit has\\n\\nspecifically held that Hall is not retroactive to cases on collateral review. See In re\\n\\nHenry, 757 F.3d 1151, 1161 (11th Cir. 2014) (explaining that Hall does not supply\\n\\na new substantive rule but rather “merely provides new procedures for ensuring\\n\\nthat States do not execute members of an already protected group”). We also note\\n\\nthat Alabama courts have found the Perkins standard to comport with Hall in that\\n\\nit does not prevent an offender from presenting the standard error of measurement\\n\\nor non-IQ evidence of intellectual disability. See Reeves v. State, 226 So. 3d 711,\\n\\n728–29 (Ala. Ct. Crim. App. 2016). Here, the state court did not unreasonably\\n\\napply Atkins when it analyzed the intellectual functioning component of\\n\\nintellectual disability.\\n\\n Jenkins also argues that the state court’s decision on this component in his\\n\\nRule 32 proceeding resulted from an unreasonable determination of the facts. He\\n\\nasserts that his IQ score of 76, when combined with the test’s standard error of\\n\\nmeasurement and considered alongside the Flynn effect (which we discuss below),\\n\\n\\n 45\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 46 of 73\\n\\n\\ndoes show that he has significantly subaverage intellectual function. For the\\n\\nreasons that follow, we disagree.\\n\\n Most fundamentally, neither of the clinicians who testified about Jenkins’s\\n\\nintellectual ability in the Rule 32 hearing opined that he was mentally retarded.\\n\\nThe closest thing Jenkins can point to is the testimony of Dr. Kirkland that his IQ\\n\\nis two standard deviations below the mean. But his entire testimony, in context, did\\n\\nnot establish intellectual disability. To the contrary: Dr. Kirkland testified that\\n\\nJenkins “scored in the range of borderline intellectual functioning which is\\n\\nbetween mild mental retardation and low average intellectual functioning.”\\n\\nFurthermore, Jenkins’s own expert Dr. Lisak fully reviewed Dr. Kirkland’s clinical\\n\\nfindings. He agreed with those findings, concluding that the IQ results show\\n\\n“borderline intelligence.” It is tremendously significant that no clinical assessment\\n\\nin the entire record of these proceedings has found that Jenkins has mental\\n\\nretardation or intellectual disability.\\n\\n Jenkins nonetheless argues that his IQ score of 76 establishes the intellectual\\n\\ncomponent of intellectual ability when the standard error of measurement and the\\n\\nFlynn effect are considered alongside that score. It was not unreasonable for the\\n\\nstate court to disagree. As we have discussed, Jenkins is not retroactively entitled\\n\\nto have the state courts take into account the margin of error of the IQ test. That\\n\\nstatistical margin of error—5 points in this case—takes into account characteristics\\n\\n\\n 46\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 47 of 73\\n\\n\\nof the test and variations due to chance in order to state how confident the tester is\\n\\nthat the measured result accurately states Jenkins’s true, unknowable IQ. A result\\n\\nof 76 with a standard error of measurement of 5 simply means that the tester is\\n\\n68% confident that Jenkins’s true IQ falls between 71 and 81, and 95% confident\\n\\nthat his true IQ falls between 66 and 86. See David H. Kaye & David A. Freedman,\\n\\nReference Guide on Statistics, in Reference Manual on Scientific Evidence 211,\\n\\n243–44 (3d ed. 2011); see also Hall, 572 U.S. at 738–39 (Alito, J., dissenting). Of\\n\\ncourse, Jenkins would like us to consider only the possibility that the test result\\n\\noverstated his IQ—but the standard error of measurement is a two-way street.\\n\\n“[T]he standard error of measurement is a bi-directional concept that does not\\n\\ncarry with it a presumption that an individual’s IQ falls to the bottom of his IQ\\n\\nrange.” Ledford v. Warden, Ga. Diag. & Clas. Prison, 818 F.3d 600, 641 (11th Cir.\\n\\n2016).\\n\\n Even if Jenkins were somehow entitled to subtract the 5-point margin of\\n\\nerror from his score and thereby lower his “true” IQ to 71, he still would not satisfy\\n\\nthe clinical intellectual component of intellectual disability. Thus, he sought in the\\n\\ndistrict court to introduce evidence about the Flynn effect in order to bring his IQ\\n\\ndown to 70 or below. As our Court has explained, the Flynn effect “recognizes the\\n\\nfact that IQ test scores have been increasing over time”:\\n\\n The Flynn effect acknowledges that as an intelligence test ages, or\\n moves farther from the date on which it was standardized, or normed,\\n 47\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 48 of 73\\n\\n\\n the mean score of the population as a whole on that assessment\\n instrument increases, thereby artificially inflating the IQ scores of\\n individual test subjects. Therefore, the IQ test scores must be\\n recalibrated to keep all test subjects on a level playing field.\\n\\nThomas, 607 F.3d at 753. We have said that “[a]n evaluator may also consider the\\n\\n‘Flynn effect’” in assessing an offender’s possible intellectual disability, id.\\n\\n(emphasis added), but neither we nor the Alabama Supreme Court nor the U.S.\\n\\nSupreme Court have said that a court must consider it in order to reasonably apply\\n\\nAtkins. See, e.g., Reeves, 226 So. 3d at 739 (“This Court has repeatedly held that a\\n\\ncircuit court is not required to accept, consider, or apply the ‘Flynn Effect’ in\\n\\ndetermining intellectual disability.”).\\n\\n More problematically, Jenkins did not raise the Flynn effect issue in the\\n\\nAlabama state courts. Jenkins asserts that he implicitly raised it as part and parcel\\n\\nof his Atkins claim during his Rule 32 appeal, but we do not agree that the\\n\\nAlabama courts had a full and fair opportunity to consider this issue, because it\\n\\nwas not explicitly presented to them. See Kelley v. Sec’y for the Dep’t of Corr., 377\\n\\nF.3d 1317, 1349–50 (11th Cir. 2004) (holding that one ineffective-assistance\\n\\nargument was unexhausted despite the petitioner’s general presentation of a\\n\\nStrickland claim in the state courts). Thus, under AEDPA, Jenkins is procedurally\\n\\nbarred from raising this unexhausted claim in his federal habeas petition. See 28\\n\\nU.S.C. § 2254(b)(1)(A), (c); Mason v. Allen, 605 F.3d 1114, 1119 (11th Cir. 2010).\\n\\nWe will allow a prisoner to overcome a procedural bar if he establishes cause for\\n 48\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 49 of 73\\n\\n\\nhis default and prejudice therefrom. See, e.g., Smith v. Jones, 256 F.3d 1135, 1138\\n\\n(11th Cir. 2001). But Jenkins cannot establish prejudice from being barred from\\n\\nasserting this claim now because the Supreme Court has not “clearly established”\\n\\nthat a court must consider the Flynn effect in assessing intellectual disability. Thus,\\n\\nunder AEDPA, it could not form the basis for a federal habeas claim. See 28\\n\\nU.S.C. § 2254(d)(1). We therefore conclude that the state court did not\\n\\nunreasonably determine the facts or unreasonably apply Atkins with respect to the\\n\\nintellectual component of intellectual disability.\\n\\n 2. Adaptive Function\\n\\n The second component of intellectual disability is “significant or substantial\\n\\ndeficits in adaptive behavior.” Perkins, 851 So. 2d at 456. Jenkins argues that the\\n\\nstate court, which mentioned only Jenkins’s successes in employment and social\\n\\nskills, see Jenkins, 972 So. 2d at 155, failed to consider his deficits. He adds that\\n\\neven the mention of employment success does not tell the whole story. Jenkins\\n\\ninsists that his jobs were menial, his employers took advantage of him, and he held\\n\\nthe jobs for only a few months each.\\n\\n But the state court’s terse discussion of Jenkins’s adaptive behavior did not\\n\\ninvolve an unreasonable determination of the facts or an unreasonable application\\n\\nof Atkins. To the contrary, its decision on this issue was reasonable in light of both\\n\\nthe record and Atkins.\\n\\n\\n 49\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 50 of 73\\n\\n\\n Atkins elaborated that the adaptive-skill component of intellectual disability\\n\\nrequires substantial present limitation in at least two of the following areas:\\n\\n“communication, self-care, home living, social skills, community use, self-\\n\\ndirection, health and safety, functional academics, leisure, and work.” Atkins, 536\\n\\nU.S. at 308 n.3 (quoting definition of American Association on Mental\\n\\nRetardation). The record supports the state court’s judgment that Jenkins does not\\n\\nhave substantial deficits in at least two areas. To be sure, we will assume that\\n\\nJenkins may be substantially limited in functional academics; Dr. Kirkland\\n\\nassessed his reading, spelling, and arithmetic skills at a third-grade level. But the\\n\\nrecord does not compel a finding of substantial present deficits in the areas of\\n\\ncommunication, self-care, home living, and work. Although Jenkins emphasizes\\n\\nhis transient employment, the record shows he was able to find work quickly, to\\n\\nwork hard to support the Seals’ household until Lonnie Seal found a job, and, later,\\n\\nto move into his own home. That record evidence is inconsistent with a substantial\\n\\ndeficit in the area of work.\\n\\n As the State observes, the facts of the crime also fail to show significant\\n\\ndeficits in the areas of communication, self-care, community use, and self-\\n\\ndirection. Jenkins was able to communicate well enough to solicit an alibi and to\\n\\nsell his car, including writing out a bill of sale. He then was able to arrange for his\\n\\nflight from Alabama, including purchasing a bus ticket and hitchhiking across the\\n\\n\\n 50\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 51 of 73\\n\\n\\ncountry. Indeed, throughout his trial and postconviction proceedings, the people\\n\\nwho knew Jenkins as an adult consistently described a man who did not have\\n\\nserious difficulties in communicating, forming relationships, working, and caring\\n\\nfor himself. No witness mentioned these kinds of difficulties in his or her\\n\\ntestimony about Jenkins. There was some evidence that Jenkins was “not\\n\\narticulate” and that his house was messy, but that testimony does not rise to the\\n\\nlevel of overwhelming evidence of substantial functional deficits that would\\n\\nsupport a conclusion that the state court unreasonably determined the facts.\\n\\n Of course, the record also contains evidence of Jenkins’s childhood\\n\\nacademic and social deficits, attributable at least in part to his parents’ neglect, but\\n\\nJenkins’s childhood is not directly relevant to our consideration of his present\\n\\nlimitations. Although we acknowledge that the Supreme Court has noted that such\\n\\n“risk factors” may support further exploration of the possibility of intellectual\\n\\ndisability, see Moore v. Texas, 137 S. Ct. 1039, 1051 (2017), the state court record\\n\\ncontains the results of that exploration. Overall, that record supports the state\\n\\ncourt’s conclusion that Jenkins does not have substantial deficits in adaptive\\n\\nbehaviors.\\n\\n 3. Juvenile Onset\\n\\n The third and final component of intellectual disability is manifestation of\\n\\nthe first and second components “before the defendant reached age 18.” Perkins,\\n\\n\\n 51\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 52 of 73\\n\\n\\n851 So. 2d at 456. Jenkins asserts, the State assumes, and we agree that the state\\n\\ncourt did not rule upon this issue, so there is nothing to which we must defer under\\n\\nAEDPA. See, e.g., Brumfield v. Cain, 135 S. Ct. 2269, 2282 (2015). Thus, we will\\n\\nreview the district court’s legal conclusions on this issue de novo, and its factual\\n\\nfindings for clear error. See, e.g., Grossman v. McDonough, 466 F.3d 1325, 1335\\n\\n(11th Cir. 2006). Jenkins argues that he has satisfied the age-of-onset component\\n\\nby showing “a wide range of evidence of deficits in both intellectual and adaptive\\n\\nfunctioning throughout Jenkins’s childhood.”\\n\\n In fact, the ample record of Jenkins’s childhood, as developed in the Rule 32\\n\\nproceeding, does not point to intellectual disability before age 18. School\\n\\nassessments at age 12 reported an IQ score of 83, an adaptive behavior rating of\\n\\n“low,” and noted overall an intellectual capacity of “average,” a diagnosis of\\n\\ndyslexia, and “poor functional skills.” The psychologist attributed Jenkins’s\\n\\ndifficulties to a combination of “[f]requent moves, changes in schools and family\\n\\ninstability” with “[l]ow ability coupled with a specific learning disability” and\\n\\n“[l]ack of responsibility for self control and poor use of time.” Although Jenkins\\n\\nfailed several grades of school, teachers additionally attributed his difficulties to\\n\\nhis behavioral problems and his high number of absences from school.\\n\\n Similarly, IQ tests ordered by the juvenile court at age 14 gave scores of 81\\n\\nand 86 on different tests, in the “[d]ull–[n]ormal” range and reflecting “a definite\\n\\n\\n 52\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 53 of 73\\n\\n\\nlearning disability.” These records all show a child with serious academic deficits\\n\\nand some intellectual and adaptive deficits, but they do not clearly show an\\n\\nintellectually disabled child. In particular, Jenkins’s childhood IQ scores of 83, 81,\\n\\nand 86 fail to point to intellectual disability or mental retardation. We are not left\\n\\nwith the definite and firm conviction that the district court erred when it found that\\n\\n“Jenkins’s [childhood] intelligence scores and adaptive functioning skills were\\n\\nattributable to his learning disability and other circumstances in his life, rather than\\n\\nmental retardation.”\\n\\n In sum, we agree with the district court that the decision of the Alabama\\n\\nCourt of Criminal Appeals that Jenkins does not have intellectual disability was\\n\\nnot contrary to or an unreasonable application of Atkins v. Virginia.\\n\\n 4. Evidentiary Hearing\\n\\n Finally, Jenkins argues that the district court erred when it denied his request\\n\\nfor an evidentiary hearing on his Atkins claim. If Jenkins diligently attempted to\\n\\npresent this claim in the state court, we review for abuse of discretion. Burgess v.\\n\\nComm’r, Ala. Dep’t of Corr., 723 F.3d 1308, 1320 (11th Cir. 2013). “The district\\n\\ncourt abuses its discretion where ‘such a hearing could enable an applicant to prove\\n\\nthe petition’s factual allegations, which, if true, would entitle the applicant to\\n\\nfederal habeas relief.’” Id. (quoting Schriro v. Landrigan, 550 U.S. 465, 474\\n\\n(2007)). If Jenkins did not diligently attempt to develop the factual basis for this\\n\\n\\n 53\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 54 of 73\\n\\n\\nclaim in the state court, absent certain circumstances, he is not entitled to a hearing\\n\\nand we will affirm the district court’s denial. 28 U.S.C. § 2254(e)(2).\\n\\n The parties strenuously dispute whether Jenkins diligently attempted to\\n\\npresent this issue in the state court. On the one hand, the State argues that Jenkins\\n\\nfailed to allege that he was mentally retarded even after the U.S. Supreme Court\\n\\ndecided Atkins in 2002. On the other hand, Jenkins asserts that he raised the\\n\\npossibility of mental retardation as early as his original Rule 32 petition in 1995.\\n\\n As we have noted, shortly after the Supreme Court decided Atkins, the\\n\\nAlabama Court of Criminal Appeals ordered supplemental briefing on the possible\\n\\nimpact of Atkins on Jenkins. Jenkins’s Rule 32 appeal was pending at that point,\\n\\nhaving been fully briefed and orally argued in the state appellate court. Having\\n\\nreviewed the state court records, we agree with the State that Jenkins did not\\n\\ndiligently attempt to present his Atkins claim in the state courts.\\n\\n “Diligence will require in the usual case that the prisoner, at a minimum,\\n\\nseek an evidentiary hearing in state court.” Williams, 529 U.S. at 437. The record\\n\\nclearly shows that Jenkins did not do so. In the initial supplemental brief on Atkins\\n\\nhe filed in 2002, Jenkins merely requested a future hearing after the Alabama\\n\\nlegislature’s enactment of a statute on executing the mentally retarded. Seventeen\\n\\nyears later, that contingency is still in the future. Only in his reply supplemental\\n\\nbrief did Jenkins then assert, for the first time ever, that he is mentally retarded.\\n\\n\\n 54\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 55 of 73\\n\\n\\n We acknowledge that Jenkins’s initial Rule 32 petition included in its\\n\\nallegations of ineffective assistance of counsel a list of unpresented mitigating\\n\\nevidence including his “developmental[] impair[ment]” and “learning disabilities,\\n\\nlow intelligence, poor comprehension and retarded socialization skills.” We\\n\\nconclude that these tangential Strickland references to his mental and social\\n\\nabilities do not amount to diligence. Thus, because Jenkins did not diligently\\n\\nattempt to develop the factual basis for his Atkins claim in the state courts, we\\n\\naffirm the district court’s denial of his request for an evidentiary hearing.\\n\\n Even if Jenkins had diligently attempted to present this claim in the state\\n\\ncourt, however, we conclude in the alternative that the district court did not abuse\\n\\nits discretion when it denied a hearing. We agree, in light of the ample evidence\\n\\nabout intellectual ability that Jenkins adduced in support of his ineffective-\\n\\nassistance claim, that such a hearing would not have enabled Jenkins to prove that\\n\\nhe is intellectually disabled. See Burgess, 723 F.3d at 1320.\\n\\n IV. CONCLUSION\\n\\n The denial of Jenkins’s petition for a writ of habeas corpus is\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 55\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 56 of 73\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4165,"cells":{"id":{"kind":"number","value":4657888,"string":"4,657,888"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"United States v. Edward Neil Feldman"},"decision_date":{"kind":"string","value":"2019-08-30"},"docket_number":{"kind":"string","value":"16-12978"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before JILL PRYOR and JULIE CARNES, Circuit Judges, and ANTOON,* District Judge.\", \"parties\": \"\", \"opinions\": [{\"author\": \"JULIE CARNES, Circuit Judge:\", \"type\": \"010combined\", \"text\": \"Case: 16-12978 Date Filed: 08/30/2019 Page: 1 of 67\\n\\n\\n [PUBLISH]\\n\\n\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 16-12978\\n ________________________\\n\\n D.C. Docket No. 8:14-cr-00521-JDW-AEP-1\\n\\n\\n\\nUSA,\\n\\n Plaintiff - Appellee,\\n\\nversus\\n\\nEDWARD NEIL FELDMAN,\\nKIM XUAN FELDMAN,\\n\\n Defendants - Appellants.\\n\\n ________________________\\n\\n Appeals from the United States District Court\\n for the Middle District of Florida\\n ________________________\\n\\n (August 30, 2019)\\n\\nBefore JILL PRYOR and JULIE CARNES, Circuit Judges, and ANTOON,*\\nDistrict Judge.\\n\\n\\n *\\n Honorable John Antoon II, United States District Judge for the Middle District of\\nFlorida, sitting by designation.\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 2 of 67\\n\\n\\n\\n\\nJULIE CARNES, Circuit Judge:\\n\\n Following a jury trial, Defendants Dr. Edward Feldman and his wife Kim\\n\\nFeldman appeal their convictions for conspiring to distribute controlled substances\\n\\nwithout a legitimate medical purpose, conspiring to commit money laundering, and\\n\\nthree counts of illegal monetary transactions. Dr. Feldman was also convicted of\\n\\nthree counts of distributing controlled substances, without a legitimate medical\\n\\npurpose, to three individuals resulting in their deaths. Both Defendants raise\\n\\nvarious challenges to their convictions on appeal. Dr. Feldman also challenges his\\n\\n300-month sentence, which incorporated a 20-year mandatory-minimum penalty\\n\\npursuant to 21 U.S.C. § 841(b)(1)(C). After careful review of the parties’ briefs\\n\\nand with the benefit of oral argument, we affirm both Defendants’ convictions, but\\n\\nvacate Dr. Feldman’s sentence.\\n\\n BACKGROUND\\n\\n I. Facts\\n\\n A. The Clinic\\n\\n This case involves the operation of a pain-management clinic in Florida:\\n\\nFeldman Orthopedic and Wellness Center. The clinic was operated by Dr.\\n\\nFeldman, a physician, along with his wife, who served as the clinic’s office\\n\\nmanager and handled all of the clinic’s money. The clinic did not accept\\n\\ninsurance. To obtain an office visit, patients paid $300 in cash or credit for their\\n\\n 2\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 3 of 67\\n\\n\\nfirst visit and $150 for each follow-up visit. Patients were provided a discount for\\n\\nreferring new patients to the clinic. Office visits with Dr. Feldman were typically\\n\\nbrief, as he usually performed cursory examinations—or sometimes no\\n\\nexamination at all.\\n\\n Dr. Feldman was the only physician at the clinic’s first location on 38th\\n\\nAvenue, where he saw approximately 40 to 100 patients per day. According to one\\n\\nformer employee, the clinic’s patients came from out of state and often appeared\\n\\n“stoned” and “drunk.” One of the clinic’s former patients, Mike Shaw, testified\\n\\nthat he came to the clinic because Dr. Feldman was a “writer,” meaning that he\\n\\nwould prescribe any medication suggested by the patient. The clinic sometimes\\n\\nreceived phone calls from concerned family members of patients suffering from\\n\\naddiction, yet Dr. Feldman did not treat those patients any differently.\\n\\n In July 2010, the clinic moved to 66th Street, where Dr. Feldman began\\n\\nseeing 75 to 150 patients per day. Due to heavy law enforcement presence at this\\n\\nlocation, the Feldmans instructed their staff to stop accepting patients from out of\\n\\nstate. They also told their staff to place notes in the patient files if they suspected\\n\\nthat someone was an undercover law enforcement officer. If a patient was flagged\\n\\nas an undercover officer, Dr. Feldman conducted a more extensive physical and\\n\\ndid not prescribe those patients any controlled substances. Based on his belief that\\n\\n\\n\\n\\n 3\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 4 of 67\\n\\n\\nsunglasses and hats could conceal recording devices, Dr. Feldman banned these\\n\\nitems at the clinic’s third and final location on Park Boulevard.\\n\\n In an effort to tighten protocol, the Feldmans also instructed staff to\\n\\ndischarge patients whose urine tested positive for cocaine or other serious drugs.\\n\\nIn addition, the clinic began utilizing the Prescription Drug Monitoring Database—\\n\\na database that shows the prescriptions a patient has previously obtained and serves\\n\\nas a tool for determining whether patients are doctor shopping. Notwithstanding\\n\\nthe above gestures, Mrs. Feldman ultimately told staff to stop conducting urinalysis\\n\\nscreenings because too many patients were being discharged based on those\\n\\nresults. Moreover, the Feldmans decided that no patient would be discharged until\\n\\nDr. Feldman had himself reviewed that patient.\\n\\n In January 2011, a second doctor, Dr. Nancy Bruemmer joined the clinic.\\n\\nDr. Bruemmer often recommended lowering patients’ dosage of medications or\\n\\ndischarging them altogether. Whenever Dr. Bruemmer made these\\n\\nrecommendations, however, Mrs. Feldman made sure that Dr. Feldman saw these\\n\\npatients before they were discharged.\\n\\n In early 2011, law enforcement officers began investigating Dr. Feldman’s\\n\\nclinic. As part of the investigation, several undercover officers posed as patients at\\n\\nthe clinic. Dr. Kevin Chaitoff, a physician who specializes in pain management\\n\\nand anesthesia, reviewed the transcripts from the undercover officers’ visits to the\\n\\n\\n 4\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 5 of 67\\n\\n\\nclinic and determined that there were several problems with the way Dr. Feldman\\n\\nhandled each visit. Specifically, Dr. Chaitoff determined that Dr. Feldman\\n\\nobtained inadequate patient histories, conducted inadequate physical examinations,\\n\\nand prescribed controlled substances without documented justification.\\n\\n In August 2012, law enforcement officers executed a search warrant at the\\n\\nclinic and seized 3,200 patient files. At the request of Brian Zdrojewski, an agent\\n\\nwith the Drug Enforcement Administration, Dr. Chaitoff randomly selected 30\\n\\npatient files to review in order to determine whether Dr. Feldman was operating\\n\\nwithin the usual course of professional practice with respect to each patient. Dr.\\n\\nChaitoff concluded that none of the patients in the files he reviewed had been\\n\\nprescribed controlled substances for a legitimate medical purpose. Dr. Chaitoff\\n\\nwas later asked to review an additional 18 patient files and he reached the same\\n\\nconclusion as to those files. The investigation revealed at least three patients of\\n\\nDr. Feldman’s who had died while under his care: Joey Mayes, Shannon Wren,\\n\\nand Ricky Gonzalez. Dr. Chaitoff reviewed the files of these patients and\\n\\ndetermined that Dr. Feldman’s treatment of them was not within the usual course\\n\\nof professional practice.\\n\\n B. Money Laundering\\n\\n The investigation also revealed that between 2010 and 2014, the Feldmans\\n\\ndeposited approximately $6,787,103.99 into 37 bank accounts—the majority of\\n\\n\\n 5\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 6 of 67\\n\\n\\nwhich came from cash deposits that were less than $10,000. John Barna, a retired\\n\\ndetective with the Pinellas County Sheriff’s Office, determined that the Feldmans\\n\\npurchased their home on Trilby Avenue, in part, with $190,000 from a savings\\n\\naccount in Dr. Feldman’s name at JPMorgan Chase Bank. Records showed that of\\n\\nthat $190,000, $150,000 was transferred from the clinic’s bank account in $50,000\\n\\nincrements. That transfer is the subject of Count 6. At issue in Count 7, is a\\n\\n$150,000 wire transfer from First Bank Creve Coeur Mark Lewis P.A. Trust\\n\\nAccount that Dr. Feldman used to fund the clinic location on Park Boulevard.\\n\\nFinally, the subject of Count 8 is a wire transfer in the amount of $187,689.50 from\\n\\na Merrill Lynch account in Dr. Feldman’s name to one in Mrs. Feldman’s name.\\n\\n II. Procedural History\\n\\n A federal grand jury subsequently issued an indictment charging the\\n\\nFeldmans with: (1) conspiring to distribute and dispense controlled substances—\\n\\nprimarily oxycodone and methadone, which are Schedule II controlled substances,\\n\\nand alprazolam and diazepam, which are Schedule IV controlled substances—not\\n\\nfor a legitimate medical purpose, in violation of 21 U.S.C. §§ 841(b)(1)(C), (b)(2)\\n\\nand 846 (Count 1); (2) conspiring to commit money laundering, in violation of 18\\n\\nU.S.C. § 1956(h) (Count 5); and (3) engaging in monetary transactions with\\n\\ncriminal derived property exceeding the value of $10,000, in violation of 18 U.S.C.\\n\\n§ 1957 (Counts 6 through 8). Dr. Feldman was also charged with three substantive\\n\\n\\n 6\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 7 of 67\\n\\n\\ncounts of distribution and dispensation of Schedule II and Schedule IV controlled\\n\\nsubstances, not for a legitimate medical purpose, resulting in the deaths of Joey\\n\\nMayes (Count 2), Ricky Gonzalez (Count 3), and Shannon Wren (Count 4), in\\n\\nviolation of § 841(a)(1), (b)(1)(C) (Counts 2 through 4).\\n\\n The Feldmans pled not guilty and proceeded to trial. The first trial ended in\\n\\na mistrial. At the second trial, the Government presented testimony from the case\\n\\nagents, undercover officers, former patients and employees, the medical examiners,\\n\\nand Dr. Chaitoff. At the close of the Government’s case-in-chief, the Feldmans\\n\\nmoved for judgment of acquittal. Dr. Feldman argued that the Government failed\\n\\nto meet its burden of proof as to each count. The district court denied the\\n\\nFeldmans’ motions for judgment of acquittal.\\n\\n Dr. Feldman called several witnesses on his behalf. Dr. Simopoulos, a pain\\n\\nmanagement physician, testified that, based on his review of the patient files, he\\n\\nbelieved that Dr. Feldman’s prescribing practices were within the usual course of\\n\\nprofessional practice. Dr. Vernard Adams, a forensic pathologist and former\\n\\nmedical examiner, opined about the cause of death for each of the three victims.\\n\\nDr. Feldman also testified. He denied that he had an agreement with Mrs. Feldman\\n\\nor anyone else at the clinic to engage in illegal activity. After the defense rested,\\n\\nthe district court denied the parties’ renewed motions for judgment of acquittal.\\n\\n\\n\\n\\n 7\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 8 of 67\\n\\n\\n The jury found the Feldmans guilty as charged. The district court sentenced\\n\\nDr. Feldman and Mrs. Feldman to imprisonment terms of 300 months and\\n\\n48 months, respectively. Dr. Feldman’s 300-month sentence incorporated the 20-\\n\\nyear mandatory minimum sentence under 21 U.S.C. § 841(b)(1)(C), based on his\\n\\nconvictions for Counts 2 through 4.1 This appeal followed.\\n\\n DISCUSSION\\n\\n I. Motion to Sever\\n\\n Mrs. Feldman argues that the district court abused its discretion by denying\\n\\nher motion to sever her trial from Dr. Feldman’s.\\n\\n We review the district court’s denial of a motion for severance for an abuse\\n\\nof discretion. United States v. Cobb, 185 F.3d 1193, 1196 (11th Cir. 1999). If it\\n\\nappears that a joint trial will prejudice a defendant or the Government, the district\\n\\ncourt may sever the defendants’ trials. Fed. R. Crim. P. 14(a). “In assessing the\\n\\nmerits of a severance motion, the district court must balance the possibility of\\n\\nprejudice to the defendant against the public interest in judicial efficiency and\\n\\neconomy.” United States v. Eyster, 948 F.2d 1196, 1213 (11th Cir. 1991). To\\n\\nobtain reversal of a district court’s denial of a severance motion, a defendant must\\n\\nmake a showing of “compelling prejudice” against which the district court offered\\n\\n\\n\\n1\\n Dr. Feldman received ten-year concurrent sentences on each of his other five counts of\\nconviction.\\n 8\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 9 of 67\\n\\n\\nno protection. United States v. Walser, 3 F.3d 380, 385 (11th Cir. 1993). “A\\n\\ndefendant can show compelling prejudice by demonstrating that the jury was\\n\\nunable to sift through the evidence and make an individualized determination as to\\n\\neach defendant.” United States v. Ramirez, 426 F.3d 1344, 1352 (11th Cir. 2005)\\n\\n(quotation marks omitted).\\n\\n Prior to trial, Mrs. Feldman moved to sever Counts 2 through 4—which\\n\\ncharged Dr. Feldman with distributing and dispensing controlled substances,\\n\\nwithout a legitimate medical purpose, resulting in death. She argued that she\\n\\nwould be substantially prejudiced by a joint trial that included those counts\\n\\nbecause they applied to Dr. Feldman only. The district court denied the motion.\\n\\n Mrs. Feldman has not met her burden of showing that the district court\\n\\nabused its discretion by denying her motion for severance. The Feldmans were\\n\\nindicted together and charged with being part of two conspiracies: (1) the\\n\\ndistribution of controlled substances without a legitimate medical purpose; and\\n\\n(2) money laundering. “It is well settled that defendants who are indicted together\\n\\nare usually tried together.” United States v. Browne, 505 F.3d 1229, 1268 (11th\\n\\nCir. 2007); see also United States v. Knowles, 66 F.3d 1146, 1158 (11th Cir. 1995)\\n\\n(“This court is reluctant to reverse a district court’s denial of severance,\\n\\nparticularly in conspiracy cases, as generally ‘persons who are charged together\\n\\nshould also be tried together.’”).\\n\\n\\n 9\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 10 of 67\\n\\n\\n Mrs. Feldman asserts that she was prejudiced by the admission of evidence\\n\\nrelating to Counts 2 through 4 because those counts applied only to Dr. Feldman.\\n\\nWe have concluded, however, that “[a] defendant does not suffer compelling\\n\\nprejudice, sufficient to mandate a severance, simply because much of the evidence\\n\\nat trial is applicable only to co-defendants.” United States v. Schlei, 122 F.3d 944,\\n\\n984 (11th Cir. 1997) (quotation marks omitted). Moreover, Mrs. Feldman has not\\n\\nmet her burden of demonstrating that the jury was not able to make an\\n\\nindividualized determination of guilt as to the evidence pertaining to her. See\\n\\nUnited States v. Francis, 131 F.3d 1452, 1459 (11th Cir. 1997) (explaining that, in\\n\\norder to establish prejudice, a defendant must show “that the jury was unable to\\n\\nmake an individualized guilt determination for each defendant”).\\n\\n Nevertheless, even if there were any potential prejudice, we have determined\\n\\nthat prejudice can be avoided where the district court instructs the jury that it\\n\\nshould consider the evidence against each defendant separately. See Schlei, 122\\n\\nF.3d at 984. Here, the district court instructed the jury that:\\n\\n [A] separate crime or offense is charged against each defendant\\n in each count of the indictment. Each charge and the evidence\\n pertaining to it should be considered separately. Also, the case of each\\n defendant should be considered separately and individually. The fact\\n that you may find one or any . . . of the defendants guilty or not guilty\\n of any of the offenses charged should not affect your verdict as to any\\n other offense or any other defendant.\\n\\n I caution you, members of the jury, that you are here to determine\\n from the evidence in this case whether each defendant is guilty or not\\n 10\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 11 of 67\\n\\n\\n guilty. Each defendant is on trial only for the specific offense alleged\\n in each count of the indictment.\\n\\n We presume that the district court’s cautionary instruction prevented any\\n\\npossible prejudice. See Francis, 131 F.3d at 1459 (explaining that “cautionary\\n\\ninstructions to the jury to consider the evidence separately are presumed to guard\\n\\nadequately against prejudice.” (quotation marks omitted)). Because Mrs. Feldman\\n\\nhas not provided any evidence showing that the jury was not able to make an\\n\\nindividualized determination of guilt, she has failed to demonstrate that she\\n\\nsuffered compelling prejudice or received an unfair trial. See Browne, 505 F.3d at\\n\\n1268 (“[Defendant] must discharge the ‘heavy burden’ of demonstrating\\n\\n‘compelling prejudice’ from the joinder.” (quotation marks omitted)).\\n\\nAccordingly, the district court did not abuse its discretion by denying her motion\\n\\nfor severance.\\n\\n II. Expert Testimony\\n\\n Mrs. Feldman’s next claim of error is that the district court abused its\\n\\ndiscretion by denying her motion to exclude Dr. Chaitoff’s testimony and by\\n\\npermitting Dr. Chaitoff to extrapolate an opinion as to all 3,200 patient files, based\\n\\non his review of a small percentage of those files.\\n\\n We review for abuse of discretion the district court’s determinations\\n\\nregarding the admissibility and reliability of expert testimony. United States v.\\n\\n\\n\\n 11\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 12 of 67\\n\\n\\nFrazier, 387 F.3d 1244, 1258 (11th Cir. 2004). Federal Rule of Evidence 702\\n\\nprovides that a witness who is qualified as an expert may provide an opinion if:\\n\\n (a) the expert’s scientific, technical, or other specialized knowledge\\n will help the trier of fact to understand the evidence or to determine a\\n fact in issue; (b) the testimony is based on sufficient facts or data;\\n (c) the testimony is the product of reliable principles and methods; and\\n (d) the expert has reliably applied the principles and methods to the\\n facts of the case.\\n\\nFed. R. Evid. 702. The district court has wide discretion in deciding how to\\n\\ndetermine reliability. United States v. Abreu, 406 F.3d 1304, 1307 (11th Cir.\\n\\n2005).\\n\\n Prior to trial, Dr. Feldman filed a motion in limine to exclude, or limit, the\\n\\ntestimony of the Government’s proposed expert, Dr. Chaitoff. Following a\\n\\nDaubert 2 hearing, a magistrate judge determined that Dr. Feldman’s objection to\\n\\nDr. Chaitoff’s testimony under Rule 702 was without merit, “at least to the extent\\n\\nthat the opinion does not extrapolate from the reviewed files to a broader\\n\\nconclusion involving files not reviewed.” The magistrate judge noted that Dr.\\n\\nChaitoff’s testimony at the hearing resolved any potential issue pertaining to\\n\\nextrapolation, as he testified that the review of one patient’s file would be\\n\\ninsufficient to draw conclusions regarding files that were not reviewed.\\n\\n\\n\\n\\n2\\n Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993).\\n 12\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 13 of 67\\n\\n\\n Mrs. Feldman now challenges for the first time on appeal an isolated\\n\\nstatement made by Dr. Chaitoff at trial. Specifically, while discussing his selection\\n\\nand review of the first 30 patient files, Dr. Chaitoff stated that, “the methodology\\n\\nwas random and the conclusions applied to all the patients. So whether I reviewed\\n\\n30 or I reviewed 3,200, I suspect my conclusions would have been identical.” As\\n\\nnoted, Mrs. Feldman did not object to this statement.\\n\\n The Federal Rules of Evidence provide that a defendant need not object to\\n\\npreserve a claim of error on appeal if the district court makes a definitive ruling on\\n\\na pretrial motion, such as a motion in limine. See Fed. R. Evid. 103(b) (“Once the\\n\\ncourt rules definitively on the record—either before or at trial—a party need not\\n\\nrenew an objection or offer of proof to preserve a claim of error for appeal.”). That\\n\\nis not what happened here, however. The magistrate judge did not expressly rule\\n\\non the merits of the extrapolation issue, stating only that the objection to Dr.\\n\\nChaitoff’s testimony pursuant to Rule 702 was without merit “at least to the extent\\n\\nthat the opinion does not extrapolate from the reviewed files to a broader\\n\\nconclusion involving files not reviewed.” Moreover, it is clear from the order that\\n\\nthe magistrate judge believed that any potential extrapolation by Dr. Chaitoff was a\\n\\nnon-issue, as he noted that Dr. Chaitoff had specifically testified that his review of\\n\\none patient’s file would be insufficient to draw conclusions as to other patient files\\n\\nthat had not been reviewed.\\n\\n\\n 13\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 14 of 67\\n\\n\\n Because the magistrate judge did not make a definitive ruling on the\\n\\nextrapolation issue, Mrs. Feldman was required to object to Dr. Chaitoff’s\\n\\nstatement to avoid plain error review. See United States v. Wilson, 788 F.3d 1298,\\n\\n1313 (11th Cir. 2015) (reviewing defendant’s argument for plain error where the\\n\\ndistrict court’s pretrial ruling was not definitive); Cf. United States v. Harris, 886\\n\\nF.3d 1120, 1127 n.2 (11th Cir. 2018) (concluding that defendant’s argument on\\n\\nappeal was not subject to plain error review—despite the defendant’s failure to\\n\\nobject at trial—because the district court definitively ruled on the motion in\\n\\nlimine). She did not object, so we review her argument for plain error. 3\\n\\n Mrs. Feldman cannot prevail under plain error review. Even if she could\\n\\nshow an error that was plain, she cannot demonstrate that her substantial rights\\n\\nwere violated. Indeed, she cannot show that she was prejudiced—much less\\n\\nsubstantially prejudiced—by one stray comment made by Dr. Chaitoff concerning\\n\\nhis suspicion as to the remaining 3,000+ patient files not reviewed. The comment\\n\\nwas made during approximately three days’ worth of testimony from Dr. Chaitoff.\\n\\nDuring that testimony, Dr. Chaitoff focused entirely on the 48 patient files that he\\n\\nreviewed and his opinion as to whether Dr. Feldman’s treatment of the specific\\n\\n\\n\\n3\\n To show plain error, the defendant must establish that there was “(1) an error (2) that is plain\\nand (3) that has affected the defendant’s substantial rights; and if the first three prongs are met,\\nthen a court may exercise its discretion to correct the error if (4) the error ‘seriously affect[s] the\\nfairness, integrity or public reputation of judicial proceedings.’” United States v. Madden, 733\\nF.3d 1314, 1320 (11th Cir. 2013) (quotation marks omitted).\\n 14\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 15 of 67\\n\\n\\npatients in those files was within the ordinary course of professional practice.\\n\\nAccordingly, Mrs. Feldman is not entitled to relief on this claim.\\n\\n III. Double Jeopardy\\n\\n Mrs. Feldman asserts that the district court’s order declaring a mistrial\\n\\nduring the first trial was not supported by manifest necessity, and thus the court\\n\\nshould have granted her motion to dismiss the indictment on double jeopardy\\n\\ngrounds. She also argues that the district court failed to comply with Federal Rule\\n\\nof Criminal Procedure 26.3, which requires the court to give the parties an\\n\\nopportunity to object or consent to an order declaring mistrial.\\n\\n We review the district court’s declaration of a mistrial on grounds of\\n\\nmanifest necessity to determine whether the court exercised sound discretion in\\n\\ndeciding that such a necessity existed. United States v. Davis, 708 F.3d 1216, 1221\\n\\n(11th Cir. 2013). And “if the court’s decision to declare a mistrial was not based\\n\\non manifest necessity, it was an abuse of discretion not to dismiss the indictment\\n\\non double jeopardy grounds.” See id.\\n\\n The Double Jeopardy Clause of the Fifth Amendment safeguards a criminal\\n\\ndefendant from being subjected to multiple prosecutions for the same offense.\\n\\nU.S. Const. amend. V. “Ordinarily, when ‘a defendant successfully seeks to avoid\\n\\nhis trial prior to its conclusion by a motion for mistrial, the Double Jeopardy\\n\\n\\n\\n\\n 15\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 16 of 67\\n\\n\\nClause is not offended by a second prosecution.’” United States v. Shelley, 405\\n\\nF.3d 1195, 1200 (11th Cir. 2005) (quotation marks omitted).\\n\\n “[D]istrict courts are permitted to declare a mistrial and discharge a jury\\n\\nonly where, ‘taking all the circumstances into consideration, there is a manifest\\n\\nnecessity for the act, or the ends of public justice would otherwise be defeated.’”\\n\\nUnited States v. Therve, 764 F.3d 1293, 1298 (11th Cir. 2014) (quotation marks\\n\\nomitted). “Our task is to decide whether the district court exercised ‘sound\\n\\ndiscretion’ in declaring a mistrial, and in order to determine that ‘we review the\\n\\nentire record in the case without limiting ourselves to the actual findings of the trial\\n\\ncourt.’” Davis, 708 F.3d at 1221 (quotation marks omitted and citation omitted).\\n\\nBut “the requirement that manifest necessity be demonstrated operates[] only when\\n\\nthe trial court has declared a mistrial without the consent of the defendant.” United\\n\\nStates v. Puleo, 817 F.2d 702, 705 (11th Cir. 1987). If the defendant consents to a\\n\\nmistrial, the Double Jeopardy Clause will not bar her retrial. Id.\\n\\n The facts surrounding the declaration of a mistrial during the first trial are as\\n\\nfollows. During the first trial, Mrs. Feldman’s attorney asked Dr. Feldman on\\n\\ncross-examination if he had ever been convicted of a felony. Dr. Feldman stated\\n\\nthat he had been convicted of one felony. Thereafter, the prosecutor asked Dr.\\n\\nFeldman whether the felony he was convicted of related to his practice as a doctor.\\n\\nDr. Feldman objected and moved for a mistrial, arguing that a curative instruction\\n\\n\\n 16\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 17 of 67\\n\\n\\nwas not going to cure the prejudice. Indeed, the defendants were united in their\\n\\nposition that a mistrial was necessary. Mrs. Feldman indicated that she had “the\\n\\nsame position as this is a conspiracy case with regard to that I agree that, I agree\\n\\nthat I don’t believe a curative instruction is going to—.” Following subsequent\\n\\ndiscussion, the district court recessed for the day. The next morning, the district\\n\\ncourt heard argument from the Government and Dr. Feldman as to whether the\\n\\nprejudicial effect of the question could be remedied by a curative instruction.\\n\\nUltimately, the district court decided that the defendants were correct in asserting\\n\\nthat a mistrial was necessary because the court concluded that it was not possible\\n\\nto cure this issue. The court thus declared a mistrial. During this lengthy\\n\\ndeliberative process by the court, Mrs. Feldman never indicated any change in her\\n\\noriginal position that a mistrial was necessary nor did she object when the district\\n\\ncourt ultimately acquiesced to the defense position that a mistrial was required.\\n\\n Instead, almost a month later—when the jury was long gone—she filed a\\n\\nmotion arguing that the court was wrong to conclude that manifest necessity\\n\\nrequired a mistrial. She further argued that she was not provided the opportunity to\\n\\nconsent or object to the order declaring a mistrial. The district court denied the\\n\\nmotion.\\n\\n Mrs. Feldman argues that the Double Jeopardy Clause barred her retrial\\n\\nbecause a mistrial was not manifestly necessary. But we need not address manifest\\n\\n\\n 17\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 18 of 67\\n\\n\\nnecessity. Because Mrs. Feldman acquiesced in the court’s decision to declare a\\n\\nmistrial, she cannot now be credibly heard to argue that the court’s decision was a\\n\\nmistake.\\n\\n Mrs. Feldman argues that she could not have agreed with the district court’s\\n\\ndecision to declare a mistrial because the district court failed to provide proper\\n\\nnotice before declaring a mistrial. Federal Rule of Criminal Procedure 26.3\\n\\nprovides that prior to ordering a mistrial, “the court must give each defendant and\\n\\nthe government an opportunity to comment on the propriety of the order, to state\\n\\nwhether that party consents or objects, and to suggest alternatives.” Fed. R. Crim.\\n\\nP. 26.3. “[C]ompliance with Rule 26.3 is only one factor to be considered in\\n\\ndetermining whether a trial judge exercised sound discretion in declaring a\\n\\nmistrial.” Davis, 708 F.3d at 1224 (citing United States v. Berroa, 374 F.3d 1053,\\n\\n1058 (11th Cir. 2004)).\\n\\n We conclude that Mrs. Feldman gave implied consent to the mistrial. The\\n\\nrecord in the present case clearly shows that Mrs. Feldman had numerous\\n\\nopportunities to comment or object to the potential declaration of a mistrial, and\\n\\nshe never gave even a hint that she objected during the lengthy proceedings\\n\\nconcerning this matter. Indeed, after Dr. Feldman moved for a mistrial, Mrs.\\n\\nFeldman agreed that the prejudice could not be cured by a curative instruction, and\\n\\nthen noted that this was a conspiracy case. When the court reconvened after the\\n\\n\\n 18\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 19 of 67\\n\\n\\novernight recess, Mrs. Feldman raised no objection to the court declaring a\\n\\nmistrial, nor did she object or argue that the court should allow the ongoing trial to\\n\\nproceed against her alone. Even as the jury was departing, she said nothing to the\\n\\ncourt indicating that a mistrial should not have been declared as to her. See Puleo,\\n\\n817 F.2d at 705 (concluding that defendant consented to mistrial when “even as the\\n\\njury was departing . . . [counsel] said nothing regarding the mistrial”).\\n\\n In deciding a Double Jeopardy challenge to a second trial, we have recently\\n\\nreconfirmed Puleo’s teaching that “a defendant’s consent to a mistrial need not be\\n\\nexpress but may always be implied from the totality of circumstances.” United\\n\\nStates v. Isaac Feldman, 931 F.3d 1245, 1256–57 (11th Cir. 2019) (quotation\\n\\nmarks omitted and alterations accepted). As with the present case, we noted in\\n\\nIsaac Feldman that, albeit aware that the district court was in the process of\\n\\ndeclaring a mistrial, the defendant “never voiced any objection to the jury’s\\n\\ndismissal . . . . The totality of these circumstances compels the conclusion that [the\\n\\ndefendant] impliedly consented to the dismissal of the original jury . . . .” Id. at\\n\\n1257.\\n\\n So it is here. Mrs. Feldman implicitly consented to the order declaring a\\n\\nmistrial. We therefore conclude that she is entitled to no relief on her claim\\n\\nchallenging the district court’s denial of her motion to dismiss the indictment on\\n\\ndouble jeopardy grounds.\\n\\n\\n 19\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 20 of 67\\n\\n\\n IV. Prosecutorial Misconduct\\n\\n Mrs. Feldman argues that the prosecutor’s closing argument was improper\\n\\nand prejudicial because the prosecutor argued facts not in evidence, put forth a new\\n\\nlegal theory of conviction not alleged in the indictment, and interjected her own\\n\\npersonal opinion to inflame the jury.\\n\\n We generally review allegations of prosecutorial misconduct de novo.\\n\\nUnited States v. Eckhardt, 466 F.3d 938, 947 (11th Cir. 2006). Where a defendant\\n\\nfails to object during a prosecutor’s closing argument, however, “relief is available\\n\\nto rectify only plain error that is so obvious that failure to correct it would\\n\\njeopardize the fairness and integrity of the trial.” United States v. Bailey, 123 F.3d\\n\\n1381, 1400 (11th Cir. 1997).\\n\\n To prevail on a prosecutorial misconduct claim, the defendant must show\\n\\nthat the conduct was improper and that it was prejudicial to his substantial rights.\\n\\nId. A defendant’s substantial rights are prejudiced if there is a reasonable\\n\\nprobability that, but for the remarks, the outcome of the trial would have been\\n\\ndifferent. United States v. Lopez, 590 F.3d 1238, 1256 (11th Cir. 2009).\\n\\n“Prosecutorial misconduct must be considered in the context of the entire trial,\\n\\nalong with any curative instruction.” Id. If the district court provided a curative\\n\\ninstruction, “we will reverse only if the evidence is so prejudicial as to be incurable\\n\\nby that measure.” Id.\\n\\n\\n 20\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 21 of 67\\n\\n\\n Mrs. Feldman raises several challenges to the prosecutor’s remarks during\\n\\nclosing argument. We conclude that only a few of those challenges warrant any\\n\\ndiscussion here. 4 First, Mrs. Feldman asserts that the prosecutor improperly\\n\\ncommented on facts that were not in evidence when discussing Dr. Feldman’s\\n\\nexpert, Dr. Adams. Specifically, she takes issue with the prosecutor’s statement\\n\\nthat when Dr. Adams was employed by the county medical examiner’s office, he\\n\\nhad found that people who died at home had died of drug overdoses. However,\\n\\nnow that he was being paid as a defense expert, he had concluded that one of the\\n\\nvictims who died at home—Shannon Wren—died from a heart attack, not a drug\\n\\noverdose. We find nothing improper about the prosecutor’s statement, as it was\\n\\nsupported by Dr. Adams’ testimony that he had previously made findings of drug\\n\\noverdose deaths where people were found dead in their homes.5\\n\\n We next address Mrs. Feldman’s assertion that the prosecutor inserted a new\\n\\ntheory of conviction not alleged in the indictment by arguing that Mrs. Feldman\\n\\n\\n\\n\\n4\\n Mrs. Feldman argues that the prosecutor’s following remarks were also improper: (1) a\\nstatement that Mayes’ sister testified that Dr. Feldman was surprised when she accused Dr.\\nFeldman of prescribing Xanax and methadone to her brother; (2) a statement implying that there\\nwere several actions the Department of Health could have taken against Dr. Feldman; and (3) a\\nstatement that Mrs. Feldman had spoken with Aimee Martin about ways to stay off the Drug\\nEnforcement Administration’s radar. We find no merit to these challenges.\\n5\\n Notably, the district court also provided the following curative instruction when Mrs. Feldman\\nobjected to the prosecutor’s statement regarding Dr. Adams’ testimony: “[t]o the extent the\\nlawyers disagree on what the facts are, members of the jury, you will resolve those\\ndisagreements. These are arguments of counsel, they are not evidence.”\\n 21\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 22 of 67\\n\\n\\nviolated an Internal Revenue Service reporting requirement by failing to disclose\\n\\nforeign bank accounts. When addressing the money laundering convictions in her\\n\\nclosing argument, the prosecutor stated that she asked defense expert, Stephen\\n\\nOscher, if he noticed whether Mrs. Feldman had disclosed to the federal\\n\\ngovernment that she had bank accounts outside of the country. After the district\\n\\ncourt overruled Mrs. Feldman’s objection to this statement, the prosecutor\\n\\nexplained that Oscher had said he did not know, he had not seen anything. The\\n\\nprosecutor then stated, “[t]hat’s also a requirement, a reporting requirement.” In\\n\\nother words, the prosecutor appeared to imply that Mrs. Feldman had violated a\\n\\nreporting requirement by failing to disclose foreign bank accounts. The\\n\\nindictment, however, charged Mrs. Feldman with violating a transaction reporting\\n\\nrequirement, not failing to report the existence of a foreign bank account.\\n\\n Nevertheless, even if we assume that this remark was improper, it does not\\n\\namount to prosecutorial misconduct because Mrs. Feldman has not met her burden\\n\\nof showing that her substantial rights were violated. In the first place, the district\\n\\ncourt provided curative instructions. See Lopez, 590 F.3d at 1256 (explaining that\\n\\n“we will reverse only if the evidence is so prejudicial as to be incurable by that\\n\\nmeasure”). The court instructed the jury that their decision must be based on\\n\\nevidence, which “includes the testimony of the witnesses and the exhibits which\\n\\nhave been admitted in the record . . . anything the lawyers say is not evidence in\\n\\n\\n 22\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 23 of 67\\n\\n\\nthe case.” The record likewise contained sufficient evidence of guilt independent\\n\\nof the challenged remarks. Id. (“When the record contains sufficient independent\\n\\nevidence of guilt, any error is harmless.” (citing Eckhardt, 466 F.3d at 947)). As\\n\\ndiscussed infra, the Government presented sufficient evidence from which a\\n\\nreasonable jury could conclude that Mrs. Feldman was guilty beyond a reasonable\\n\\ndoubt. In short, the challenged comment was a small portion of the 17-day trial in\\n\\nwhich the Government presented substantial evidence of guilt against Mrs.\\n\\nFeldman.\\n\\n Finally, we reject Mrs. Feldman’s argument that the prosecutor’s statement\\n\\nthat Wren had a “butt-load” of drugs in his body was unfairly prejudicial. Because\\n\\nMrs. Feldman failed to object to this statement, our review is limited to plain error.\\n\\nMrs. Feldman has failed to show error, let alone plain error. See United States v.\\n\\nBailey, 123 F.3d 1381, 1400 (11th Cir. 1997) (“When a defendant fails to object to\\n\\nthe prosecutor’s closing argument, relief is available to rectify only plain error that\\n\\nis so obvious that failure to correct it would jeopardize the fairness and integrity of\\n\\nthe trial.”). We have stated that “there is no prohibition on colorful and perhaps\\n\\nflamboyant remarks if they relate to the evidence adduced at trial.” Id. (quotation\\n\\nmarks omitted). We acknowledge that the prosecutor could have described the\\n\\nquantity of drugs in Wren’s body in a less graphic manner. However, the truth is\\n\\nthat Wren did have a great deal of drugs in his system, and ultimately he died.\\n\\n\\n 23\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 24 of 67\\n\\n\\nThus, the prosecutor’s description was consistent with the evidence presented at\\n\\ntrial. In short, the remark challenged by Mrs. Feldman does not amount to error,\\n\\nlet alone plain error.\\n\\n As a final matter, Mrs. Feldman asserts that the alleged errors, when viewed\\n\\nindividually or cumulatively, prejudiced her substantial rights and deprived her of\\n\\na fair trial. “The cumulative error doctrine provides that an aggregation of non-\\n\\nreversible errors . . . can yield a denial of the constitutional right to a fair trial,\\n\\nwhich calls for reversal.” United States v. Capers, 708 F.3d 1286, 1299 (11th Cir.\\n\\n2013) (quotation marks omitted). Because Mrs. Feldman has established no error,\\n\\nshe cannot prevail under the cumulative error doctrine. See United States v.\\n\\nGamory, 635 F.3d 480, 497 (11th Cir. 2011) (“Where there is no error or only a\\n\\nsingle error, there can be no cumulative error.”).\\n\\n V. Jury Instruction\\n\\n Dr. Feldman argues that the district court abused its discretion by denying\\n\\nhis proposed jury instruction on his theory of defense.6 The specific instruction\\n\\nDr. Feldman requested was the following:\\n\\n It is a physician’s duty and obligation to try to relieve a patient’s pain.\\n Therefore, it is ethical and medically justifiable for a physician to\\n prescribe controlled substances for a patient suffering from chronic\\n moderate to severe pain, even if the patient has developed a tolerance\\n or addiction to those substances. It is also ethical and medically\\n justifiable for such treatment to be performed for the purpose of\\n\\n6\\n Mrs. Feldman adopts this argument.\\n 24\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 25 of 67\\n\\n\\n relieving chronic moderate to severe pain in a patient, regardless of the\\n patient’s [addiction] history.\\n\\nThe district court denied the instruction as argumentative, citing this Court’s\\n\\ndecision in United States v. Paradies, 98 F.3d 1266, 1287 (11th Cir. 1996), where\\n\\nwe rejected a defendant’s proposed jury instruction because it was partisan and\\n\\nsought to have the court parrot the factual findings that the defendant was\\n\\nadvocating to the jury.\\n\\n We review the district court’s refusal to give a requested jury instruction for\\n\\nan abuse of discretion. United States v. Dohan, 508 F.3d 989, 993 (11th Cir.\\n\\n2007). The district court’s refusal to give a requested instruction is reversible error\\n\\nwhere the instruction “(1) was correct, (2) was not substantially covered by the\\n\\ncharge actually given, and (3) dealt with some point in the trial so important that\\n\\nfailure to give the requested instruction seriously impaired the defendant’s ability\\n\\nto conduct his defense.” Eckhardt, 466 F.3d at 947–48.\\n\\n Here, the district court did not abuse its discretion by refusing to give Dr.\\n\\nFeldman’s proposed jury instruction. In the first place, Dr. Feldman has not cited\\n\\nto any legal authority demonstrating that his requested instruction was a correct\\n\\nstatement of the law. Although he cites to Florida Statute § 765.1103—a statutory\\n\\nprovision governing pain management with respect to palliative care—this statute\\n\\ndoes not state that it is ethical for a doctor to prescribe controlled substances for\\n\\nchronic pain, regardless of a patient’s tolerance or addiction issues. See Fla. Stat.\\n 25\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 26 of 67\\n\\n\\n§ 765.1103 (providing that “practitioners . . . must, as appropriate, comply with a\\n\\nrequest for pain management or palliative care from a patient under their care”).\\n\\n Further, Dr. Feldman’s proposed jury instruction essentially asked the\\n\\ndistrict court to instruct the jury about the facts Dr. Feldman hoped the jury would\\n\\nfind: that it was ethically and medically justifiable for Dr. Feldman to prescribe\\n\\ncontrolled substances without respect to the patient’s medical history or addiction\\n\\nissues. See United States v. Maxwell, 579 F.3d 1282, 1304 (11th Cir. 2009)\\n\\n(explaining that the district court is not obliged to present jury instructions that\\n\\n“contain partisan and argumentative statements of law and fact”); Paradies, 98\\n\\nF.3d at 1287 (same). The proposed instruction was plainly argumentative and the\\n\\ncourt was not required to give a theory-of-defense instruction that “was more in the\\n\\nnature of a jury argument than a charge.” United States v. Barham, 595 F.2d 231,\\n\\n245 (5th Cir. 1979). In short, the district court’s refusal to give Dr. Feldman’s\\n\\nproposed instruction was not an abuse of discretion.\\n\\n VI. Sufficiency of the Evidence as to the Conspiracy Counts\\n\\n Both Feldmans challenge the sufficiency of the evidence presented to\\n\\nsupport each of their conspiracy convictions. We review challenges to the\\n\\nsufficiency of the evidence de novo, “viewing the evidence in the light most\\n\\nfavorable to the government and drawing all reasonable inferences and credibility\\n\\nchoices in favor of the jury’s verdict.” United States v. Taylor, 480 F.3d 1025,\\n\\n\\n 26\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 27 of 67\\n\\n\\n1026 (11th Cir. 2007). “If a reasonable jury could conclude that the evidence\\n\\nestablishes guilt beyond a reasonable doubt, we will affirm the verdict.” Browne,\\n\\n505 F.3d at 1253. “Because we recognize that ‘the jury is free to choose between\\n\\nor among the reasonable conclusions to be drawn from the evidence presented at\\n\\ntrial,’ our sufficiency review requires only that a guilty verdict be reasonable, not\\n\\ninevitable, based on the evidence presented at trial.” Id. (quotation marks omitted).\\n\\n A. Conspiracy to distribute controlled substances without a\\n legitimate medical purpose (Count 1)\\n\\n Both defendants challenge their conviction for conspiring to distribute or\\n\\ndispense controlled substances without a legitimate medical purpose. To establish\\n\\na conviction for illegal distribution or dispensation of controlled substances under\\n\\n21 U.S.C. § 841(a)(1), the Government must prove that the defendant distributed\\n\\nor “dispensed controlled substances for other than legitimate medical purposes in\\n\\nthe usual course of professional [medical] practice, and that he did so knowingly\\n\\nand intentionally.” United States v. Azmat, 805 F.3d 1018, 1035 (11th Cir. 2015)\\n\\n(quotation marks omitted). In order to establish a conspiracy conviction, the\\n\\nGovernment must show that: (1) “an agreement existed between two or more\\n\\npersons to commit a crime” and (2) “the defendants knowingly and voluntarily\\n\\njoined or participated in the conspiracy.” United States v. Seher, 562 F.3d 1344,\\n\\n1364 (11th Cir. 2009) (quotation marks omitted). The existence of an agreement\\n\\nmay “be proved by inferences from the conduct of the alleged participants or from\\n 27\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 28 of 67\\n\\n\\ncircumstantial evidence of a scheme.” Azmat, 805 F.3d at 1035 (quotation marks\\n\\nomitted). “A conspiracy conviction will be upheld if ‘the circumstances\\n\\nsurrounding a person’s presence at the scene of conspiratorial activity are so\\n\\nobvious that knowledge of its character can fairly be attributed to him.’” Id.\\n\\n Here, the Feldmans do not challenge the jury’s finding that Dr. Feldman was\\n\\ndispensing controlled substances outside the usual course of professional practice.\\n\\nInstead, their arguments focus on whether Mrs. Feldman conspired with Dr.\\n\\nFeldman to do so. We conclude that the Government presented ample evidence\\n\\nfrom which a reasonable jury could conclude that the Feldmans had an agreement\\n\\nto illegally distribute and dispense prescription medications.\\n\\n The evidence presented at trial showed that Mrs. Feldman was extensively\\n\\ninvolved in the operation of the clinic. She wore a white doctor’s coat, she sat in\\n\\non office visits, she encouraged new patient referrals, and she managed all of the\\n\\nclinic’s money. Because staff members were instructed to copy the sign-in sheets\\n\\nwhen providing Mrs. Feldman with any cash payments, the jury could infer that\\n\\nMrs. Feldman was aware of the large number of patients seen at the clinic each\\n\\nday. According to former patient Charlene Riffle, Mrs. Feldman also wrote out\\n\\nprescriptions ordered by Dr. Feldman while he sat at his desk eating fruit.\\n\\n The Government presented sufficient evidence from which a reasonable jury\\n\\ncould infer not only that Mrs. Feldman was aware that Dr. Feldman was\\n\\n\\n 28\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 29 of 67\\n\\n\\nprescribing controlled substances outside the usual course of medicine but also that\\n\\nshe knowingly participated in his efforts to do so. Mrs. Feldman asked staff\\n\\nmembers to alert her when law enforcement officers were present in the parking lot\\n\\nand, on those days, she parked in the back of the office. Multiple staff members\\n\\nalso testified that Mrs. Feldman directed them to put post-it notes in patient files\\n\\nfor anyone whom they believed was an undercover law enforcement officer. She\\n\\ntold them that post-it notes were important, so they could be removed easily and\\n\\nnot become a permanent item in the file.\\n\\n When the clinic began discharging too many patients, Mrs. Feldman\\n\\ninstructed staff to stop conducting urinalysis screenings. As two experts testified,\\n\\nthe urinalysis screenings are instrumental to determining whether a patient is\\n\\ntaking the medication as prescribed or diverting it to others. And, on at least one\\n\\noccasion, Aimee Martin, a former patient and friend of the Feldmans, asked Mrs.\\n\\nFeldman if she could fill her prescription early because it had run out. Mrs.\\n\\nFeldman suggested that Martin instead take the medication of her husband—who\\n\\nwas also one of Dr. Feldman’s patients—thereby encouraging the use of controlled\\n\\nsubstances outside the usual course of professional practice.\\n\\n Additionally, Mrs. Feldman told former employee Melinda Detwiler that she\\n\\nand Dr. Feldman had moved their assets into her name and the names of her\\n\\nchildren, in case anything happened. Finally, as to Dr. Feldman, he testified that\\n\\n\\n 29\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 30 of 67\\n\\n\\nhe did not have any agreement with Mrs. Feldman to illegally distribute controlled\\n\\nsubstances, but the jury obviously rejected this testimony. We may therefore\\n\\nconsider his disbelieved testimony as substantive evidence of his guilt. See United\\n\\nStates v. Woodard, 459 F.3d 1078, 1087 (11th Cir. 2006) (“[A] defendant’s\\n\\ntestimony—if disbelieved by the jury—may be considered substantive evidence of\\n\\nguilt.”).\\n\\n Although Mrs. Feldman asserts that there was an innocent explanation for\\n\\nher instruction to staff to stop conducing urinalysis screenings—the test was\\n\\npurportedly too costly—Detwiler testified that the reason Mrs. Feldman instructed\\n\\nstaff to stop doing the test was because it was resulting in too many patients being\\n\\ndischarged. Further, Dr. Feldman admitted on cross-examination that the\\n\\nurinalysis screening did not cost the clinic a lot of money, as the screening done in\\n\\nthe clinic was included in the cost of the office visit and the patient bore the cost of\\n\\nany urinalysis screening sent to the laboratory. We are likewise unpersuaded by\\n\\nMrs. Feldman’s argument that she made the comment to Detwiler about shifting\\n\\nher assets because she wanted to have everything in place if Dr. Feldman passed\\n\\naway. Indeed, Detwiler testified that Mrs. Feldman made the comment during a\\n\\nconversation about clinic patients who were being investigated by law\\n\\nenforcement. Viewing all of the evidence in the light most favorable to the\\n\\nGovernment, a reasonable jury could conclude that the Feldmans conspired and\\n\\n\\n 30\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 31 of 67\\n\\n\\nworked together to illegally dispense controlled substances without a legitimate\\n\\nmedical purpose.\\n\\n B. Money Laundering (Counts 5 through 8)\\n\\n The Feldmans also argue that the Government presented insufficient\\n\\nevidence to establish that they conspired to commit money laundering or that they\\n\\ncommitted any substantive money laundering offenses.\\n\\n As to the conspiracy charge alleged in Count 5, the Government needed to\\n\\nprove two elements to establish a conspiracy under § 1956(h): “(1) an agreement\\n\\nbetween two or more persons to commit a money-laundering offense; and\\n\\n(2) knowing and voluntary participation in that agreement by the defendant.”\\n\\nUnited States v. Moran, 778 F.3d 942, 962 (11th Cir. 2015) (quotation marks\\n\\nomitted). An essential element of money laundering conspiracy is that the\\n\\ndefendant “knew that the funds involved in the transactions represented the\\n\\nproceeds of unlawful activity.” See United States v. Awan, 966 F.2d 1415, 1434\\n\\n(11th Cir. 1992).\\n\\n Relying on their argument that the Government presented insufficient\\n\\nevidence as to the drug conspiracy alleged in Count 1, the Feldmans argue that the\\n\\nevidence presented at trial was insufficient to prove beyond a reasonable doubt that\\n\\nthey knew the funds involved in the financial transactions at issue involved\\n\\ncriminal proceeds. However, based on our conclusion that sufficient evidence\\n\\n\\n 31\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 32 of 67\\n\\n\\nsupported the drug conspiracy conviction, it necessarily follows that the Feldmans\\n\\nknew that the transactions involved the proceeds of unlawful activity.\\n\\n Dr. Feldman’s argument that the Government failed to present sufficient\\n\\nevidence showing that the Feldmans conducted these financial transactions with\\n\\nthe intent to conceal the criminal proceeds is also without merit. Concealment was\\n\\nonly one of the objects of the conspiracy. Indeed, Count 5 alleged that the\\n\\nFeldmans conspired to conduct a financial transaction with criminal proceeds from\\n\\nthe drug conspiracy (1) to promote the carrying on of unlawful activity, (2) to\\n\\nconceal the proceeds, (3) to avoid transaction reporting requirements, and (4) to\\n\\nengage in monetary transactions in property derived from a specified unlawful\\n\\nactivity. The jury determined that the Feldmans were guilty of all four objects of\\n\\nthe conspiracy. Further, Dr. Feldman does not challenge the jury’s finding as to\\n\\nthose other three objects, and he has therefore abandoned any challenge he may\\n\\nhave had. See United States v. Jernigan, 341 F.3d 1273, 1283 n.8 (11th Cir. 2003)\\n\\n(stating that an issue not briefed on appeal is considered abandoned). Because the\\n\\nFeldmans argue only that the Government failed to establish their knowledge of\\n\\nthe specified unlawful activity—the drug conspiracy—they are not entitled to relief\\n\\nfrom their money laundering convictions. United States v. Ross, 131 F.3d 970, 984\\n\\n(11th Cir. 1997) (“A guilty verdict in a multi-object conspiracy will be upheld if\\n\\nthe evidence is sufficient to support a conviction of any of the alleged objects.”).\\n\\n\\n 32\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 33 of 67\\n\\n\\n VII. Sufficiency of Evidence as to the Convictions Based on the\\n Dispensation of Controlled Substances Without a Legitimate\\n Medical Purpose that Resulted in Death (Counts 2 through 4)\\n\\n In Counts 2 through 4, Dr. Feldman was charged with distribution of\\n\\ncontrolled substances without a legitimate medical purpose resulting in the deaths\\n\\nof three people: Joey Mayes (Count 2), Ricky Gonzalez (Count 3), and Shannon\\n\\nWren (Count 4). The jury convicted Dr. Feldman on all counts, finding that each\\n\\nvictim died as a result of ingesting drugs that Dr. Feldman had prescribed. On\\n\\nappeal, Dr. Feldman challenges the sufficiency of the evidence showing that\\n\\nSchedule II drugs he prescribed caused each victim’s death: a fact that triggers a\\n\\n20-year mandatory minimum sentence under the Controlled Substances Act.\\n\\n To establish a conviction for unlawful dispensation by a physician under\\n\\n21 U.S.C. § 841(a)(1), the Government must establish that the defendant\\n\\nknowingly and intentionally distributed or dispensed controlled substances other\\n\\nthan for a legitimate medical purpose and outside the ordinary course of\\n\\nprofessional practice. 21 U.S.C. § 841(a)(1); Azmat, 805 F.3d at 1035. There is a\\n\\npenalty-enhancement provision for § 841(a), which provides that a defendant shall\\n\\nbe sentenced to a term of not less than 20 years’ imprisonment, or more than life, if\\n\\nhe dispensed a Schedule I or II drug and death or serious bodily injury “results\\n\\nfrom the use of such substance.” 21 U.S.C. § 841(b)(1)(C). “Because the ‘death\\n\\nresults’ enhancement increase[s] the minimum and maximum sentences to which\\n\\n\\n 33\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 34 of 67\\n\\n\\n[a defendant is] exposed, it is an element that must be submitted to the jury and\\n\\nfound beyond a reasonable doubt.” Burrage v. United States, 571 U.S. 204, 210\\n\\n(2014) (citing Alleyne v. United States, 570 U.S. 99, 115–16 (2013), and Apprendi\\n\\nv. New Jersey, 530 U.S. 466, 490 (2000)).\\n\\n Here, Dr. Feldman’s sufficiency-of-the-evidence argument proceeds in two\\n\\nparts. First, he argues that the evidence was insufficient to prove that it was he\\n\\nwho prescribed the specific drugs that were ingested by each victim. Second, he\\n\\nargues that the evidence was insufficient to prove that the Schedule II drugs\\n\\n(oxycodone and methadone)—prescribed by him and ingested by the victims—\\n\\nwere a but-for cause of each victim’s death. We disagree on both fronts and\\n\\naddress each issue in turn.\\n\\n A. Drugs Ingested by the Deceased Were Prescribed by Dr.\\n Feldman\\n\\n Evidence presented at trial fully supported the jury’s finding that, as charged\\n\\nin the indictment, the specific drugs ingested by each victim (Mayes, Gonzalez,\\n\\nand Wren) were prescribed by Dr. Feldman.\\n\\n 1. Ricky Gonzalez\\n\\n The trial evidence was sufficient to show that Dr. Feldman prescribed the\\n\\nspecific drugs ingested by Ricky Gonzalez. On October 5, 2010, only three days\\n\\nbefore his death, Gonzalez received a prescription from Dr. Feldman for 240 30-\\n\\nmilligram pills of oxycodone and 90 10-milligram pills of diazepam (Valium).\\n 34\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 35 of 67\\n\\n\\nGonzalez’s girlfriend, Casey Osteen, testified that she saw Gonzalez take\\n\\noxycodone and diazepam from prescriptions dispensed by Dr. Feldman the night\\n\\nbefore he died. When she last saw Gonzalez at approximately 11:00 p.m. that\\n\\nnight, he appeared “lethargic” and “groggy.”\\n\\n Gonzalez was found dead in his front yard the next day, October 8, 2010.\\n\\nOfficers found the bottle of diazepam prescribed by Dr. Feldman in Gonzalez’s\\n\\npants pocket. The bottle contained only 8 of the 90 10-milligram pills of diazepam\\n\\nprescribed by Dr. Feldman, as well as 35 other diazepam pills of varying dosages.\\n\\nThe bottle of oxycodone prescribed by Dr. Feldman three days earlier was found\\n\\nempty and smashed near a neighbor’s truck. Finally, the toxicology reports\\n\\nconfirmed that Gonzalez had oxycodone and diazepam—the same two drugs\\n\\nprescribed by Dr. Feldman—in his system. Based on this evidence, a reasonable\\n\\njury could infer that the drugs ingested by Gonzalez were prescribed by Dr.\\n\\nFeldman.\\n\\n 2. Joey Mayes\\n\\n The evidence was also sufficient to demonstrate that Dr. Feldman prescribed\\n\\nthe specific drugs Mayes ingested. Mayes first visited Dr. Feldman on March 12,\\n\\n2010 complaining of back pain. On this visit, Dr. Feldman prescribed Mayes 240\\n\\n30-milligram tablets of oxycodone, 60 2-milligram tablets of alprazolam (Xanax),\\n\\nand 60 2-milligram tablets of methadone. Less than one week later, on March 18,\\n\\n\\n 35\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 36 of 67\\n\\n\\n2010, officers responded to a call from Mayes’ residence reporting his sudden\\n\\ndeath from a possible drug overdose. Mayes’ mother testified at trial that she\\n\\nobserved Mayes taking the medications prescribed by Dr. Feldman the day before\\n\\nhe died. Officers found multiple pill bottles in Mayes’ room, including empty pill\\n\\nbottles for the oxycodone and alprazolam prescriptions dispensed by Dr. Feldman\\n\\non March 12. A subsequent toxicology report revealed that, at the time of death,\\n\\nMayes had in his system methadone or methadone metabolites, oxycodone, and\\n\\nalprazolam: the same three drugs that Dr. Feldman had prescribed.\\n\\n Based on this evidence, a reasonable jury could find beyond a reasonable\\n\\ndoubt that the controlled substances ingested by Mayes were prescribed by Dr.\\n\\nFeldman. Indeed, the controlled substances found in Mayes’ system were the same\\n\\ntype of substances prescribed by Dr. Feldman just one week before. In addition,\\n\\nthe pill bottles containing those prescriptions were found empty in Mayes’ room.\\n\\nSee United States v. Merrill, 513 F.3d 1293, 1299 (11th Cir. 2008) (concluding that\\n\\nsufficient evidence established that a doctor’s recent prescriptions led to the\\n\\npatients’ deaths given the short time period between the prescription and the fact\\n\\nthat the defendant “prescribed for each of these patients the exact type of drug the\\n\\nmedical examiner found to have caused their deaths”).\\n\\n\\n\\n\\n 36\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 37 of 67\\n\\n\\n 3. Shannon Wren\\n\\n Finally, the evidence was sufficient to prove that Dr. Feldman had\\n\\nprescribed the specific drugs Wren ingested. Wren was initially seen by Dr.\\n\\nFeldman on May 18, 2011, at which time Dr. Feldman prescribed Wren 120 15-\\n\\nmilligram tablets of oxycodone, 60 10-milligram tablets of diazepam, and 60 25-\\n\\nmilligram tablets of amitriptyline.\\n\\n Approximately one week later, on May 26, 2011, Wren was seen vomiting\\n\\nand coughing by his girlfriend as she left for work. Later that day, Wren was\\n\\nfound dead in his master bathroom. The first officer on the scene found multiple\\n\\npill bottles in the bathroom. 7 Included in those medications were the containers of\\n\\namitriptyline that Dr. Feldman had prescribed on May 18, which contained only 34\\n\\npills out of the 60-count prescription. Subsequently, Wren’s father handed over to\\n\\nan investigating agent the bottles of oxycodone and diazepam prescribed by Dr.\\n\\nFeldman, which were empty at that time.\\n\\n The subsequent autopsy revealed oxycodone, diazepam and its metabolites,\\n\\nand amitriptyline and its metabolites in Wren’s system. During Wren’s May 18\\n\\nvisit, Dr. Feldman had prescribed oxycodone, diazepam, and amitriptyline, which\\n\\n\\n\\n\\n7\\n These bottles contained melatonin, gabapentin, fluoxetine, metoprolol, trazodone, anastrozole,\\ninjectable testosterone, citalopram, adrenal tea capsule, and amitriptyline. Of these substances,\\nonly amitriptyline was found in Wren’s system during the autopsy. As noted in text, Dr.\\nFeldman had recently prescribed amitriptyline.\\n 37\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 38 of 67\\n\\n\\nwere the same drugs found in Wren’s system after he died. As with the other two\\n\\nvictims, a reasonable jury could conclude that the drugs that Wren ingested were\\n\\nprescribed by Dr. Feldman.\\n\\n In short, with respect to each victim, a jury could readily find beyond a\\n\\nreasonable doubt that the controlled substances the victim had ingested were\\n\\nprescribed by Dr. Feldman.\\n\\n B. Sufficient Evidence Was Presented at Trial From Which a Jury\\n Could Conclude That But for Ingestion of the Schedule II\\n Drugs Prescribed by Dr. Feldman, the Victims Would Not Have\\n Died\\n\\n As noted, the Government had sought an enhancement of Dr. Feldman’s\\n\\nsentence on these three substantive drug counts. To repeat, the enhancement\\n\\nprovision in question imposes a 20-year mandatory minimum sentence and an\\n\\nincreased maximum sentence for a defendant who has distributed a Schedule I or II\\n\\ndrug when death or serious bodily injury “results from the use of such substance.”\\n\\n21 U.S.C. § 841(b)(1)(C). On appeal, Dr. Feldman argues that there was\\n\\ninsufficient evidence to prove that the Schedule II drugs he prescribed caused the\\n\\nvictims’ deaths.\\n\\n As to the role played by a Schedule II drug, the jury found in its special\\n\\nverdict that “but for” the Schedule II and Schedule IV drugs prescribed to each\\n\\nvictim by Dr. Feldman, death would not have resulted. Dr. Feldman argues,\\n\\nhowever, that to convict a person based on death resulting from distribution of a\\n 38\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 39 of 67\\n\\n\\nSchedule II drug, the Government must prove that the Schedule II drug was the\\n\\n“sole,” the “only,” or the “independent” cause of each victim’s death, and that the\\n\\nGovernment failed to do so. More particularly, he argues that because each victim\\n\\nalso had a non-Schedule II drug in their bodies—which was also prescribed by Dr.\\n\\nFeldman—and because the non-Schedule II drug potentially contributed to the\\n\\nvictim’s death, the Schedule II drug cannot be said to have caused the particular\\n\\nvictim’s death.\\n\\n In arguing that the Government must prove that the Schedule II drug alone\\n\\ncaused the death of the victim, Dr. Feldman relies on Burrage v. United States, 571\\n\\nU.S. 204 (2014), which examined the causality requirement of § 841(b)(1)(C) of\\n\\nthe Controlled Substances Act. As we explain below, his argument misreads\\n\\nBurrage, which neither requires proof that a Schedule I or II drug “alone” caused a\\n\\nvictim’s death nor precludes application of § 841(b)(1)(C)’s mandatory minimum\\n\\nsentence in mixed-drug intoxication cases. Instead, Burrage held only that, “where\\n\\nuse of the drug distributed by the defendant is not an independently sufficient\\n\\ncause of the victim’s death or serious bodily injury, a defendant cannot be liable\\n\\nunder the penalty enhancement provision of 21 U.S.C. § 841(b)(1)(C) unless such\\n\\nuse is a but-for cause of the death or injury.” Id. at 218–19 (emphasis added).\\n\\nApplying that standard, we find that sufficient evidence of but-for causation was\\n\\npresented at trial.\\n\\n\\n 39\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 40 of 67\\n\\n\\n 1. Burrage\\n\\n In Burrage, the victim had died after ingesting alprazolam, clonazepam,\\n\\noxycodone, hydrocodone, and heroin. 571 U.S. at 207. The defendant had\\n\\nprovided the victim only with the heroin, which is a Schedule I drug, and he was\\n\\ntried based on that distribution, as well as on the enhancement penalty triggered by\\n\\n§ 841(b)(1)(C) when death results from distribution of a Schedule I or II drug. Id.\\n\\nat 206–07. Neither expert witness could say whether the victim would have lived\\n\\nhad he not taken the heroin; in other words, he might have died even without the\\n\\nheroin, given the other drugs he had taken. Id. at 207. This failure of proof meant\\n\\nthat the Government had not proved that but for the heroin, the victim would have\\n\\nlived. Id. Nevertheless, the district court concluded that the Government was not\\n\\nrequired to prove that the heroin was a “but-for” cause of the victim’s death, and it\\n\\ninstructed the jury that it could find that death resulted from the defendant’s heroin\\n\\ndistribution if it merely found that the heroin was a “contributing cause” of the\\n\\nvictim’s death. Id. at 207–08.\\n\\n The Supreme Court disagreed with the causation standard adopted by the\\n\\ndistrict court. Because § 841(b)(1)(C) does not define the phrase “results from,”\\n\\nthe Court applied its ordinary meaning, concluding that the phrase imposed the\\n\\n“traditional” requirement of “actual causality.” Id. at 210–11. Proving “actual\\n\\ncausality,” the Court explained, generally requires proof that, at a minimum, “the\\n\\n\\n 40\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 41 of 67\\n\\n\\nharm would not have occurred in the absence of—that is, but for—the defendant’s\\n\\nconduct.” Id. at 211 (quotation marks omitted). In other words, a factor can be\\n\\nconsidered an “actual cause” if it was necessary for bringing about that result—that\\n\\nis, if the result would not have occurred without (or “but for”) the influence of the\\n\\nfactor.\\n\\n To illustrate how this but-for notion of “actual causation” applies, the Court\\n\\nused three hypotheticals. Id. at 211–12. In the first hypothetical, “A shoots B,\\n\\nwho is hit and dies.” Id. at 211. In this case, only one factor is in play—A’s\\n\\nconduct in shooting B. The Court explained that, in this example, A actually\\n\\ncaused B’s death because B would not have died “but for” A’s conduct. Id. In\\n\\nother words, A’s conduct was necessary for producing B’s death: had A not shot\\n\\nB, B would have lived.\\n\\n The Court made clear, though, that but-for causality does not require that a\\n\\nsingle factor alone produce the particular result. As the Court explained, “[t]he\\n\\nsame conclusion follows if the predicate act combines with other factors to\\n\\nproduce the result, so long as the other factors alone would not have done so—if,\\n\\nso to speak, it was the straw that broke the camel’s back.” Id. (emphasis added).\\n\\nTo illustrate how a factor can be a but-for cause (that is, a necessary factor for\\n\\nproducing a result) even when other factors are themselves but-for causes, the\\n\\nCourt posed a second hypothetical.\\n\\n\\n 41\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 42 of 67\\n\\n\\n In this hypothetical, a “man debilitated by multiple diseases” ingests poison\\n\\nand dies. Id. Here, the Court explained, the poison “is a but-for cause of his death\\n\\neven if those diseases played a part in his demise, so long as, without the\\n\\nincremental effect of the poison, he would have lived.” Id. Thus, both the poison\\n\\nand the debilitating diseases are but-for causes of the death if each factor played a\\n\\nnecessary role in producing that outcome. Further, the poison and the debilitating\\n\\ndiseases would each constitute a but-for cause of the man’s death—“the straw that\\n\\nbroke the camel’s back”—if the man would have lived had he (1) only had\\n\\ndebilitating diseases, but not taken the poison or (2) only taken the poison, but not\\n\\nhad debilitating diseases. In other words, had one of these factors been absent, the\\n\\ndecedent would not have died. If, however, a debilitating disease was\\n\\n“independently sufficient”—that is, enough by itself—to kill the man at the same\\n\\ntime that he was ingesting the poison, the “incremental effect” of the poison was\\n\\nnot necessary to produce the resulting death. Id. at 211, 218–19. In that case, the\\n\\npoison would not be considered a but-for cause of the man dying; it was not “so to\\n\\nspeak, [] the straw that broke the camel’s back,” because the debilitating disease\\n\\nhad performed that function.8 Id. at 211.\\n\\n\\n8\\n The Supreme Court considered, but did not decide, the question of causation when a victim\\ndies following two concurrent events, each of which would have been enough on its own to\\ncause death. The Court posited a scenario in which “A stabs B, inflicting a fatal wound; while at\\nthe same moment X, acting independently, shoots B in the head also inflicting a fatal wound; and\\nB dies from the combined effects of the two wounds.” Burrage, 571 U.S. at 215 (alterations\\n\\n 42\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 43 of 67\\n\\n\\n The Court further elaborated on the but-for-causation principle in a third\\n\\nhypothetical, asking the reader to consider a baseball game where the visiting\\n\\nteam’s leadoff batter hits a home run in the first inning and the visiting team\\n\\nultimately wins 1 to 0. Id. at 211–12. The Court noted that one could accurately\\n\\nsay that the home run actually caused the victory because the visiting team would\\n\\nnot have won but for the home run—that is, the home run was necessary for the 1-\\n\\n0 victory to occur. Id. The Court explained that the home run was no less a but-\\n\\nfor cause of the victory just because it was only one of many but-for causes that\\n\\nwere each necessary for the visiting team to win. Id. at 212. As the Court put it,\\n\\n“[i]t is beside the point that the victory also resulted from a host of other necessary\\n\\ncauses, such as skillful pitching, the coach’s decision to put the leadoff batter in the\\n\\nlineup, and the league’s decision to schedule the game.” Id. (emphasis in original).\\n\\nThis is so because whether a factor is a but-for cause turns only on whether the\\n\\n\\n\\n\\naccepted) (citation omitted). In other words, B would have died had he only been stabbed.\\nLikewise, he would have died had he only been shot. Each was an independent sufficient cause\\nof his death.\\n\\nQuoting from the noted criminal law treatise, LaFave, Substantive Criminal Law at 468, the\\nSupreme Court noted that the treatise recognizes a special rule for such scenarios providing that\\n“A will generally be liable for homicide even though his conduct was not a but-for cause of B’s\\ndeath (since B would have died from X’s actions in any event).” Id. The Court acknowledged,\\nhowever, that: “We need not accept or reject the special rule developed for [the above scenario],\\nsince there was no evidence here that [the decedent’s] heroin use was an independently sufficient\\ncause of his death. No expert was prepared to say that [the decedent] would have died from the\\nheroin use alone.” Id.\\n 43\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 44 of 67\\n\\n\\nresult would have occurred without it; that is, whether that individual factor made a\\n\\ndifference to the outcome.\\n\\n “By contrast,” the Court says, “it makes little sense to say that an event\\n\\nresulted from or was the outcome of some earlier action if the action merely played\\n\\na nonessential contributing role in producing the event.” Id. Thus, the Court\\n\\nexplains, if we modify the baseball hypothetical such that the visiting team’s\\n\\nleadoff batter hits a home run in the first inning, but the visiting team ultimately\\n\\nwins by 5 to 2 rather than 1 to 0, the home run was not a but-for cause of the\\n\\nvictory. Id. This is so because the visiting team would have won even without the\\n\\nhome run: the visiting team had a total of 5 runs and only needed 3 to win,\\n\\nmeaning that the other runs that the visiting team scored were collectively\\n\\nsufficient on their own to cause the victory without needing the incremental effect\\n\\nof the leadoff hitter’s home run. In this scenario, the leadoff home run played a\\n\\nnonessential role in producing the outcome. It did not “[break] the camel’s back.”\\n\\nId. at 211.\\n\\n 2. Application of Burrage to Multi-Drug-Usage Scenarios\\n\\n In order to determine when one can properly conclude that death has\\n\\nresulted from the ingestion of a Schedule I or II drug, which event triggers\\n\\n§ 841(b)(1)(C)’s sentence enhancement, we move away from our national pastime\\n\\n\\n\\n\\n 44\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 45 of 67\\n\\n\\nto examine how Burrage applies to various multi-drug use scenarios involving the\\n\\ningestion of a Schedule I or II drug together with non-Schedule I or II drugs.\\n\\n First, we posit a deceased who ingested heroin, a Schedule I drug, in an\\n\\namount large enough to constitute a fatal dose for this person had the heroin been\\n\\ntaken alone. In this hypothetical, we further assume that the decedent had also\\n\\ntaken Valium, a Schedule IV drug, 9 but that the Valium, by itself, would not have\\n\\ncaused death. Figuring out causation here is simple. Because the drug user has\\n\\ndied from this fatal dose of heroin, we can readily say that the heroin was a but-for\\n\\ncause of his death. In other words, because the decedent would have lived had he\\n\\nnot ingested the heroin, use of the heroin was necessary for producing the death.\\n\\n Indeed, if the prosecution proved that the drug user took a fatal dose of\\n\\nheroin that by itself would have been enough to produce death, this evidence\\n\\nsatisfies a higher standard for proving actual causation than even Burrage’s but-for\\n\\ntest requires. In this situation, the heroin was not only a necessary, but-for cause of\\n\\nthe death, but it was also independently sufficient, by itself, to cause the user’s\\n\\ndeath, even without the influence of any other factors, including the Valium.\\n\\n Second, we change the facts a bit to assume that the deceased not only\\n\\ningested enough heroin that by itself would be fatal, but he also took enough\\n\\n\\n\\n9\\n Valium, whose generic version is called diazepam, is a Schedule IV drug that does not give\\nrise to an enhancement if death results.\\n 45\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 46 of 67\\n\\n\\nValium that, had the Valium been taken alone, it would likewise have caused\\n\\ndeath. In other words, each of the two drugs independently would have caused\\n\\ndeath. Although the Supreme Court expressly declined to reach this particular\\n\\nscenario,10 there is a strong argument that the heroin should be considered to have\\n\\ncaused the death for purposes of the sentencing enhancement. If actual causation\\n\\ncan be established by proving a factor was necessary to cause a result (meaning it\\n\\nsatisfies the but-for test), it logically follows that actual causation can also be\\n\\nestablished by satisfying a higher standard for proving that the factor was\\n\\nsufficient, by itself, to cause the result. It would create an extremely odd result if\\n\\nthe purveyor of a Schedule I or II drug that was strong enough by itself to cause\\n\\ndeath could escape liability merely because the victim also happened to have taken\\n\\na non-Schedule I or II drug that was itself independently sufficient to cause death.\\n\\nAnd the Supreme Court in Burrage indicated that it did not require satisfaction of\\n\\nthe but-for test when the drug at issue is an independently sufficient cause of the\\n\\nvictim’s death. Burrage, 571 U.S. at 218–19.\\n\\n Our third hypothetical takes us into a scenario in which the heroin is not a\\n\\nbut-for cause. In this hypothetical, where again the deceased has taken both heroin\\n\\nand Valium, we assume that the heroin was not enough, by itself, to induce death,\\n\\nbut that the Valium dosage was strong enough to have caused the decedent to die,\\n\\n\\n10\\n See discussion at n.8, supra.\\n 46\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 47 of 67\\n\\n\\neven had he not ingested any heroin. In this scenario, use of the heroin does not\\n\\nmeet the but-for test because it had no impact on what ultimately happened to the\\n\\ndrug user. He would have died—whether or not he had used heroin—because the\\n\\nSchedule IV Valium was going to kill him anyway.\\n\\n In our fourth and final hypothetical, we assume the heroin would not have\\n\\ncaused the deceased’s death had it been taken alone, but likewise the deceased\\n\\nwould not have died had he taken just the Valium. Yet, by virtue of the fact that he\\n\\ningested the heroin along with the Valium, he died. Here, the heroin can be\\n\\nconsidered the straw that broke the camel’s back. Stated another way, had the\\n\\nvictim never ingested the heroin, the decedent would still be alive. We conclude\\n\\nthat, consistent with Burrage, the but-for test for the heroin is met in this scenario.\\n\\n Dr. Feldman, however, disagrees with our analysis, at least as to this fourth\\n\\nscenario. In the present case, the evidence established that each victim had\\n\\ningested two types of drugs prescribed by the doctor: Schedule II drugs\\n\\n(methadone and oxycodone for one victim and oxycodone for the other two\\n\\nvictims) and a non-Schedule II drug (Xanax as to one victim and Valium as to the\\n\\nother two victims). Further, the evidence supported an inference either that the\\n\\nSchedule II drugs were independently sufficient to cause each victim’s death or, at\\n\\nthe very least, that neither of the non-Schedule II drugs would have caused death\\n\\nhad the decedent not ingested the Schedule II drug(s). Thus, the evidence clearly\\n\\n\\n 47\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 48 of 67\\n\\n\\nestablished that, had each respective victim not used the Schedule II drug in\\n\\nquestion, he would not have died. In other words, but for the Schedule II drugs,\\n\\nthe victims would still be alive.\\n\\n It is Dr. Feldman’s position, though, that causation cannot be established\\n\\nunless the Schedule II drug was the sole or only cause of death. According to the\\n\\ndoctor, if the non-Schedule II drugs (that he had also prescribed together with the\\n\\nSchedule II drugs) played any role in causing the victim’s death, it can no longer\\n\\nbe said that use of the Schedule II drugs resulted in death, even though the\\n\\nSchedule II drugs have met the but-for test. That certainly sounds like an odd\\n\\nconclusion. And it is. Burrage does not require proof that a Schedule I or II drug\\n\\nbe the “only” cause of a victim’s death. Indeed, the Court explicitly rejected that\\n\\ncharacterization of but-for causation, stating that a factor can be a but-for cause\\n\\neven if it “combines with other factors to produce the result,” and that whether “a\\n\\nhost of other necessary causes” might also contribute to the result is “beside the\\n\\npoint.” Id. at 211–12 (emphasis in original). Moreover, a rule requiring proof that\\n\\na drug is the one-and-only but-for cause would be tantamount to requiring a\\n\\nshowing that the drug was “independently sufficient” to cause the death—a bar\\n\\nthat is much higher that the “minimum” showing of but-for causation that Burrage\\n\\nrequired. Id. at 211 (emphasis in original). Notably, Burrage expressly held that\\n\\nthe Government need not prove that a Schedule I or II drug was “an independently\\n\\n\\n 48\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 49 of 67\\n\\n\\nsufficient cause of the victim’s death or serious bodily injury,” stating that only\\n\\nbut-for causation is required, and implying that we can dispense with the but-for\\n\\ntest when a drug is an independently sufficient cause of death. Id. at 218–19.\\n\\n Likewise, Dr. Feldman’s argument that evidence of “mixed-drug\\n\\nintoxication” necessarily precludes a finding that a victim died as a result of a\\n\\nSchedule I or II drug finds no support in Burrage. In Burrage, the medical experts\\n\\ntestified that the victim had died from “mixed drug intoxication with heroin,\\n\\noxycodone, alprazolam, and clonazepam all playing a contributing role.” Id. at\\n\\n207 (quotation marks omitted). Neither expert could say, however, “whether [the\\n\\nvictim] would have lived had he not taken the heroin.” Id. In concluding that\\n\\ninsufficient evidence supported the defendant’s conviction, the Court held that\\n\\nproving actual causation required more than showing that a Schedule I or II drug\\n\\n“contributes to an aggregate force (such as mixed-drug intoxication) that is itself a\\n\\nbut-for cause of death.” Id. at 214. But this statement in no way suggests that a\\n\\ndefendant is off the hook anytime a Schedule I or II drug that he distributed ends\\n\\nup in a deadly drug cocktail that kills a victim. Instead, the Court’s statement was\\n\\nintended to clarify that the Government could not establish that a Schedule I or II\\n\\ndrug actually caused a death without evidence that the Schedule I or II drug\\n\\nitself—as opposed to the drug cocktail as a whole—made a difference to the\\n\\noutcome. Id. at 214, 219 (concluding that the lack of “evidence that [the victim]\\n\\n\\n 49\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 50 of 67\\n\\n\\nwould have lived but for his heroin use” required reversal of the defendant’s\\n\\nconviction). In other words, to prove but-for causation, the Government must drill\\n\\ndown and show that, but for the Schedule I or II drug’s inclusion in the drug\\n\\ncocktail, the victim would have lived. If the Government makes that showing, it is\\n\\n“beside the point” that other drugs were necessary to make the drug cocktail\\n\\ndeadly. See id. at 211–12, 218–19.\\n\\n Indeed, a sister circuit, the Sixth Circuit, has reached the same conclusion.\\n\\nIn United States v. Volkman, 797 F.3d 377, 396 (6th Cir. 2015), the defendant\\n\\ndoctor had been convicted of the enhancement provision based on his dispensing\\n\\nof the Schedule II drug oxycodone to the victim. He argued that because the\\n\\ndecedent died based on the “combined effects of oxycodone, diazepam, and\\n\\nalprazolam,” the Government had failed to prove that the death resulted from the\\n\\ningestion of oxycodone. Id. at 395. Relying on Burrage, the Sixth Circuit panel\\n\\nrejected this argument, holding that “[t]he Government was not required to prove,\\n\\nhowever, that oxycodone was [the victim’s] only cause of death” because “but-for\\n\\ncausation exists where a particular controlled substance—here, oxycodone—\\n\\n‘combines with other factors’—here, inter alia, diazepam and alprazolam—to\\n\\nresult in death,” and “[t]he Government presented sufficient oxycodone-specific\\n\\nevidence for a rational jury to find that, ‘without the incremental effect’ of the\\n\\noxycodone, [the victim] would not have died.” Id. So it is here. As Volkman\\n\\n\\n 50\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 51 of 67\\n\\n\\nexplained, “the existence of other potential contributing causes of death is\\n\\nirrelevant so long as the Government presented sufficient evidence that [the\\n\\nSchedule II controlled substance] was a but-for cause of [the victim’s] death.” Id.\\n\\nat 395–96; see also Roundtree v. United States, 885 F.3d 1095, 1098 (8th Cir.\\n\\n2018) (concluding that where an expert testified that heroin and alcohol had\\n\\nworked together “synergistically” to cause death and where alcohol was not\\n\\nenough by itself to cause death, no reasonable jury could find that the heroin was\\n\\nmerely a contributory factor, as opposed to a but-for cause).\\n\\n With these principles in mind, we conclude that the Government presented\\n\\nsufficient evidence for each of Counts 2–4 to prove beyond a reasonable doubt that\\n\\na Schedule II drug that Dr. Feldman prescribed (methadone or oxycodone) was, at\\n\\na minimum, a but-for cause of each victim’s death.\\n\\n 3. Ricky Gonzalez\\n\\n The evidence presented at trial was sufficient to support a conclusion that a\\n\\nSchedule II drug—oxycodone—was a but-for cause of Gonzalez’s death. First,\\n\\nevidence provided by the defense expert, Dr. Adams, indicated that there was\\n\\nenough oxycodone in Gonzalez’s system to kill a person who did not have a high\\n\\ndegree of tolerance. In other words, use of oxycodone was a potentially sufficient\\n\\ncause, by itself, to cause death.\\n\\n\\n\\n\\n 51\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 52 of 67\\n\\n\\n But whether or not oxycodone was a sufficient cause, there was ample\\n\\nevidence from which a jury could conclude that but for Gonzalez’s use of\\n\\noxycodone, he would not have died. Dr. Thogmartin, the medical examiner who\\n\\nperformed Gonzalez’s autopsy, observed that Gonzalez had pulmonary edema, or\\n\\nfluid-filled lungs, an occurrence often seen in drug overdoses. Dr. Thogmartin\\n\\nconcluded that Gonzalez’s cause of death was oxycodone toxicity with a\\n\\ncontributory condition of diazepam (Valium) intoxication. 11\\n\\n Noting that Gonzalez had a significant amount of oxycodone and diazepam\\n\\nin his system, the doctor explained that the oxycodone, a Schedule II drug, had\\n\\n“poisoned” Gonzalez and the diazepam, a Schedule IV drug, had reacted\\n\\n“synergistically” to speed up his death. He explained that in an overdose scenario\\n\\ninvolving the mixture of an opiate, such as oxycodone, and a benzodiazepine, such\\n\\nas diazepam, the “main player” in causing death is the opiate, with the\\n\\nbenzodiazepine acting as only a “minor leaguer,” a “contributory cause.” As to\\n\\nhow he would characterize such an interaction, he determined that the oxycodone\\n\\nwas “the big player” and therefore the “cause of death,” and that the diazepam\\n\\n“was not lethal in and of itself, but a helper.” Indeed, in explaining that\\n\\n“oxycodone is the big player,” the doctor doubted that diazepam alone would have\\n\\n\\n\\n11\\n Again, the jury concluded that Dr. Feldman had prescribed both the oxycodone and diazepam\\nfound in Gonzalez’s system.\\n 52\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 53 of 67\\n\\n\\nkilled Gonzalez,” noting that “[d]iazepam is really hard to overdose on. . . . I mean,\\n\\nyou’d have to like eat a ton and even then you may not die.” While he thought\\n\\nthere was enough oxycodone here to kill Gonzalez by itself, he couldn’t say for\\n\\ncertain that Gonzalez would have died had he not also ingested the diazepam, as\\n\\nthat is the typical scenario he has observed. Nevertheless, Dr. Thogmartin\\n\\nconcluded, with a reasonable degree of medical certainty, that if it were not for the\\n\\noxycodone he ingested, Gonzalez would “not have died this day of this death.” In\\n\\nshort, the doctor concluded that Gonzalez died of oxycodone intoxication, with\\n\\ndeath having been sped up by the concurrent ingestion of the diazepam.\\n\\n In short, at the least, a jury could reasonably conclude that the oxycodone\\n\\nwas a necessary, but-for cause of his death. In other words, Gonzalez would have\\n\\nremained alive had he not ingested this Schedule II drug. Even assuming that the\\n\\ndiazepam may have also constituted a necessary, but-for cause of death, that fact\\n\\ndoes not deprive the oxycodone of its own “but-for” status. To argue that it does is\\n\\nto conclude that the oxycodone had to be a sufficient cause of death: that is, that\\n\\nthe drug had to, by itself, be capable of causing death. As explained above, this is\\n\\nnot how Burrage defined a but-for cause and indeed Burrage made clear that there\\n\\ncan be multiple but-for causes of an event. Accordingly, a reasonable jury could\\n\\nconclude beyond a reasonable doubt that Gonzalez died as a result of a Schedule II\\n\\ndrug that Dr. Feldman prescribed: namely, oxycodone.\\n\\n\\n 53\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 54 of 67\\n\\n\\n 4. Joey Mayes\\n\\n The toxicology report indicated that Mayes had ingested two Schedule II\\n\\ndrugs—methadone and oxycodone—and one Schedule IV drug—alprazolam\\n\\n(Xanax). Dr. Feldman had prescribed all three drugs. There was testimony from\\n\\nboth the Government’s expert, who was the medical examiner who performed the\\n\\nautopsy, and from the defense expert that there was enough methadone in Mayes’\\n\\nsystem to kill him, even had other drugs not been ingested.\\n\\n As for the defense expert, Dr. Adams testified that Mayes’ cause of death\\n\\nwas “intoxication by methadone,” which was “the only drug on board capable of\\n\\ncausing a cardiac death.” From that testimony, alone, a jury could have found, at a\\n\\nminimum, that methadone played a necessary (but-for) role in Mayes’ death,\\n\\nwithout which he would have lived. Moreover, because Dr. Adams’ testimony\\n\\nexculpated the Schedule IV alprazolam, as well as the Schedule II oxycodone, a\\n\\njury would have been entitled to find that methadone was also an “independently\\n\\nsufficient cause” of Mayes’ death. Burrage, 571 U.S. at 218–19.\\n\\n As for the Government’s expert, in his autopsy report, Dr. Wilson had listed\\n\\nthe cause of death as multidrug toxicity involving the above-listed drugs. He did\\n\\nnot stop there, however. Dr. Wilson further testified that Mayes had “particularly\\n\\nelevated” levels of methadone in his system in an amount that was enough by itself\\n\\nto kill him, and he made clear that the methadone was the critical drug necessary to\\n\\n\\n 54\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 55 of 67\\n\\n\\nhave caused Mayes’ death. Further, given the high level of methadone, Mayes\\n\\nwould have died even had he received quicker medical attention. In addition to the\\n\\nmethadone, Dr. Wilson also stated that there was likely enough oxycodone (a\\n\\nSchedule II drug), by itself, to kill Mayes.\\n\\n And while the doctor acknowledged that his autopsy report had used the\\n\\nphrase “multidrug toxicity,” explaining that in his opinion each drug “had a role”\\n\\nin killing Mayes, acting “synergistically” with one another, he reiterated that the\\n\\nmethadone was “the major drug that was involved in Mr. Mayes’ death, with\\n\\ndistributing effects from alprazolam as well as oxycodone.”\\n\\n Based on this testimony, a reasonable jury could conclude that each of the\\n\\nSchedule II drugs (methadone and oxycodone) played, at a minimum, a necessary\\n\\n“but-for” role in causing Mayes’ death. As discussed above, that another drug (the\\n\\nalprazolam) may have also played a necessary role in causing Mayes’ death is\\n\\n“beside the point.” Id. at 211–12.\\n\\n 5. Shannon Wren\\n\\n As to Shannon Wren, defense expert Dr. Adams opined that the cause of\\n\\ndeath for this decedent was a heart attack, not the drugs ingested by Wren. The\\n\\nmedical examiner, Dr. Wilson, disagreed. Instead, Dr. Wilson concluded that\\n\\nWren died from multidrug toxicity. He testified that the oxycodone in Wren’s\\n\\nsystem would have been enough, by itself, to kill him, and that the diazepam would\\n\\n\\n 55\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 56 of 67\\n\\n\\nnot have been sufficient to kill Wren, by itself. As to listing the ultimate cause of\\n\\ndeath as multidrug toxicity, he explained that each of the drugs acted with the\\n\\nothers to cause death. 12\\n\\n In short, taking the evidence in the light most favorable to the Government,\\n\\nthis evidence was sufficient to support an inference that the oxycodone was a\\n\\nsufficient cause of Wren’s death, as well as a necessary, “but-for” cause.\\n\\n Dr. Feldman points out that his expert, Dr. Adams, disagreed with Dr.\\n\\nWilson’s conclusion that Wren died from multidrug toxicity. Instead, Dr. Adams\\n\\nbelieved that Wren’s enlarged heart and the fact that he was found dead on the\\n\\nbathroom floor indicated that he had died from sudden cardiac arrest, rather than\\n\\nfrom a drug overdose. Dr. Adams did not perform the autopsy, however; Dr.\\n\\nWilson did. Further, Dr. Wilson considered Wren’s enlarged heart and\\n\\nnevertheless concluded that there was no evidence of heart attack because there\\n\\nwas “still enough . . . reservoir opening of the coronary arteries to give enough\\n\\nblood to the heart.” It was within the province of the jury to choose among the\\n\\nreasonable constructions of the evidence and weigh Dr. Wilson’s testimony more\\n\\nheavily than Dr. Adams’. See United States v. Ndiaye, 434 F.3d 1270, 1294 (11th\\n\\n\\n\\n12\\n The drugs found in Wren’s system included oxycodone, diazepam, amitriptyline, and\\nmetabolites of the latter two drugs. Like diazepam, amitriptyline (more familiarly known as the\\nantidepressant Elavil) is a Schedule IV drug. The defense expert indicated that this Elavil was in\\n“an innocuous concentration,” and the Government’s expert did not focus on it at all. Nor was\\nthis drug listed on the special verdict form.\\n 56\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 57 of 67\\n\\n\\nCir. 2006) (“Because the jury is free to choose among reasonable constructions of\\n\\nthe evidence, the evidence may be sufficient even if it is not entirely inconsistent\\n\\nwith conclusions other than guilt.”).\\n\\n In sum, when viewed in the light most favorable to the Government, there\\n\\nwas sufficient evidence from which a reasonable jury could conclude that a\\n\\nSchedule II drug prescribed by Dr. Feldman—methadone and oxycodone in\\n\\nMayes’ case, and oxycodone in Gonzalez’s and Wren’s cases—was, at the very\\n\\nleast, a but-for cause of each of the victims’ deaths, if not an independently\\n\\nsufficient cause.\\n\\n VIII. Alleged Alleyne Error\\n Dr. Feldman asserts that even if there was sufficient evidence to support the\\n\\nsentencing enhancement for Counts 2–4, its application in this case was\\n\\nunconstitutional under Alleyne v. United States, 570 U.S. 99 (2013). The\\n\\nGovernment does not dispute that Dr. Feldman adequately preserved this issue\\n\\nbelow, and we therefore review his claims of Alleyne error de novo. United States\\n\\nv. King, 751 F.3d 1268, 1278–79 (11th Cir. 2014).\\n\\n As noted above, in Counts 2–4, Dr. Feldman was convicted of violating\\n\\n21 U.S.C. § 841(a)(1) based on his unlawful dispensing of oxycodone (a Schedule\\n\\nII drug), methadone (a Schedule II drug), alprazolam (a Schedule IV drug), and\\n\\ndiazepam (a Schedule IV drug) to the three individuals identified in those counts.\\n\\n\\n 57\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 58 of 67\\n\\n\\nThe jury rendered that verdict by answering in the affirmative the special verdict\\n\\nform’s 13 first two questions for each count, finding that Dr. Feldman was guilty of\\n\\ndistributing and dispensing the Schedule II and Schedule IV drugs identified in the\\n\\nindictment. Although Dr. Feldman challenged the sufficiency of the evidence\\n\\nsupporting his conviction for conspiracy to distribute these substances and argued\\n\\nthat insufficient evidence supported a jury finding that the Schedule II drugs\\n\\nresulted in death, he has not challenged his substantive distribution convictions.\\n\\nAnd even had he done so, we would have rejected that challenge. There is ample\\n\\nevidence to support the jury’s conclusion that Dr. Feldman prescribed all of these\\n\\nsubstances to the particular victims and that he did so not for a legitimate medical\\n\\npurpose and not in the usual course of professional practice.\\n\\n The maximum penalty for an “unenhanced” conviction based on distribution\\n\\nof a Schedule II drug is 20 years’ imprisonment. 21 U.S.C. § 841(b)(1)(C). Thus,\\n\\nwithout any other enhancements, Dr. Feldman was subject to a 20-year maximum\\n\\nsentence, with no mandatory-minimum provision, for each conviction set out in\\n\\nCounts 2–4.\\n\\n The Government, however, sought an enhanced sentence, alleging that each\\n\\nof the victims identified for purposes of Counts 2–4 had died as a result of their\\n\\n\\n13\\n The special verdict form for each victim included four overarching questions, with two of the\\nquestions including sub-parts that asked the jury to make findings regarding each specified\\ncontrolled substance.\\n 58\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 59 of 67\\n\\n\\ningestion of the Schedule II drugs prescribed by Dr. Feldman. Under the\\n\\nControlled Substances Act’s sentence-enhancement provision, if death results from\\n\\nthe use of a Schedule II substance, the person convicted of distributing that drug is\\n\\nsubject to a mandatory-minimum 20-year sentence, with a statutory maximum of\\n\\nlife imprisonment. Id. § 841(b)(1)(C). Death resulting from ingestion of a\\n\\nSchedule IV substance, however, results in no enhancement.\\n\\n The district court determined that the jury had convicted Dr. Feldman of\\n\\nconduct justifying the above enhancement. As a result, Dr. Feldman was subject to\\n\\na mandatory-minimum 20-year sentence, and the district court ultimately decided\\n\\nthat a 300-month sentence (25-years) was a reasonable sentence based on the\\n\\nconduct underlying these three convictions.14\\n\\n “Because the ‘death results’ enhancement increase[s] the minimum and\\n\\nmaximum sentences to which [a defendant is] exposed, it is an element that must\\n\\nbe submitted to the jury and found beyond a reasonable doubt.” Burrage, 571 U.S.\\n\\nat 210 (citing Alleyne, 570 U.S. at 115–16, and Apprendi, 530 U.S. at 490). Dr.\\n\\nFeldman asserts that the jury did not convict him of distributing a Schedule II\\n\\nsubstance from which death has resulted, which, if correct, would mean that the\\n\\n\\n14\\n Dr. Feldman received a 10-year sentence for his convictions for conspiracy to distribute\\ncontrolled substances and conspiracy to commit money laundering, a 10-year sentence for his\\nsubstantive money laundering convictions, and a 25-year sentence for the enhanced distribution\\nof a Schedule II substance, use of which had resulted in death. All sentences ran concurrently\\nwith each other.\\n\\n 59\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 60 of 67\\n\\n\\ndistrict court violated Alleyne when it sentenced Dr. Feldman as if the jury had so\\n\\nfound. Specifically, and relying again on Burrage, Dr. Feldman argues that the\\n\\njury’s special verdict established only a finding that a combination of Schedule II\\n\\nand Schedule IV drugs in the aggregate killed each victim, not that the victims\\n\\nwould have lived “but for” their use of the Schedule II drugs. Accordingly, Dr.\\n\\nFeldman contends that the jury verdict was insufficient to trigger a 20-year\\n\\nmandatory minimum sentence enhancement or a statutory maximum exceeding\\n\\ntwenty years.\\n\\n To the extent that Dr. Feldman’s Alleyne argument mirrors his sufficiency-\\n\\nof-the-evidence argument, we reject it as a misreading of Burrage. As discussed\\n\\nabove, Burrage did not hold that a Schedule I or II drug must be the “only” cause\\n\\nof a victim’s death. Indeed, Burrage reached the opposite conclusion, explaining\\n\\nthat a factor can be a but-for cause even if it “combines with other factors to\\n\\nproduce the result, so long as the other factors alone would not have” produced the\\n\\nsame result. Id. at 211. Under Burrage, but-for causality simply turns on whether\\n\\nthe victim would have died “without the incremental effect” of a Schedule I or II\\n\\ndrug. Id. In other words, a Schedule I or II drug is a but-for cause of a victim’s\\n\\ndeath if it made a necessary contribution to the death—“if, so to speak, it was the\\n\\nstraw that broke the camel’s back”—such that the victim would have lived if he\\n\\nhad not taken the Schedule I or II drug. Id. The existence of other necessary\\n\\n\\n 60\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 61 of 67\\n\\n\\ncauses, such as other drugs in a deadly cocktail, is “beside the point.” Id. at 212.\\n\\nFor a sentencing court to impose an enhanced sentence under § 841(b)(1)(C), a\\n\\njury need only find that a Schedule I or II drug was a but-for cause of death or\\n\\nserious bodily injury. Id. at 210–11, 218–19.\\n\\n As discussed at length above, we conclude that the evidence was sufficient\\n\\nto prove beyond a reasonable doubt that but for ingestion of the Schedule II drugs\\n\\ndispensed by Dr. Feldman, the victims would not have died. Nevertheless, we\\n\\nagree with Dr. Feldman that the special verdict in this case failed to establish that\\n\\nthe jury actually found that the Schedule II drugs were but-for causes of the\\n\\nvictims’ deaths. Accordingly, we conclude that the district court erred under\\n\\nAlleyne by imposing an enhanced sentence under § 841(b)(1)(C). We explain our\\n\\nreasons for reaching this result.\\n\\n The third and fourth questions on the verdict form for each of Counts 2, 3,\\n\\nand 4 focused on the deaths of Mayes, Gonzalez, and Wren, respectively.\\n\\nAggregating all the drugs dispensed by Dr. Feldman to the victim—both Schedule\\n\\nII and Schedule IV—the third question asked whether the jury found that the\\n\\nvictim would not have died but for his use of the listed drugs. 15 The question for\\n\\n\\n\\n15\\n Asking for a “yes” or “no” answer, the question read: “We, the jury, find beyond a\\nreasonable doubt that the death of [the victim] resulted from the use of one or more of the\\nfollowing controlled substances: [identifying Schedule II and Schedule IV controlled substances\\ndispensed to victim], that is, but for his ingestion of the controlled substances charged, he would\\nnot have died.”\\n 61\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 62 of 67\\n\\n\\nCount 2 (Mayes) listed as the relevant drugs oxycodone and methadone, which are\\n\\nSchedule II drugs, as well as alprazolam, which is a Schedule IV drug. The\\n\\nquestions for Counts 3 and 4 (Gonzalez and Wren) each listed as the relevant drugs\\n\\noxycodone, a Schedule II drug, and diazepam, a Schedule IV drug. For each\\n\\ncount, the jury responded “Yes,” finding that the victim would not have died but\\n\\nfor his use of the listed controlled substances.\\n\\n These findings clearly do not support the district court’s application of\\n\\n§ 841(b)(1)(C)’s mandatory-minimum sentence. This is so because, at most, the\\n\\njury’s “Yes” responses to these questions represented a finding only that the\\n\\nSchedule II and Schedule IV drugs in the aggregate caused the deaths. And\\n\\nBurrage held that it is insufficient merely to show that a Schedule I or II drug was\\n\\npart of a mixed-drug cocktail, which, as “an aggregate force,” was “itself a but-for\\n\\ncause of death.” Id. at 214. Instead, Burrage requires a jury finding that a\\n\\nSchedule I or II drug was itself a but-for cause of the victim’s death—that is,\\n\\nwithout the specific incremental effect of the Schedule I or II drug, the victim\\n\\nwould have lived. See id. at 207, 211, 218–19 (reversing the defendant’s\\n\\nconviction where the evidence showed only that the victim died as a result of a\\n\\ncocktail of drugs, including heroin, oxycodone, alprazolam, and clonazepam, and\\n\\nthe Government conceded that no evidence showed the victim would have lived\\n\\nbut for his heroin use).\\n\\n\\n 62\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 63 of 67\\n\\n\\n The fourth question on Dr. Feldman’s special verdict form purportedly\\n\\nattempted to disaggregate the drugs to seek a finding as to which specific drugs\\n\\ncaused the death. Considering all the circumstances, however, we conclude that\\n\\nthis question failed to elicit the jury’s finding on what was the relevant question:\\n\\nwhether the Schedule II drugs were but-for causes of the victims’ deaths.\\n\\nSpecifically, this final question for each of Counts 2–4 asked whether the jury\\n\\nfound “beyond a reasonable doubt that the death of [the victim] resulted from the\\n\\nuse of the following,” then it listed separately each specific drug dispensed to the\\n\\nvictim: both Schedule II and Schedule IV substances. (emphasis added). The jury\\n\\nanswered “Yes” for each drug listed. In contrast with the prior question, though,\\n\\nthis fourth question entirely omitted any mention of but-for causation and failed to\\n\\nmake clear to the jury that absent a finding that the victim would not have died had\\n\\nhe not used a particular drug, the jury could not conclude that the victim’s death\\n\\nresulted from that drug.\\n\\n We cannot conclude based on the jury’s response to this final question that it\\n\\nfound beyond a reasonable doubt that a Schedule II drug Dr. Feldman prescribed—\\n\\noxycodone and methadone for Count 2, and oxycodone for each of Counts 3 and\\n\\n4—was a but-for cause of the victims’ deaths. First, the verdict form’s failure to\\n\\nspell out what “resulted from” meant in the critical and final question raises a\\n\\nsignificant concern that the jury was unaware of that phrase’s meaning as it\\n\\n\\n 63\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 64 of 67\\n\\n\\npertains to whether death resulted from a specific drug. And because the first\\n\\n“results in” question explained that “resulted from” meant that the victim would\\n\\nnot have died but for ingestion of the drugs in the aggregate, but the second\\n\\nquestion omitted an explanation of what “resulted from” meant, the jury could\\n\\nhave concluded that it was once again being asked to find whether all the drugs\\n\\nresulted in each victim’s death, rather than whether each victim would have lived\\n\\nbut for his ingestion of a Schedule II drug in particular.\\n\\n Even with the omission of the “but-for” language in this pivotal, final\\n\\nquestion, we might well reach a different result had the district court’s instructions\\n\\nexplained to the jury that, in order to conclude that death had resulted from a\\n\\nparticular Schedule II drug, it must find that but for ingestion of that drug, death\\n\\nwould not have occurred. Unfortunately, the court’s instructions did not do so.\\n\\nThe court instructed the jury only that it “must find beyond a reasonable doubt that\\n\\nbut for the decedent’s use of the charged controlled substances distributed . . . the\\n\\ndecedent would not have died.” (emphasis added). This plural reference to\\n\\ncontrolled substances in the instructions necessarily included the Schedule IV\\n\\nsubstances, which of course cannot trigger an enhancement. And it replicates the\\n\\nfirst interrogatory on the special verdict form, which directed the jury to make a\\n\\nfinding only as to whether the drugs in the aggregate caused each victim’s death.\\n\\n\\n\\n\\n 64\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 65 of 67\\n\\n\\n Thus, the court’s instructions failed to make clear that the jury needed to\\n\\nmake a specific finding as to whether the victim would have lived but for his use of\\n\\nthe Schedule II controlled substance. Indeed, the instructions never indicated that\\n\\nthe Schedule II drugs had any special significance in the case, as the court merely\\n\\ninstructed the jury that “a more serious offense is committed” whenever “death or\\n\\nserious injury is a consequence of the decedent’s use of a controlled substance.”\\n\\n(emphasis added). Nor did the prosecutor or defense attorney point out in closing\\n\\narguments the specific determination that the jury was being called on to make.\\n\\n Given this context, we cannot conclude that the jury found beyond a\\n\\nreasonable doubt that the victims would have lived but for their ingestion of the\\n\\nSchedule II drugs. Admittedly, the jury’s finding that the victims’ deaths “resulted\\n\\nfrom” each of the Schedule II and Schedule IV drugs is not inconsistent with a\\n\\nfinding that the Schedule II drugs were but-for causes of the deaths. As we have\\n\\nexplained at length, there can be many necessary causes of a result, and, if a\\n\\nSchedule II drug was a necessary cause of a victim’s death, it would have been\\n\\n“beside the point” that a Schedule IV drug also happened to occupy the same\\n\\nstatus. Id. at 211–12.\\n\\n But the verdict rendered is likewise consistent with a finding that the\\n\\nSchedule II drugs were nonessential “contributing cause[s],” which did not make a\\n\\ndifference to the ultimate outcome but merely “contribute[d] to an aggregate force\\n\\n\\n 65\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 66 of 67\\n\\n\\n(such as mixed-drug intoxication) that is itself a but-for cause of death.” Id. at 208,\\n\\n214. The latter finding, of course, would be insufficient to support a 20-year\\n\\nmandatory minimum sentence under Burrage. Id. at 214–16. And because we\\n\\ncannot determine which of the two findings the special verdict reflects, the jury’s\\n\\nverdict does not constitute the necessary finding that but for ingestion of a\\n\\nSchedule II substance, the victim would have lived.16 That being so, the district\\n\\ncourt’s application of § 841(b)(1)(C)’s sentencing enhancement cannot stand. 17\\n\\n Accordingly, we reverse the district court’s application of § 841(b)(1)(C)’s\\n\\n20-year mandatory minimum sentence on Counts 2–4 and remand the case for the\\n\\ndistrict court to resentence Dr. Feldman “to a term of imprisonment of not more\\n\\nthan 20 years” as to each of these counts. 21 U.S.C. § 841(b)(1)(C) (stating that\\n\\n“such person shall be sentenced to a term of imprisonment of not more than\\n\\n20 years and if death or serious bodily injury results from the use of such substance\\n\\nshall be sentenced to a term of imprisonment of not less than twenty years or more\\n\\nthan life”).\\n\\n\\n\\n\\n16\\n We emphasize that we are not saying that the jury acquitted Dr. Feldman of the charge that his\\ndispensing of a Schedule II drug resulted in death. Rather, the outcome of our ruling is the same\\nas it would have been had the Government never sought a specific finding from the jury on this\\nissue.\\n17\\n We are aware that an Alleyne error can be subject to harmless error review. United States v.\\nKing, 751 F.3d 1268, 1279 (11th Cir. 2014). The Government, however, has not argued that any\\nAlleyne error here would be harmless beyond a reasonable doubt.\\n 66\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 67 of 67\\n\\n\\n CONCLUSION\\n\\n For the above reasons, we AFFIRM the Feldmans’ convictions, REVERSE\\n\\nDr. Feldman’s sentences on Counts 2–4, and REMAND for the district court to\\n\\nconduct a new sentencing hearing for Dr. Feldman.\\n\\n\\n\\n\\n 67\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435141/\", \"author_raw\": \"JULIE CARNES, Circuit Judge:\"}]}"},"judge1":{"kind":"string","value":"JILL PRYOR"},"judge2":{"kind":"string","value":"JULIE CARNES"},"judge3":{"kind":"string","value":"ANTOON"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"JILL PRYOR\": \", Circuit\", \"JULIE CARNES\": \", Circuit\", \"ANTOON\": \", District\"}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4657888/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: United States v. Edward Neil Feldman\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-30\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nCase: 16-12978 Date Filed: 08/30/2019 Page: 1 of 67\n\n\n [PUBLISH]\n\n\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n ________________________\n\n No. 16-12978\n ________________________\n\n D.C. Docket No. 8:14-cr-00521-JDW-AEP-1\n\n\n\nUSA,\n\n Plaintiff - Appellee,\n\nversus\n\nEDWARD NEIL FELDMAN,\nKIM XUAN FELDMAN,\n\n Defendants - Appellants.\n\n ________________________\n\n Appeals from the United States District Court\n for the Middle District of Florida\n ________________________\n\n (August 30, 2019)\n\nBefore JILL PRYOR and JULIE CARNES, Circuit Judges, and ANTOON,*\nDistrict Judge.\n\n\n *\n Honorable John Antoon II, United States District Judge for the Middle District of\nFlorida, sitting by designation.\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 2 of 67\n\n\n\n\nJULIE CARNES, Circuit Judge:\n\n Following a jury trial, Defendants Dr. Edward Feldman and his wife Kim\n\nFeldman appeal their convictions for conspiring to distribute controlled substances\n\nwithout a legitimate medical purpose, conspiring to commit money laundering, and\n\nthree counts of illegal monetary transactions. Dr. Feldman was also convicted of\n\nthree counts of distributing controlled substances, without a legitimate medical\n\npurpose, to three individuals resulting in their deaths. Both Defendants raise\n\nvarious challenges to their convictions on appeal. Dr. Feldman also challenges his\n\n300-month sentence, which incorporated a 20-year mandatory-minimum penalty\n\npursuant to 21 U.S.C. § 841(b)(1)(C). After careful review of the parties’ briefs\n\nand with the benefit of oral argument, we affirm both Defendants’ convictions, but\n\nvacate Dr. Feldman’s sentence.\n\n BACKGROUND\n\n I. Facts\n\n A. The Clinic\n\n This case involves the operation of a pain-management clinic in Florida:\n\nFeldman Orthopedic and Wellness Center. The clinic was operated by Dr.\n\nFeldman, a physician, along with his wife, who served as the clinic’s office\n\nmanager and handled all of the clinic’s money. The clinic did not accept\n\ninsurance. To obtain an office visit, patients paid $300 in cash or credit for their\n\n 2\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 3 of 67\n\n\nfirst visit and $150 for each follow-up visit. Patients were provided a discount for\n\nreferring new patients to the clinic. Office visits with Dr. Feldman were typically\n\nbrief, as he usually performed cursory examinations—or sometimes no\n\nexamination at all.\n\n Dr. Feldman was the only physician at the clinic’s first location on 38th\n\nAvenue, where he saw approximately 40 to 100 patients per day. According to one\n\nformer employee, the clinic’s patients came from out of state and often appeared\n\n“stoned” and “drunk.” One of the clinic’s former patients, Mike Shaw, testified\n\nthat he came to the clinic because Dr. Feldman was a “writer,” meaning that he\n\nwould prescribe any medication suggested by the patient. The clinic sometimes\n\nreceived phone calls from concerned family members of patients suffering from\n\naddiction, yet Dr. Feldman did not treat those patients any differently.\n\n In July 2010, the clinic moved to 66th Street, where Dr. Feldman began\n\nseeing 75 to 150 patients per day. Due to heavy law enforcement presence at this\n\nlocation, the Feldmans instructed their staff to stop accepting patients from out of\n\nstate. They also told their staff to place notes in the patient files if they suspected\n\nthat someone was an undercover law enforcement officer. If a patient was flagged\n\nas an undercover officer, Dr. Feldman conducted a more extensive physical and\n\ndid not prescribe those patients any controlled substances. Based on his belief that\n\n\n\n\n 3\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 4 of 67\n\n\nsunglasses and hats could conceal recording devices, Dr. Feldman banned these\n\nitems at the clinic’s third and final location on Park Boulevard.\n\n In an effort to tighten protocol, the Feldmans also instructed staff to\n\ndischarge patients whose urine tested positive for cocaine or other serious drugs.\n\nIn addition, the clinic began utilizing the Prescription Drug Monitoring Database—\n\na database that shows the prescriptions a patient has previously obtained and serves\n\nas a tool for determining whether patients are doctor shopping. Notwithstanding\n\nthe above gestures, Mrs. Feldman ultimately told staff to stop conducting urinalysis\n\nscreenings because too many patients were being discharged based on those\n\nresults. Moreover, the Feldmans decided that no patient would be discharged until\n\nDr. Feldman had himself reviewed that patient.\n\n In January 2011, a second doctor, Dr. Nancy Bruemmer joined the clinic.\n\nDr. Bruemmer often recommended lowering patients’ dosage of medications or\n\ndischarging them altogether. Whenever Dr. Bruemmer made these\n\nrecommendations, however, Mrs. Feldman made sure that Dr. Feldman saw these\n\npatients before they were discharged.\n\n In early 2011, law enforcement officers began investigating Dr. Feldman’s\n\nclinic. As part of the investigation, several undercover officers posed as patients at\n\nthe clinic. Dr. Kevin Chaitoff, a physician who specializes in pain management\n\nand anesthesia, reviewed the transcripts from the undercover officers’ visits to the\n\n\n 4\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 5 of 67\n\n\nclinic and determined that there were several problems with the way Dr. Feldman\n\nhandled each visit. Specifically, Dr. Chaitoff determined that Dr. Feldman\n\nobtained inadequate patient histories, conducted inadequate physical examinations,\n\nand prescribed controlled substances without documented justification.\n\n In August 2012, law enforcement officers executed a search warrant at the\n\nclinic and seized 3,200 patient files. At the request of Brian Zdrojewski, an agent\n\nwith the Drug Enforcement Administration, Dr. Chaitoff randomly selected 30\n\npatient files to review in order to determine whether Dr. Feldman was operating\n\nwithin the usual course of professional practice with respect to each patient. Dr.\n\nChaitoff concluded that none of the patients in the files he reviewed had been\n\nprescribed controlled substances for a legitimate medical purpose. Dr. Chaitoff\n\nwas later asked to review an additional 18 patient files and he reached the same\n\nconclusion as to those files. The investigation revealed at least three patients of\n\nDr. Feldman’s who had died while under his care: Joey Mayes, Shannon Wren,\n\nand Ricky Gonzalez. Dr. Chaitoff reviewed the files of these patients and\n\ndetermined that Dr. Feldman’s treatment of them was not within the usual course\n\nof professional practice.\n\n B. Money Laundering\n\n The investigation also revealed that between 2010 and 2014, the Feldmans\n\ndeposited approximately $6,787,103.99 into 37 bank accounts—the majority of\n\n\n 5\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 6 of 67\n\n\nwhich came from cash deposits that were less than $10,000. John Barna, a retired\n\ndetective with the Pinellas County Sheriff’s Office, determined that the Feldmans\n\npurchased their home on Trilby Avenue, in part, with $190,000 from a savings\n\naccount in Dr. Feldman’s name at JPMorgan Chase Bank. Records showed that of\n\nthat $190,000, $150,000 was transferred from the clinic’s bank account in $50,000\n\nincrements. That transfer is the subject of Count 6. At issue in Count 7, is a\n\n$150,000 wire transfer from First Bank Creve Coeur Mark Lewis P.A. Trust\n\nAccount that Dr. Feldman used to fund the clinic location on Park Boulevard.\n\nFinally, the subject of Count 8 is a wire transfer in the amount of $187,689.50 from\n\na Merrill Lynch account in Dr. Feldman’s name to one in Mrs. Feldman’s name.\n\n II. Procedural History\n\n A federal grand jury subsequently issued an indictment charging the\n\nFeldmans with: (1) conspiring to distribute and dispense controlled substances—\n\nprimarily oxycodone and methadone, which are Schedule II controlled substances,\n\nand alprazolam and diazepam, which are Schedule IV controlled substances—not\n\nfor a legitimate medical purpose, in violation of 21 U.S.C. §§ 841(b)(1)(C), (b)(2)\n\nand 846 (Count 1); (2) conspiring to commit money laundering, in violation of 18\n\nU.S.C. § 1956(h) (Count 5); and (3) engaging in monetary transactions with\n\ncriminal derived property exceeding the value of $10,000, in violation of 18 U.S.C.\n\n§ 1957 (Counts 6 through 8). Dr. Feldman was also charged with three substantive\n\n\n 6\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 7 of 67\n\n\ncounts of distribution and dispensation of Schedule II and Schedule IV controlled\n\nsubstances, not for a legitimate medical purpose, resulting in the deaths of Joey\n\nMayes (Count 2), Ricky Gonzalez (Count 3), and Shannon Wren (Count 4), in\n\nviolation of § 841(a)(1), (b)(1)(C) (Counts 2 through 4).\n\n The Feldmans pled not guilty and proceeded to trial. The first trial ended in\n\na mistrial. At the second trial, the Government presented testimony from the case\n\nagents, undercover officers, former patients and employees, the medical examiners,\n\nand Dr. Chaitoff. At the close of the Government’s case-in-chief, the Feldmans\n\nmoved for judgment of acquittal. Dr. Feldman argued that the Government failed\n\nto meet its burden of proof as to each count. The district court denied the\n\nFeldmans’ motions for judgment of acquittal.\n\n Dr. Feldman called several witnesses on his behalf. Dr. Simopoulos, a pain\n\nmanagement physician, testified that, based on his review of the patient files, he\n\nbelieved that Dr. Feldman’s prescribing practices were within the usual course of\n\nprofessional practice. Dr. Vernard Adams, a forensic pathologist and former\n\nmedical examiner, opined about the cause of death for each of the three victims.\n\nDr. Feldman also testified. He denied that he had an agreement with Mrs. Feldman\n\nor anyone else at the clinic to engage in illegal activity. After the defense rested,\n\nthe district court denied the parties’ renewed motions for judgment of acquittal.\n\n\n\n\n 7\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 8 of 67\n\n\n The jury found the Feldmans guilty as charged. The district court sentenced\n\nDr. Feldman and Mrs. Feldman to imprisonment terms of 300 months and\n\n48 months, respectively. Dr. Feldman’s 300-month sentence incorporated the 20-\n\nyear mandatory minimum sentence under 21 U.S.C. § 841(b)(1)(C), based on his\n\nconvictions for Counts 2 through 4.1 This appeal followed.\n\n DISCUSSION\n\n I. Motion to Sever\n\n Mrs. Feldman argues that the district court abused its discretion by denying\n\nher motion to sever her trial from Dr. Feldman’s.\n\n We review the district court’s denial of a motion for severance for an abuse\n\nof discretion. United States v. Cobb, 185 F.3d 1193, 1196 (11th Cir. 1999). If it\n\nappears that a joint trial will prejudice a defendant or the Government, the district\n\ncourt may sever the defendants’ trials. Fed. R. Crim. P. 14(a). “In assessing the\n\nmerits of a severance motion, the district court must balance the possibility of\n\nprejudice to the defendant against the public interest in judicial efficiency and\n\neconomy.” United States v. Eyster, 948 F.2d 1196, 1213 (11th Cir. 1991). To\n\nobtain reversal of a district court’s denial of a severance motion, a defendant must\n\nmake a showing of “compelling prejudice” against which the district court offered\n\n\n\n1\n Dr. Feldman received ten-year concurrent sentences on each of his other five counts of\nconviction.\n 8\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 9 of 67\n\n\nno protection. United States v. Walser, 3 F.3d 380, 385 (11th Cir. 1993). “A\n\ndefendant can show compelling prejudice by demonstrating that the jury was\n\nunable to sift through the evidence and make an individualized determination as to\n\neach defendant.” United States v. Ramirez, 426 F.3d 1344, 1352 (11th Cir. 2005)\n\n(quotation marks omitted).\n\n Prior to trial, Mrs. Feldman moved to sever Counts 2 through 4—which\n\ncharged Dr. Feldman with distributing and dispensing controlled substances,\n\nwithout a legitimate medical purpose, resulting in death. She argued that she\n\nwould be substantially prejudiced by a joint trial that included those counts\n\nbecause they applied to Dr. Feldman only. The district court denied the motion.\n\n Mrs. Feldman has not met her burden of showing that the district court\n\nabused its discretion by denying her motion for severance. The Feldmans were\n\nindicted together and charged with being part of two conspiracies: (1) the\n\ndistribution of controlled substances without a legitimate medical purpose; and\n\n(2) money laundering. “It is well settled that defendants who are indicted together\n\nare usually tried together.” United States v. Browne, 505 F.3d 1229, 1268 (11th\n\nCir. 2007); see also United States v. Knowles, 66 F.3d 1146, 1158 (11th Cir. 1995)\n\n(“This court is reluctant to reverse a district court’s denial of severance,\n\nparticularly in conspiracy cases, as generally ‘persons who are charged together\n\nshould also be tried together.’”).\n\n\n 9\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 10 of 67\n\n\n Mrs. Feldman asserts that she was prejudiced by the admission of evidence\n\nrelating to Counts 2 through 4 because those counts applied only to Dr. Feldman.\n\nWe have concluded, however, that “[a] defendant does not suffer compelling\n\nprejudice, sufficient to mandate a severance, simply because much of the evidence\n\nat trial is applicable only to co-defendants.” United States v. Schlei, 122 F.3d 944,\n\n984 (11th Cir. 1997) (quotation marks omitted). Moreover, Mrs. Feldman has not\n\nmet her burden of demonstrating that the jury was not able to make an\n\nindividualized determination of guilt as to the evidence pertaining to her. See\n\nUnited States v. Francis, 131 F.3d 1452, 1459 (11th Cir. 1997) (explaining that, in\n\norder to establish prejudice, a defendant must show “that the jury was unable to\n\nmake an individualized guilt determination for each defendant”).\n\n Nevertheless, even if there were any potential prejudice, we have determined\n\nthat prejudice can be avoided where the district court instructs the jury that it\n\nshould consider the evidence against each defendant separately. See Schlei, 122\n\nF.3d at 984. Here, the district court instructed the jury that:\n\n [A] separate crime or offense is charged against each defendant\n in each count of the indictment. Each charge and the evidence\n pertaining to it should be considered separately. Also, the case of each\n defendant should be considered separately and individually. The fact\n that you may find one or any . . . of the defendants guilty or not guilty\n of any of the offenses charged should not affect your verdict as to any\n other offense or any other defendant.\n\n I caution you, members of the jury, that you are here to determine\n from the evidence in this case whether each defendant is guilty or not\n 10\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 11 of 67\n\n\n guilty. Each defendant is on trial only for the specific offense alleged\n in each count of the indictment.\n\n We presume that the district court’s cautionary instruction prevented any\n\npossible prejudice. See Francis, 131 F.3d at 1459 (explaining that “cautionary\n\ninstructions to the jury to consider the evidence separately are presumed to guard\n\nadequately against prejudice.” (quotation marks omitted)). Because Mrs. Feldman\n\nhas not provided any evidence showing that the jury was not able to make an\n\nindividualized determination of guilt, she has failed to demonstrate that she\n\nsuffered compelling prejudice or received an unfair trial. See Browne, 505 F.3d at\n\n1268 (“[Defendant] must discharge the ‘heavy burden’ of demonstrating\n\n‘compelling prejudice’ from the joinder.” (quotation marks omitted)).\n\nAccordingly, the district court did not abuse its discretion by denying her motion\n\nfor severance.\n\n II. Expert Testimony\n\n Mrs. Feldman’s next claim of error is that the district court abused its\n\ndiscretion by denying her motion to exclude Dr. Chaitoff’s testimony and by\n\npermitting Dr. Chaitoff to extrapolate an opinion as to all 3,200 patient files, based\n\non his review of a small percentage of those files.\n\n We review for abuse of discretion the district court’s determinations\n\nregarding the admissibility and reliability of expert testimony. United States v.\n\n\n\n 11\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 12 of 67\n\n\nFrazier, 387 F.3d 1244, 1258 (11th Cir. 2004). Federal Rule of Evidence 702\n\nprovides that a witness who is qualified as an expert may provide an opinion if:\n\n (a) the expert’s scientific, technical, or other specialized knowledge\n will help the trier of fact to understand the evidence or to determine a\n fact in issue; (b) the testimony is based on sufficient facts or data;\n (c) the testimony is the product of reliable principles and methods; and\n (d) the expert has reliably applied the principles and methods to the\n facts of the case.\n\nFed. R. Evid. 702. The district court has wide discretion in deciding how to\n\ndetermine reliability. United States v. Abreu, 406 F.3d 1304, 1307 (11th Cir.\n\n2005).\n\n Prior to trial, Dr. Feldman filed a motion in limine to exclude, or limit, the\n\ntestimony of the Government’s proposed expert, Dr. Chaitoff. Following a\n\nDaubert 2 hearing, a magistrate judge determined that Dr. Feldman’s objection to\n\nDr. Chaitoff’s testimony under Rule 702 was without merit, “at least to the extent\n\nthat the opinion does not extrapolate from the reviewed files to a broader\n\nconclusion involving files not reviewed.” The magistrate judge noted that Dr.\n\nChaitoff’s testimony at the hearing resolved any potential issue pertaining to\n\nextrapolation, as he testified that the review of one patient’s file would be\n\ninsufficient to draw conclusions regarding files that were not reviewed.\n\n\n\n\n2\n Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993).\n 12\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 13 of 67\n\n\n Mrs. Feldman now challenges for the first time on appeal an isolated\n\nstatement made by Dr. Chaitoff at trial. Specifically, while discussing his selection\n\nand review of the first 30 patient files, Dr. Chaitoff stated that, “the methodology\n\nwas random and the conclusions applied to all the patients. So whether I reviewed\n\n30 or I reviewed 3,200, I suspect my conclusions would have been identical.” As\n\nnoted, Mrs. Feldman did not object to this statement.\n\n The Federal Rules of Evidence provide that a defendant need not object to\n\npreserve a claim of error on appeal if the district court makes a definitive ruling on\n\na pretrial motion, such as a motion in limine. See Fed. R. Evid. 103(b) (“Once the\n\ncourt rules definitively on the record—either before or at trial—a party need not\n\nrenew an objection or offer of proof to preserve a claim of error for appeal.”). That\n\nis not what happened here, however. The magistrate judge did not expressly rule\n\non the merits of the extrapolation issue, stating only that the objection to Dr.\n\nChaitoff’s testimony pursuant to Rule 702 was without merit “at least to the extent\n\nthat the opinion does not extrapolate from the reviewed files to a broader\n\nconclusion involving files not reviewed.” Moreover, it is clear from the order that\n\nthe magistrate judge believed that any potential extrapolation by Dr. Chaitoff was a\n\nnon-issue, as he noted that Dr. Chaitoff had specifically testified that his review of\n\none patient’s file would be insufficient to draw conclusions as to other patient files\n\nthat had not been reviewed.\n\n\n 13\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 14 of 67\n\n\n Because the magistrate judge did not make a definitive ruling on the\n\nextrapolation issue, Mrs. Feldman was required to object to Dr. Chaitoff’s\n\nstatement to avoid plain error review. See United States v. Wilson, 788 F.3d 1298,\n\n1313 (11th Cir. 2015) (reviewing defendant’s argument for plain error where the\n\ndistrict court’s pretrial ruling was not definitive); Cf. United States v. Harris, 886\n\nF.3d 1120, 1127 n.2 (11th Cir. 2018) (concluding that defendant’s argument on\n\nappeal was not subject to plain error review—despite the defendant’s failure to\n\nobject at trial—because the district court definitively ruled on the motion in\n\nlimine). She did not object, so we review her argument for plain error. 3\n\n Mrs. Feldman cannot prevail under plain error review. Even if she could\n\nshow an error that was plain, she cannot demonstrate that her substantial rights\n\nwere violated. Indeed, she cannot show that she was prejudiced—much less\n\nsubstantially prejudiced—by one stray comment made by Dr. Chaitoff concerning\n\nhis suspicion as to the remaining 3,000+ patient files not reviewed. The comment\n\nwas made during approximately three days’ worth of testimony from Dr. Chaitoff.\n\nDuring that testimony, Dr. Chaitoff focused entirely on the 48 patient files that he\n\nreviewed and his opinion as to whether Dr. Feldman’s treatment of the specific\n\n\n\n3\n To show plain error, the defendant must establish that there was “(1) an error (2) that is plain\nand (3) that has affected the defendant’s substantial rights; and if the first three prongs are met,\nthen a court may exercise its discretion to correct the error if (4) the error ‘seriously affect[s] the\nfairness, integrity or public reputation of judicial proceedings.’” United States v. Madden, 733\nF.3d 1314, 1320 (11th Cir. 2013) (quotation marks omitted).\n 14\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 15 of 67\n\n\npatients in those files was within the ordinary course of professional practice.\n\nAccordingly, Mrs. Feldman is not entitled to relief on this claim.\n\n III. Double Jeopardy\n\n Mrs. Feldman asserts that the district court’s order declaring a mistrial\n\nduring the first trial was not supported by manifest necessity, and thus the court\n\nshould have granted her motion to dismiss the indictment on double jeopardy\n\ngrounds. She also argues that the district court failed to comply with Federal Rule\n\nof Criminal Procedure 26.3, which requires the court to give the parties an\n\nopportunity to object or consent to an order declaring mistrial.\n\n We review the district court’s declaration of a mistrial on grounds of\n\nmanifest necessity to determine whether the court exercised sound discretion in\n\ndeciding that such a necessity existed. United States v. Davis, 708 F.3d 1216, 1221\n\n(11th Cir. 2013). And “if the court’s decision to declare a mistrial was not based\n\non manifest necessity, it was an abuse of discretion not to dismiss the indictment\n\non double jeopardy grounds.” See id.\n\n The Double Jeopardy Clause of the Fifth Amendment safeguards a criminal\n\ndefendant from being subjected to multiple prosecutions for the same offense.\n\nU.S. Const. amend. V. “Ordinarily, when ‘a defendant successfully seeks to avoid\n\nhis trial prior to its conclusion by a motion for mistrial, the Double Jeopardy\n\n\n\n\n 15\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 16 of 67\n\n\nClause is not offended by a second prosecution.’” United States v. Shelley, 405\n\nF.3d 1195, 1200 (11th Cir. 2005) (quotation marks omitted).\n\n “[D]istrict courts are permitted to declare a mistrial and discharge a jury\n\nonly where, ‘taking all the circumstances into consideration, there is a manifest\n\nnecessity for the act, or the ends of public justice would otherwise be defeated.’”\n\nUnited States v. Therve, 764 F.3d 1293, 1298 (11th Cir. 2014) (quotation marks\n\nomitted). “Our task is to decide whether the district court exercised ‘sound\n\ndiscretion’ in declaring a mistrial, and in order to determine that ‘we review the\n\nentire record in the case without limiting ourselves to the actual findings of the trial\n\ncourt.’” Davis, 708 F.3d at 1221 (quotation marks omitted and citation omitted).\n\nBut “the requirement that manifest necessity be demonstrated operates[] only when\n\nthe trial court has declared a mistrial without the consent of the defendant.” United\n\nStates v. Puleo, 817 F.2d 702, 705 (11th Cir. 1987). If the defendant consents to a\n\nmistrial, the Double Jeopardy Clause will not bar her retrial. Id.\n\n The facts surrounding the declaration of a mistrial during the first trial are as\n\nfollows. During the first trial, Mrs. Feldman’s attorney asked Dr. Feldman on\n\ncross-examination if he had ever been convicted of a felony. Dr. Feldman stated\n\nthat he had been convicted of one felony. Thereafter, the prosecutor asked Dr.\n\nFeldman whether the felony he was convicted of related to his practice as a doctor.\n\nDr. Feldman objected and moved for a mistrial, arguing that a curative instruction\n\n\n 16\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 17 of 67\n\n\nwas not going to cure the prejudice. Indeed, the defendants were united in their\n\nposition that a mistrial was necessary. Mrs. Feldman indicated that she had “the\n\nsame position as this is a conspiracy case with regard to that I agree that, I agree\n\nthat I don’t believe a curative instruction is going to—.” Following subsequent\n\ndiscussion, the district court recessed for the day. The next morning, the district\n\ncourt heard argument from the Government and Dr. Feldman as to whether the\n\nprejudicial effect of the question could be remedied by a curative instruction.\n\nUltimately, the district court decided that the defendants were correct in asserting\n\nthat a mistrial was necessary because the court concluded that it was not possible\n\nto cure this issue. The court thus declared a mistrial. During this lengthy\n\ndeliberative process by the court, Mrs. Feldman never indicated any change in her\n\noriginal position that a mistrial was necessary nor did she object when the district\n\ncourt ultimately acquiesced to the defense position that a mistrial was required.\n\n Instead, almost a month later—when the jury was long gone—she filed a\n\nmotion arguing that the court was wrong to conclude that manifest necessity\n\nrequired a mistrial. She further argued that she was not provided the opportunity to\n\nconsent or object to the order declaring a mistrial. The district court denied the\n\nmotion.\n\n Mrs. Feldman argues that the Double Jeopardy Clause barred her retrial\n\nbecause a mistrial was not manifestly necessary. But we need not address manifest\n\n\n 17\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 18 of 67\n\n\nnecessity. Because Mrs. Feldman acquiesced in the court’s decision to declare a\n\nmistrial, she cannot now be credibly heard to argue that the court’s decision was a\n\nmistake.\n\n Mrs. Feldman argues that she could not have agreed with the district court’s\n\ndecision to declare a mistrial because the district court failed to provide proper\n\nnotice before declaring a mistrial. Federal Rule of Criminal Procedure 26.3\n\nprovides that prior to ordering a mistrial, “the court must give each defendant and\n\nthe government an opportunity to comment on the propriety of the order, to state\n\nwhether that party consents or objects, and to suggest alternatives.” Fed. R. Crim.\n\nP. 26.3. “[C]ompliance with Rule 26.3 is only one factor to be considered in\n\ndetermining whether a trial judge exercised sound discretion in declaring a\n\nmistrial.” Davis, 708 F.3d at 1224 (citing United States v. Berroa, 374 F.3d 1053,\n\n1058 (11th Cir. 2004)).\n\n We conclude that Mrs. Feldman gave implied consent to the mistrial. The\n\nrecord in the present case clearly shows that Mrs. Feldman had numerous\n\nopportunities to comment or object to the potential declaration of a mistrial, and\n\nshe never gave even a hint that she objected during the lengthy proceedings\n\nconcerning this matter. Indeed, after Dr. Feldman moved for a mistrial, Mrs.\n\nFeldman agreed that the prejudice could not be cured by a curative instruction, and\n\nthen noted that this was a conspiracy case. When the court reconvened after the\n\n\n 18\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 19 of 67\n\n\novernight recess, Mrs. Feldman raised no objection to the court declaring a\n\nmistrial, nor did she object or argue that the court should allow the ongoing trial to\n\nproceed against her alone. Even as the jury was departing, she said nothing to the\n\ncourt indicating that a mistrial should not have been declared as to her. See Puleo,\n\n817 F.2d at 705 (concluding that defendant consented to mistrial when “even as the\n\njury was departing . . . [counsel] said nothing regarding the mistrial”).\n\n In deciding a Double Jeopardy challenge to a second trial, we have recently\n\nreconfirmed Puleo’s teaching that “a defendant’s consent to a mistrial need not be\n\nexpress but may always be implied from the totality of circumstances.” United\n\nStates v. Isaac Feldman, 931 F.3d 1245, 1256–57 (11th Cir. 2019) (quotation\n\nmarks omitted and alterations accepted). As with the present case, we noted in\n\nIsaac Feldman that, albeit aware that the district court was in the process of\n\ndeclaring a mistrial, the defendant “never voiced any objection to the jury’s\n\ndismissal . . . . The totality of these circumstances compels the conclusion that [the\n\ndefendant] impliedly consented to the dismissal of the original jury . . . .” Id. at\n\n1257.\n\n So it is here. Mrs. Feldman implicitly consented to the order declaring a\n\nmistrial. We therefore conclude that she is entitled to no relief on her claim\n\nchallenging the district court’s denial of her motion to dismiss the indictment on\n\ndouble jeopardy grounds.\n\n\n 19\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 20 of 67\n\n\n IV. Prosecutorial Misconduct\n\n Mrs. Feldman argues that the prosecutor’s closing argument was improper\n\nand prejudicial because the prosecutor argued facts not in evidence, put forth a new\n\nlegal theory of conviction not alleged in the indictment, and interjected her own\n\npersonal opinion to inflame the jury.\n\n We generally review allegations of prosecutorial misconduct de novo.\n\nUnited States v. Eckhardt, 466 F.3d 938, 947 (11th Cir. 2006). Where a defendant\n\nfails to object during a prosecutor’s closing argument, however, “relief is available\n\nto rectify only plain error that is so obvious that failure to correct it would\n\njeopardize the fairness and integrity of the trial.” United States v. Bailey, 123 F.3d\n\n1381, 1400 (11th Cir. 1997).\n\n To prevail on a prosecutorial misconduct claim, the defendant must show\n\nthat the conduct was improper and that it was prejudicial to his substantial rights.\n\nId. A defendant’s substantial rights are prejudiced if there is a reasonable\n\nprobability that, but for the remarks, the outcome of the trial would have been\n\ndifferent. United States v. Lopez, 590 F.3d 1238, 1256 (11th Cir. 2009).\n\n“Prosecutorial misconduct must be considered in the context of the entire trial,\n\nalong with any curative instruction.” Id. If the district court provided a curative\n\ninstruction, “we will reverse only if the evidence is so prejudicial as to be incurable\n\nby that measure.” Id.\n\n\n 20\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 21 of 67\n\n\n Mrs. Feldman raises several challenges to the prosecutor’s remarks during\n\nclosing argument. We conclude that only a few of those challenges warrant any\n\ndiscussion here. 4 First, Mrs. Feldman asserts that the prosecutor improperly\n\ncommented on facts that were not in evidence when discussing Dr. Feldman’s\n\nexpert, Dr. Adams. Specifically, she takes issue with the prosecutor’s statement\n\nthat when Dr. Adams was employed by the county medical examiner’s office, he\n\nhad found that people who died at home had died of drug overdoses. However,\n\nnow that he was being paid as a defense expert, he had concluded that one of the\n\nvictims who died at home—Shannon Wren—died from a heart attack, not a drug\n\noverdose. We find nothing improper about the prosecutor’s statement, as it was\n\nsupported by Dr. Adams’ testimony that he had previously made findings of drug\n\noverdose deaths where people were found dead in their homes.5\n\n We next address Mrs. Feldman’s assertion that the prosecutor inserted a new\n\ntheory of conviction not alleged in the indictment by arguing that Mrs. Feldman\n\n\n\n\n4\n Mrs. Feldman argues that the prosecutor’s following remarks were also improper: (1) a\nstatement that Mayes’ sister testified that Dr. Feldman was surprised when she accused Dr.\nFeldman of prescribing Xanax and methadone to her brother; (2) a statement implying that there\nwere several actions the Department of Health could have taken against Dr. Feldman; and (3) a\nstatement that Mrs. Feldman had spoken with Aimee Martin about ways to stay off the Drug\nEnforcement Administration’s radar. We find no merit to these challenges.\n5\n Notably, the district court also provided the following curative instruction when Mrs. Feldman\nobjected to the prosecutor’s statement regarding Dr. Adams’ testimony: “[t]o the extent the\nlawyers disagree on what the facts are, members of the jury, you will resolve those\ndisagreements. These are arguments of counsel, they are not evidence.”\n 21\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 22 of 67\n\n\nviolated an Internal Revenue Service reporting requirement by failing to disclose\n\nforeign bank accounts. When addressing the money laundering convictions in her\n\nclosing argument, the prosecutor stated that she asked defense expert, Stephen\n\nOscher, if he noticed whether Mrs. Feldman had disclosed to the federal\n\ngovernment that she had bank accounts outside of the country. After the district\n\ncourt overruled Mrs. Feldman’s objection to this statement, the prosecutor\n\nexplained that Oscher had said he did not know, he had not seen anything. The\n\nprosecutor then stated, “[t]hat’s also a requirement, a reporting requirement.” In\n\nother words, the prosecutor appeared to imply that Mrs. Feldman had violated a\n\nreporting requirement by failing to disclose foreign bank accounts. The\n\nindictment, however, charged Mrs. Feldman with violating a transaction reporting\n\nrequirement, not failing to report the existence of a foreign bank account.\n\n Nevertheless, even if we assume that this remark was improper, it does not\n\namount to prosecutorial misconduct because Mrs. Feldman has not met her burden\n\nof showing that her substantial rights were violated. In the first place, the district\n\ncourt provided curative instructions. See Lopez, 590 F.3d at 1256 (explaining that\n\n“we will reverse only if the evidence is so prejudicial as to be incurable by that\n\nmeasure”). The court instructed the jury that their decision must be based on\n\nevidence, which “includes the testimony of the witnesses and the exhibits which\n\nhave been admitted in the record . . . anything the lawyers say is not evidence in\n\n\n 22\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 23 of 67\n\n\nthe case.” The record likewise contained sufficient evidence of guilt independent\n\nof the challenged remarks. Id. (“When the record contains sufficient independent\n\nevidence of guilt, any error is harmless.” (citing Eckhardt, 466 F.3d at 947)). As\n\ndiscussed infra, the Government presented sufficient evidence from which a\n\nreasonable jury could conclude that Mrs. Feldman was guilty beyond a reasonable\n\ndoubt. In short, the challenged comment was a small portion of the 17-day trial in\n\nwhich the Government presented substantial evidence of guilt against Mrs.\n\nFeldman.\n\n Finally, we reject Mrs. Feldman’s argument that the prosecutor’s statement\n\nthat Wren had a “butt-load” of drugs in his body was unfairly prejudicial. Because\n\nMrs. Feldman failed to object to this statement, our review is limited to plain error.\n\nMrs. Feldman has failed to show error, let alone plain error. See United States v.\n\nBailey, 123 F.3d 1381, 1400 (11th Cir. 1997) (“When a defendant fails to object to\n\nthe prosecutor’s closing argument, relief is available to rectify only plain error that\n\nis so obvious that failure to correct it would jeopardize the fairness and integrity of\n\nthe trial.”). We have stated that “there is no prohibition on colorful and perhaps\n\nflamboyant remarks if they relate to the evidence adduced at trial.” Id. (quotation\n\nmarks omitted). We acknowledge that the prosecutor could have described the\n\nquantity of drugs in Wren’s body in a less graphic manner. However, the truth is\n\nthat Wren did have a great deal of drugs in his system, and ultimately he died.\n\n\n 23\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 24 of 67\n\n\nThus, the prosecutor’s description was consistent with the evidence presented at\n\ntrial. In short, the remark challenged by Mrs. Feldman does not amount to error,\n\nlet alone plain error.\n\n As a final matter, Mrs. Feldman asserts that the alleged errors, when viewed\n\nindividually or cumulatively, prejudiced her substantial rights and deprived her of\n\na fair trial. “The cumulative error doctrine provides that an aggregation of non-\n\nreversible errors . . . can yield a denial of the constitutional right to a fair trial,\n\nwhich calls for reversal.” United States v. Capers, 708 F.3d 1286, 1299 (11th Cir.\n\n2013) (quotation marks omitted). Because Mrs. Feldman has established no error,\n\nshe cannot prevail under the cumulative error doctrine. See United States v.\n\nGamory, 635 F.3d 480, 497 (11th Cir. 2011) (“Where there is no error or only a\n\nsingle error, there can be no cumulative error.”).\n\n V. Jury Instruction\n\n Dr. Feldman argues that the district court abused its discretion by denying\n\nhis proposed jury instruction on his theory of defense.6 The specific instruction\n\nDr. Feldman requested was the following:\n\n It is a physician’s duty and obligation to try to relieve a patient’s pain.\n Therefore, it is ethical and medically justifiable for a physician to\n prescribe controlled substances for a patient suffering from chronic\n moderate to severe pain, even if the patient has developed a tolerance\n or addiction to those substances. It is also ethical and medically\n justifiable for such treatment to be performed for the purpose of\n\n6\n Mrs. Feldman adopts this argument.\n 24\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 25 of 67\n\n\n relieving chronic moderate to severe pain in a patient, regardless of the\n patient’s [addiction] history.\n\nThe district court denied the instruction as argumentative, citing this Court’s\n\ndecision in United States v. Paradies, 98 F.3d 1266, 1287 (11th Cir. 1996), where\n\nwe rejected a defendant’s proposed jury instruction because it was partisan and\n\nsought to have the court parrot the factual findings that the defendant was\n\nadvocating to the jury.\n\n We review the district court’s refusal to give a requested jury instruction for\n\nan abuse of discretion. United States v. Dohan, 508 F.3d 989, 993 (11th Cir.\n\n2007). The district court’s refusal to give a requested instruction is reversible error\n\nwhere the instruction “(1) was correct, (2) was not substantially covered by the\n\ncharge actually given, and (3) dealt with some point in the trial so important that\n\nfailure to give the requested instruction seriously impaired the defendant’s ability\n\nto conduct his defense.” Eckhardt, 466 F.3d at 947–48.\n\n Here, the district court did not abuse its discretion by refusing to give Dr.\n\nFeldman’s proposed jury instruction. In the first place, Dr. Feldman has not cited\n\nto any legal authority demonstrating that his requested instruction was a correct\n\nstatement of the law. Although he cites to Florida Statute § 765.1103—a statutory\n\nprovision governing pain management with respect to palliative care—this statute\n\ndoes not state that it is ethical for a doctor to prescribe controlled substances for\n\nchronic pain, regardless of a patient’s tolerance or addiction issues. See Fla. Stat.\n 25\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 26 of 67\n\n\n§ 765.1103 (providing that “practitioners . . . must, as appropriate, comply with a\n\nrequest for pain management or palliative care from a patient under their care”).\n\n Further, Dr. Feldman’s proposed jury instruction essentially asked the\n\ndistrict court to instruct the jury about the facts Dr. Feldman hoped the jury would\n\nfind: that it was ethically and medically justifiable for Dr. Feldman to prescribe\n\ncontrolled substances without respect to the patient’s medical history or addiction\n\nissues. See United States v. Maxwell, 579 F.3d 1282, 1304 (11th Cir. 2009)\n\n(explaining that the district court is not obliged to present jury instructions that\n\n“contain partisan and argumentative statements of law and fact”); Paradies, 98\n\nF.3d at 1287 (same). The proposed instruction was plainly argumentative and the\n\ncourt was not required to give a theory-of-defense instruction that “was more in the\n\nnature of a jury argument than a charge.” United States v. Barham, 595 F.2d 231,\n\n245 (5th Cir. 1979). In short, the district court’s refusal to give Dr. Feldman’s\n\nproposed instruction was not an abuse of discretion.\n\n VI. Sufficiency of the Evidence as to the Conspiracy Counts\n\n Both Feldmans challenge the sufficiency of the evidence presented to\n\nsupport each of their conspiracy convictions. We review challenges to the\n\nsufficiency of the evidence de novo, “viewing the evidence in the light most\n\nfavorable to the government and drawing all reasonable inferences and credibility\n\nchoices in favor of the jury’s verdict.” United States v. Taylor, 480 F.3d 1025,\n\n\n 26\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 27 of 67\n\n\n1026 (11th Cir. 2007). “If a reasonable jury could conclude that the evidence\n\nestablishes guilt beyond a reasonable doubt, we will affirm the verdict.” Browne,\n\n505 F.3d at 1253. “Because we recognize that ‘the jury is free to choose between\n\nor among the reasonable conclusions to be drawn from the evidence presented at\n\ntrial,’ our sufficiency review requires only that a guilty verdict be reasonable, not\n\ninevitable, based on the evidence presented at trial.” Id. (quotation marks omitted).\n\n A. Conspiracy to distribute controlled substances without a\n legitimate medical purpose (Count 1)\n\n Both defendants challenge their conviction for conspiring to distribute or\n\ndispense controlled substances without a legitimate medical purpose. To establish\n\na conviction for illegal distribution or dispensation of controlled substances under\n\n21 U.S.C. § 841(a)(1), the Government must prove that the defendant distributed\n\nor “dispensed controlled substances for other than legitimate medical purposes in\n\nthe usual course of professional [medical] practice, and that he did so knowingly\n\nand intentionally.” United States v. Azmat, 805 F.3d 1018, 1035 (11th Cir. 2015)\n\n(quotation marks omitted). In order to establish a conspiracy conviction, the\n\nGovernment must show that: (1) “an agreement existed between two or more\n\npersons to commit a crime” and (2) “the defendants knowingly and voluntarily\n\njoined or participated in the conspiracy.” United States v. Seher, 562 F.3d 1344,\n\n1364 (11th Cir. 2009) (quotation marks omitted). The existence of an agreement\n\nmay “be proved by inferences from the conduct of the alleged participants or from\n 27\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 28 of 67\n\n\ncircumstantial evidence of a scheme.” Azmat, 805 F.3d at 1035 (quotation marks\n\nomitted). “A conspiracy conviction will be upheld if ‘the circumstances\n\nsurrounding a person’s presence at the scene of conspiratorial activity are so\n\nobvious that knowledge of its character can fairly be attributed to him.’” Id.\n\n Here, the Feldmans do not challenge the jury’s finding that Dr. Feldman was\n\ndispensing controlled substances outside the usual course of professional practice.\n\nInstead, their arguments focus on whether Mrs. Feldman conspired with Dr.\n\nFeldman to do so. We conclude that the Government presented ample evidence\n\nfrom which a reasonable jury could conclude that the Feldmans had an agreement\n\nto illegally distribute and dispense prescription medications.\n\n The evidence presented at trial showed that Mrs. Feldman was extensively\n\ninvolved in the operation of the clinic. She wore a white doctor’s coat, she sat in\n\non office visits, she encouraged new patient referrals, and she managed all of the\n\nclinic’s money. Because staff members were instructed to copy the sign-in sheets\n\nwhen providing Mrs. Feldman with any cash payments, the jury could infer that\n\nMrs. Feldman was aware of the large number of patients seen at the clinic each\n\nday. According to former patient Charlene Riffle, Mrs. Feldman also wrote out\n\nprescriptions ordered by Dr. Feldman while he sat at his desk eating fruit.\n\n The Government presented sufficient evidence from which a reasonable jury\n\ncould infer not only that Mrs. Feldman was aware that Dr. Feldman was\n\n\n 28\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 29 of 67\n\n\nprescribing controlled substances outside the usual course of medicine but also that\n\nshe knowingly participated in his efforts to do so. Mrs. Feldman asked staff\n\nmembers to alert her when law enforcement officers were present in the parking lot\n\nand, on those days, she parked in the back of the office. Multiple staff members\n\nalso testified that Mrs. Feldman directed them to put post-it notes in patient files\n\nfor anyone whom they believed was an undercover law enforcement officer. She\n\ntold them that post-it notes were important, so they could be removed easily and\n\nnot become a permanent item in the file.\n\n When the clinic began discharging too many patients, Mrs. Feldman\n\ninstructed staff to stop conducting urinalysis screenings. As two experts testified,\n\nthe urinalysis screenings are instrumental to determining whether a patient is\n\ntaking the medication as prescribed or diverting it to others. And, on at least one\n\noccasion, Aimee Martin, a former patient and friend of the Feldmans, asked Mrs.\n\nFeldman if she could fill her prescription early because it had run out. Mrs.\n\nFeldman suggested that Martin instead take the medication of her husband—who\n\nwas also one of Dr. Feldman’s patients—thereby encouraging the use of controlled\n\nsubstances outside the usual course of professional practice.\n\n Additionally, Mrs. Feldman told former employee Melinda Detwiler that she\n\nand Dr. Feldman had moved their assets into her name and the names of her\n\nchildren, in case anything happened. Finally, as to Dr. Feldman, he testified that\n\n\n 29\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 30 of 67\n\n\nhe did not have any agreement with Mrs. Feldman to illegally distribute controlled\n\nsubstances, but the jury obviously rejected this testimony. We may therefore\n\nconsider his disbelieved testimony as substantive evidence of his guilt. See United\n\nStates v. Woodard, 459 F.3d 1078, 1087 (11th Cir. 2006) (“[A] defendant’s\n\ntestimony—if disbelieved by the jury—may be considered substantive evidence of\n\nguilt.”).\n\n Although Mrs. Feldman asserts that there was an innocent explanation for\n\nher instruction to staff to stop conducing urinalysis screenings—the test was\n\npurportedly too costly—Detwiler testified that the reason Mrs. Feldman instructed\n\nstaff to stop doing the test was because it was resulting in too many patients being\n\ndischarged. Further, Dr. Feldman admitted on cross-examination that the\n\nurinalysis screening did not cost the clinic a lot of money, as the screening done in\n\nthe clinic was included in the cost of the office visit and the patient bore the cost of\n\nany urinalysis screening sent to the laboratory. We are likewise unpersuaded by\n\nMrs. Feldman’s argument that she made the comment to Detwiler about shifting\n\nher assets because she wanted to have everything in place if Dr. Feldman passed\n\naway. Indeed, Detwiler testified that Mrs. Feldman made the comment during a\n\nconversation about clinic patients who were being investigated by law\n\nenforcement. Viewing all of the evidence in the light most favorable to the\n\nGovernment, a reasonable jury could conclude that the Feldmans conspired and\n\n\n 30\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 31 of 67\n\n\nworked together to illegally dispense controlled substances without a legitimate\n\nmedical purpose.\n\n B. Money Laundering (Counts 5 through 8)\n\n The Feldmans also argue that the Government presented insufficient\n\nevidence to establish that they conspired to commit money laundering or that they\n\ncommitted any substantive money laundering offenses.\n\n As to the conspiracy charge alleged in Count 5, the Government needed to\n\nprove two elements to establish a conspiracy under § 1956(h): “(1) an agreement\n\nbetween two or more persons to commit a money-laundering offense; and\n\n(2) knowing and voluntary participation in that agreement by the defendant.”\n\nUnited States v. Moran, 778 F.3d 942, 962 (11th Cir. 2015) (quotation marks\n\nomitted). An essential element of money laundering conspiracy is that the\n\ndefendant “knew that the funds involved in the transactions represented the\n\nproceeds of unlawful activity.” See United States v. Awan, 966 F.2d 1415, 1434\n\n(11th Cir. 1992).\n\n Relying on their argument that the Government presented insufficient\n\nevidence as to the drug conspiracy alleged in Count 1, the Feldmans argue that the\n\nevidence presented at trial was insufficient to prove beyond a reasonable doubt that\n\nthey knew the funds involved in the financial transactions at issue involved\n\ncriminal proceeds. However, based on our conclusion that sufficient evidence\n\n\n 31\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 32 of 67\n\n\nsupported the drug conspiracy conviction, it necessarily follows that the Feldmans\n\nknew that the transactions involved the proceeds of unlawful activity.\n\n Dr. Feldman’s argument that the Government failed to present sufficient\n\nevidence showing that the Feldmans conducted these financial transactions with\n\nthe intent to conceal the criminal proceeds is also without merit. Concealment was\n\nonly one of the objects of the conspiracy. Indeed, Count 5 alleged that the\n\nFeldmans conspired to conduct a financial transaction with criminal proceeds from\n\nthe drug conspiracy (1) to promote the carrying on of unlawful activity, (2) to\n\nconceal the proceeds, (3) to avoid transaction reporting requirements, and (4) to\n\nengage in monetary transactions in property derived from a specified unlawful\n\nactivity. The jury determined that the Feldmans were guilty of all four objects of\n\nthe conspiracy. Further, Dr. Feldman does not challenge the jury’s finding as to\n\nthose other three objects, and he has therefore abandoned any challenge he may\n\nhave had. See United States v. Jernigan, 341 F.3d 1273, 1283 n.8 (11th Cir. 2003)\n\n(stating that an issue not briefed on appeal is considered abandoned). Because the\n\nFeldmans argue only that the Government failed to establish their knowledge of\n\nthe specified unlawful activity—the drug conspiracy—they are not entitled to relief\n\nfrom their money laundering convictions. United States v. Ross, 131 F.3d 970, 984\n\n(11th Cir. 1997) (“A guilty verdict in a multi-object conspiracy will be upheld if\n\nthe evidence is sufficient to support a conviction of any of the alleged objects.”).\n\n\n 32\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 33 of 67\n\n\n VII. Sufficiency of Evidence as to the Convictions Based on the\n Dispensation of Controlled Substances Without a Legitimate\n Medical Purpose that Resulted in Death (Counts 2 through 4)\n\n In Counts 2 through 4, Dr. Feldman was charged with distribution of\n\ncontrolled substances without a legitimate medical purpose resulting in the deaths\n\nof three people: Joey Mayes (Count 2), Ricky Gonzalez (Count 3), and Shannon\n\nWren (Count 4). The jury convicted Dr. Feldman on all counts, finding that each\n\nvictim died as a result of ingesting drugs that Dr. Feldman had prescribed. On\n\nappeal, Dr. Feldman challenges the sufficiency of the evidence showing that\n\nSchedule II drugs he prescribed caused each victim’s death: a fact that triggers a\n\n20-year mandatory minimum sentence under the Controlled Substances Act.\n\n To establish a conviction for unlawful dispensation by a physician under\n\n21 U.S.C. § 841(a)(1), the Government must establish that the defendant\n\nknowingly and intentionally distributed or dispensed controlled substances other\n\nthan for a legitimate medical purpose and outside the ordinary course of\n\nprofessional practice. 21 U.S.C. § 841(a)(1); Azmat, 805 F.3d at 1035. There is a\n\npenalty-enhancement provision for § 841(a), which provides that a defendant shall\n\nbe sentenced to a term of not less than 20 years’ imprisonment, or more than life, if\n\nhe dispensed a Schedule I or II drug and death or serious bodily injury “results\n\nfrom the use of such substance.” 21 U.S.C. § 841(b)(1)(C). “Because the ‘death\n\nresults’ enhancement increase[s] the minimum and maximum sentences to which\n\n\n 33\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 34 of 67\n\n\n[a defendant is] exposed, it is an element that must be submitted to the jury and\n\nfound beyond a reasonable doubt.” Burrage v. United States, 571 U.S. 204, 210\n\n(2014) (citing Alleyne v. United States, 570 U.S. 99, 115–16 (2013), and Apprendi\n\nv. New Jersey, 530 U.S. 466, 490 (2000)).\n\n Here, Dr. Feldman’s sufficiency-of-the-evidence argument proceeds in two\n\nparts. First, he argues that the evidence was insufficient to prove that it was he\n\nwho prescribed the specific drugs that were ingested by each victim. Second, he\n\nargues that the evidence was insufficient to prove that the Schedule II drugs\n\n(oxycodone and methadone)—prescribed by him and ingested by the victims—\n\nwere a but-for cause of each victim’s death. We disagree on both fronts and\n\naddress each issue in turn.\n\n A. Drugs Ingested by the Deceased Were Prescribed by Dr.\n Feldman\n\n Evidence presented at trial fully supported the jury’s finding that, as charged\n\nin the indictment, the specific drugs ingested by each victim (Mayes, Gonzalez,\n\nand Wren) were prescribed by Dr. Feldman.\n\n 1. Ricky Gonzalez\n\n The trial evidence was sufficient to show that Dr. Feldman prescribed the\n\nspecific drugs ingested by Ricky Gonzalez. On October 5, 2010, only three days\n\nbefore his death, Gonzalez received a prescription from Dr. Feldman for 240 30-\n\nmilligram pills of oxycodone and 90 10-milligram pills of diazepam (Valium).\n 34\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 35 of 67\n\n\nGonzalez’s girlfriend, Casey Osteen, testified that she saw Gonzalez take\n\noxycodone and diazepam from prescriptions dispensed by Dr. Feldman the night\n\nbefore he died. When she last saw Gonzalez at approximately 11:00 p.m. that\n\nnight, he appeared “lethargic” and “groggy.”\n\n Gonzalez was found dead in his front yard the next day, October 8, 2010.\n\nOfficers found the bottle of diazepam prescribed by Dr. Feldman in Gonzalez’s\n\npants pocket. The bottle contained only 8 of the 90 10-milligram pills of diazepam\n\nprescribed by Dr. Feldman, as well as 35 other diazepam pills of varying dosages.\n\nThe bottle of oxycodone prescribed by Dr. Feldman three days earlier was found\n\nempty and smashed near a neighbor’s truck. Finally, the toxicology reports\n\nconfirmed that Gonzalez had oxycodone and diazepam—the same two drugs\n\nprescribed by Dr. Feldman—in his system. Based on this evidence, a reasonable\n\njury could infer that the drugs ingested by Gonzalez were prescribed by Dr.\n\nFeldman.\n\n 2. Joey Mayes\n\n The evidence was also sufficient to demonstrate that Dr. Feldman prescribed\n\nthe specific drugs Mayes ingested. Mayes first visited Dr. Feldman on March 12,\n\n2010 complaining of back pain. On this visit, Dr. Feldman prescribed Mayes 240\n\n30-milligram tablets of oxycodone, 60 2-milligram tablets of alprazolam (Xanax),\n\nand 60 2-milligram tablets of methadone. Less than one week later, on March 18,\n\n\n 35\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 36 of 67\n\n\n2010, officers responded to a call from Mayes’ residence reporting his sudden\n\ndeath from a possible drug overdose. Mayes’ mother testified at trial that she\n\nobserved Mayes taking the medications prescribed by Dr. Feldman the day before\n\nhe died. Officers found multiple pill bottles in Mayes’ room, including empty pill\n\nbottles for the oxycodone and alprazolam prescriptions dispensed by Dr. Feldman\n\non March 12. A subsequent toxicology report revealed that, at the time of death,\n\nMayes had in his system methadone or methadone metabolites, oxycodone, and\n\nalprazolam: the same three drugs that Dr. Feldman had prescribed.\n\n Based on this evidence, a reasonable jury could find beyond a reasonable\n\ndoubt that the controlled substances ingested by Mayes were prescribed by Dr.\n\nFeldman. Indeed, the controlled substances found in Mayes’ system were the same\n\ntype of substances prescribed by Dr. Feldman just one week before. In addition,\n\nthe pill bottles containing those prescriptions were found empty in Mayes’ room.\n\nSee United States v. Merrill, 513 F.3d 1293, 1299 (11th Cir. 2008) (concluding that\n\nsufficient evidence established that a doctor’s recent prescriptions led to the\n\npatients’ deaths given the short time period between the prescription and the fact\n\nthat the defendant “prescribed for each of these patients the exact type of drug the\n\nmedical examiner found to have caused their deaths”).\n\n\n\n\n 36\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 37 of 67\n\n\n 3. Shannon Wren\n\n Finally, the evidence was sufficient to prove that Dr. Feldman had\n\nprescribed the specific drugs Wren ingested. Wren was initially seen by Dr.\n\nFeldman on May 18, 2011, at which time Dr. Feldman prescribed Wren 120 15-\n\nmilligram tablets of oxycodone, 60 10-milligram tablets of diazepam, and 60 25-\n\nmilligram tablets of amitriptyline.\n\n Approximately one week later, on May 26, 2011, Wren was seen vomiting\n\nand coughing by his girlfriend as she left for work. Later that day, Wren was\n\nfound dead in his master bathroom. The first officer on the scene found multiple\n\npill bottles in the bathroom. 7 Included in those medications were the containers of\n\namitriptyline that Dr. Feldman had prescribed on May 18, which contained only 34\n\npills out of the 60-count prescription. Subsequently, Wren’s father handed over to\n\nan investigating agent the bottles of oxycodone and diazepam prescribed by Dr.\n\nFeldman, which were empty at that time.\n\n The subsequent autopsy revealed oxycodone, diazepam and its metabolites,\n\nand amitriptyline and its metabolites in Wren’s system. During Wren’s May 18\n\nvisit, Dr. Feldman had prescribed oxycodone, diazepam, and amitriptyline, which\n\n\n\n\n7\n These bottles contained melatonin, gabapentin, fluoxetine, metoprolol, trazodone, anastrozole,\ninjectable testosterone, citalopram, adrenal tea capsule, and amitriptyline. Of these substances,\nonly amitriptyline was found in Wren’s system during the autopsy. As noted in text, Dr.\nFeldman had recently prescribed amitriptyline.\n 37\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 38 of 67\n\n\nwere the same drugs found in Wren’s system after he died. As with the other two\n\nvictims, a reasonable jury could conclude that the drugs that Wren ingested were\n\nprescribed by Dr. Feldman.\n\n In short, with respect to each victim, a jury could readily find beyond a\n\nreasonable doubt that the controlled substances the victim had ingested were\n\nprescribed by Dr. Feldman.\n\n B. Sufficient Evidence Was Presented at Trial From Which a Jury\n Could Conclude That But for Ingestion of the Schedule II\n Drugs Prescribed by Dr. Feldman, the Victims Would Not Have\n Died\n\n As noted, the Government had sought an enhancement of Dr. Feldman’s\n\nsentence on these three substantive drug counts. To repeat, the enhancement\n\nprovision in question imposes a 20-year mandatory minimum sentence and an\n\nincreased maximum sentence for a defendant who has distributed a Schedule I or II\n\ndrug when death or serious bodily injury “results from the use of such substance.”\n\n21 U.S.C. § 841(b)(1)(C). On appeal, Dr. Feldman argues that there was\n\ninsufficient evidence to prove that the Schedule II drugs he prescribed caused the\n\nvictims’ deaths.\n\n As to the role played by a Schedule II drug, the jury found in its special\n\nverdict that “but for” the Schedule II and Schedule IV drugs prescribed to each\n\nvictim by Dr. Feldman, death would not have resulted. Dr. Feldman argues,\n\nhowever, that to convict a person based on death resulting from distribution of a\n 38\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 39 of 67\n\n\nSchedule II drug, the Government must prove that the Schedule II drug was the\n\n“sole,” the “only,” or the “independent” cause of each victim’s death, and that the\n\nGovernment failed to do so. More particularly, he argues that because each victim\n\nalso had a non-Schedule II drug in their bodies—which was also prescribed by Dr.\n\nFeldman—and because the non-Schedule II drug potentially contributed to the\n\nvictim’s death, the Schedule II drug cannot be said to have caused the particular\n\nvictim’s death.\n\n In arguing that the Government must prove that the Schedule II drug alone\n\ncaused the death of the victim, Dr. Feldman relies on Burrage v. United States, 571\n\nU.S. 204 (2014), which examined the causality requirement of § 841(b)(1)(C) of\n\nthe Controlled Substances Act. As we explain below, his argument misreads\n\nBurrage, which neither requires proof that a Schedule I or II drug “alone” caused a\n\nvictim’s death nor precludes application of § 841(b)(1)(C)’s mandatory minimum\n\nsentence in mixed-drug intoxication cases. Instead, Burrage held only that, “where\n\nuse of the drug distributed by the defendant is not an independently sufficient\n\ncause of the victim’s death or serious bodily injury, a defendant cannot be liable\n\nunder the penalty enhancement provision of 21 U.S.C. § 841(b)(1)(C) unless such\n\nuse is a but-for cause of the death or injury.” Id. at 218–19 (emphasis added).\n\nApplying that standard, we find that sufficient evidence of but-for causation was\n\npresented at trial.\n\n\n 39\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 40 of 67\n\n\n 1. Burrage\n\n In Burrage, the victim had died after ingesting alprazolam, clonazepam,\n\noxycodone, hydrocodone, and heroin. 571 U.S. at 207. The defendant had\n\nprovided the victim only with the heroin, which is a Schedule I drug, and he was\n\ntried based on that distribution, as well as on the enhancement penalty triggered by\n\n§ 841(b)(1)(C) when death results from distribution of a Schedule I or II drug. Id.\n\nat 206–07. Neither expert witness could say whether the victim would have lived\n\nhad he not taken the heroin; in other words, he might have died even without the\n\nheroin, given the other drugs he had taken. Id. at 207. This failure of proof meant\n\nthat the Government had not proved that but for the heroin, the victim would have\n\nlived. Id. Nevertheless, the district court concluded that the Government was not\n\nrequired to prove that the heroin was a “but-for” cause of the victim’s death, and it\n\ninstructed the jury that it could find that death resulted from the defendant’s heroin\n\ndistribution if it merely found that the heroin was a “contributing cause” of the\n\nvictim’s death. Id. at 207–08.\n\n The Supreme Court disagreed with the causation standard adopted by the\n\ndistrict court. Because § 841(b)(1)(C) does not define the phrase “results from,”\n\nthe Court applied its ordinary meaning, concluding that the phrase imposed the\n\n“traditional” requirement of “actual causality.” Id. at 210–11. Proving “actual\n\ncausality,” the Court explained, generally requires proof that, at a minimum, “the\n\n\n 40\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 41 of 67\n\n\nharm would not have occurred in the absence of—that is, but for—the defendant’s\n\nconduct.” Id. at 211 (quotation marks omitted). In other words, a factor can be\n\nconsidered an “actual cause” if it was necessary for bringing about that result—that\n\nis, if the result would not have occurred without (or “but for”) the influence of the\n\nfactor.\n\n To illustrate how this but-for notion of “actual causation” applies, the Court\n\nused three hypotheticals. Id. at 211–12. In the first hypothetical, “A shoots B,\n\nwho is hit and dies.” Id. at 211. In this case, only one factor is in play—A’s\n\nconduct in shooting B. The Court explained that, in this example, A actually\n\ncaused B’s death because B would not have died “but for” A’s conduct. Id. In\n\nother words, A’s conduct was necessary for producing B’s death: had A not shot\n\nB, B would have lived.\n\n The Court made clear, though, that but-for causality does not require that a\n\nsingle factor alone produce the particular result. As the Court explained, “[t]he\n\nsame conclusion follows if the predicate act combines with other factors to\n\nproduce the result, so long as the other factors alone would not have done so—if,\n\nso to speak, it was the straw that broke the camel’s back.” Id. (emphasis added).\n\nTo illustrate how a factor can be a but-for cause (that is, a necessary factor for\n\nproducing a result) even when other factors are themselves but-for causes, the\n\nCourt posed a second hypothetical.\n\n\n 41\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 42 of 67\n\n\n In this hypothetical, a “man debilitated by multiple diseases” ingests poison\n\nand dies. Id. Here, the Court explained, the poison “is a but-for cause of his death\n\neven if those diseases played a part in his demise, so long as, without the\n\nincremental effect of the poison, he would have lived.” Id. Thus, both the poison\n\nand the debilitating diseases are but-for causes of the death if each factor played a\n\nnecessary role in producing that outcome. Further, the poison and the debilitating\n\ndiseases would each constitute a but-for cause of the man’s death—“the straw that\n\nbroke the camel’s back”—if the man would have lived had he (1) only had\n\ndebilitating diseases, but not taken the poison or (2) only taken the poison, but not\n\nhad debilitating diseases. In other words, had one of these factors been absent, the\n\ndecedent would not have died. If, however, a debilitating disease was\n\n“independently sufficient”—that is, enough by itself—to kill the man at the same\n\ntime that he was ingesting the poison, the “incremental effect” of the poison was\n\nnot necessary to produce the resulting death. Id. at 211, 218–19. In that case, the\n\npoison would not be considered a but-for cause of the man dying; it was not “so to\n\nspeak, [] the straw that broke the camel’s back,” because the debilitating disease\n\nhad performed that function.8 Id. at 211.\n\n\n8\n The Supreme Court considered, but did not decide, the question of causation when a victim\ndies following two concurrent events, each of which would have been enough on its own to\ncause death. The Court posited a scenario in which “A stabs B, inflicting a fatal wound; while at\nthe same moment X, acting independently, shoots B in the head also inflicting a fatal wound; and\nB dies from the combined effects of the two wounds.” Burrage, 571 U.S. at 215 (alterations\n\n 42\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 43 of 67\n\n\n The Court further elaborated on the but-for-causation principle in a third\n\nhypothetical, asking the reader to consider a baseball game where the visiting\n\nteam’s leadoff batter hits a home run in the first inning and the visiting team\n\nultimately wins 1 to 0. Id. at 211–12. The Court noted that one could accurately\n\nsay that the home run actually caused the victory because the visiting team would\n\nnot have won but for the home run—that is, the home run was necessary for the 1-\n\n0 victory to occur. Id. The Court explained that the home run was no less a but-\n\nfor cause of the victory just because it was only one of many but-for causes that\n\nwere each necessary for the visiting team to win. Id. at 212. As the Court put it,\n\n“[i]t is beside the point that the victory also resulted from a host of other necessary\n\ncauses, such as skillful pitching, the coach’s decision to put the leadoff batter in the\n\nlineup, and the league’s decision to schedule the game.” Id. (emphasis in original).\n\nThis is so because whether a factor is a but-for cause turns only on whether the\n\n\n\n\naccepted) (citation omitted). In other words, B would have died had he only been stabbed.\nLikewise, he would have died had he only been shot. Each was an independent sufficient cause\nof his death.\n\nQuoting from the noted criminal law treatise, LaFave, Substantive Criminal Law at 468, the\nSupreme Court noted that the treatise recognizes a special rule for such scenarios providing that\n“A will generally be liable for homicide even though his conduct was not a but-for cause of B’s\ndeath (since B would have died from X’s actions in any event).” Id. The Court acknowledged,\nhowever, that: “We need not accept or reject the special rule developed for [the above scenario],\nsince there was no evidence here that [the decedent’s] heroin use was an independently sufficient\ncause of his death. No expert was prepared to say that [the decedent] would have died from the\nheroin use alone.” Id.\n 43\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 44 of 67\n\n\nresult would have occurred without it; that is, whether that individual factor made a\n\ndifference to the outcome.\n\n “By contrast,” the Court says, “it makes little sense to say that an event\n\nresulted from or was the outcome of some earlier action if the action merely played\n\na nonessential contributing role in producing the event.” Id. Thus, the Court\n\nexplains, if we modify the baseball hypothetical such that the visiting team’s\n\nleadoff batter hits a home run in the first inning, but the visiting team ultimately\n\nwins by 5 to 2 rather than 1 to 0, the home run was not a but-for cause of the\n\nvictory. Id. This is so because the visiting team would have won even without the\n\nhome run: the visiting team had a total of 5 runs and only needed 3 to win,\n\nmeaning that the other runs that the visiting team scored were collectively\n\nsufficient on their own to cause the victory without needing the incremental effect\n\nof the leadoff hitter’s home run. In this scenario, the leadoff home run played a\n\nnonessential role in producing the outcome. It did not “[break] the camel’s back.”\n\nId. at 211.\n\n 2. Application of Burrage to Multi-Drug-Usage Scenarios\n\n In order to determine when one can properly conclude that death has\n\nresulted from the ingestion of a Schedule I or II drug, which event triggers\n\n§ 841(b)(1)(C)’s sentence enhancement, we move away from our national pastime\n\n\n\n\n 44\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 45 of 67\n\n\nto examine how Burrage applies to various multi-drug use scenarios involving the\n\ningestion of a Schedule I or II drug together with non-Schedule I or II drugs.\n\n First, we posit a deceased who ingested heroin, a Schedule I drug, in an\n\namount large enough to constitute a fatal dose for this person had the heroin been\n\ntaken alone. In this hypothetical, we further assume that the decedent had also\n\ntaken Valium, a Schedule IV drug, 9 but that the Valium, by itself, would not have\n\ncaused death. Figuring out causation here is simple. Because the drug user has\n\ndied from this fatal dose of heroin, we can readily say that the heroin was a but-for\n\ncause of his death. In other words, because the decedent would have lived had he\n\nnot ingested the heroin, use of the heroin was necessary for producing the death.\n\n Indeed, if the prosecution proved that the drug user took a fatal dose of\n\nheroin that by itself would have been enough to produce death, this evidence\n\nsatisfies a higher standard for proving actual causation than even Burrage’s but-for\n\ntest requires. In this situation, the heroin was not only a necessary, but-for cause of\n\nthe death, but it was also independently sufficient, by itself, to cause the user’s\n\ndeath, even without the influence of any other factors, including the Valium.\n\n Second, we change the facts a bit to assume that the deceased not only\n\ningested enough heroin that by itself would be fatal, but he also took enough\n\n\n\n9\n Valium, whose generic version is called diazepam, is a Schedule IV drug that does not give\nrise to an enhancement if death results.\n 45\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 46 of 67\n\n\nValium that, had the Valium been taken alone, it would likewise have caused\n\ndeath. In other words, each of the two drugs independently would have caused\n\ndeath. Although the Supreme Court expressly declined to reach this particular\n\nscenario,10 there is a strong argument that the heroin should be considered to have\n\ncaused the death for purposes of the sentencing enhancement. If actual causation\n\ncan be established by proving a factor was necessary to cause a result (meaning it\n\nsatisfies the but-for test), it logically follows that actual causation can also be\n\nestablished by satisfying a higher standard for proving that the factor was\n\nsufficient, by itself, to cause the result. It would create an extremely odd result if\n\nthe purveyor of a Schedule I or II drug that was strong enough by itself to cause\n\ndeath could escape liability merely because the victim also happened to have taken\n\na non-Schedule I or II drug that was itself independently sufficient to cause death.\n\nAnd the Supreme Court in Burrage indicated that it did not require satisfaction of\n\nthe but-for test when the drug at issue is an independently sufficient cause of the\n\nvictim’s death. Burrage, 571 U.S. at 218–19.\n\n Our third hypothetical takes us into a scenario in which the heroin is not a\n\nbut-for cause. In this hypothetical, where again the deceased has taken both heroin\n\nand Valium, we assume that the heroin was not enough, by itself, to induce death,\n\nbut that the Valium dosage was strong enough to have caused the decedent to die,\n\n\n10\n See discussion at n.8, supra.\n 46\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 47 of 67\n\n\neven had he not ingested any heroin. In this scenario, use of the heroin does not\n\nmeet the but-for test because it had no impact on what ultimately happened to the\n\ndrug user. He would have died—whether or not he had used heroin—because the\n\nSchedule IV Valium was going to kill him anyway.\n\n In our fourth and final hypothetical, we assume the heroin would not have\n\ncaused the deceased’s death had it been taken alone, but likewise the deceased\n\nwould not have died had he taken just the Valium. Yet, by virtue of the fact that he\n\ningested the heroin along with the Valium, he died. Here, the heroin can be\n\nconsidered the straw that broke the camel’s back. Stated another way, had the\n\nvictim never ingested the heroin, the decedent would still be alive. We conclude\n\nthat, consistent with Burrage, the but-for test for the heroin is met in this scenario.\n\n Dr. Feldman, however, disagrees with our analysis, at least as to this fourth\n\nscenario. In the present case, the evidence established that each victim had\n\ningested two types of drugs prescribed by the doctor: Schedule II drugs\n\n(methadone and oxycodone for one victim and oxycodone for the other two\n\nvictims) and a non-Schedule II drug (Xanax as to one victim and Valium as to the\n\nother two victims). Further, the evidence supported an inference either that the\n\nSchedule II drugs were independently sufficient to cause each victim’s death or, at\n\nthe very least, that neither of the non-Schedule II drugs would have caused death\n\nhad the decedent not ingested the Schedule II drug(s). Thus, the evidence clearly\n\n\n 47\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 48 of 67\n\n\nestablished that, had each respective victim not used the Schedule II drug in\n\nquestion, he would not have died. In other words, but for the Schedule II drugs,\n\nthe victims would still be alive.\n\n It is Dr. Feldman’s position, though, that causation cannot be established\n\nunless the Schedule II drug was the sole or only cause of death. According to the\n\ndoctor, if the non-Schedule II drugs (that he had also prescribed together with the\n\nSchedule II drugs) played any role in causing the victim’s death, it can no longer\n\nbe said that use of the Schedule II drugs resulted in death, even though the\n\nSchedule II drugs have met the but-for test. That certainly sounds like an odd\n\nconclusion. And it is. Burrage does not require proof that a Schedule I or II drug\n\nbe the “only” cause of a victim’s death. Indeed, the Court explicitly rejected that\n\ncharacterization of but-for causation, stating that a factor can be a but-for cause\n\neven if it “combines with other factors to produce the result,” and that whether “a\n\nhost of other necessary causes” might also contribute to the result is “beside the\n\npoint.” Id. at 211–12 (emphasis in original). Moreover, a rule requiring proof that\n\na drug is the one-and-only but-for cause would be tantamount to requiring a\n\nshowing that the drug was “independently sufficient” to cause the death—a bar\n\nthat is much higher that the “minimum” showing of but-for causation that Burrage\n\nrequired. Id. at 211 (emphasis in original). Notably, Burrage expressly held that\n\nthe Government need not prove that a Schedule I or II drug was “an independently\n\n\n 48\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 49 of 67\n\n\nsufficient cause of the victim’s death or serious bodily injury,” stating that only\n\nbut-for causation is required, and implying that we can dispense with the but-for\n\ntest when a drug is an independently sufficient cause of death. Id. at 218–19.\n\n Likewise, Dr. Feldman’s argument that evidence of “mixed-drug\n\nintoxication” necessarily precludes a finding that a victim died as a result of a\n\nSchedule I or II drug finds no support in Burrage. In Burrage, the medical experts\n\ntestified that the victim had died from “mixed drug intoxication with heroin,\n\noxycodone, alprazolam, and clonazepam all playing a contributing role.” Id. at\n\n207 (quotation marks omitted). Neither expert could say, however, “whether [the\n\nvictim] would have lived had he not taken the heroin.” Id. In concluding that\n\ninsufficient evidence supported the defendant’s conviction, the Court held that\n\nproving actual causation required more than showing that a Schedule I or II drug\n\n“contributes to an aggregate force (such as mixed-drug intoxication) that is itself a\n\nbut-for cause of death.” Id. at 214. But this statement in no way suggests that a\n\ndefendant is off the hook anytime a Schedule I or II drug that he distributed ends\n\nup in a deadly drug cocktail that kills a victim. Instead, the Court’s statement was\n\nintended to clarify that the Government could not establish that a Schedule I or II\n\ndrug actually caused a death without evidence that the Schedule I or II drug\n\nitself—as opposed to the drug cocktail as a whole—made a difference to the\n\noutcome. Id. at 214, 219 (concluding that the lack of “evidence that [the victim]\n\n\n 49\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 50 of 67\n\n\nwould have lived but for his heroin use” required reversal of the defendant’s\n\nconviction). In other words, to prove but-for causation, the Government must drill\n\ndown and show that, but for the Schedule I or II drug’s inclusion in the drug\n\ncocktail, the victim would have lived. If the Government makes that showing, it is\n\n“beside the point” that other drugs were necessary to make the drug cocktail\n\ndeadly. See id. at 211–12, 218–19.\n\n Indeed, a sister circuit, the Sixth Circuit, has reached the same conclusion.\n\nIn United States v. Volkman, 797 F.3d 377, 396 (6th Cir. 2015), the defendant\n\ndoctor had been convicted of the enhancement provision based on his dispensing\n\nof the Schedule II drug oxycodone to the victim. He argued that because the\n\ndecedent died based on the “combined effects of oxycodone, diazepam, and\n\nalprazolam,” the Government had failed to prove that the death resulted from the\n\ningestion of oxycodone. Id. at 395. Relying on Burrage, the Sixth Circuit panel\n\nrejected this argument, holding that “[t]he Government was not required to prove,\n\nhowever, that oxycodone was [the victim’s] only cause of death” because “but-for\n\ncausation exists where a particular controlled substance—here, oxycodone—\n\n‘combines with other factors’—here, inter alia, diazepam and alprazolam—to\n\nresult in death,” and “[t]he Government presented sufficient oxycodone-specific\n\nevidence for a rational jury to find that, ‘without the incremental effect’ of the\n\noxycodone, [the victim] would not have died.” Id. So it is here. As Volkman\n\n\n 50\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 51 of 67\n\n\nexplained, “the existence of other potential contributing causes of death is\n\nirrelevant so long as the Government presented sufficient evidence that [the\n\nSchedule II controlled substance] was a but-for cause of [the victim’s] death.” Id.\n\nat 395–96; see also Roundtree v. United States, 885 F.3d 1095, 1098 (8th Cir.\n\n2018) (concluding that where an expert testified that heroin and alcohol had\n\nworked together “synergistically” to cause death and where alcohol was not\n\nenough by itself to cause death, no reasonable jury could find that the heroin was\n\nmerely a contributory factor, as opposed to a but-for cause).\n\n With these principles in mind, we conclude that the Government presented\n\nsufficient evidence for each of Counts 2–4 to prove beyond a reasonable doubt that\n\na Schedule II drug that Dr. Feldman prescribed (methadone or oxycodone) was, at\n\na minimum, a but-for cause of each victim’s death.\n\n 3. Ricky Gonzalez\n\n The evidence presented at trial was sufficient to support a conclusion that a\n\nSchedule II drug—oxycodone—was a but-for cause of Gonzalez’s death. First,\n\nevidence provided by the defense expert, Dr. Adams, indicated that there was\n\nenough oxycodone in Gonzalez’s system to kill a person who did not have a high\n\ndegree of tolerance. In other words, use of oxycodone was a potentially sufficient\n\ncause, by itself, to cause death.\n\n\n\n\n 51\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 52 of 67\n\n\n But whether or not oxycodone was a sufficient cause, there was ample\n\nevidence from which a jury could conclude that but for Gonzalez’s use of\n\noxycodone, he would not have died. Dr. Thogmartin, the medical examiner who\n\nperformed Gonzalez’s autopsy, observed that Gonzalez had pulmonary edema, or\n\nfluid-filled lungs, an occurrence often seen in drug overdoses. Dr. Thogmartin\n\nconcluded that Gonzalez’s cause of death was oxycodone toxicity with a\n\ncontributory condition of diazepam (Valium) intoxication. 11\n\n Noting that Gonzalez had a significant amount of oxycodone and diazepam\n\nin his system, the doctor explained that the oxycodone, a Schedule II drug, had\n\n“poisoned” Gonzalez and the diazepam, a Schedule IV drug, had reacted\n\n“synergistically” to speed up his death. He explained that in an overdose scenario\n\ninvolving the mixture of an opiate, such as oxycodone, and a benzodiazepine, such\n\nas diazepam, the “main player” in causing death is the opiate, with the\n\nbenzodiazepine acting as only a “minor leaguer,” a “contributory cause.” As to\n\nhow he would characterize such an interaction, he determined that the oxycodone\n\nwas “the big player” and therefore the “cause of death,” and that the diazepam\n\n“was not lethal in and of itself, but a helper.” Indeed, in explaining that\n\n“oxycodone is the big player,” the doctor doubted that diazepam alone would have\n\n\n\n11\n Again, the jury concluded that Dr. Feldman had prescribed both the oxycodone and diazepam\nfound in Gonzalez’s system.\n 52\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 53 of 67\n\n\nkilled Gonzalez,” noting that “[d]iazepam is really hard to overdose on. . . . I mean,\n\nyou’d have to like eat a ton and even then you may not die.” While he thought\n\nthere was enough oxycodone here to kill Gonzalez by itself, he couldn’t say for\n\ncertain that Gonzalez would have died had he not also ingested the diazepam, as\n\nthat is the typical scenario he has observed. Nevertheless, Dr. Thogmartin\n\nconcluded, with a reasonable degree of medical certainty, that if it were not for the\n\noxycodone he ingested, Gonzalez would “not have died this day of this death.” In\n\nshort, the doctor concluded that Gonzalez died of oxycodone intoxication, with\n\ndeath having been sped up by the concurrent ingestion of the diazepam.\n\n In short, at the least, a jury could reasonably conclude that the oxycodone\n\nwas a necessary, but-for cause of his death. In other words, Gonzalez would have\n\nremained alive had he not ingested this Schedule II drug. Even assuming that the\n\ndiazepam may have also constituted a necessary, but-for cause of death, that fact\n\ndoes not deprive the oxycodone of its own “but-for” status. To argue that it does is\n\nto conclude that the oxycodone had to be a sufficient cause of death: that is, that\n\nthe drug had to, by itself, be capable of causing death. As explained above, this is\n\nnot how Burrage defined a but-for cause and indeed Burrage made clear that there\n\ncan be multiple but-for causes of an event. Accordingly, a reasonable jury could\n\nconclude beyond a reasonable doubt that Gonzalez died as a result of a Schedule II\n\ndrug that Dr. Feldman prescribed: namely, oxycodone.\n\n\n 53\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 54 of 67\n\n\n 4. Joey Mayes\n\n The toxicology report indicated that Mayes had ingested two Schedule II\n\ndrugs—methadone and oxycodone—and one Schedule IV drug—alprazolam\n\n(Xanax). Dr. Feldman had prescribed all three drugs. There was testimony from\n\nboth the Government’s expert, who was the medical examiner who performed the\n\nautopsy, and from the defense expert that there was enough methadone in Mayes’\n\nsystem to kill him, even had other drugs not been ingested.\n\n As for the defense expert, Dr. Adams testified that Mayes’ cause of death\n\nwas “intoxication by methadone,” which was “the only drug on board capable of\n\ncausing a cardiac death.” From that testimony, alone, a jury could have found, at a\n\nminimum, that methadone played a necessary (but-for) role in Mayes’ death,\n\nwithout which he would have lived. Moreover, because Dr. Adams’ testimony\n\nexculpated the Schedule IV alprazolam, as well as the Schedule II oxycodone, a\n\njury would have been entitled to find that methadone was also an “independently\n\nsufficient cause” of Mayes’ death. Burrage, 571 U.S. at 218–19.\n\n As for the Government’s expert, in his autopsy report, Dr. Wilson had listed\n\nthe cause of death as multidrug toxicity involving the above-listed drugs. He did\n\nnot stop there, however. Dr. Wilson further testified that Mayes had “particularly\n\nelevated” levels of methadone in his system in an amount that was enough by itself\n\nto kill him, and he made clear that the methadone was the critical drug necessary to\n\n\n 54\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 55 of 67\n\n\nhave caused Mayes’ death. Further, given the high level of methadone, Mayes\n\nwould have died even had he received quicker medical attention. In addition to the\n\nmethadone, Dr. Wilson also stated that there was likely enough oxycodone (a\n\nSchedule II drug), by itself, to kill Mayes.\n\n And while the doctor acknowledged that his autopsy report had used the\n\nphrase “multidrug toxicity,” explaining that in his opinion each drug “had a role”\n\nin killing Mayes, acting “synergistically” with one another, he reiterated that the\n\nmethadone was “the major drug that was involved in Mr. Mayes’ death, with\n\ndistributing effects from alprazolam as well as oxycodone.”\n\n Based on this testimony, a reasonable jury could conclude that each of the\n\nSchedule II drugs (methadone and oxycodone) played, at a minimum, a necessary\n\n“but-for” role in causing Mayes’ death. As discussed above, that another drug (the\n\nalprazolam) may have also played a necessary role in causing Mayes’ death is\n\n“beside the point.” Id. at 211–12.\n\n 5. Shannon Wren\n\n As to Shannon Wren, defense expert Dr. Adams opined that the cause of\n\ndeath for this decedent was a heart attack, not the drugs ingested by Wren. The\n\nmedical examiner, Dr. Wilson, disagreed. Instead, Dr. Wilson concluded that\n\nWren died from multidrug toxicity. He testified that the oxycodone in Wren’s\n\nsystem would have been enough, by itself, to kill him, and that the diazepam would\n\n\n 55\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 56 of 67\n\n\nnot have been sufficient to kill Wren, by itself. As to listing the ultimate cause of\n\ndeath as multidrug toxicity, he explained that each of the drugs acted with the\n\nothers to cause death. 12\n\n In short, taking the evidence in the light most favorable to the Government,\n\nthis evidence was sufficient to support an inference that the oxycodone was a\n\nsufficient cause of Wren’s death, as well as a necessary, “but-for” cause.\n\n Dr. Feldman points out that his expert, Dr. Adams, disagreed with Dr.\n\nWilson’s conclusion that Wren died from multidrug toxicity. Instead, Dr. Adams\n\nbelieved that Wren’s enlarged heart and the fact that he was found dead on the\n\nbathroom floor indicated that he had died from sudden cardiac arrest, rather than\n\nfrom a drug overdose. Dr. Adams did not perform the autopsy, however; Dr.\n\nWilson did. Further, Dr. Wilson considered Wren’s enlarged heart and\n\nnevertheless concluded that there was no evidence of heart attack because there\n\nwas “still enough . . . reservoir opening of the coronary arteries to give enough\n\nblood to the heart.” It was within the province of the jury to choose among the\n\nreasonable constructions of the evidence and weigh Dr. Wilson’s testimony more\n\nheavily than Dr. Adams’. See United States v. Ndiaye, 434 F.3d 1270, 1294 (11th\n\n\n\n12\n The drugs found in Wren’s system included oxycodone, diazepam, amitriptyline, and\nmetabolites of the latter two drugs. Like diazepam, amitriptyline (more familiarly known as the\nantidepressant Elavil) is a Schedule IV drug. The defense expert indicated that this Elavil was in\n“an innocuous concentration,” and the Government’s expert did not focus on it at all. Nor was\nthis drug listed on the special verdict form.\n 56\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 57 of 67\n\n\nCir. 2006) (“Because the jury is free to choose among reasonable constructions of\n\nthe evidence, the evidence may be sufficient even if it is not entirely inconsistent\n\nwith conclusions other than guilt.”).\n\n In sum, when viewed in the light most favorable to the Government, there\n\nwas sufficient evidence from which a reasonable jury could conclude that a\n\nSchedule II drug prescribed by Dr. Feldman—methadone and oxycodone in\n\nMayes’ case, and oxycodone in Gonzalez’s and Wren’s cases—was, at the very\n\nleast, a but-for cause of each of the victims’ deaths, if not an independently\n\nsufficient cause.\n\n VIII. Alleged Alleyne Error\n Dr. Feldman asserts that even if there was sufficient evidence to support the\n\nsentencing enhancement for Counts 2–4, its application in this case was\n\nunconstitutional under Alleyne v. United States, 570 U.S. 99 (2013). The\n\nGovernment does not dispute that Dr. Feldman adequately preserved this issue\n\nbelow, and we therefore review his claims of Alleyne error de novo. United States\n\nv. King, 751 F.3d 1268, 1278–79 (11th Cir. 2014).\n\n As noted above, in Counts 2–4, Dr. Feldman was convicted of violating\n\n21 U.S.C. § 841(a)(1) based on his unlawful dispensing of oxycodone (a Schedule\n\nII drug), methadone (a Schedule II drug), alprazolam (a Schedule IV drug), and\n\ndiazepam (a Schedule IV drug) to the three individuals identified in those counts.\n\n\n 57\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 58 of 67\n\n\nThe jury rendered that verdict by answering in the affirmative the special verdict\n\nform’s 13 first two questions for each count, finding that Dr. Feldman was guilty of\n\ndistributing and dispensing the Schedule II and Schedule IV drugs identified in the\n\nindictment. Although Dr. Feldman challenged the sufficiency of the evidence\n\nsupporting his conviction for conspiracy to distribute these substances and argued\n\nthat insufficient evidence supported a jury finding that the Schedule II drugs\n\nresulted in death, he has not challenged his substantive distribution convictions.\n\nAnd even had he done so, we would have rejected that challenge. There is ample\n\nevidence to support the jury’s conclusion that Dr. Feldman prescribed all of these\n\nsubstances to the particular victims and that he did so not for a legitimate medical\n\npurpose and not in the usual course of professional practice.\n\n The maximum penalty for an “unenhanced” conviction based on distribution\n\nof a Schedule II drug is 20 years’ imprisonment. 21 U.S.C. § 841(b)(1)(C). Thus,\n\nwithout any other enhancements, Dr. Feldman was subject to a 20-year maximum\n\nsentence, with no mandatory-minimum provision, for each conviction set out in\n\nCounts 2–4.\n\n The Government, however, sought an enhanced sentence, alleging that each\n\nof the victims identified for purposes of Counts 2–4 had died as a result of their\n\n\n13\n The special verdict form for each victim included four overarching questions, with two of the\nquestions including sub-parts that asked the jury to make findings regarding each specified\ncontrolled substance.\n 58\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 59 of 67\n\n\ningestion of the Schedule II drugs prescribed by Dr. Feldman. Under the\n\nControlled Substances Act’s sentence-enhancement provision, if death results from\n\nthe use of a Schedule II substance, the person convicted of distributing that drug is\n\nsubject to a mandatory-minimum 20-year sentence, with a statutory maximum of\n\nlife imprisonment. Id. § 841(b)(1)(C). Death resulting from ingestion of a\n\nSchedule IV substance, however, results in no enhancement.\n\n The district court determined that the jury had convicted Dr. Feldman of\n\nconduct justifying the above enhancement. As a result, Dr. Feldman was subject to\n\na mandatory-minimum 20-year sentence, and the district court ultimately decided\n\nthat a 300-month sentence (25-years) was a reasonable sentence based on the\n\nconduct underlying these three convictions.14\n\n “Because the ‘death results’ enhancement increase[s] the minimum and\n\nmaximum sentences to which [a defendant is] exposed, it is an element that must\n\nbe submitted to the jury and found beyond a reasonable doubt.” Burrage, 571 U.S.\n\nat 210 (citing Alleyne, 570 U.S. at 115–16, and Apprendi, 530 U.S. at 490). Dr.\n\nFeldman asserts that the jury did not convict him of distributing a Schedule II\n\nsubstance from which death has resulted, which, if correct, would mean that the\n\n\n14\n Dr. Feldman received a 10-year sentence for his convictions for conspiracy to distribute\ncontrolled substances and conspiracy to commit money laundering, a 10-year sentence for his\nsubstantive money laundering convictions, and a 25-year sentence for the enhanced distribution\nof a Schedule II substance, use of which had resulted in death. All sentences ran concurrently\nwith each other.\n\n 59\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 60 of 67\n\n\ndistrict court violated Alleyne when it sentenced Dr. Feldman as if the jury had so\n\nfound. Specifically, and relying again on Burrage, Dr. Feldman argues that the\n\njury’s special verdict established only a finding that a combination of Schedule II\n\nand Schedule IV drugs in the aggregate killed each victim, not that the victims\n\nwould have lived “but for” their use of the Schedule II drugs. Accordingly, Dr.\n\nFeldman contends that the jury verdict was insufficient to trigger a 20-year\n\nmandatory minimum sentence enhancement or a statutory maximum exceeding\n\ntwenty years.\n\n To the extent that Dr. Feldman’s Alleyne argument mirrors his sufficiency-\n\nof-the-evidence argument, we reject it as a misreading of Burrage. As discussed\n\nabove, Burrage did not hold that a Schedule I or II drug must be the “only” cause\n\nof a victim’s death. Indeed, Burrage reached the opposite conclusion, explaining\n\nthat a factor can be a but-for cause even if it “combines with other factors to\n\nproduce the result, so long as the other factors alone would not have” produced the\n\nsame result. Id. at 211. Under Burrage, but-for causality simply turns on whether\n\nthe victim would have died “without the incremental effect” of a Schedule I or II\n\ndrug. Id. In other words, a Schedule I or II drug is a but-for cause of a victim’s\n\ndeath if it made a necessary contribution to the death—“if, so to speak, it was the\n\nstraw that broke the camel’s back”—such that the victim would have lived if he\n\nhad not taken the Schedule I or II drug. Id. The existence of other necessary\n\n\n 60\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 61 of 67\n\n\ncauses, such as other drugs in a deadly cocktail, is “beside the point.” Id. at 212.\n\nFor a sentencing court to impose an enhanced sentence under § 841(b)(1)(C), a\n\njury need only find that a Schedule I or II drug was a but-for cause of death or\n\nserious bodily injury. Id. at 210–11, 218–19.\n\n As discussed at length above, we conclude that the evidence was sufficient\n\nto prove beyond a reasonable doubt that but for ingestion of the Schedule II drugs\n\ndispensed by Dr. Feldman, the victims would not have died. Nevertheless, we\n\nagree with Dr. Feldman that the special verdict in this case failed to establish that\n\nthe jury actually found that the Schedule II drugs were but-for causes of the\n\nvictims’ deaths. Accordingly, we conclude that the district court erred under\n\nAlleyne by imposing an enhanced sentence under § 841(b)(1)(C). We explain our\n\nreasons for reaching this result.\n\n The third and fourth questions on the verdict form for each of Counts 2, 3,\n\nand 4 focused on the deaths of Mayes, Gonzalez, and Wren, respectively.\n\nAggregating all the drugs dispensed by Dr. Feldman to the victim—both Schedule\n\nII and Schedule IV—the third question asked whether the jury found that the\n\nvictim would not have died but for his use of the listed drugs. 15 The question for\n\n\n\n15\n Asking for a “yes” or “no” answer, the question read: “We, the jury, find beyond a\nreasonable doubt that the death of [the victim] resulted from the use of one or more of the\nfollowing controlled substances: [identifying Schedule II and Schedule IV controlled substances\ndispensed to victim], that is, but for his ingestion of the controlled substances charged, he would\nnot have died.”\n 61\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 62 of 67\n\n\nCount 2 (Mayes) listed as the relevant drugs oxycodone and methadone, which are\n\nSchedule II drugs, as well as alprazolam, which is a Schedule IV drug. The\n\nquestions for Counts 3 and 4 (Gonzalez and Wren) each listed as the relevant drugs\n\noxycodone, a Schedule II drug, and diazepam, a Schedule IV drug. For each\n\ncount, the jury responded “Yes,” finding that the victim would not have died but\n\nfor his use of the listed controlled substances.\n\n These findings clearly do not support the district court’s application of\n\n§ 841(b)(1)(C)’s mandatory-minimum sentence. This is so because, at most, the\n\njury’s “Yes” responses to these questions represented a finding only that the\n\nSchedule II and Schedule IV drugs in the aggregate caused the deaths. And\n\nBurrage held that it is insufficient merely to show that a Schedule I or II drug was\n\npart of a mixed-drug cocktail, which, as “an aggregate force,” was “itself a but-for\n\ncause of death.” Id. at 214. Instead, Burrage requires a jury finding that a\n\nSchedule I or II drug was itself a but-for cause of the victim’s death—that is,\n\nwithout the specific incremental effect of the Schedule I or II drug, the victim\n\nwould have lived. See id. at 207, 211, 218–19 (reversing the defendant’s\n\nconviction where the evidence showed only that the victim died as a result of a\n\ncocktail of drugs, including heroin, oxycodone, alprazolam, and clonazepam, and\n\nthe Government conceded that no evidence showed the victim would have lived\n\nbut for his heroin use).\n\n\n 62\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 63 of 67\n\n\n The fourth question on Dr. Feldman’s special verdict form purportedly\n\nattempted to disaggregate the drugs to seek a finding as to which specific drugs\n\ncaused the death. Considering all the circumstances, however, we conclude that\n\nthis question failed to elicit the jury’s finding on what was the relevant question:\n\nwhether the Schedule II drugs were but-for causes of the victims’ deaths.\n\nSpecifically, this final question for each of Counts 2–4 asked whether the jury\n\nfound “beyond a reasonable doubt that the death of [the victim] resulted from the\n\nuse of the following,” then it listed separately each specific drug dispensed to the\n\nvictim: both Schedule II and Schedule IV substances. (emphasis added). The jury\n\nanswered “Yes” for each drug listed. In contrast with the prior question, though,\n\nthis fourth question entirely omitted any mention of but-for causation and failed to\n\nmake clear to the jury that absent a finding that the victim would not have died had\n\nhe not used a particular drug, the jury could not conclude that the victim’s death\n\nresulted from that drug.\n\n We cannot conclude based on the jury’s response to this final question that it\n\nfound beyond a reasonable doubt that a Schedule II drug Dr. Feldman prescribed—\n\noxycodone and methadone for Count 2, and oxycodone for each of Counts 3 and\n\n4—was a but-for cause of the victims’ deaths. First, the verdict form’s failure to\n\nspell out what “resulted from” meant in the critical and final question raises a\n\nsignificant concern that the jury was unaware of that phrase’s meaning as it\n\n\n 63\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 64 of 67\n\n\npertains to whether death resulted from a specific drug. And because the first\n\n“results in” question explained that “resulted from” meant that the victim would\n\nnot have died but for ingestion of the drugs in the aggregate, but the second\n\nquestion omitted an explanation of what “resulted from” meant, the jury could\n\nhave concluded that it was once again being asked to find whether all the drugs\n\nresulted in each victim’s death, rather than whether each victim would have lived\n\nbut for his ingestion of a Schedule II drug in particular.\n\n Even with the omission of the “but-for” language in this pivotal, final\n\nquestion, we might well reach a different result had the district court’s instructions\n\nexplained to the jury that, in order to conclude that death had resulted from a\n\nparticular Schedule II drug, it must find that but for ingestion of that drug, death\n\nwould not have occurred. Unfortunately, the court’s instructions did not do so.\n\nThe court instructed the jury only that it “must find beyond a reasonable doubt that\n\nbut for the decedent’s use of the charged controlled substances distributed . . . the\n\ndecedent would not have died.” (emphasis added). This plural reference to\n\ncontrolled substances in the instructions necessarily included the Schedule IV\n\nsubstances, which of course cannot trigger an enhancement. And it replicates the\n\nfirst interrogatory on the special verdict form, which directed the jury to make a\n\nfinding only as to whether the drugs in the aggregate caused each victim’s death.\n\n\n\n\n 64\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 65 of 67\n\n\n Thus, the court’s instructions failed to make clear that the jury needed to\n\nmake a specific finding as to whether the victim would have lived but for his use of\n\nthe Schedule II controlled substance. Indeed, the instructions never indicated that\n\nthe Schedule II drugs had any special significance in the case, as the court merely\n\ninstructed the jury that “a more serious offense is committed” whenever “death or\n\nserious injury is a consequence of the decedent’s use of a controlled substance.”\n\n(emphasis added). Nor did the prosecutor or defense attorney point out in closing\n\narguments the specific determination that the jury was being called on to make.\n\n Given this context, we cannot conclude that the jury found beyond a\n\nreasonable doubt that the victims would have lived but for their ingestion of the\n\nSchedule II drugs. Admittedly, the jury’s finding that the victims’ deaths “resulted\n\nfrom” each of the Schedule II and Schedule IV drugs is not inconsistent with a\n\nfinding that the Schedule II drugs were but-for causes of the deaths. As we have\n\nexplained at length, there can be many necessary causes of a result, and, if a\n\nSchedule II drug was a necessary cause of a victim’s death, it would have been\n\n“beside the point” that a Schedule IV drug also happened to occupy the same\n\nstatus. Id. at 211–12.\n\n But the verdict rendered is likewise consistent with a finding that the\n\nSchedule II drugs were nonessential “contributing cause[s],” which did not make a\n\ndifference to the ultimate outcome but merely “contribute[d] to an aggregate force\n\n\n 65\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 66 of 67\n\n\n(such as mixed-drug intoxication) that is itself a but-for cause of death.” Id. at 208,\n\n214. The latter finding, of course, would be insufficient to support a 20-year\n\nmandatory minimum sentence under Burrage. Id. at 214–16. And because we\n\ncannot determine which of the two findings the special verdict reflects, the jury’s\n\nverdict does not constitute the necessary finding that but for ingestion of a\n\nSchedule II substance, the victim would have lived.16 That being so, the district\n\ncourt’s application of § 841(b)(1)(C)’s sentencing enhancement cannot stand. 17\n\n Accordingly, we reverse the district court’s application of § 841(b)(1)(C)’s\n\n20-year mandatory minimum sentence on Counts 2–4 and remand the case for the\n\ndistrict court to resentence Dr. Feldman “to a term of imprisonment of not more\n\nthan 20 years” as to each of these counts. 21 U.S.C. § 841(b)(1)(C) (stating that\n\n“such person shall be sentenced to a term of imprisonment of not more than\n\n20 years and if death or serious bodily injury results from the use of such substance\n\nshall be sentenced to a term of imprisonment of not less than twenty years or more\n\nthan life”).\n\n\n\n\n16\n We emphasize that we are not saying that the jury acquitted Dr. Feldman of the charge that his\ndispensing of a Schedule II drug resulted in death. Rather, the outcome of our ruling is the same\nas it would have been had the Government never sought a specific finding from the jury on this\nissue.\n17\n We are aware that an Alleyne error can be subject to harmless error review. United States v.\nKing, 751 F.3d 1268, 1279 (11th Cir. 2014). The Government, however, has not argued that any\nAlleyne error here would be harmless beyond a reasonable doubt.\n 66\n\f Case: 16-12978 Date Filed: 08/30/2019 Page: 67 of 67\n\n\n CONCLUSION\n\n For the above reasons, we AFFIRM the Feldmans’ convictions, REVERSE\n\nDr. Feldman’s sentences on Counts 2–4, and REMAND for the district court to\n\nconduct a new sentencing hearing for Dr. Feldman.\n\n\n\n\n 67\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: United States v. Edward Neil Feldman\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-30\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 16-12978 Date Filed: 08/30/2019 Page: 1 of 67\\n\\n\\n [PUBLISH]\\n\\n\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n ________________________\\n\\n No. 16-12978\\n ________________________\\n\\n D.C. Docket No. 8:14-cr-00521-JDW-AEP-1\\n\\n\\n\\nUSA,\\n\\n Plaintiff - Appellee,\\n\\nversus\\n\\nEDWARD NEIL FELDMAN,\\nKIM XUAN FELDMAN,\\n\\n Defendants - Appellants.\\n\\n ________________________\\n\\n Appeals from the United States District Court\\n for the Middle District of Florida\\n ________________________\\n\\n (August 30, 2019)\\n\\nBefore JILL PRYOR and JULIE CARNES, Circuit Judges, and ANTOON,*\\nDistrict Judge.\\n\\n\\n *\\n Honorable John Antoon II, United States District Judge for the Middle District of\\nFlorida, sitting by designation.\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 2 of 67\\n\\n\\n\\n\\nJULIE CARNES, Circuit Judge:\\n\\n Following a jury trial, Defendants Dr. Edward Feldman and his wife Kim\\n\\nFeldman appeal their convictions for conspiring to distribute controlled substances\\n\\nwithout a legitimate medical purpose, conspiring to commit money laundering, and\\n\\nthree counts of illegal monetary transactions. Dr. Feldman was also convicted of\\n\\nthree counts of distributing controlled substances, without a legitimate medical\\n\\npurpose, to three individuals resulting in their deaths. Both Defendants raise\\n\\nvarious challenges to their convictions on appeal. Dr. Feldman also challenges his\\n\\n300-month sentence, which incorporated a 20-year mandatory-minimum penalty\\n\\npursuant to 21 U.S.C. § 841(b)(1)(C). After careful review of the parties’ briefs\\n\\nand with the benefit of oral argument, we affirm both Defendants’ convictions, but\\n\\nvacate Dr. Feldman’s sentence.\\n\\n BACKGROUND\\n\\n I. Facts\\n\\n A. The Clinic\\n\\n This case involves the operation of a pain-management clinic in Florida:\\n\\nFeldman Orthopedic and Wellness Center. The clinic was operated by Dr.\\n\\nFeldman, a physician, along with his wife, who served as the clinic’s office\\n\\nmanager and handled all of the clinic’s money. The clinic did not accept\\n\\ninsurance. To obtain an office visit, patients paid $300 in cash or credit for their\\n\\n 2\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 3 of 67\\n\\n\\nfirst visit and $150 for each follow-up visit. Patients were provided a discount for\\n\\nreferring new patients to the clinic. Office visits with Dr. Feldman were typically\\n\\nbrief, as he usually performed cursory examinations—or sometimes no\\n\\nexamination at all.\\n\\n Dr. Feldman was the only physician at the clinic’s first location on 38th\\n\\nAvenue, where he saw approximately 40 to 100 patients per day. According to one\\n\\nformer employee, the clinic’s patients came from out of state and often appeared\\n\\n“stoned” and “drunk.” One of the clinic’s former patients, Mike Shaw, testified\\n\\nthat he came to the clinic because Dr. Feldman was a “writer,” meaning that he\\n\\nwould prescribe any medication suggested by the patient. The clinic sometimes\\n\\nreceived phone calls from concerned family members of patients suffering from\\n\\naddiction, yet Dr. Feldman did not treat those patients any differently.\\n\\n In July 2010, the clinic moved to 66th Street, where Dr. Feldman began\\n\\nseeing 75 to 150 patients per day. Due to heavy law enforcement presence at this\\n\\nlocation, the Feldmans instructed their staff to stop accepting patients from out of\\n\\nstate. They also told their staff to place notes in the patient files if they suspected\\n\\nthat someone was an undercover law enforcement officer. If a patient was flagged\\n\\nas an undercover officer, Dr. Feldman conducted a more extensive physical and\\n\\ndid not prescribe those patients any controlled substances. Based on his belief that\\n\\n\\n\\n\\n 3\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 4 of 67\\n\\n\\nsunglasses and hats could conceal recording devices, Dr. Feldman banned these\\n\\nitems at the clinic’s third and final location on Park Boulevard.\\n\\n In an effort to tighten protocol, the Feldmans also instructed staff to\\n\\ndischarge patients whose urine tested positive for cocaine or other serious drugs.\\n\\nIn addition, the clinic began utilizing the Prescription Drug Monitoring Database—\\n\\na database that shows the prescriptions a patient has previously obtained and serves\\n\\nas a tool for determining whether patients are doctor shopping. Notwithstanding\\n\\nthe above gestures, Mrs. Feldman ultimately told staff to stop conducting urinalysis\\n\\nscreenings because too many patients were being discharged based on those\\n\\nresults. Moreover, the Feldmans decided that no patient would be discharged until\\n\\nDr. Feldman had himself reviewed that patient.\\n\\n In January 2011, a second doctor, Dr. Nancy Bruemmer joined the clinic.\\n\\nDr. Bruemmer often recommended lowering patients’ dosage of medications or\\n\\ndischarging them altogether. Whenever Dr. Bruemmer made these\\n\\nrecommendations, however, Mrs. Feldman made sure that Dr. Feldman saw these\\n\\npatients before they were discharged.\\n\\n In early 2011, law enforcement officers began investigating Dr. Feldman’s\\n\\nclinic. As part of the investigation, several undercover officers posed as patients at\\n\\nthe clinic. Dr. Kevin Chaitoff, a physician who specializes in pain management\\n\\nand anesthesia, reviewed the transcripts from the undercover officers’ visits to the\\n\\n\\n 4\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 5 of 67\\n\\n\\nclinic and determined that there were several problems with the way Dr. Feldman\\n\\nhandled each visit. Specifically, Dr. Chaitoff determined that Dr. Feldman\\n\\nobtained inadequate patient histories, conducted inadequate physical examinations,\\n\\nand prescribed controlled substances without documented justification.\\n\\n In August 2012, law enforcement officers executed a search warrant at the\\n\\nclinic and seized 3,200 patient files. At the request of Brian Zdrojewski, an agent\\n\\nwith the Drug Enforcement Administration, Dr. Chaitoff randomly selected 30\\n\\npatient files to review in order to determine whether Dr. Feldman was operating\\n\\nwithin the usual course of professional practice with respect to each patient. Dr.\\n\\nChaitoff concluded that none of the patients in the files he reviewed had been\\n\\nprescribed controlled substances for a legitimate medical purpose. Dr. Chaitoff\\n\\nwas later asked to review an additional 18 patient files and he reached the same\\n\\nconclusion as to those files. The investigation revealed at least three patients of\\n\\nDr. Feldman’s who had died while under his care: Joey Mayes, Shannon Wren,\\n\\nand Ricky Gonzalez. Dr. Chaitoff reviewed the files of these patients and\\n\\ndetermined that Dr. Feldman’s treatment of them was not within the usual course\\n\\nof professional practice.\\n\\n B. Money Laundering\\n\\n The investigation also revealed that between 2010 and 2014, the Feldmans\\n\\ndeposited approximately $6,787,103.99 into 37 bank accounts—the majority of\\n\\n\\n 5\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 6 of 67\\n\\n\\nwhich came from cash deposits that were less than $10,000. John Barna, a retired\\n\\ndetective with the Pinellas County Sheriff’s Office, determined that the Feldmans\\n\\npurchased their home on Trilby Avenue, in part, with $190,000 from a savings\\n\\naccount in Dr. Feldman’s name at JPMorgan Chase Bank. Records showed that of\\n\\nthat $190,000, $150,000 was transferred from the clinic’s bank account in $50,000\\n\\nincrements. That transfer is the subject of Count 6. At issue in Count 7, is a\\n\\n$150,000 wire transfer from First Bank Creve Coeur Mark Lewis P.A. Trust\\n\\nAccount that Dr. Feldman used to fund the clinic location on Park Boulevard.\\n\\nFinally, the subject of Count 8 is a wire transfer in the amount of $187,689.50 from\\n\\na Merrill Lynch account in Dr. Feldman’s name to one in Mrs. Feldman’s name.\\n\\n II. Procedural History\\n\\n A federal grand jury subsequently issued an indictment charging the\\n\\nFeldmans with: (1) conspiring to distribute and dispense controlled substances—\\n\\nprimarily oxycodone and methadone, which are Schedule II controlled substances,\\n\\nand alprazolam and diazepam, which are Schedule IV controlled substances—not\\n\\nfor a legitimate medical purpose, in violation of 21 U.S.C. §§ 841(b)(1)(C), (b)(2)\\n\\nand 846 (Count 1); (2) conspiring to commit money laundering, in violation of 18\\n\\nU.S.C. § 1956(h) (Count 5); and (3) engaging in monetary transactions with\\n\\ncriminal derived property exceeding the value of $10,000, in violation of 18 U.S.C.\\n\\n§ 1957 (Counts 6 through 8). Dr. Feldman was also charged with three substantive\\n\\n\\n 6\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 7 of 67\\n\\n\\ncounts of distribution and dispensation of Schedule II and Schedule IV controlled\\n\\nsubstances, not for a legitimate medical purpose, resulting in the deaths of Joey\\n\\nMayes (Count 2), Ricky Gonzalez (Count 3), and Shannon Wren (Count 4), in\\n\\nviolation of § 841(a)(1), (b)(1)(C) (Counts 2 through 4).\\n\\n The Feldmans pled not guilty and proceeded to trial. The first trial ended in\\n\\na mistrial. At the second trial, the Government presented testimony from the case\\n\\nagents, undercover officers, former patients and employees, the medical examiners,\\n\\nand Dr. Chaitoff. At the close of the Government’s case-in-chief, the Feldmans\\n\\nmoved for judgment of acquittal. Dr. Feldman argued that the Government failed\\n\\nto meet its burden of proof as to each count. The district court denied the\\n\\nFeldmans’ motions for judgment of acquittal.\\n\\n Dr. Feldman called several witnesses on his behalf. Dr. Simopoulos, a pain\\n\\nmanagement physician, testified that, based on his review of the patient files, he\\n\\nbelieved that Dr. Feldman’s prescribing practices were within the usual course of\\n\\nprofessional practice. Dr. Vernard Adams, a forensic pathologist and former\\n\\nmedical examiner, opined about the cause of death for each of the three victims.\\n\\nDr. Feldman also testified. He denied that he had an agreement with Mrs. Feldman\\n\\nor anyone else at the clinic to engage in illegal activity. After the defense rested,\\n\\nthe district court denied the parties’ renewed motions for judgment of acquittal.\\n\\n\\n\\n\\n 7\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 8 of 67\\n\\n\\n The jury found the Feldmans guilty as charged. The district court sentenced\\n\\nDr. Feldman and Mrs. Feldman to imprisonment terms of 300 months and\\n\\n48 months, respectively. Dr. Feldman’s 300-month sentence incorporated the 20-\\n\\nyear mandatory minimum sentence under 21 U.S.C. § 841(b)(1)(C), based on his\\n\\nconvictions for Counts 2 through 4.1 This appeal followed.\\n\\n DISCUSSION\\n\\n I. Motion to Sever\\n\\n Mrs. Feldman argues that the district court abused its discretion by denying\\n\\nher motion to sever her trial from Dr. Feldman’s.\\n\\n We review the district court’s denial of a motion for severance for an abuse\\n\\nof discretion. United States v. Cobb, 185 F.3d 1193, 1196 (11th Cir. 1999). If it\\n\\nappears that a joint trial will prejudice a defendant or the Government, the district\\n\\ncourt may sever the defendants’ trials. Fed. R. Crim. P. 14(a). “In assessing the\\n\\nmerits of a severance motion, the district court must balance the possibility of\\n\\nprejudice to the defendant against the public interest in judicial efficiency and\\n\\neconomy.” United States v. Eyster, 948 F.2d 1196, 1213 (11th Cir. 1991). To\\n\\nobtain reversal of a district court’s denial of a severance motion, a defendant must\\n\\nmake a showing of “compelling prejudice” against which the district court offered\\n\\n\\n\\n1\\n Dr. Feldman received ten-year concurrent sentences on each of his other five counts of\\nconviction.\\n 8\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 9 of 67\\n\\n\\nno protection. United States v. Walser, 3 F.3d 380, 385 (11th Cir. 1993). “A\\n\\ndefendant can show compelling prejudice by demonstrating that the jury was\\n\\nunable to sift through the evidence and make an individualized determination as to\\n\\neach defendant.” United States v. Ramirez, 426 F.3d 1344, 1352 (11th Cir. 2005)\\n\\n(quotation marks omitted).\\n\\n Prior to trial, Mrs. Feldman moved to sever Counts 2 through 4—which\\n\\ncharged Dr. Feldman with distributing and dispensing controlled substances,\\n\\nwithout a legitimate medical purpose, resulting in death. She argued that she\\n\\nwould be substantially prejudiced by a joint trial that included those counts\\n\\nbecause they applied to Dr. Feldman only. The district court denied the motion.\\n\\n Mrs. Feldman has not met her burden of showing that the district court\\n\\nabused its discretion by denying her motion for severance. The Feldmans were\\n\\nindicted together and charged with being part of two conspiracies: (1) the\\n\\ndistribution of controlled substances without a legitimate medical purpose; and\\n\\n(2) money laundering. “It is well settled that defendants who are indicted together\\n\\nare usually tried together.” United States v. Browne, 505 F.3d 1229, 1268 (11th\\n\\nCir. 2007); see also United States v. Knowles, 66 F.3d 1146, 1158 (11th Cir. 1995)\\n\\n(“This court is reluctant to reverse a district court’s denial of severance,\\n\\nparticularly in conspiracy cases, as generally ‘persons who are charged together\\n\\nshould also be tried together.’”).\\n\\n\\n 9\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 10 of 67\\n\\n\\n Mrs. Feldman asserts that she was prejudiced by the admission of evidence\\n\\nrelating to Counts 2 through 4 because those counts applied only to Dr. Feldman.\\n\\nWe have concluded, however, that “[a] defendant does not suffer compelling\\n\\nprejudice, sufficient to mandate a severance, simply because much of the evidence\\n\\nat trial is applicable only to co-defendants.” United States v. Schlei, 122 F.3d 944,\\n\\n984 (11th Cir. 1997) (quotation marks omitted). Moreover, Mrs. Feldman has not\\n\\nmet her burden of demonstrating that the jury was not able to make an\\n\\nindividualized determination of guilt as to the evidence pertaining to her. See\\n\\nUnited States v. Francis, 131 F.3d 1452, 1459 (11th Cir. 1997) (explaining that, in\\n\\norder to establish prejudice, a defendant must show “that the jury was unable to\\n\\nmake an individualized guilt determination for each defendant”).\\n\\n Nevertheless, even if there were any potential prejudice, we have determined\\n\\nthat prejudice can be avoided where the district court instructs the jury that it\\n\\nshould consider the evidence against each defendant separately. See Schlei, 122\\n\\nF.3d at 984. Here, the district court instructed the jury that:\\n\\n [A] separate crime or offense is charged against each defendant\\n in each count of the indictment. Each charge and the evidence\\n pertaining to it should be considered separately. Also, the case of each\\n defendant should be considered separately and individually. The fact\\n that you may find one or any . . . of the defendants guilty or not guilty\\n of any of the offenses charged should not affect your verdict as to any\\n other offense or any other defendant.\\n\\n I caution you, members of the jury, that you are here to determine\\n from the evidence in this case whether each defendant is guilty or not\\n 10\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 11 of 67\\n\\n\\n guilty. Each defendant is on trial only for the specific offense alleged\\n in each count of the indictment.\\n\\n We presume that the district court’s cautionary instruction prevented any\\n\\npossible prejudice. See Francis, 131 F.3d at 1459 (explaining that “cautionary\\n\\ninstructions to the jury to consider the evidence separately are presumed to guard\\n\\nadequately against prejudice.” (quotation marks omitted)). Because Mrs. Feldman\\n\\nhas not provided any evidence showing that the jury was not able to make an\\n\\nindividualized determination of guilt, she has failed to demonstrate that she\\n\\nsuffered compelling prejudice or received an unfair trial. See Browne, 505 F.3d at\\n\\n1268 (“[Defendant] must discharge the ‘heavy burden’ of demonstrating\\n\\n‘compelling prejudice’ from the joinder.” (quotation marks omitted)).\\n\\nAccordingly, the district court did not abuse its discretion by denying her motion\\n\\nfor severance.\\n\\n II. Expert Testimony\\n\\n Mrs. Feldman’s next claim of error is that the district court abused its\\n\\ndiscretion by denying her motion to exclude Dr. Chaitoff’s testimony and by\\n\\npermitting Dr. Chaitoff to extrapolate an opinion as to all 3,200 patient files, based\\n\\non his review of a small percentage of those files.\\n\\n We review for abuse of discretion the district court’s determinations\\n\\nregarding the admissibility and reliability of expert testimony. United States v.\\n\\n\\n\\n 11\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 12 of 67\\n\\n\\nFrazier, 387 F.3d 1244, 1258 (11th Cir. 2004). Federal Rule of Evidence 702\\n\\nprovides that a witness who is qualified as an expert may provide an opinion if:\\n\\n (a) the expert’s scientific, technical, or other specialized knowledge\\n will help the trier of fact to understand the evidence or to determine a\\n fact in issue; (b) the testimony is based on sufficient facts or data;\\n (c) the testimony is the product of reliable principles and methods; and\\n (d) the expert has reliably applied the principles and methods to the\\n facts of the case.\\n\\nFed. R. Evid. 702. The district court has wide discretion in deciding how to\\n\\ndetermine reliability. United States v. Abreu, 406 F.3d 1304, 1307 (11th Cir.\\n\\n2005).\\n\\n Prior to trial, Dr. Feldman filed a motion in limine to exclude, or limit, the\\n\\ntestimony of the Government’s proposed expert, Dr. Chaitoff. Following a\\n\\nDaubert 2 hearing, a magistrate judge determined that Dr. Feldman’s objection to\\n\\nDr. Chaitoff’s testimony under Rule 702 was without merit, “at least to the extent\\n\\nthat the opinion does not extrapolate from the reviewed files to a broader\\n\\nconclusion involving files not reviewed.” The magistrate judge noted that Dr.\\n\\nChaitoff’s testimony at the hearing resolved any potential issue pertaining to\\n\\nextrapolation, as he testified that the review of one patient’s file would be\\n\\ninsufficient to draw conclusions regarding files that were not reviewed.\\n\\n\\n\\n\\n2\\n Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993).\\n 12\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 13 of 67\\n\\n\\n Mrs. Feldman now challenges for the first time on appeal an isolated\\n\\nstatement made by Dr. Chaitoff at trial. Specifically, while discussing his selection\\n\\nand review of the first 30 patient files, Dr. Chaitoff stated that, “the methodology\\n\\nwas random and the conclusions applied to all the patients. So whether I reviewed\\n\\n30 or I reviewed 3,200, I suspect my conclusions would have been identical.” As\\n\\nnoted, Mrs. Feldman did not object to this statement.\\n\\n The Federal Rules of Evidence provide that a defendant need not object to\\n\\npreserve a claim of error on appeal if the district court makes a definitive ruling on\\n\\na pretrial motion, such as a motion in limine. See Fed. R. Evid. 103(b) (“Once the\\n\\ncourt rules definitively on the record—either before or at trial—a party need not\\n\\nrenew an objection or offer of proof to preserve a claim of error for appeal.”). That\\n\\nis not what happened here, however. The magistrate judge did not expressly rule\\n\\non the merits of the extrapolation issue, stating only that the objection to Dr.\\n\\nChaitoff’s testimony pursuant to Rule 702 was without merit “at least to the extent\\n\\nthat the opinion does not extrapolate from the reviewed files to a broader\\n\\nconclusion involving files not reviewed.” Moreover, it is clear from the order that\\n\\nthe magistrate judge believed that any potential extrapolation by Dr. Chaitoff was a\\n\\nnon-issue, as he noted that Dr. Chaitoff had specifically testified that his review of\\n\\none patient’s file would be insufficient to draw conclusions as to other patient files\\n\\nthat had not been reviewed.\\n\\n\\n 13\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 14 of 67\\n\\n\\n Because the magistrate judge did not make a definitive ruling on the\\n\\nextrapolation issue, Mrs. Feldman was required to object to Dr. Chaitoff’s\\n\\nstatement to avoid plain error review. See United States v. Wilson, 788 F.3d 1298,\\n\\n1313 (11th Cir. 2015) (reviewing defendant’s argument for plain error where the\\n\\ndistrict court’s pretrial ruling was not definitive); Cf. United States v. Harris, 886\\n\\nF.3d 1120, 1127 n.2 (11th Cir. 2018) (concluding that defendant’s argument on\\n\\nappeal was not subject to plain error review—despite the defendant’s failure to\\n\\nobject at trial—because the district court definitively ruled on the motion in\\n\\nlimine). She did not object, so we review her argument for plain error. 3\\n\\n Mrs. Feldman cannot prevail under plain error review. Even if she could\\n\\nshow an error that was plain, she cannot demonstrate that her substantial rights\\n\\nwere violated. Indeed, she cannot show that she was prejudiced—much less\\n\\nsubstantially prejudiced—by one stray comment made by Dr. Chaitoff concerning\\n\\nhis suspicion as to the remaining 3,000+ patient files not reviewed. The comment\\n\\nwas made during approximately three days’ worth of testimony from Dr. Chaitoff.\\n\\nDuring that testimony, Dr. Chaitoff focused entirely on the 48 patient files that he\\n\\nreviewed and his opinion as to whether Dr. Feldman’s treatment of the specific\\n\\n\\n\\n3\\n To show plain error, the defendant must establish that there was “(1) an error (2) that is plain\\nand (3) that has affected the defendant’s substantial rights; and if the first three prongs are met,\\nthen a court may exercise its discretion to correct the error if (4) the error ‘seriously affect[s] the\\nfairness, integrity or public reputation of judicial proceedings.’” United States v. Madden, 733\\nF.3d 1314, 1320 (11th Cir. 2013) (quotation marks omitted).\\n 14\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 15 of 67\\n\\n\\npatients in those files was within the ordinary course of professional practice.\\n\\nAccordingly, Mrs. Feldman is not entitled to relief on this claim.\\n\\n III. Double Jeopardy\\n\\n Mrs. Feldman asserts that the district court’s order declaring a mistrial\\n\\nduring the first trial was not supported by manifest necessity, and thus the court\\n\\nshould have granted her motion to dismiss the indictment on double jeopardy\\n\\ngrounds. She also argues that the district court failed to comply with Federal Rule\\n\\nof Criminal Procedure 26.3, which requires the court to give the parties an\\n\\nopportunity to object or consent to an order declaring mistrial.\\n\\n We review the district court’s declaration of a mistrial on grounds of\\n\\nmanifest necessity to determine whether the court exercised sound discretion in\\n\\ndeciding that such a necessity existed. United States v. Davis, 708 F.3d 1216, 1221\\n\\n(11th Cir. 2013). And “if the court’s decision to declare a mistrial was not based\\n\\non manifest necessity, it was an abuse of discretion not to dismiss the indictment\\n\\non double jeopardy grounds.” See id.\\n\\n The Double Jeopardy Clause of the Fifth Amendment safeguards a criminal\\n\\ndefendant from being subjected to multiple prosecutions for the same offense.\\n\\nU.S. Const. amend. V. “Ordinarily, when ‘a defendant successfully seeks to avoid\\n\\nhis trial prior to its conclusion by a motion for mistrial, the Double Jeopardy\\n\\n\\n\\n\\n 15\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 16 of 67\\n\\n\\nClause is not offended by a second prosecution.’” United States v. Shelley, 405\\n\\nF.3d 1195, 1200 (11th Cir. 2005) (quotation marks omitted).\\n\\n “[D]istrict courts are permitted to declare a mistrial and discharge a jury\\n\\nonly where, ‘taking all the circumstances into consideration, there is a manifest\\n\\nnecessity for the act, or the ends of public justice would otherwise be defeated.’”\\n\\nUnited States v. Therve, 764 F.3d 1293, 1298 (11th Cir. 2014) (quotation marks\\n\\nomitted). “Our task is to decide whether the district court exercised ‘sound\\n\\ndiscretion’ in declaring a mistrial, and in order to determine that ‘we review the\\n\\nentire record in the case without limiting ourselves to the actual findings of the trial\\n\\ncourt.’” Davis, 708 F.3d at 1221 (quotation marks omitted and citation omitted).\\n\\nBut “the requirement that manifest necessity be demonstrated operates[] only when\\n\\nthe trial court has declared a mistrial without the consent of the defendant.” United\\n\\nStates v. Puleo, 817 F.2d 702, 705 (11th Cir. 1987). If the defendant consents to a\\n\\nmistrial, the Double Jeopardy Clause will not bar her retrial. Id.\\n\\n The facts surrounding the declaration of a mistrial during the first trial are as\\n\\nfollows. During the first trial, Mrs. Feldman’s attorney asked Dr. Feldman on\\n\\ncross-examination if he had ever been convicted of a felony. Dr. Feldman stated\\n\\nthat he had been convicted of one felony. Thereafter, the prosecutor asked Dr.\\n\\nFeldman whether the felony he was convicted of related to his practice as a doctor.\\n\\nDr. Feldman objected and moved for a mistrial, arguing that a curative instruction\\n\\n\\n 16\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 17 of 67\\n\\n\\nwas not going to cure the prejudice. Indeed, the defendants were united in their\\n\\nposition that a mistrial was necessary. Mrs. Feldman indicated that she had “the\\n\\nsame position as this is a conspiracy case with regard to that I agree that, I agree\\n\\nthat I don’t believe a curative instruction is going to—.” Following subsequent\\n\\ndiscussion, the district court recessed for the day. The next morning, the district\\n\\ncourt heard argument from the Government and Dr. Feldman as to whether the\\n\\nprejudicial effect of the question could be remedied by a curative instruction.\\n\\nUltimately, the district court decided that the defendants were correct in asserting\\n\\nthat a mistrial was necessary because the court concluded that it was not possible\\n\\nto cure this issue. The court thus declared a mistrial. During this lengthy\\n\\ndeliberative process by the court, Mrs. Feldman never indicated any change in her\\n\\noriginal position that a mistrial was necessary nor did she object when the district\\n\\ncourt ultimately acquiesced to the defense position that a mistrial was required.\\n\\n Instead, almost a month later—when the jury was long gone—she filed a\\n\\nmotion arguing that the court was wrong to conclude that manifest necessity\\n\\nrequired a mistrial. She further argued that she was not provided the opportunity to\\n\\nconsent or object to the order declaring a mistrial. The district court denied the\\n\\nmotion.\\n\\n Mrs. Feldman argues that the Double Jeopardy Clause barred her retrial\\n\\nbecause a mistrial was not manifestly necessary. But we need not address manifest\\n\\n\\n 17\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 18 of 67\\n\\n\\nnecessity. Because Mrs. Feldman acquiesced in the court’s decision to declare a\\n\\nmistrial, she cannot now be credibly heard to argue that the court’s decision was a\\n\\nmistake.\\n\\n Mrs. Feldman argues that she could not have agreed with the district court’s\\n\\ndecision to declare a mistrial because the district court failed to provide proper\\n\\nnotice before declaring a mistrial. Federal Rule of Criminal Procedure 26.3\\n\\nprovides that prior to ordering a mistrial, “the court must give each defendant and\\n\\nthe government an opportunity to comment on the propriety of the order, to state\\n\\nwhether that party consents or objects, and to suggest alternatives.” Fed. R. Crim.\\n\\nP. 26.3. “[C]ompliance with Rule 26.3 is only one factor to be considered in\\n\\ndetermining whether a trial judge exercised sound discretion in declaring a\\n\\nmistrial.” Davis, 708 F.3d at 1224 (citing United States v. Berroa, 374 F.3d 1053,\\n\\n1058 (11th Cir. 2004)).\\n\\n We conclude that Mrs. Feldman gave implied consent to the mistrial. The\\n\\nrecord in the present case clearly shows that Mrs. Feldman had numerous\\n\\nopportunities to comment or object to the potential declaration of a mistrial, and\\n\\nshe never gave even a hint that she objected during the lengthy proceedings\\n\\nconcerning this matter. Indeed, after Dr. Feldman moved for a mistrial, Mrs.\\n\\nFeldman agreed that the prejudice could not be cured by a curative instruction, and\\n\\nthen noted that this was a conspiracy case. When the court reconvened after the\\n\\n\\n 18\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 19 of 67\\n\\n\\novernight recess, Mrs. Feldman raised no objection to the court declaring a\\n\\nmistrial, nor did she object or argue that the court should allow the ongoing trial to\\n\\nproceed against her alone. Even as the jury was departing, she said nothing to the\\n\\ncourt indicating that a mistrial should not have been declared as to her. See Puleo,\\n\\n817 F.2d at 705 (concluding that defendant consented to mistrial when “even as the\\n\\njury was departing . . . [counsel] said nothing regarding the mistrial”).\\n\\n In deciding a Double Jeopardy challenge to a second trial, we have recently\\n\\nreconfirmed Puleo’s teaching that “a defendant’s consent to a mistrial need not be\\n\\nexpress but may always be implied from the totality of circumstances.” United\\n\\nStates v. Isaac Feldman, 931 F.3d 1245, 1256–57 (11th Cir. 2019) (quotation\\n\\nmarks omitted and alterations accepted). As with the present case, we noted in\\n\\nIsaac Feldman that, albeit aware that the district court was in the process of\\n\\ndeclaring a mistrial, the defendant “never voiced any objection to the jury’s\\n\\ndismissal . . . . The totality of these circumstances compels the conclusion that [the\\n\\ndefendant] impliedly consented to the dismissal of the original jury . . . .” Id. at\\n\\n1257.\\n\\n So it is here. Mrs. Feldman implicitly consented to the order declaring a\\n\\nmistrial. We therefore conclude that she is entitled to no relief on her claim\\n\\nchallenging the district court’s denial of her motion to dismiss the indictment on\\n\\ndouble jeopardy grounds.\\n\\n\\n 19\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 20 of 67\\n\\n\\n IV. Prosecutorial Misconduct\\n\\n Mrs. Feldman argues that the prosecutor’s closing argument was improper\\n\\nand prejudicial because the prosecutor argued facts not in evidence, put forth a new\\n\\nlegal theory of conviction not alleged in the indictment, and interjected her own\\n\\npersonal opinion to inflame the jury.\\n\\n We generally review allegations of prosecutorial misconduct de novo.\\n\\nUnited States v. Eckhardt, 466 F.3d 938, 947 (11th Cir. 2006). Where a defendant\\n\\nfails to object during a prosecutor’s closing argument, however, “relief is available\\n\\nto rectify only plain error that is so obvious that failure to correct it would\\n\\njeopardize the fairness and integrity of the trial.” United States v. Bailey, 123 F.3d\\n\\n1381, 1400 (11th Cir. 1997).\\n\\n To prevail on a prosecutorial misconduct claim, the defendant must show\\n\\nthat the conduct was improper and that it was prejudicial to his substantial rights.\\n\\nId. A defendant’s substantial rights are prejudiced if there is a reasonable\\n\\nprobability that, but for the remarks, the outcome of the trial would have been\\n\\ndifferent. United States v. Lopez, 590 F.3d 1238, 1256 (11th Cir. 2009).\\n\\n“Prosecutorial misconduct must be considered in the context of the entire trial,\\n\\nalong with any curative instruction.” Id. If the district court provided a curative\\n\\ninstruction, “we will reverse only if the evidence is so prejudicial as to be incurable\\n\\nby that measure.” Id.\\n\\n\\n 20\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 21 of 67\\n\\n\\n Mrs. Feldman raises several challenges to the prosecutor’s remarks during\\n\\nclosing argument. We conclude that only a few of those challenges warrant any\\n\\ndiscussion here. 4 First, Mrs. Feldman asserts that the prosecutor improperly\\n\\ncommented on facts that were not in evidence when discussing Dr. Feldman’s\\n\\nexpert, Dr. Adams. Specifically, she takes issue with the prosecutor’s statement\\n\\nthat when Dr. Adams was employed by the county medical examiner’s office, he\\n\\nhad found that people who died at home had died of drug overdoses. However,\\n\\nnow that he was being paid as a defense expert, he had concluded that one of the\\n\\nvictims who died at home—Shannon Wren—died from a heart attack, not a drug\\n\\noverdose. We find nothing improper about the prosecutor’s statement, as it was\\n\\nsupported by Dr. Adams’ testimony that he had previously made findings of drug\\n\\noverdose deaths where people were found dead in their homes.5\\n\\n We next address Mrs. Feldman’s assertion that the prosecutor inserted a new\\n\\ntheory of conviction not alleged in the indictment by arguing that Mrs. Feldman\\n\\n\\n\\n\\n4\\n Mrs. Feldman argues that the prosecutor’s following remarks were also improper: (1) a\\nstatement that Mayes’ sister testified that Dr. Feldman was surprised when she accused Dr.\\nFeldman of prescribing Xanax and methadone to her brother; (2) a statement implying that there\\nwere several actions the Department of Health could have taken against Dr. Feldman; and (3) a\\nstatement that Mrs. Feldman had spoken with Aimee Martin about ways to stay off the Drug\\nEnforcement Administration’s radar. We find no merit to these challenges.\\n5\\n Notably, the district court also provided the following curative instruction when Mrs. Feldman\\nobjected to the prosecutor’s statement regarding Dr. Adams’ testimony: “[t]o the extent the\\nlawyers disagree on what the facts are, members of the jury, you will resolve those\\ndisagreements. These are arguments of counsel, they are not evidence.”\\n 21\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 22 of 67\\n\\n\\nviolated an Internal Revenue Service reporting requirement by failing to disclose\\n\\nforeign bank accounts. When addressing the money laundering convictions in her\\n\\nclosing argument, the prosecutor stated that she asked defense expert, Stephen\\n\\nOscher, if he noticed whether Mrs. Feldman had disclosed to the federal\\n\\ngovernment that she had bank accounts outside of the country. After the district\\n\\ncourt overruled Mrs. Feldman’s objection to this statement, the prosecutor\\n\\nexplained that Oscher had said he did not know, he had not seen anything. The\\n\\nprosecutor then stated, “[t]hat’s also a requirement, a reporting requirement.” In\\n\\nother words, the prosecutor appeared to imply that Mrs. Feldman had violated a\\n\\nreporting requirement by failing to disclose foreign bank accounts. The\\n\\nindictment, however, charged Mrs. Feldman with violating a transaction reporting\\n\\nrequirement, not failing to report the existence of a foreign bank account.\\n\\n Nevertheless, even if we assume that this remark was improper, it does not\\n\\namount to prosecutorial misconduct because Mrs. Feldman has not met her burden\\n\\nof showing that her substantial rights were violated. In the first place, the district\\n\\ncourt provided curative instructions. See Lopez, 590 F.3d at 1256 (explaining that\\n\\n“we will reverse only if the evidence is so prejudicial as to be incurable by that\\n\\nmeasure”). The court instructed the jury that their decision must be based on\\n\\nevidence, which “includes the testimony of the witnesses and the exhibits which\\n\\nhave been admitted in the record . . . anything the lawyers say is not evidence in\\n\\n\\n 22\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 23 of 67\\n\\n\\nthe case.” The record likewise contained sufficient evidence of guilt independent\\n\\nof the challenged remarks. Id. (“When the record contains sufficient independent\\n\\nevidence of guilt, any error is harmless.” (citing Eckhardt, 466 F.3d at 947)). As\\n\\ndiscussed infra, the Government presented sufficient evidence from which a\\n\\nreasonable jury could conclude that Mrs. Feldman was guilty beyond a reasonable\\n\\ndoubt. In short, the challenged comment was a small portion of the 17-day trial in\\n\\nwhich the Government presented substantial evidence of guilt against Mrs.\\n\\nFeldman.\\n\\n Finally, we reject Mrs. Feldman’s argument that the prosecutor’s statement\\n\\nthat Wren had a “butt-load” of drugs in his body was unfairly prejudicial. Because\\n\\nMrs. Feldman failed to object to this statement, our review is limited to plain error.\\n\\nMrs. Feldman has failed to show error, let alone plain error. See United States v.\\n\\nBailey, 123 F.3d 1381, 1400 (11th Cir. 1997) (“When a defendant fails to object to\\n\\nthe prosecutor’s closing argument, relief is available to rectify only plain error that\\n\\nis so obvious that failure to correct it would jeopardize the fairness and integrity of\\n\\nthe trial.”). We have stated that “there is no prohibition on colorful and perhaps\\n\\nflamboyant remarks if they relate to the evidence adduced at trial.” Id. (quotation\\n\\nmarks omitted). We acknowledge that the prosecutor could have described the\\n\\nquantity of drugs in Wren’s body in a less graphic manner. However, the truth is\\n\\nthat Wren did have a great deal of drugs in his system, and ultimately he died.\\n\\n\\n 23\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 24 of 67\\n\\n\\nThus, the prosecutor’s description was consistent with the evidence presented at\\n\\ntrial. In short, the remark challenged by Mrs. Feldman does not amount to error,\\n\\nlet alone plain error.\\n\\n As a final matter, Mrs. Feldman asserts that the alleged errors, when viewed\\n\\nindividually or cumulatively, prejudiced her substantial rights and deprived her of\\n\\na fair trial. “The cumulative error doctrine provides that an aggregation of non-\\n\\nreversible errors . . . can yield a denial of the constitutional right to a fair trial,\\n\\nwhich calls for reversal.” United States v. Capers, 708 F.3d 1286, 1299 (11th Cir.\\n\\n2013) (quotation marks omitted). Because Mrs. Feldman has established no error,\\n\\nshe cannot prevail under the cumulative error doctrine. See United States v.\\n\\nGamory, 635 F.3d 480, 497 (11th Cir. 2011) (“Where there is no error or only a\\n\\nsingle error, there can be no cumulative error.”).\\n\\n V. Jury Instruction\\n\\n Dr. Feldman argues that the district court abused its discretion by denying\\n\\nhis proposed jury instruction on his theory of defense.6 The specific instruction\\n\\nDr. Feldman requested was the following:\\n\\n It is a physician’s duty and obligation to try to relieve a patient’s pain.\\n Therefore, it is ethical and medically justifiable for a physician to\\n prescribe controlled substances for a patient suffering from chronic\\n moderate to severe pain, even if the patient has developed a tolerance\\n or addiction to those substances. It is also ethical and medically\\n justifiable for such treatment to be performed for the purpose of\\n\\n6\\n Mrs. Feldman adopts this argument.\\n 24\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 25 of 67\\n\\n\\n relieving chronic moderate to severe pain in a patient, regardless of the\\n patient’s [addiction] history.\\n\\nThe district court denied the instruction as argumentative, citing this Court’s\\n\\ndecision in United States v. Paradies, 98 F.3d 1266, 1287 (11th Cir. 1996), where\\n\\nwe rejected a defendant’s proposed jury instruction because it was partisan and\\n\\nsought to have the court parrot the factual findings that the defendant was\\n\\nadvocating to the jury.\\n\\n We review the district court’s refusal to give a requested jury instruction for\\n\\nan abuse of discretion. United States v. Dohan, 508 F.3d 989, 993 (11th Cir.\\n\\n2007). The district court’s refusal to give a requested instruction is reversible error\\n\\nwhere the instruction “(1) was correct, (2) was not substantially covered by the\\n\\ncharge actually given, and (3) dealt with some point in the trial so important that\\n\\nfailure to give the requested instruction seriously impaired the defendant’s ability\\n\\nto conduct his defense.” Eckhardt, 466 F.3d at 947–48.\\n\\n Here, the district court did not abuse its discretion by refusing to give Dr.\\n\\nFeldman’s proposed jury instruction. In the first place, Dr. Feldman has not cited\\n\\nto any legal authority demonstrating that his requested instruction was a correct\\n\\nstatement of the law. Although he cites to Florida Statute § 765.1103—a statutory\\n\\nprovision governing pain management with respect to palliative care—this statute\\n\\ndoes not state that it is ethical for a doctor to prescribe controlled substances for\\n\\nchronic pain, regardless of a patient’s tolerance or addiction issues. See Fla. Stat.\\n 25\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 26 of 67\\n\\n\\n§ 765.1103 (providing that “practitioners . . . must, as appropriate, comply with a\\n\\nrequest for pain management or palliative care from a patient under their care”).\\n\\n Further, Dr. Feldman’s proposed jury instruction essentially asked the\\n\\ndistrict court to instruct the jury about the facts Dr. Feldman hoped the jury would\\n\\nfind: that it was ethically and medically justifiable for Dr. Feldman to prescribe\\n\\ncontrolled substances without respect to the patient’s medical history or addiction\\n\\nissues. See United States v. Maxwell, 579 F.3d 1282, 1304 (11th Cir. 2009)\\n\\n(explaining that the district court is not obliged to present jury instructions that\\n\\n“contain partisan and argumentative statements of law and fact”); Paradies, 98\\n\\nF.3d at 1287 (same). The proposed instruction was plainly argumentative and the\\n\\ncourt was not required to give a theory-of-defense instruction that “was more in the\\n\\nnature of a jury argument than a charge.” United States v. Barham, 595 F.2d 231,\\n\\n245 (5th Cir. 1979). In short, the district court’s refusal to give Dr. Feldman’s\\n\\nproposed instruction was not an abuse of discretion.\\n\\n VI. Sufficiency of the Evidence as to the Conspiracy Counts\\n\\n Both Feldmans challenge the sufficiency of the evidence presented to\\n\\nsupport each of their conspiracy convictions. We review challenges to the\\n\\nsufficiency of the evidence de novo, “viewing the evidence in the light most\\n\\nfavorable to the government and drawing all reasonable inferences and credibility\\n\\nchoices in favor of the jury’s verdict.” United States v. Taylor, 480 F.3d 1025,\\n\\n\\n 26\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 27 of 67\\n\\n\\n1026 (11th Cir. 2007). “If a reasonable jury could conclude that the evidence\\n\\nestablishes guilt beyond a reasonable doubt, we will affirm the verdict.” Browne,\\n\\n505 F.3d at 1253. “Because we recognize that ‘the jury is free to choose between\\n\\nor among the reasonable conclusions to be drawn from the evidence presented at\\n\\ntrial,’ our sufficiency review requires only that a guilty verdict be reasonable, not\\n\\ninevitable, based on the evidence presented at trial.” Id. (quotation marks omitted).\\n\\n A. Conspiracy to distribute controlled substances without a\\n legitimate medical purpose (Count 1)\\n\\n Both defendants challenge their conviction for conspiring to distribute or\\n\\ndispense controlled substances without a legitimate medical purpose. To establish\\n\\na conviction for illegal distribution or dispensation of controlled substances under\\n\\n21 U.S.C. § 841(a)(1), the Government must prove that the defendant distributed\\n\\nor “dispensed controlled substances for other than legitimate medical purposes in\\n\\nthe usual course of professional [medical] practice, and that he did so knowingly\\n\\nand intentionally.” United States v. Azmat, 805 F.3d 1018, 1035 (11th Cir. 2015)\\n\\n(quotation marks omitted). In order to establish a conspiracy conviction, the\\n\\nGovernment must show that: (1) “an agreement existed between two or more\\n\\npersons to commit a crime” and (2) “the defendants knowingly and voluntarily\\n\\njoined or participated in the conspiracy.” United States v. Seher, 562 F.3d 1344,\\n\\n1364 (11th Cir. 2009) (quotation marks omitted). The existence of an agreement\\n\\nmay “be proved by inferences from the conduct of the alleged participants or from\\n 27\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 28 of 67\\n\\n\\ncircumstantial evidence of a scheme.” Azmat, 805 F.3d at 1035 (quotation marks\\n\\nomitted). “A conspiracy conviction will be upheld if ‘the circumstances\\n\\nsurrounding a person’s presence at the scene of conspiratorial activity are so\\n\\nobvious that knowledge of its character can fairly be attributed to him.’” Id.\\n\\n Here, the Feldmans do not challenge the jury’s finding that Dr. Feldman was\\n\\ndispensing controlled substances outside the usual course of professional practice.\\n\\nInstead, their arguments focus on whether Mrs. Feldman conspired with Dr.\\n\\nFeldman to do so. We conclude that the Government presented ample evidence\\n\\nfrom which a reasonable jury could conclude that the Feldmans had an agreement\\n\\nto illegally distribute and dispense prescription medications.\\n\\n The evidence presented at trial showed that Mrs. Feldman was extensively\\n\\ninvolved in the operation of the clinic. She wore a white doctor’s coat, she sat in\\n\\non office visits, she encouraged new patient referrals, and she managed all of the\\n\\nclinic’s money. Because staff members were instructed to copy the sign-in sheets\\n\\nwhen providing Mrs. Feldman with any cash payments, the jury could infer that\\n\\nMrs. Feldman was aware of the large number of patients seen at the clinic each\\n\\nday. According to former patient Charlene Riffle, Mrs. Feldman also wrote out\\n\\nprescriptions ordered by Dr. Feldman while he sat at his desk eating fruit.\\n\\n The Government presented sufficient evidence from which a reasonable jury\\n\\ncould infer not only that Mrs. Feldman was aware that Dr. Feldman was\\n\\n\\n 28\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 29 of 67\\n\\n\\nprescribing controlled substances outside the usual course of medicine but also that\\n\\nshe knowingly participated in his efforts to do so. Mrs. Feldman asked staff\\n\\nmembers to alert her when law enforcement officers were present in the parking lot\\n\\nand, on those days, she parked in the back of the office. Multiple staff members\\n\\nalso testified that Mrs. Feldman directed them to put post-it notes in patient files\\n\\nfor anyone whom they believed was an undercover law enforcement officer. She\\n\\ntold them that post-it notes were important, so they could be removed easily and\\n\\nnot become a permanent item in the file.\\n\\n When the clinic began discharging too many patients, Mrs. Feldman\\n\\ninstructed staff to stop conducting urinalysis screenings. As two experts testified,\\n\\nthe urinalysis screenings are instrumental to determining whether a patient is\\n\\ntaking the medication as prescribed or diverting it to others. And, on at least one\\n\\noccasion, Aimee Martin, a former patient and friend of the Feldmans, asked Mrs.\\n\\nFeldman if she could fill her prescription early because it had run out. Mrs.\\n\\nFeldman suggested that Martin instead take the medication of her husband—who\\n\\nwas also one of Dr. Feldman’s patients—thereby encouraging the use of controlled\\n\\nsubstances outside the usual course of professional practice.\\n\\n Additionally, Mrs. Feldman told former employee Melinda Detwiler that she\\n\\nand Dr. Feldman had moved their assets into her name and the names of her\\n\\nchildren, in case anything happened. Finally, as to Dr. Feldman, he testified that\\n\\n\\n 29\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 30 of 67\\n\\n\\nhe did not have any agreement with Mrs. Feldman to illegally distribute controlled\\n\\nsubstances, but the jury obviously rejected this testimony. We may therefore\\n\\nconsider his disbelieved testimony as substantive evidence of his guilt. See United\\n\\nStates v. Woodard, 459 F.3d 1078, 1087 (11th Cir. 2006) (“[A] defendant’s\\n\\ntestimony—if disbelieved by the jury—may be considered substantive evidence of\\n\\nguilt.”).\\n\\n Although Mrs. Feldman asserts that there was an innocent explanation for\\n\\nher instruction to staff to stop conducing urinalysis screenings—the test was\\n\\npurportedly too costly—Detwiler testified that the reason Mrs. Feldman instructed\\n\\nstaff to stop doing the test was because it was resulting in too many patients being\\n\\ndischarged. Further, Dr. Feldman admitted on cross-examination that the\\n\\nurinalysis screening did not cost the clinic a lot of money, as the screening done in\\n\\nthe clinic was included in the cost of the office visit and the patient bore the cost of\\n\\nany urinalysis screening sent to the laboratory. We are likewise unpersuaded by\\n\\nMrs. Feldman’s argument that she made the comment to Detwiler about shifting\\n\\nher assets because she wanted to have everything in place if Dr. Feldman passed\\n\\naway. Indeed, Detwiler testified that Mrs. Feldman made the comment during a\\n\\nconversation about clinic patients who were being investigated by law\\n\\nenforcement. Viewing all of the evidence in the light most favorable to the\\n\\nGovernment, a reasonable jury could conclude that the Feldmans conspired and\\n\\n\\n 30\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 31 of 67\\n\\n\\nworked together to illegally dispense controlled substances without a legitimate\\n\\nmedical purpose.\\n\\n B. Money Laundering (Counts 5 through 8)\\n\\n The Feldmans also argue that the Government presented insufficient\\n\\nevidence to establish that they conspired to commit money laundering or that they\\n\\ncommitted any substantive money laundering offenses.\\n\\n As to the conspiracy charge alleged in Count 5, the Government needed to\\n\\nprove two elements to establish a conspiracy under § 1956(h): “(1) an agreement\\n\\nbetween two or more persons to commit a money-laundering offense; and\\n\\n(2) knowing and voluntary participation in that agreement by the defendant.”\\n\\nUnited States v. Moran, 778 F.3d 942, 962 (11th Cir. 2015) (quotation marks\\n\\nomitted). An essential element of money laundering conspiracy is that the\\n\\ndefendant “knew that the funds involved in the transactions represented the\\n\\nproceeds of unlawful activity.” See United States v. Awan, 966 F.2d 1415, 1434\\n\\n(11th Cir. 1992).\\n\\n Relying on their argument that the Government presented insufficient\\n\\nevidence as to the drug conspiracy alleged in Count 1, the Feldmans argue that the\\n\\nevidence presented at trial was insufficient to prove beyond a reasonable doubt that\\n\\nthey knew the funds involved in the financial transactions at issue involved\\n\\ncriminal proceeds. However, based on our conclusion that sufficient evidence\\n\\n\\n 31\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 32 of 67\\n\\n\\nsupported the drug conspiracy conviction, it necessarily follows that the Feldmans\\n\\nknew that the transactions involved the proceeds of unlawful activity.\\n\\n Dr. Feldman’s argument that the Government failed to present sufficient\\n\\nevidence showing that the Feldmans conducted these financial transactions with\\n\\nthe intent to conceal the criminal proceeds is also without merit. Concealment was\\n\\nonly one of the objects of the conspiracy. Indeed, Count 5 alleged that the\\n\\nFeldmans conspired to conduct a financial transaction with criminal proceeds from\\n\\nthe drug conspiracy (1) to promote the carrying on of unlawful activity, (2) to\\n\\nconceal the proceeds, (3) to avoid transaction reporting requirements, and (4) to\\n\\nengage in monetary transactions in property derived from a specified unlawful\\n\\nactivity. The jury determined that the Feldmans were guilty of all four objects of\\n\\nthe conspiracy. Further, Dr. Feldman does not challenge the jury’s finding as to\\n\\nthose other three objects, and he has therefore abandoned any challenge he may\\n\\nhave had. See United States v. Jernigan, 341 F.3d 1273, 1283 n.8 (11th Cir. 2003)\\n\\n(stating that an issue not briefed on appeal is considered abandoned). Because the\\n\\nFeldmans argue only that the Government failed to establish their knowledge of\\n\\nthe specified unlawful activity—the drug conspiracy—they are not entitled to relief\\n\\nfrom their money laundering convictions. United States v. Ross, 131 F.3d 970, 984\\n\\n(11th Cir. 1997) (“A guilty verdict in a multi-object conspiracy will be upheld if\\n\\nthe evidence is sufficient to support a conviction of any of the alleged objects.”).\\n\\n\\n 32\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 33 of 67\\n\\n\\n VII. Sufficiency of Evidence as to the Convictions Based on the\\n Dispensation of Controlled Substances Without a Legitimate\\n Medical Purpose that Resulted in Death (Counts 2 through 4)\\n\\n In Counts 2 through 4, Dr. Feldman was charged with distribution of\\n\\ncontrolled substances without a legitimate medical purpose resulting in the deaths\\n\\nof three people: Joey Mayes (Count 2), Ricky Gonzalez (Count 3), and Shannon\\n\\nWren (Count 4). The jury convicted Dr. Feldman on all counts, finding that each\\n\\nvictim died as a result of ingesting drugs that Dr. Feldman had prescribed. On\\n\\nappeal, Dr. Feldman challenges the sufficiency of the evidence showing that\\n\\nSchedule II drugs he prescribed caused each victim’s death: a fact that triggers a\\n\\n20-year mandatory minimum sentence under the Controlled Substances Act.\\n\\n To establish a conviction for unlawful dispensation by a physician under\\n\\n21 U.S.C. § 841(a)(1), the Government must establish that the defendant\\n\\nknowingly and intentionally distributed or dispensed controlled substances other\\n\\nthan for a legitimate medical purpose and outside the ordinary course of\\n\\nprofessional practice. 21 U.S.C. § 841(a)(1); Azmat, 805 F.3d at 1035. There is a\\n\\npenalty-enhancement provision for § 841(a), which provides that a defendant shall\\n\\nbe sentenced to a term of not less than 20 years’ imprisonment, or more than life, if\\n\\nhe dispensed a Schedule I or II drug and death or serious bodily injury “results\\n\\nfrom the use of such substance.” 21 U.S.C. § 841(b)(1)(C). “Because the ‘death\\n\\nresults’ enhancement increase[s] the minimum and maximum sentences to which\\n\\n\\n 33\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 34 of 67\\n\\n\\n[a defendant is] exposed, it is an element that must be submitted to the jury and\\n\\nfound beyond a reasonable doubt.” Burrage v. United States, 571 U.S. 204, 210\\n\\n(2014) (citing Alleyne v. United States, 570 U.S. 99, 115–16 (2013), and Apprendi\\n\\nv. New Jersey, 530 U.S. 466, 490 (2000)).\\n\\n Here, Dr. Feldman’s sufficiency-of-the-evidence argument proceeds in two\\n\\nparts. First, he argues that the evidence was insufficient to prove that it was he\\n\\nwho prescribed the specific drugs that were ingested by each victim. Second, he\\n\\nargues that the evidence was insufficient to prove that the Schedule II drugs\\n\\n(oxycodone and methadone)—prescribed by him and ingested by the victims—\\n\\nwere a but-for cause of each victim’s death. We disagree on both fronts and\\n\\naddress each issue in turn.\\n\\n A. Drugs Ingested by the Deceased Were Prescribed by Dr.\\n Feldman\\n\\n Evidence presented at trial fully supported the jury’s finding that, as charged\\n\\nin the indictment, the specific drugs ingested by each victim (Mayes, Gonzalez,\\n\\nand Wren) were prescribed by Dr. Feldman.\\n\\n 1. Ricky Gonzalez\\n\\n The trial evidence was sufficient to show that Dr. Feldman prescribed the\\n\\nspecific drugs ingested by Ricky Gonzalez. On October 5, 2010, only three days\\n\\nbefore his death, Gonzalez received a prescription from Dr. Feldman for 240 30-\\n\\nmilligram pills of oxycodone and 90 10-milligram pills of diazepam (Valium).\\n 34\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 35 of 67\\n\\n\\nGonzalez’s girlfriend, Casey Osteen, testified that she saw Gonzalez take\\n\\noxycodone and diazepam from prescriptions dispensed by Dr. Feldman the night\\n\\nbefore he died. When she last saw Gonzalez at approximately 11:00 p.m. that\\n\\nnight, he appeared “lethargic” and “groggy.”\\n\\n Gonzalez was found dead in his front yard the next day, October 8, 2010.\\n\\nOfficers found the bottle of diazepam prescribed by Dr. Feldman in Gonzalez’s\\n\\npants pocket. The bottle contained only 8 of the 90 10-milligram pills of diazepam\\n\\nprescribed by Dr. Feldman, as well as 35 other diazepam pills of varying dosages.\\n\\nThe bottle of oxycodone prescribed by Dr. Feldman three days earlier was found\\n\\nempty and smashed near a neighbor’s truck. Finally, the toxicology reports\\n\\nconfirmed that Gonzalez had oxycodone and diazepam—the same two drugs\\n\\nprescribed by Dr. Feldman—in his system. Based on this evidence, a reasonable\\n\\njury could infer that the drugs ingested by Gonzalez were prescribed by Dr.\\n\\nFeldman.\\n\\n 2. Joey Mayes\\n\\n The evidence was also sufficient to demonstrate that Dr. Feldman prescribed\\n\\nthe specific drugs Mayes ingested. Mayes first visited Dr. Feldman on March 12,\\n\\n2010 complaining of back pain. On this visit, Dr. Feldman prescribed Mayes 240\\n\\n30-milligram tablets of oxycodone, 60 2-milligram tablets of alprazolam (Xanax),\\n\\nand 60 2-milligram tablets of methadone. Less than one week later, on March 18,\\n\\n\\n 35\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 36 of 67\\n\\n\\n2010, officers responded to a call from Mayes’ residence reporting his sudden\\n\\ndeath from a possible drug overdose. Mayes’ mother testified at trial that she\\n\\nobserved Mayes taking the medications prescribed by Dr. Feldman the day before\\n\\nhe died. Officers found multiple pill bottles in Mayes’ room, including empty pill\\n\\nbottles for the oxycodone and alprazolam prescriptions dispensed by Dr. Feldman\\n\\non March 12. A subsequent toxicology report revealed that, at the time of death,\\n\\nMayes had in his system methadone or methadone metabolites, oxycodone, and\\n\\nalprazolam: the same three drugs that Dr. Feldman had prescribed.\\n\\n Based on this evidence, a reasonable jury could find beyond a reasonable\\n\\ndoubt that the controlled substances ingested by Mayes were prescribed by Dr.\\n\\nFeldman. Indeed, the controlled substances found in Mayes’ system were the same\\n\\ntype of substances prescribed by Dr. Feldman just one week before. In addition,\\n\\nthe pill bottles containing those prescriptions were found empty in Mayes’ room.\\n\\nSee United States v. Merrill, 513 F.3d 1293, 1299 (11th Cir. 2008) (concluding that\\n\\nsufficient evidence established that a doctor’s recent prescriptions led to the\\n\\npatients’ deaths given the short time period between the prescription and the fact\\n\\nthat the defendant “prescribed for each of these patients the exact type of drug the\\n\\nmedical examiner found to have caused their deaths”).\\n\\n\\n\\n\\n 36\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 37 of 67\\n\\n\\n 3. Shannon Wren\\n\\n Finally, the evidence was sufficient to prove that Dr. Feldman had\\n\\nprescribed the specific drugs Wren ingested. Wren was initially seen by Dr.\\n\\nFeldman on May 18, 2011, at which time Dr. Feldman prescribed Wren 120 15-\\n\\nmilligram tablets of oxycodone, 60 10-milligram tablets of diazepam, and 60 25-\\n\\nmilligram tablets of amitriptyline.\\n\\n Approximately one week later, on May 26, 2011, Wren was seen vomiting\\n\\nand coughing by his girlfriend as she left for work. Later that day, Wren was\\n\\nfound dead in his master bathroom. The first officer on the scene found multiple\\n\\npill bottles in the bathroom. 7 Included in those medications were the containers of\\n\\namitriptyline that Dr. Feldman had prescribed on May 18, which contained only 34\\n\\npills out of the 60-count prescription. Subsequently, Wren’s father handed over to\\n\\nan investigating agent the bottles of oxycodone and diazepam prescribed by Dr.\\n\\nFeldman, which were empty at that time.\\n\\n The subsequent autopsy revealed oxycodone, diazepam and its metabolites,\\n\\nand amitriptyline and its metabolites in Wren’s system. During Wren’s May 18\\n\\nvisit, Dr. Feldman had prescribed oxycodone, diazepam, and amitriptyline, which\\n\\n\\n\\n\\n7\\n These bottles contained melatonin, gabapentin, fluoxetine, metoprolol, trazodone, anastrozole,\\ninjectable testosterone, citalopram, adrenal tea capsule, and amitriptyline. Of these substances,\\nonly amitriptyline was found in Wren’s system during the autopsy. As noted in text, Dr.\\nFeldman had recently prescribed amitriptyline.\\n 37\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 38 of 67\\n\\n\\nwere the same drugs found in Wren’s system after he died. As with the other two\\n\\nvictims, a reasonable jury could conclude that the drugs that Wren ingested were\\n\\nprescribed by Dr. Feldman.\\n\\n In short, with respect to each victim, a jury could readily find beyond a\\n\\nreasonable doubt that the controlled substances the victim had ingested were\\n\\nprescribed by Dr. Feldman.\\n\\n B. Sufficient Evidence Was Presented at Trial From Which a Jury\\n Could Conclude That But for Ingestion of the Schedule II\\n Drugs Prescribed by Dr. Feldman, the Victims Would Not Have\\n Died\\n\\n As noted, the Government had sought an enhancement of Dr. Feldman’s\\n\\nsentence on these three substantive drug counts. To repeat, the enhancement\\n\\nprovision in question imposes a 20-year mandatory minimum sentence and an\\n\\nincreased maximum sentence for a defendant who has distributed a Schedule I or II\\n\\ndrug when death or serious bodily injury “results from the use of such substance.”\\n\\n21 U.S.C. § 841(b)(1)(C). On appeal, Dr. Feldman argues that there was\\n\\ninsufficient evidence to prove that the Schedule II drugs he prescribed caused the\\n\\nvictims’ deaths.\\n\\n As to the role played by a Schedule II drug, the jury found in its special\\n\\nverdict that “but for” the Schedule II and Schedule IV drugs prescribed to each\\n\\nvictim by Dr. Feldman, death would not have resulted. Dr. Feldman argues,\\n\\nhowever, that to convict a person based on death resulting from distribution of a\\n 38\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 39 of 67\\n\\n\\nSchedule II drug, the Government must prove that the Schedule II drug was the\\n\\n“sole,” the “only,” or the “independent” cause of each victim’s death, and that the\\n\\nGovernment failed to do so. More particularly, he argues that because each victim\\n\\nalso had a non-Schedule II drug in their bodies—which was also prescribed by Dr.\\n\\nFeldman—and because the non-Schedule II drug potentially contributed to the\\n\\nvictim’s death, the Schedule II drug cannot be said to have caused the particular\\n\\nvictim’s death.\\n\\n In arguing that the Government must prove that the Schedule II drug alone\\n\\ncaused the death of the victim, Dr. Feldman relies on Burrage v. United States, 571\\n\\nU.S. 204 (2014), which examined the causality requirement of § 841(b)(1)(C) of\\n\\nthe Controlled Substances Act. As we explain below, his argument misreads\\n\\nBurrage, which neither requires proof that a Schedule I or II drug “alone” caused a\\n\\nvictim’s death nor precludes application of § 841(b)(1)(C)’s mandatory minimum\\n\\nsentence in mixed-drug intoxication cases. Instead, Burrage held only that, “where\\n\\nuse of the drug distributed by the defendant is not an independently sufficient\\n\\ncause of the victim’s death or serious bodily injury, a defendant cannot be liable\\n\\nunder the penalty enhancement provision of 21 U.S.C. § 841(b)(1)(C) unless such\\n\\nuse is a but-for cause of the death or injury.” Id. at 218–19 (emphasis added).\\n\\nApplying that standard, we find that sufficient evidence of but-for causation was\\n\\npresented at trial.\\n\\n\\n 39\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 40 of 67\\n\\n\\n 1. Burrage\\n\\n In Burrage, the victim had died after ingesting alprazolam, clonazepam,\\n\\noxycodone, hydrocodone, and heroin. 571 U.S. at 207. The defendant had\\n\\nprovided the victim only with the heroin, which is a Schedule I drug, and he was\\n\\ntried based on that distribution, as well as on the enhancement penalty triggered by\\n\\n§ 841(b)(1)(C) when death results from distribution of a Schedule I or II drug. Id.\\n\\nat 206–07. Neither expert witness could say whether the victim would have lived\\n\\nhad he not taken the heroin; in other words, he might have died even without the\\n\\nheroin, given the other drugs he had taken. Id. at 207. This failure of proof meant\\n\\nthat the Government had not proved that but for the heroin, the victim would have\\n\\nlived. Id. Nevertheless, the district court concluded that the Government was not\\n\\nrequired to prove that the heroin was a “but-for” cause of the victim’s death, and it\\n\\ninstructed the jury that it could find that death resulted from the defendant’s heroin\\n\\ndistribution if it merely found that the heroin was a “contributing cause” of the\\n\\nvictim’s death. Id. at 207–08.\\n\\n The Supreme Court disagreed with the causation standard adopted by the\\n\\ndistrict court. Because § 841(b)(1)(C) does not define the phrase “results from,”\\n\\nthe Court applied its ordinary meaning, concluding that the phrase imposed the\\n\\n“traditional” requirement of “actual causality.” Id. at 210–11. Proving “actual\\n\\ncausality,” the Court explained, generally requires proof that, at a minimum, “the\\n\\n\\n 40\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 41 of 67\\n\\n\\nharm would not have occurred in the absence of—that is, but for—the defendant’s\\n\\nconduct.” Id. at 211 (quotation marks omitted). In other words, a factor can be\\n\\nconsidered an “actual cause” if it was necessary for bringing about that result—that\\n\\nis, if the result would not have occurred without (or “but for”) the influence of the\\n\\nfactor.\\n\\n To illustrate how this but-for notion of “actual causation” applies, the Court\\n\\nused three hypotheticals. Id. at 211–12. In the first hypothetical, “A shoots B,\\n\\nwho is hit and dies.” Id. at 211. In this case, only one factor is in play—A’s\\n\\nconduct in shooting B. The Court explained that, in this example, A actually\\n\\ncaused B’s death because B would not have died “but for” A’s conduct. Id. In\\n\\nother words, A’s conduct was necessary for producing B’s death: had A not shot\\n\\nB, B would have lived.\\n\\n The Court made clear, though, that but-for causality does not require that a\\n\\nsingle factor alone produce the particular result. As the Court explained, “[t]he\\n\\nsame conclusion follows if the predicate act combines with other factors to\\n\\nproduce the result, so long as the other factors alone would not have done so—if,\\n\\nso to speak, it was the straw that broke the camel’s back.” Id. (emphasis added).\\n\\nTo illustrate how a factor can be a but-for cause (that is, a necessary factor for\\n\\nproducing a result) even when other factors are themselves but-for causes, the\\n\\nCourt posed a second hypothetical.\\n\\n\\n 41\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 42 of 67\\n\\n\\n In this hypothetical, a “man debilitated by multiple diseases” ingests poison\\n\\nand dies. Id. Here, the Court explained, the poison “is a but-for cause of his death\\n\\neven if those diseases played a part in his demise, so long as, without the\\n\\nincremental effect of the poison, he would have lived.” Id. Thus, both the poison\\n\\nand the debilitating diseases are but-for causes of the death if each factor played a\\n\\nnecessary role in producing that outcome. Further, the poison and the debilitating\\n\\ndiseases would each constitute a but-for cause of the man’s death—“the straw that\\n\\nbroke the camel’s back”—if the man would have lived had he (1) only had\\n\\ndebilitating diseases, but not taken the poison or (2) only taken the poison, but not\\n\\nhad debilitating diseases. In other words, had one of these factors been absent, the\\n\\ndecedent would not have died. If, however, a debilitating disease was\\n\\n“independently sufficient”—that is, enough by itself—to kill the man at the same\\n\\ntime that he was ingesting the poison, the “incremental effect” of the poison was\\n\\nnot necessary to produce the resulting death. Id. at 211, 218–19. In that case, the\\n\\npoison would not be considered a but-for cause of the man dying; it was not “so to\\n\\nspeak, [] the straw that broke the camel’s back,” because the debilitating disease\\n\\nhad performed that function.8 Id. at 211.\\n\\n\\n8\\n The Supreme Court considered, but did not decide, the question of causation when a victim\\ndies following two concurrent events, each of which would have been enough on its own to\\ncause death. The Court posited a scenario in which “A stabs B, inflicting a fatal wound; while at\\nthe same moment X, acting independently, shoots B in the head also inflicting a fatal wound; and\\nB dies from the combined effects of the two wounds.” Burrage, 571 U.S. at 215 (alterations\\n\\n 42\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 43 of 67\\n\\n\\n The Court further elaborated on the but-for-causation principle in a third\\n\\nhypothetical, asking the reader to consider a baseball game where the visiting\\n\\nteam’s leadoff batter hits a home run in the first inning and the visiting team\\n\\nultimately wins 1 to 0. Id. at 211–12. The Court noted that one could accurately\\n\\nsay that the home run actually caused the victory because the visiting team would\\n\\nnot have won but for the home run—that is, the home run was necessary for the 1-\\n\\n0 victory to occur. Id. The Court explained that the home run was no less a but-\\n\\nfor cause of the victory just because it was only one of many but-for causes that\\n\\nwere each necessary for the visiting team to win. Id. at 212. As the Court put it,\\n\\n“[i]t is beside the point that the victory also resulted from a host of other necessary\\n\\ncauses, such as skillful pitching, the coach’s decision to put the leadoff batter in the\\n\\nlineup, and the league’s decision to schedule the game.” Id. (emphasis in original).\\n\\nThis is so because whether a factor is a but-for cause turns only on whether the\\n\\n\\n\\n\\naccepted) (citation omitted). In other words, B would have died had he only been stabbed.\\nLikewise, he would have died had he only been shot. Each was an independent sufficient cause\\nof his death.\\n\\nQuoting from the noted criminal law treatise, LaFave, Substantive Criminal Law at 468, the\\nSupreme Court noted that the treatise recognizes a special rule for such scenarios providing that\\n“A will generally be liable for homicide even though his conduct was not a but-for cause of B’s\\ndeath (since B would have died from X’s actions in any event).” Id. The Court acknowledged,\\nhowever, that: “We need not accept or reject the special rule developed for [the above scenario],\\nsince there was no evidence here that [the decedent’s] heroin use was an independently sufficient\\ncause of his death. No expert was prepared to say that [the decedent] would have died from the\\nheroin use alone.” Id.\\n 43\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 44 of 67\\n\\n\\nresult would have occurred without it; that is, whether that individual factor made a\\n\\ndifference to the outcome.\\n\\n “By contrast,” the Court says, “it makes little sense to say that an event\\n\\nresulted from or was the outcome of some earlier action if the action merely played\\n\\na nonessential contributing role in producing the event.” Id. Thus, the Court\\n\\nexplains, if we modify the baseball hypothetical such that the visiting team’s\\n\\nleadoff batter hits a home run in the first inning, but the visiting team ultimately\\n\\nwins by 5 to 2 rather than 1 to 0, the home run was not a but-for cause of the\\n\\nvictory. Id. This is so because the visiting team would have won even without the\\n\\nhome run: the visiting team had a total of 5 runs and only needed 3 to win,\\n\\nmeaning that the other runs that the visiting team scored were collectively\\n\\nsufficient on their own to cause the victory without needing the incremental effect\\n\\nof the leadoff hitter’s home run. In this scenario, the leadoff home run played a\\n\\nnonessential role in producing the outcome. It did not “[break] the camel’s back.”\\n\\nId. at 211.\\n\\n 2. Application of Burrage to Multi-Drug-Usage Scenarios\\n\\n In order to determine when one can properly conclude that death has\\n\\nresulted from the ingestion of a Schedule I or II drug, which event triggers\\n\\n§ 841(b)(1)(C)’s sentence enhancement, we move away from our national pastime\\n\\n\\n\\n\\n 44\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 45 of 67\\n\\n\\nto examine how Burrage applies to various multi-drug use scenarios involving the\\n\\ningestion of a Schedule I or II drug together with non-Schedule I or II drugs.\\n\\n First, we posit a deceased who ingested heroin, a Schedule I drug, in an\\n\\namount large enough to constitute a fatal dose for this person had the heroin been\\n\\ntaken alone. In this hypothetical, we further assume that the decedent had also\\n\\ntaken Valium, a Schedule IV drug, 9 but that the Valium, by itself, would not have\\n\\ncaused death. Figuring out causation here is simple. Because the drug user has\\n\\ndied from this fatal dose of heroin, we can readily say that the heroin was a but-for\\n\\ncause of his death. In other words, because the decedent would have lived had he\\n\\nnot ingested the heroin, use of the heroin was necessary for producing the death.\\n\\n Indeed, if the prosecution proved that the drug user took a fatal dose of\\n\\nheroin that by itself would have been enough to produce death, this evidence\\n\\nsatisfies a higher standard for proving actual causation than even Burrage’s but-for\\n\\ntest requires. In this situation, the heroin was not only a necessary, but-for cause of\\n\\nthe death, but it was also independently sufficient, by itself, to cause the user’s\\n\\ndeath, even without the influence of any other factors, including the Valium.\\n\\n Second, we change the facts a bit to assume that the deceased not only\\n\\ningested enough heroin that by itself would be fatal, but he also took enough\\n\\n\\n\\n9\\n Valium, whose generic version is called diazepam, is a Schedule IV drug that does not give\\nrise to an enhancement if death results.\\n 45\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 46 of 67\\n\\n\\nValium that, had the Valium been taken alone, it would likewise have caused\\n\\ndeath. In other words, each of the two drugs independently would have caused\\n\\ndeath. Although the Supreme Court expressly declined to reach this particular\\n\\nscenario,10 there is a strong argument that the heroin should be considered to have\\n\\ncaused the death for purposes of the sentencing enhancement. If actual causation\\n\\ncan be established by proving a factor was necessary to cause a result (meaning it\\n\\nsatisfies the but-for test), it logically follows that actual causation can also be\\n\\nestablished by satisfying a higher standard for proving that the factor was\\n\\nsufficient, by itself, to cause the result. It would create an extremely odd result if\\n\\nthe purveyor of a Schedule I or II drug that was strong enough by itself to cause\\n\\ndeath could escape liability merely because the victim also happened to have taken\\n\\na non-Schedule I or II drug that was itself independently sufficient to cause death.\\n\\nAnd the Supreme Court in Burrage indicated that it did not require satisfaction of\\n\\nthe but-for test when the drug at issue is an independently sufficient cause of the\\n\\nvictim’s death. Burrage, 571 U.S. at 218–19.\\n\\n Our third hypothetical takes us into a scenario in which the heroin is not a\\n\\nbut-for cause. In this hypothetical, where again the deceased has taken both heroin\\n\\nand Valium, we assume that the heroin was not enough, by itself, to induce death,\\n\\nbut that the Valium dosage was strong enough to have caused the decedent to die,\\n\\n\\n10\\n See discussion at n.8, supra.\\n 46\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 47 of 67\\n\\n\\neven had he not ingested any heroin. In this scenario, use of the heroin does not\\n\\nmeet the but-for test because it had no impact on what ultimately happened to the\\n\\ndrug user. He would have died—whether or not he had used heroin—because the\\n\\nSchedule IV Valium was going to kill him anyway.\\n\\n In our fourth and final hypothetical, we assume the heroin would not have\\n\\ncaused the deceased’s death had it been taken alone, but likewise the deceased\\n\\nwould not have died had he taken just the Valium. Yet, by virtue of the fact that he\\n\\ningested the heroin along with the Valium, he died. Here, the heroin can be\\n\\nconsidered the straw that broke the camel’s back. Stated another way, had the\\n\\nvictim never ingested the heroin, the decedent would still be alive. We conclude\\n\\nthat, consistent with Burrage, the but-for test for the heroin is met in this scenario.\\n\\n Dr. Feldman, however, disagrees with our analysis, at least as to this fourth\\n\\nscenario. In the present case, the evidence established that each victim had\\n\\ningested two types of drugs prescribed by the doctor: Schedule II drugs\\n\\n(methadone and oxycodone for one victim and oxycodone for the other two\\n\\nvictims) and a non-Schedule II drug (Xanax as to one victim and Valium as to the\\n\\nother two victims). Further, the evidence supported an inference either that the\\n\\nSchedule II drugs were independently sufficient to cause each victim’s death or, at\\n\\nthe very least, that neither of the non-Schedule II drugs would have caused death\\n\\nhad the decedent not ingested the Schedule II drug(s). Thus, the evidence clearly\\n\\n\\n 47\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 48 of 67\\n\\n\\nestablished that, had each respective victim not used the Schedule II drug in\\n\\nquestion, he would not have died. In other words, but for the Schedule II drugs,\\n\\nthe victims would still be alive.\\n\\n It is Dr. Feldman’s position, though, that causation cannot be established\\n\\nunless the Schedule II drug was the sole or only cause of death. According to the\\n\\ndoctor, if the non-Schedule II drugs (that he had also prescribed together with the\\n\\nSchedule II drugs) played any role in causing the victim’s death, it can no longer\\n\\nbe said that use of the Schedule II drugs resulted in death, even though the\\n\\nSchedule II drugs have met the but-for test. That certainly sounds like an odd\\n\\nconclusion. And it is. Burrage does not require proof that a Schedule I or II drug\\n\\nbe the “only” cause of a victim’s death. Indeed, the Court explicitly rejected that\\n\\ncharacterization of but-for causation, stating that a factor can be a but-for cause\\n\\neven if it “combines with other factors to produce the result,” and that whether “a\\n\\nhost of other necessary causes” might also contribute to the result is “beside the\\n\\npoint.” Id. at 211–12 (emphasis in original). Moreover, a rule requiring proof that\\n\\na drug is the one-and-only but-for cause would be tantamount to requiring a\\n\\nshowing that the drug was “independently sufficient” to cause the death—a bar\\n\\nthat is much higher that the “minimum” showing of but-for causation that Burrage\\n\\nrequired. Id. at 211 (emphasis in original). Notably, Burrage expressly held that\\n\\nthe Government need not prove that a Schedule I or II drug was “an independently\\n\\n\\n 48\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 49 of 67\\n\\n\\nsufficient cause of the victim’s death or serious bodily injury,” stating that only\\n\\nbut-for causation is required, and implying that we can dispense with the but-for\\n\\ntest when a drug is an independently sufficient cause of death. Id. at 218–19.\\n\\n Likewise, Dr. Feldman’s argument that evidence of “mixed-drug\\n\\nintoxication” necessarily precludes a finding that a victim died as a result of a\\n\\nSchedule I or II drug finds no support in Burrage. In Burrage, the medical experts\\n\\ntestified that the victim had died from “mixed drug intoxication with heroin,\\n\\noxycodone, alprazolam, and clonazepam all playing a contributing role.” Id. at\\n\\n207 (quotation marks omitted). Neither expert could say, however, “whether [the\\n\\nvictim] would have lived had he not taken the heroin.” Id. In concluding that\\n\\ninsufficient evidence supported the defendant’s conviction, the Court held that\\n\\nproving actual causation required more than showing that a Schedule I or II drug\\n\\n“contributes to an aggregate force (such as mixed-drug intoxication) that is itself a\\n\\nbut-for cause of death.” Id. at 214. But this statement in no way suggests that a\\n\\ndefendant is off the hook anytime a Schedule I or II drug that he distributed ends\\n\\nup in a deadly drug cocktail that kills a victim. Instead, the Court’s statement was\\n\\nintended to clarify that the Government could not establish that a Schedule I or II\\n\\ndrug actually caused a death without evidence that the Schedule I or II drug\\n\\nitself—as opposed to the drug cocktail as a whole—made a difference to the\\n\\noutcome. Id. at 214, 219 (concluding that the lack of “evidence that [the victim]\\n\\n\\n 49\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 50 of 67\\n\\n\\nwould have lived but for his heroin use” required reversal of the defendant’s\\n\\nconviction). In other words, to prove but-for causation, the Government must drill\\n\\ndown and show that, but for the Schedule I or II drug’s inclusion in the drug\\n\\ncocktail, the victim would have lived. If the Government makes that showing, it is\\n\\n“beside the point” that other drugs were necessary to make the drug cocktail\\n\\ndeadly. See id. at 211–12, 218–19.\\n\\n Indeed, a sister circuit, the Sixth Circuit, has reached the same conclusion.\\n\\nIn United States v. Volkman, 797 F.3d 377, 396 (6th Cir. 2015), the defendant\\n\\ndoctor had been convicted of the enhancement provision based on his dispensing\\n\\nof the Schedule II drug oxycodone to the victim. He argued that because the\\n\\ndecedent died based on the “combined effects of oxycodone, diazepam, and\\n\\nalprazolam,” the Government had failed to prove that the death resulted from the\\n\\ningestion of oxycodone. Id. at 395. Relying on Burrage, the Sixth Circuit panel\\n\\nrejected this argument, holding that “[t]he Government was not required to prove,\\n\\nhowever, that oxycodone was [the victim’s] only cause of death” because “but-for\\n\\ncausation exists where a particular controlled substance—here, oxycodone—\\n\\n‘combines with other factors’—here, inter alia, diazepam and alprazolam—to\\n\\nresult in death,” and “[t]he Government presented sufficient oxycodone-specific\\n\\nevidence for a rational jury to find that, ‘without the incremental effect’ of the\\n\\noxycodone, [the victim] would not have died.” Id. So it is here. As Volkman\\n\\n\\n 50\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 51 of 67\\n\\n\\nexplained, “the existence of other potential contributing causes of death is\\n\\nirrelevant so long as the Government presented sufficient evidence that [the\\n\\nSchedule II controlled substance] was a but-for cause of [the victim’s] death.” Id.\\n\\nat 395–96; see also Roundtree v. United States, 885 F.3d 1095, 1098 (8th Cir.\\n\\n2018) (concluding that where an expert testified that heroin and alcohol had\\n\\nworked together “synergistically” to cause death and where alcohol was not\\n\\nenough by itself to cause death, no reasonable jury could find that the heroin was\\n\\nmerely a contributory factor, as opposed to a but-for cause).\\n\\n With these principles in mind, we conclude that the Government presented\\n\\nsufficient evidence for each of Counts 2–4 to prove beyond a reasonable doubt that\\n\\na Schedule II drug that Dr. Feldman prescribed (methadone or oxycodone) was, at\\n\\na minimum, a but-for cause of each victim’s death.\\n\\n 3. Ricky Gonzalez\\n\\n The evidence presented at trial was sufficient to support a conclusion that a\\n\\nSchedule II drug—oxycodone—was a but-for cause of Gonzalez’s death. First,\\n\\nevidence provided by the defense expert, Dr. Adams, indicated that there was\\n\\nenough oxycodone in Gonzalez’s system to kill a person who did not have a high\\n\\ndegree of tolerance. In other words, use of oxycodone was a potentially sufficient\\n\\ncause, by itself, to cause death.\\n\\n\\n\\n\\n 51\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 52 of 67\\n\\n\\n But whether or not oxycodone was a sufficient cause, there was ample\\n\\nevidence from which a jury could conclude that but for Gonzalez’s use of\\n\\noxycodone, he would not have died. Dr. Thogmartin, the medical examiner who\\n\\nperformed Gonzalez’s autopsy, observed that Gonzalez had pulmonary edema, or\\n\\nfluid-filled lungs, an occurrence often seen in drug overdoses. Dr. Thogmartin\\n\\nconcluded that Gonzalez’s cause of death was oxycodone toxicity with a\\n\\ncontributory condition of diazepam (Valium) intoxication. 11\\n\\n Noting that Gonzalez had a significant amount of oxycodone and diazepam\\n\\nin his system, the doctor explained that the oxycodone, a Schedule II drug, had\\n\\n“poisoned” Gonzalez and the diazepam, a Schedule IV drug, had reacted\\n\\n“synergistically” to speed up his death. He explained that in an overdose scenario\\n\\ninvolving the mixture of an opiate, such as oxycodone, and a benzodiazepine, such\\n\\nas diazepam, the “main player” in causing death is the opiate, with the\\n\\nbenzodiazepine acting as only a “minor leaguer,” a “contributory cause.” As to\\n\\nhow he would characterize such an interaction, he determined that the oxycodone\\n\\nwas “the big player” and therefore the “cause of death,” and that the diazepam\\n\\n“was not lethal in and of itself, but a helper.” Indeed, in explaining that\\n\\n“oxycodone is the big player,” the doctor doubted that diazepam alone would have\\n\\n\\n\\n11\\n Again, the jury concluded that Dr. Feldman had prescribed both the oxycodone and diazepam\\nfound in Gonzalez’s system.\\n 52\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 53 of 67\\n\\n\\nkilled Gonzalez,” noting that “[d]iazepam is really hard to overdose on. . . . I mean,\\n\\nyou’d have to like eat a ton and even then you may not die.” While he thought\\n\\nthere was enough oxycodone here to kill Gonzalez by itself, he couldn’t say for\\n\\ncertain that Gonzalez would have died had he not also ingested the diazepam, as\\n\\nthat is the typical scenario he has observed. Nevertheless, Dr. Thogmartin\\n\\nconcluded, with a reasonable degree of medical certainty, that if it were not for the\\n\\noxycodone he ingested, Gonzalez would “not have died this day of this death.” In\\n\\nshort, the doctor concluded that Gonzalez died of oxycodone intoxication, with\\n\\ndeath having been sped up by the concurrent ingestion of the diazepam.\\n\\n In short, at the least, a jury could reasonably conclude that the oxycodone\\n\\nwas a necessary, but-for cause of his death. In other words, Gonzalez would have\\n\\nremained alive had he not ingested this Schedule II drug. Even assuming that the\\n\\ndiazepam may have also constituted a necessary, but-for cause of death, that fact\\n\\ndoes not deprive the oxycodone of its own “but-for” status. To argue that it does is\\n\\nto conclude that the oxycodone had to be a sufficient cause of death: that is, that\\n\\nthe drug had to, by itself, be capable of causing death. As explained above, this is\\n\\nnot how Burrage defined a but-for cause and indeed Burrage made clear that there\\n\\ncan be multiple but-for causes of an event. Accordingly, a reasonable jury could\\n\\nconclude beyond a reasonable doubt that Gonzalez died as a result of a Schedule II\\n\\ndrug that Dr. Feldman prescribed: namely, oxycodone.\\n\\n\\n 53\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 54 of 67\\n\\n\\n 4. Joey Mayes\\n\\n The toxicology report indicated that Mayes had ingested two Schedule II\\n\\ndrugs—methadone and oxycodone—and one Schedule IV drug—alprazolam\\n\\n(Xanax). Dr. Feldman had prescribed all three drugs. There was testimony from\\n\\nboth the Government’s expert, who was the medical examiner who performed the\\n\\nautopsy, and from the defense expert that there was enough methadone in Mayes’\\n\\nsystem to kill him, even had other drugs not been ingested.\\n\\n As for the defense expert, Dr. Adams testified that Mayes’ cause of death\\n\\nwas “intoxication by methadone,” which was “the only drug on board capable of\\n\\ncausing a cardiac death.” From that testimony, alone, a jury could have found, at a\\n\\nminimum, that methadone played a necessary (but-for) role in Mayes’ death,\\n\\nwithout which he would have lived. Moreover, because Dr. Adams’ testimony\\n\\nexculpated the Schedule IV alprazolam, as well as the Schedule II oxycodone, a\\n\\njury would have been entitled to find that methadone was also an “independently\\n\\nsufficient cause” of Mayes’ death. Burrage, 571 U.S. at 218–19.\\n\\n As for the Government’s expert, in his autopsy report, Dr. Wilson had listed\\n\\nthe cause of death as multidrug toxicity involving the above-listed drugs. He did\\n\\nnot stop there, however. Dr. Wilson further testified that Mayes had “particularly\\n\\nelevated” levels of methadone in his system in an amount that was enough by itself\\n\\nto kill him, and he made clear that the methadone was the critical drug necessary to\\n\\n\\n 54\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 55 of 67\\n\\n\\nhave caused Mayes’ death. Further, given the high level of methadone, Mayes\\n\\nwould have died even had he received quicker medical attention. In addition to the\\n\\nmethadone, Dr. Wilson also stated that there was likely enough oxycodone (a\\n\\nSchedule II drug), by itself, to kill Mayes.\\n\\n And while the doctor acknowledged that his autopsy report had used the\\n\\nphrase “multidrug toxicity,” explaining that in his opinion each drug “had a role”\\n\\nin killing Mayes, acting “synergistically” with one another, he reiterated that the\\n\\nmethadone was “the major drug that was involved in Mr. Mayes’ death, with\\n\\ndistributing effects from alprazolam as well as oxycodone.”\\n\\n Based on this testimony, a reasonable jury could conclude that each of the\\n\\nSchedule II drugs (methadone and oxycodone) played, at a minimum, a necessary\\n\\n“but-for” role in causing Mayes’ death. As discussed above, that another drug (the\\n\\nalprazolam) may have also played a necessary role in causing Mayes’ death is\\n\\n“beside the point.” Id. at 211–12.\\n\\n 5. Shannon Wren\\n\\n As to Shannon Wren, defense expert Dr. Adams opined that the cause of\\n\\ndeath for this decedent was a heart attack, not the drugs ingested by Wren. The\\n\\nmedical examiner, Dr. Wilson, disagreed. Instead, Dr. Wilson concluded that\\n\\nWren died from multidrug toxicity. He testified that the oxycodone in Wren’s\\n\\nsystem would have been enough, by itself, to kill him, and that the diazepam would\\n\\n\\n 55\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 56 of 67\\n\\n\\nnot have been sufficient to kill Wren, by itself. As to listing the ultimate cause of\\n\\ndeath as multidrug toxicity, he explained that each of the drugs acted with the\\n\\nothers to cause death. 12\\n\\n In short, taking the evidence in the light most favorable to the Government,\\n\\nthis evidence was sufficient to support an inference that the oxycodone was a\\n\\nsufficient cause of Wren’s death, as well as a necessary, “but-for” cause.\\n\\n Dr. Feldman points out that his expert, Dr. Adams, disagreed with Dr.\\n\\nWilson’s conclusion that Wren died from multidrug toxicity. Instead, Dr. Adams\\n\\nbelieved that Wren’s enlarged heart and the fact that he was found dead on the\\n\\nbathroom floor indicated that he had died from sudden cardiac arrest, rather than\\n\\nfrom a drug overdose. Dr. Adams did not perform the autopsy, however; Dr.\\n\\nWilson did. Further, Dr. Wilson considered Wren’s enlarged heart and\\n\\nnevertheless concluded that there was no evidence of heart attack because there\\n\\nwas “still enough . . . reservoir opening of the coronary arteries to give enough\\n\\nblood to the heart.” It was within the province of the jury to choose among the\\n\\nreasonable constructions of the evidence and weigh Dr. Wilson’s testimony more\\n\\nheavily than Dr. Adams’. See United States v. Ndiaye, 434 F.3d 1270, 1294 (11th\\n\\n\\n\\n12\\n The drugs found in Wren’s system included oxycodone, diazepam, amitriptyline, and\\nmetabolites of the latter two drugs. Like diazepam, amitriptyline (more familiarly known as the\\nantidepressant Elavil) is a Schedule IV drug. The defense expert indicated that this Elavil was in\\n“an innocuous concentration,” and the Government’s expert did not focus on it at all. Nor was\\nthis drug listed on the special verdict form.\\n 56\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 57 of 67\\n\\n\\nCir. 2006) (“Because the jury is free to choose among reasonable constructions of\\n\\nthe evidence, the evidence may be sufficient even if it is not entirely inconsistent\\n\\nwith conclusions other than guilt.”).\\n\\n In sum, when viewed in the light most favorable to the Government, there\\n\\nwas sufficient evidence from which a reasonable jury could conclude that a\\n\\nSchedule II drug prescribed by Dr. Feldman—methadone and oxycodone in\\n\\nMayes’ case, and oxycodone in Gonzalez’s and Wren’s cases—was, at the very\\n\\nleast, a but-for cause of each of the victims’ deaths, if not an independently\\n\\nsufficient cause.\\n\\n VIII. Alleged Alleyne Error\\n Dr. Feldman asserts that even if there was sufficient evidence to support the\\n\\nsentencing enhancement for Counts 2–4, its application in this case was\\n\\nunconstitutional under Alleyne v. United States, 570 U.S. 99 (2013). The\\n\\nGovernment does not dispute that Dr. Feldman adequately preserved this issue\\n\\nbelow, and we therefore review his claims of Alleyne error de novo. United States\\n\\nv. King, 751 F.3d 1268, 1278–79 (11th Cir. 2014).\\n\\n As noted above, in Counts 2–4, Dr. Feldman was convicted of violating\\n\\n21 U.S.C. § 841(a)(1) based on his unlawful dispensing of oxycodone (a Schedule\\n\\nII drug), methadone (a Schedule II drug), alprazolam (a Schedule IV drug), and\\n\\ndiazepam (a Schedule IV drug) to the three individuals identified in those counts.\\n\\n\\n 57\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 58 of 67\\n\\n\\nThe jury rendered that verdict by answering in the affirmative the special verdict\\n\\nform’s 13 first two questions for each count, finding that Dr. Feldman was guilty of\\n\\ndistributing and dispensing the Schedule II and Schedule IV drugs identified in the\\n\\nindictment. Although Dr. Feldman challenged the sufficiency of the evidence\\n\\nsupporting his conviction for conspiracy to distribute these substances and argued\\n\\nthat insufficient evidence supported a jury finding that the Schedule II drugs\\n\\nresulted in death, he has not challenged his substantive distribution convictions.\\n\\nAnd even had he done so, we would have rejected that challenge. There is ample\\n\\nevidence to support the jury’s conclusion that Dr. Feldman prescribed all of these\\n\\nsubstances to the particular victims and that he did so not for a legitimate medical\\n\\npurpose and not in the usual course of professional practice.\\n\\n The maximum penalty for an “unenhanced” conviction based on distribution\\n\\nof a Schedule II drug is 20 years’ imprisonment. 21 U.S.C. § 841(b)(1)(C). Thus,\\n\\nwithout any other enhancements, Dr. Feldman was subject to a 20-year maximum\\n\\nsentence, with no mandatory-minimum provision, for each conviction set out in\\n\\nCounts 2–4.\\n\\n The Government, however, sought an enhanced sentence, alleging that each\\n\\nof the victims identified for purposes of Counts 2–4 had died as a result of their\\n\\n\\n13\\n The special verdict form for each victim included four overarching questions, with two of the\\nquestions including sub-parts that asked the jury to make findings regarding each specified\\ncontrolled substance.\\n 58\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 59 of 67\\n\\n\\ningestion of the Schedule II drugs prescribed by Dr. Feldman. Under the\\n\\nControlled Substances Act’s sentence-enhancement provision, if death results from\\n\\nthe use of a Schedule II substance, the person convicted of distributing that drug is\\n\\nsubject to a mandatory-minimum 20-year sentence, with a statutory maximum of\\n\\nlife imprisonment. Id. § 841(b)(1)(C). Death resulting from ingestion of a\\n\\nSchedule IV substance, however, results in no enhancement.\\n\\n The district court determined that the jury had convicted Dr. Feldman of\\n\\nconduct justifying the above enhancement. As a result, Dr. Feldman was subject to\\n\\na mandatory-minimum 20-year sentence, and the district court ultimately decided\\n\\nthat a 300-month sentence (25-years) was a reasonable sentence based on the\\n\\nconduct underlying these three convictions.14\\n\\n “Because the ‘death results’ enhancement increase[s] the minimum and\\n\\nmaximum sentences to which [a defendant is] exposed, it is an element that must\\n\\nbe submitted to the jury and found beyond a reasonable doubt.” Burrage, 571 U.S.\\n\\nat 210 (citing Alleyne, 570 U.S. at 115–16, and Apprendi, 530 U.S. at 490). Dr.\\n\\nFeldman asserts that the jury did not convict him of distributing a Schedule II\\n\\nsubstance from which death has resulted, which, if correct, would mean that the\\n\\n\\n14\\n Dr. Feldman received a 10-year sentence for his convictions for conspiracy to distribute\\ncontrolled substances and conspiracy to commit money laundering, a 10-year sentence for his\\nsubstantive money laundering convictions, and a 25-year sentence for the enhanced distribution\\nof a Schedule II substance, use of which had resulted in death. All sentences ran concurrently\\nwith each other.\\n\\n 59\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 60 of 67\\n\\n\\ndistrict court violated Alleyne when it sentenced Dr. Feldman as if the jury had so\\n\\nfound. Specifically, and relying again on Burrage, Dr. Feldman argues that the\\n\\njury’s special verdict established only a finding that a combination of Schedule II\\n\\nand Schedule IV drugs in the aggregate killed each victim, not that the victims\\n\\nwould have lived “but for” their use of the Schedule II drugs. Accordingly, Dr.\\n\\nFeldman contends that the jury verdict was insufficient to trigger a 20-year\\n\\nmandatory minimum sentence enhancement or a statutory maximum exceeding\\n\\ntwenty years.\\n\\n To the extent that Dr. Feldman’s Alleyne argument mirrors his sufficiency-\\n\\nof-the-evidence argument, we reject it as a misreading of Burrage. As discussed\\n\\nabove, Burrage did not hold that a Schedule I or II drug must be the “only” cause\\n\\nof a victim’s death. Indeed, Burrage reached the opposite conclusion, explaining\\n\\nthat a factor can be a but-for cause even if it “combines with other factors to\\n\\nproduce the result, so long as the other factors alone would not have” produced the\\n\\nsame result. Id. at 211. Under Burrage, but-for causality simply turns on whether\\n\\nthe victim would have died “without the incremental effect” of a Schedule I or II\\n\\ndrug. Id. In other words, a Schedule I or II drug is a but-for cause of a victim’s\\n\\ndeath if it made a necessary contribution to the death—“if, so to speak, it was the\\n\\nstraw that broke the camel’s back”—such that the victim would have lived if he\\n\\nhad not taken the Schedule I or II drug. Id. The existence of other necessary\\n\\n\\n 60\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 61 of 67\\n\\n\\ncauses, such as other drugs in a deadly cocktail, is “beside the point.” Id. at 212.\\n\\nFor a sentencing court to impose an enhanced sentence under § 841(b)(1)(C), a\\n\\njury need only find that a Schedule I or II drug was a but-for cause of death or\\n\\nserious bodily injury. Id. at 210–11, 218–19.\\n\\n As discussed at length above, we conclude that the evidence was sufficient\\n\\nto prove beyond a reasonable doubt that but for ingestion of the Schedule II drugs\\n\\ndispensed by Dr. Feldman, the victims would not have died. Nevertheless, we\\n\\nagree with Dr. Feldman that the special verdict in this case failed to establish that\\n\\nthe jury actually found that the Schedule II drugs were but-for causes of the\\n\\nvictims’ deaths. Accordingly, we conclude that the district court erred under\\n\\nAlleyne by imposing an enhanced sentence under § 841(b)(1)(C). We explain our\\n\\nreasons for reaching this result.\\n\\n The third and fourth questions on the verdict form for each of Counts 2, 3,\\n\\nand 4 focused on the deaths of Mayes, Gonzalez, and Wren, respectively.\\n\\nAggregating all the drugs dispensed by Dr. Feldman to the victim—both Schedule\\n\\nII and Schedule IV—the third question asked whether the jury found that the\\n\\nvictim would not have died but for his use of the listed drugs. 15 The question for\\n\\n\\n\\n15\\n Asking for a “yes” or “no” answer, the question read: “We, the jury, find beyond a\\nreasonable doubt that the death of [the victim] resulted from the use of one or more of the\\nfollowing controlled substances: [identifying Schedule II and Schedule IV controlled substances\\ndispensed to victim], that is, but for his ingestion of the controlled substances charged, he would\\nnot have died.”\\n 61\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 62 of 67\\n\\n\\nCount 2 (Mayes) listed as the relevant drugs oxycodone and methadone, which are\\n\\nSchedule II drugs, as well as alprazolam, which is a Schedule IV drug. The\\n\\nquestions for Counts 3 and 4 (Gonzalez and Wren) each listed as the relevant drugs\\n\\noxycodone, a Schedule II drug, and diazepam, a Schedule IV drug. For each\\n\\ncount, the jury responded “Yes,” finding that the victim would not have died but\\n\\nfor his use of the listed controlled substances.\\n\\n These findings clearly do not support the district court’s application of\\n\\n§ 841(b)(1)(C)’s mandatory-minimum sentence. This is so because, at most, the\\n\\njury’s “Yes” responses to these questions represented a finding only that the\\n\\nSchedule II and Schedule IV drugs in the aggregate caused the deaths. And\\n\\nBurrage held that it is insufficient merely to show that a Schedule I or II drug was\\n\\npart of a mixed-drug cocktail, which, as “an aggregate force,” was “itself a but-for\\n\\ncause of death.” Id. at 214. Instead, Burrage requires a jury finding that a\\n\\nSchedule I or II drug was itself a but-for cause of the victim’s death—that is,\\n\\nwithout the specific incremental effect of the Schedule I or II drug, the victim\\n\\nwould have lived. See id. at 207, 211, 218–19 (reversing the defendant’s\\n\\nconviction where the evidence showed only that the victim died as a result of a\\n\\ncocktail of drugs, including heroin, oxycodone, alprazolam, and clonazepam, and\\n\\nthe Government conceded that no evidence showed the victim would have lived\\n\\nbut for his heroin use).\\n\\n\\n 62\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 63 of 67\\n\\n\\n The fourth question on Dr. Feldman’s special verdict form purportedly\\n\\nattempted to disaggregate the drugs to seek a finding as to which specific drugs\\n\\ncaused the death. Considering all the circumstances, however, we conclude that\\n\\nthis question failed to elicit the jury’s finding on what was the relevant question:\\n\\nwhether the Schedule II drugs were but-for causes of the victims’ deaths.\\n\\nSpecifically, this final question for each of Counts 2–4 asked whether the jury\\n\\nfound “beyond a reasonable doubt that the death of [the victim] resulted from the\\n\\nuse of the following,” then it listed separately each specific drug dispensed to the\\n\\nvictim: both Schedule II and Schedule IV substances. (emphasis added). The jury\\n\\nanswered “Yes” for each drug listed. In contrast with the prior question, though,\\n\\nthis fourth question entirely omitted any mention of but-for causation and failed to\\n\\nmake clear to the jury that absent a finding that the victim would not have died had\\n\\nhe not used a particular drug, the jury could not conclude that the victim’s death\\n\\nresulted from that drug.\\n\\n We cannot conclude based on the jury’s response to this final question that it\\n\\nfound beyond a reasonable doubt that a Schedule II drug Dr. Feldman prescribed—\\n\\noxycodone and methadone for Count 2, and oxycodone for each of Counts 3 and\\n\\n4—was a but-for cause of the victims’ deaths. First, the verdict form’s failure to\\n\\nspell out what “resulted from” meant in the critical and final question raises a\\n\\nsignificant concern that the jury was unaware of that phrase’s meaning as it\\n\\n\\n 63\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 64 of 67\\n\\n\\npertains to whether death resulted from a specific drug. And because the first\\n\\n“results in” question explained that “resulted from” meant that the victim would\\n\\nnot have died but for ingestion of the drugs in the aggregate, but the second\\n\\nquestion omitted an explanation of what “resulted from” meant, the jury could\\n\\nhave concluded that it was once again being asked to find whether all the drugs\\n\\nresulted in each victim’s death, rather than whether each victim would have lived\\n\\nbut for his ingestion of a Schedule II drug in particular.\\n\\n Even with the omission of the “but-for” language in this pivotal, final\\n\\nquestion, we might well reach a different result had the district court’s instructions\\n\\nexplained to the jury that, in order to conclude that death had resulted from a\\n\\nparticular Schedule II drug, it must find that but for ingestion of that drug, death\\n\\nwould not have occurred. Unfortunately, the court’s instructions did not do so.\\n\\nThe court instructed the jury only that it “must find beyond a reasonable doubt that\\n\\nbut for the decedent’s use of the charged controlled substances distributed . . . the\\n\\ndecedent would not have died.” (emphasis added). This plural reference to\\n\\ncontrolled substances in the instructions necessarily included the Schedule IV\\n\\nsubstances, which of course cannot trigger an enhancement. And it replicates the\\n\\nfirst interrogatory on the special verdict form, which directed the jury to make a\\n\\nfinding only as to whether the drugs in the aggregate caused each victim’s death.\\n\\n\\n\\n\\n 64\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 65 of 67\\n\\n\\n Thus, the court’s instructions failed to make clear that the jury needed to\\n\\nmake a specific finding as to whether the victim would have lived but for his use of\\n\\nthe Schedule II controlled substance. Indeed, the instructions never indicated that\\n\\nthe Schedule II drugs had any special significance in the case, as the court merely\\n\\ninstructed the jury that “a more serious offense is committed” whenever “death or\\n\\nserious injury is a consequence of the decedent’s use of a controlled substance.”\\n\\n(emphasis added). Nor did the prosecutor or defense attorney point out in closing\\n\\narguments the specific determination that the jury was being called on to make.\\n\\n Given this context, we cannot conclude that the jury found beyond a\\n\\nreasonable doubt that the victims would have lived but for their ingestion of the\\n\\nSchedule II drugs. Admittedly, the jury’s finding that the victims’ deaths “resulted\\n\\nfrom” each of the Schedule II and Schedule IV drugs is not inconsistent with a\\n\\nfinding that the Schedule II drugs were but-for causes of the deaths. As we have\\n\\nexplained at length, there can be many necessary causes of a result, and, if a\\n\\nSchedule II drug was a necessary cause of a victim’s death, it would have been\\n\\n“beside the point” that a Schedule IV drug also happened to occupy the same\\n\\nstatus. Id. at 211–12.\\n\\n But the verdict rendered is likewise consistent with a finding that the\\n\\nSchedule II drugs were nonessential “contributing cause[s],” which did not make a\\n\\ndifference to the ultimate outcome but merely “contribute[d] to an aggregate force\\n\\n\\n 65\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 66 of 67\\n\\n\\n(such as mixed-drug intoxication) that is itself a but-for cause of death.” Id. at 208,\\n\\n214. The latter finding, of course, would be insufficient to support a 20-year\\n\\nmandatory minimum sentence under Burrage. Id. at 214–16. And because we\\n\\ncannot determine which of the two findings the special verdict reflects, the jury’s\\n\\nverdict does not constitute the necessary finding that but for ingestion of a\\n\\nSchedule II substance, the victim would have lived.16 That being so, the district\\n\\ncourt’s application of § 841(b)(1)(C)’s sentencing enhancement cannot stand. 17\\n\\n Accordingly, we reverse the district court’s application of § 841(b)(1)(C)’s\\n\\n20-year mandatory minimum sentence on Counts 2–4 and remand the case for the\\n\\ndistrict court to resentence Dr. Feldman “to a term of imprisonment of not more\\n\\nthan 20 years” as to each of these counts. 21 U.S.C. § 841(b)(1)(C) (stating that\\n\\n“such person shall be sentenced to a term of imprisonment of not more than\\n\\n20 years and if death or serious bodily injury results from the use of such substance\\n\\nshall be sentenced to a term of imprisonment of not less than twenty years or more\\n\\nthan life”).\\n\\n\\n\\n\\n16\\n We emphasize that we are not saying that the jury acquitted Dr. Feldman of the charge that his\\ndispensing of a Schedule II drug resulted in death. Rather, the outcome of our ruling is the same\\nas it would have been had the Government never sought a specific finding from the jury on this\\nissue.\\n17\\n We are aware that an Alleyne error can be subject to harmless error review. United States v.\\nKing, 751 F.3d 1268, 1279 (11th Cir. 2014). The Government, however, has not argued that any\\nAlleyne error here would be harmless beyond a reasonable doubt.\\n 66\\n\\f Case: 16-12978 Date Filed: 08/30/2019 Page: 67 of 67\\n\\n\\n CONCLUSION\\n\\n For the above reasons, we AFFIRM the Feldmans’ convictions, REVERSE\\n\\nDr. Feldman’s sentences on Counts 2–4, and REMAND for the district court to\\n\\nconduct a new sentencing hearing for Dr. Feldman.\\n\\n\\n\\n\\n 67\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4166,"cells":{"id":{"kind":"number","value":4657889,"string":"4,657,889"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"Mark Jenkins v. Commissioner, Alabama Department of Corrections"},"decision_date":{"kind":"string","value":"2019-08-30"},"docket_number":{"kind":"string","value":"17-12524"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Eleventh Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before TJOFLAT, WILSON, and BRANCH, Circuit Judges.\", \"parties\": \"\", \"opinions\": [{\"author\": \"BRANCH, Circuit Judge:\", \"type\": \"010combined\", \"text\": \"Case: 17-12524 Date Filed: 08/30/2019 Page: 1 of 73\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n\\n\\n No. 17-12524\\n\\n\\n\\n D.C. Docket No. 4:08-cv-00869-VEH\\n\\nMARK ALLEN JENKINS,\\n\\n Petitioner - Appellant,\\n\\n versus\\n\\n\\n\\nCOMMISSIONER, ALABAMA DEPARTMENT OF CORRECTIONS,\\n\\n Respondent - Appellee.\\n\\n\\n\\n Appeal from the United States District Court\\n for the Northern District of Alabama\\n\\n\\n (August 30, 2019)\\n\\nBefore TJOFLAT, WILSON, and BRANCH, Circuit Judges.\\n\\nBRANCH, Circuit Judge:\\n\\n Mark Allen Jenkins, an Alabama prisoner sentenced to death for the 1989\\n\\nmurder of Tammy Ruth Hogeland, appeals the district court’s denial of his petition\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 2 of 73\\n\\n\\nfor a writ of habeas corpus, 28 U.S.C. § 2254. Before us are Jenkins’s arguments\\n\\nthat he received ineffective assistance of counsel during the penalty phase of his\\n\\ntrial and that he is intellectually disabled and therefore ineligible for the death\\n\\npenalty. After careful consideration, and with the benefit of oral argument, we\\n\\naffirm the denial of Jenkins’s habeas petition on both grounds.\\n\\n I. BACKGROUND\\n\\n A. The Crime and Arrest\\n\\n We summarize the following background narrative from opinions of the\\n\\nAlabama Court of Criminal Appeals and our own review of the record. See Jenkins\\n\\nv. State, 627 So. 2d 1034, 1037–40 (Ala. Ct. Crim. App. 1992); Jenkins v. State,\\n\\n972 So. 2d 111, 119–20 (Ala. Ct. Crim. App. 2004). The events leading up to the\\n\\nmurder began the evening of April 17, 1989. Jenkins was at the home of his\\n\\nacquaintance Christine Nicholas. He had met her several months earlier at the\\n\\nOmelet Shoppe restaurant where she worked. Jenkins was very intoxicated and\\n\\ntried to seduce her. When she resisted, Jenkins got “real mad” and asked her\\n\\nseveral times what she would do if someone came up from behind her and grabbed\\n\\nher.\\n\\n At around 1 a.m., Jenkins and Nicholas drove to the Riverchase Omelet\\n\\nShoppe. Jenkins went inside and talked with one of the waitresses, Frieda Vines.\\n\\nThe manager, Douglas Thrash, recognized Jenkins as a regular customer who knew\\n\\n\\n 2\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 3 of 73\\n\\n\\nall of the waitresses. Thrash overheard part of Jenkins’s conversation with Vines\\n\\nand heard mention of the Omelet Shoppe location near the Birmingham Airport.\\n\\nJenkins and Nicholas then returned to Nicholas’s home. Around 2 a.m., Nicholas’s\\n\\nmother asked Jenkins to leave. He did so, falling down some steps and ramming\\n\\nhis car into another vehicle in the process.\\n\\n Meanwhile, 23-year-old Tammy Hogeland was making her way to work at\\n\\nthe Omelet Shoppe. Her sister Wendy, along with Tammy’s young son, picked\\n\\nTammy up from her college classes and drove her to the Riverchase Omelet\\n\\nShoppe a little after 9:30 p.m. Sometime after 10:00, Tammy was sent to work at\\n\\nthe airport Omelet Shoppe on Tenth Avenue, since that location was unexpectedly\\n\\nshorthanded. Tammy was wearing her watch, a necklace with the words “special\\n\\nsister,” her class ring, and her diamond engagement ring. Tammy was working as a\\n\\ncook that night and was wearing a blue apron, black pants, a white shirt, and white\\n\\nshoes. She and Sarah Harris were the only employees working at the airport\\n\\nOmelet Shoppe that night.\\n\\n At about 2 a.m. on April 18, Harris saw Jenkins—whom she did not know—\\n\\ndrive up in a red sports car. His arrival was memorable to Harris because the car\\n\\nnearly jumped the curb and crashed through the restaurant’s glass wall. Jenkins got\\n\\nout of the car and came into the restaurant, appearing to be intoxicated. He walked\\n\\nover to Hogeland and began talking to her. Harris later saw Jenkins and Hogeland\\n\\n\\n 3\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 4 of 73\\n\\n\\ndrive away together in the red sports car. That was the last time anyone who knew\\n\\nHogeland ever saw her. Hogeland left behind her cigarettes, lighter, purse, and\\n\\npaycheck. She left without telling anyone, which she had never done before.\\n\\nAlthough Harris saw them drive off, she could not later say whether Hogeland left\\n\\nwith Jenkins willingly or was instead abducted.\\n\\n At around 5 a.m. that morning, Bobby and Geraldine Coe had stopped to\\n\\nbuy gas at a Chevron station on I-59 northeast of Birmingham when they saw\\n\\nJenkins drive up in the red sports car. The Coes noticed a female who appeared to\\n\\nbe “passed out” in the front passenger seat, but they could not say whether she was\\n\\nalive or dead. While Bobby Coe was pumping gas, Jenkins asked him for some\\n\\ncigarettes and said, “Looks like it’s been a long night and it looks like it’s going to\\n\\nbe a long day.” Jenkins then said “God bless you” before asking directions to I-\\n\\n459. Coe gave him directions and got back in his car. As Coe pulled out onto the\\n\\ninterstate, he saw Jenkins follow him in the red sports car. Coe then saw the car\\n\\nflash its lights, slow down, and pull to the side of the road between mile markers\\n\\n151 and 152.\\n\\n At 8 a.m., Wendy Hogeland learned that Tammy was not at the restaurant,\\n\\nand their mother called the police. Meanwhile, Jenkins went to the home of Steve\\n\\nMusser, who noticed that he was wearing the same clothes as the day before.\\n\\nJenkins told Musser that his car had been stolen the night before and asked him if\\n\\n\\n 4\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 5 of 73\\n\\n\\nhe would say that he had been with him all night. Musser refused. Christine\\n\\nNicholas also saw Jenkins in a grocery store that morning. He was looking at a\\n\\nnewspaper, making a phone call, and attempting to sell his Buick. Nicholas loaned\\n\\nhim $4 for gas. At around 10 o’clock that morning, Jenkins sold his car to Michael\\n\\nBrooks, a mechanic at a local Chevron station. Jenkins had explained that he\\n\\nneeded the money so he could visit his sick mother in California. Another\\n\\nmechanic at the service station drove Jenkins to the Greyhound bus station later\\n\\nthat morning.\\n\\n Jenkins awoke the next day on the Greyhound bus in Houston and was\\n\\nejected from the bus because his fare was used up. He then hitchhiked from\\n\\nHouston to Phoenix, then to San Diego and Los Angeles. Jenkins was first\\n\\nidentified as a suspect in Hogeland’s disappearance by Officer Mike Weems of the\\n\\nHoover Police Department on April 19. Weems ate dinner at the Riverchase\\n\\nOmelet Shoppe just about every day and had talked with Hogeland just as often.\\n\\nHe also knew Jenkins from the restaurant. Weems learned of Hogeland’s\\n\\ndisappearance from the other waitresses. Remembering how Jenkins often talked to\\n\\nHogeland and passed her notes,1 he gave Jenkins’s name to the missing persons\\n\\ninvestigator. The investigator also learned that a red Mazda RX-7 sports car, which\\n\\n\\n\\n1\\n The jury was not permitted to hear, and the record does not reveal, what exactly Jenkins had\\nsaid to Hogeland that made Officer Weems suspect him in her disappearance.\\n 5\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 6 of 73\\n\\n\\nhad been reported stolen from the service station where Jenkins worked, had been\\n\\nrecovered on I-459.\\n\\n The afternoon of April 21, a truck driver who had happened to stop on I-59\\n\\nnear mile marker 151 discovered Hogeland’s body off the side of the road. She was\\n\\nnaked, wearing only a watch, and was so badly decomposed that the body had to\\n\\nbe identified by dental records. From the fractured hyoid bone, it was determined\\n\\nthat she was manually strangled to death. Also found at the scene were Hogeland’s\\n\\napron, shoes, bra, panties, pants, and hair net, as well as some beer cans, a Mazda\\n\\nRX-7 owner’s manual, and other items later determined to have come from the red\\n\\nMazda. Her necklace and rings were never found.\\n\\n Alabama authorities issued a warrant for Jenkins’s arrest and also obtained a\\n\\nfederal fugitive warrant. An FBI agent arrested Jenkins on May 1 in front of his\\n\\nuncle’s house in Wilmington, California. Jenkins’s uncle later gave the police three\\n\\nbags of Jenkins’s personal effects. Fibers recovered from Hogeland’s body and\\n\\nclothing matched those of the Mazda, and hairs matching Hogeland’s were found\\n\\nin the Mazda. Similarly, fibers from Jenkins’s clothing placed him in the Mazda,\\n\\nand fibers of his jeans were found on Hogeland’s apron. A bootprint near\\n\\nHogeland’s body matched a combat boot from among the belongings Jenkins’s\\n\\nuncle gave to the police, and Jenkins’s roommate said the boots looked just like\\n\\nthose that Jenkins wore.\\n\\n\\n 6\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 7 of 73\\n\\n\\n A business card like those that had been in the Mazda’s glove box,\\n\\nbelonging to an owner of the Mazda, was recovered from Jenkins’s wallet after his\\n\\narrest in California. Later, Jenkins’s cellmate in the Alabama jail reported that\\n\\nJenkins told him “he had done the crime” and was worried that the couple from the\\n\\ngas station would identify him or that the police would find his fingerprint on a\\n\\nbeer can at the scene. A St. Clair County grand jury indicted Jenkins for capital\\n\\nmurder.\\n\\n B. The Trial and Sentencing\\n\\n Jenkins was represented at trial in 1991 by attorneys Douglas Scofield and\\n\\nStan Downey. Scofield, the lead attorney, had been practicing criminal defense in\\n\\nBirmingham since 1984 but had never before tried a capital murder case. Jenkins’s\\n\\nlandlord and his grandmother had arranged the referral and representation before\\n\\nthe court appointed him. Scofield agreed to the appointment on the condition that a\\n\\nSt. Clair County lawyer also be appointed. That lawyer was Downey, who had\\n\\nmore experience than Scofield but whose capital experience was limited to one\\n\\ncase that had not gone to trial. The two attorneys had agreed that Downey would\\n\\nbe responsible for jury selection, but Scofield eventually took over during the voir\\n\\ndire. Per their agreement, Scofield was responsible for the guilt-phase trial,\\n\\nalthough Downey participated by interviewing Jenkins several times. Scofield also\\n\\nmet with Jenkins around a dozen times. Jenkins had told him about his background\\n\\n\\n 7\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 8 of 73\\n\\n\\nand childhood, including the fact that he had had a difficult childhood and had\\n\\nlived on the streets since age 11.\\n\\n Throughout the weeklong trial, Scofield pursued an outright acquittal. In\\n\\nparticular, he challenged Harris’s identification of Jenkins as the man who drove\\n\\naway with Hogeland, and he challenged the medical examiner’s opinions about the\\n\\ntime and cause of Hogeland’s death. In his closing argument, Scofield took a wide-\\n\\nranging approach to undermining the State’s wholly circumstantial case against\\n\\nJenkins. He attacked the State’s timeline of the crime, questioning whether Jenkins\\n\\nin his intoxicated state would have had the time and the ability to leave Nicholas’s\\n\\nhouse, steal the Mazda, and abduct and kill Hogeland. He reminded the jury of the\\n\\nweakness of Harris’s identification of Jenkins and the Coes’ identification of the\\n\\nwoman in the car. He pointed out that no physical evidence definitively linked\\n\\nJenkins to the Mazda, to Hogeland’s body, or to her jewelry, which was never\\n\\nrecovered. And he argued that the evidence provided no plausible motive for\\n\\nJenkins to have kidnapped, robbed, or killed Hogeland. In sum, he urged the jury\\n\\nto find reasonable doubt and acquit Jenkins. At the very least, he argued, they\\n\\nshould find that the State failed to prove Jenkins’s intent to rob and kidnap\\n\\nHogeland, which were necessary to convict Jenkins of capital murder. He further\\n\\nargued that, if the jury thought Jenkins killed Hogeland as a crime of passion, they\\n\\ncould convict him only of manslaughter. The jury deliberated for three hours and\\n\\n\\n 8\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 9 of 73\\n\\n\\nfifteen minutes before convicting Jenkins of capital murder, Ala. Code § 13A-5-\\n\\n40(a)(1) & (2), i.e., murder committed in the course of kidnapping and robbery.\\n\\n The penalty phase took place that same afternoon. A week before trial,\\n\\nScofield and Downey had discussed the penalty phase, and Scofield understood\\n\\nthat Downey would be responsible for handling the penalty phase. Nonetheless, the\\n\\nwitness who was present to testify on Jenkins’s behalf during the penalty phase\\n\\nwas better acquainted with Scofield, so Downey suggested that afternoon that\\n\\nScofield conduct the direct examination. Scofield urged Downey to interview the\\n\\nwitness quickly and prepare to examine him, and Downey did so.\\n\\n Downey handled the penalty presentation to the jury. The State did not\\n\\npresent any additional aggravation evidence, resting on the evidence presented\\n\\nduring the guilt phase. One witness, Lonnie Seal, testified on Jenkins’s behalf in\\n\\norder, in Downey’s words, to “reveal another side of Mark Jenkins to you that you\\n\\ndon’t know anything about.” Seal had met Jenkins in Fontana, California, in 1988\\n\\nwhen they both worked at the same garage. Seal and his wife became friends with\\n\\nJenkins, and he visited in their home. When the Seals decided to move to Alabama\\n\\na few months later, Jenkins agreed to help with their move and to live with the\\n\\nSeals in Alabama for a time. It took Lonnie Seal several weeks to find a job, but\\n\\nJenkins found a job in two days and contributed toward the family’s rent and\\n\\n\\n\\n\\n 9\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 10 of 73\\n\\n\\ngroceries. Jenkins continued to visit the Seals after he moved into his own home,\\n\\nand he would always bring a small gift for and play with the Seals’ infant son.\\n\\n Downey’s closing argument to the jury requested a sentence of life without\\n\\nparole. Downey reminded the jurors that they had agreed during voir dire that\\n\\ndeath was not always the appropriate punishment for homicide. Quoting the Bible,2\\n\\nhe reminded the jury that the Old Testament permitted capital punishment only\\n\\nwhen supported by the testimony of two or three eyewitnesses, whereas in this case\\n\\nthere had been only circumstantial evidence. He asserted from Seal’s testimony\\n\\nthat “Mark had a side of him that was generous, that was giving, and that was\\n\\nkind,” and argued that the deterrent and retributive purposes of capital punishment\\n\\nwould not be served in this case.\\n\\n The jury deliberated for 50 minutes before returning its nonbinding\\n\\nrecommendation of death by a vote of 10 to 2. See Ala. Code § 13A-5-46(e) (1981)\\n\\n(amended 2017 to make jury’s verdict binding). After a separate hearing, the court\\n\\nfound the two statutory aggravating factors for which the State had argued: that the\\n\\nmurder was committed during the commission of a robbery, and that it was\\n\\ncommitted during the commission of a kidnapping. See Ala. Code § 13A-5-49(4).\\n\\n\\n\\n\\n2\\n “If anyone kills a person, the murderer shall be put to death on the evidence of witnesses. But\\nno person shall be put to death on the testimony of one witness.” Numbers 35:30 (ESV). Downey\\nalso paraphrased Deuteronomy 17:6 as “On the evidence of two witnesses or three witnesses, he\\nis [sic] shall be put to death. He shall not be put to death on the evidence of one witness.”\\n 10\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 11 of 73\\n\\n\\nThe court also found two statutory mitigating factors: that Jenkins had no\\n\\nsignificant history of prior criminal activity, and his age at the time of the crime\\n\\n(21). See id. § 13A-5-51(1), (7). The court considered but rejected the statutory\\n\\nmitigation that, due to his consumption of alcohol, Jenkins’s capacity to appreciate\\n\\nthe criminality of his conduct or to conform his conduct to the requirements of law\\n\\nwas substantially impaired. See id. § 13A-5-51(6). The court found that the\\n\\naggravating circumstances outweighed the mitigating circumstances and, in\\n\\naccordance with the jury’s recommendation, imposed a sentence of death.\\n\\n C. The State and Federal Post-Conviction Proceedings\\n\\n On direct appeal, the Alabama Court of Criminal Appeals affirmed Jenkins’s\\n\\nconviction and death sentence. Jenkins v. State, 627 So. 2d 1034 (Ala. Ct. Crim.\\n\\nApp. 1992), aff’d, 627 So. 2d 1054 (Ala. 1993), cert. denied, 511 U.S. 1012\\n\\n(1994). In 1995, Jenkins filed his first petition for state postconviction relief, Ala.\\n\\nR. Crim. P. 32,3 alleging, among many other things, that he received ineffective\\n\\nassistance of counsel during his penalty phase. The Rule 32 trial court conducted a\\n\\nthree-day hearing at which evidence was presented about the performance of\\n\\nJenkins’s attorneys.\\n\\n\\n\\n\\n3\\n “Subject to the limitations of Rule 32.2, any defendant who has been convicted of a criminal\\noffense may institute a proceeding in the court of original conviction to secure appropriate relief\\non the ground that . . . [t]he constitution of the United States or of the State of Alabama requires\\na new trial, a new sentence proceeding, or other relief. . . .” Ala. R. Crim. P. 32.1.\\n 11\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 12 of 73\\n\\n\\n Scofield testified about the representation that he and Downey provided. In\\n\\nrelevant part, Scofield asserted that he and Downey had “decided that [developing\\n\\nmitigation for the penalty phase] would be something that Stan [Downey] would\\n\\nhandle.” Downey later told Scofield that he hadn’t done any such investigation, but\\n\\nScofield admitted that, “[a]part from what he has told me, I don’t know what he\\n\\nhas done.” Scofield testified that Downey “didn’t appear prepared” for the penalty\\n\\nphase, based on Downey’s last-minute suggestion that Scofield examine Seal since\\n\\nhe had talked with him before.\\n\\n In preparation for the penalty phase, Scofield himself did not interview any\\n\\nmember of Jenkins’s family, but he did not know whether Downey interviewed\\n\\nany family members. Scofield did interview Jenkins’s landlord and some other\\n\\nacquaintances but did not do so specifically for the purpose of developing\\n\\nmitigation. He had met with Jenkins around a dozen times, and Jenkins had told\\n\\nhim his life story including his experience of abuse as a child. In hindsight,\\n\\nScofield said he had a “lack of appreciation of mitigating circumstances.” He\\n\\nrepeatedly denied in his Rule 32 testimony that his omissions during the penalty\\n\\nphase were strategic: “I didn’t understand what I was doing. . . . I just didn’t\\n\\nunderstand what I needed to do.”\\n\\n For reasons that remain unclear, Downey—who is since deceased—did not\\n\\ntestify or submit an affidavit in the Rule 32 proceeding. Scofield answered,\\n\\n\\n 12\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 13 of 73\\n\\n\\n“Obviously not,” when asked whether he had personal knowledge of everything\\n\\nDowney may or may not have done to prepare for the penalty phase. The record of\\n\\nthe extent of Downey’s pretrial preparation is thus limited to Downey’s detailed\\n\\nfee declaration, which he submitted to the State of Alabama in documentation of\\n\\nthe compensation due him as appointed counsel. See Ala. Code § 15-12-21. In it,\\n\\nDowney reported that he spent 71 hours4 on out-of-court preparation for trial. The\\n\\nreport details 35 hours of interviews with Jenkins in 18 separate visits to the jail\\n\\nover the course of 17 months, as well as one substantive conversation with\\n\\nJenkins’s grandmother.5\\n\\n The Rule 32 trial court also heard testimony from four family members and\\n\\none friend of Jenkins’s. The family members presented for the first time a stark\\n\\ndepiction of Jenkins as an unloved and abused child in California. Jenkins’s older\\n\\nbrother Michael Jenkins testified that all of his siblings have different fathers and\\n\\nthat his parents often used drugs and fought. Michael testified that Jenkins was\\n\\nbeaten frequently as a child by his stepfather, if not every day then every other day\\n\\n\\n\\n4\\n In Jenkins’s state postconviction appeal, the Alabama Court of Criminal Appeals erroneously\\nstated that Downey’s fee declaration reports 171 hours of pretrial preparation. See Jenkins, 972\\nSo. 2d at 137. We discuss the effect of this discrepancy infra note 7.\\n5\\n The state appellate court also noted that Downey had moved for a continuance in October 1989\\nso that he could conduct “[f]urther discovery and investigation (including a possible trip to\\nCalifornia).” Id. We note that Scofield had orally suggested to the trial court in September 1989\\nthat more time might be needed with respect to “evidence about the defendant” and witnesses in\\nCalifornia, as well as possible psychological evaluation.\\n 13\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 14 of 73\\n\\n\\nor every three days. Jenkins was beaten all over his body with a variety of\\n\\nimplements. Michael was beaten too. Jenkins’s stepfather would lock Jenkins in\\n\\nhis bedroom for a couple of hours at a time, sometimes without dinner. Some of\\n\\nthe beatings were prompted by Jenkins’s bedwetting or soiling. Michael testified\\n\\nthat Jenkins’s stepfather sometimes made Jenkins wear his soiled underwear on his\\n\\nhead and even forced him to eat his own feces a few times. Michael testified that\\n\\nJenkins’s bladder and bowel control problems began after a camping trip he took\\n\\nwith his stepfather and stepgrandfather.\\n\\n Michael testified that, as children, he and Jenkins were made to work on car\\n\\nengines late into the night and to shovel horse manure daily. The family moved\\n\\nfrequently and the children attended many different schools as a result. Michael\\n\\ntestified that he was Jenkins’s only friend as a child and that Jenkins did poorly in\\n\\nschool. Jenkins began running away from home as a young teenager and lived with\\n\\nvarious other family members or on the streets. Michael said that Jenkins was a\\n\\ngood brother to him and had asked him to testify at his trial, but he never heard\\n\\nfrom his attorneys or found out when the trial was. Michael admitted to having\\n\\ncognitive and memory problems due to a head injury years earlier.\\n\\n Jenkins’s cousin Tammy Pitts testified that, as an infant, Jenkins’s mother\\n\\nbeat and neglected him, leaving him dirty and in soiled diapers. His stepfather also\\n\\nbeat Jenkins daily until he left home at age 13, sometimes bruising him so badly\\n\\n\\n 14\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 15 of 73\\n\\n\\nthat he was laid up in bed for days or weeks. Jenkins had problems with\\n\\nbedwetting and soiling, and his stepfather made him wear soiled clothes to school.\\n\\nJenkins would be locked in his filthy bedroom around the clock except to attend\\n\\nschool and do chores. He was usually not allowed to eat dinner with the rest of the\\n\\nfamily and subsisted on scraps and dog food. The entire family belittled Jenkins\\n\\nbecause his biological father was Hispanic and his complexion was therefore\\n\\ndarker. Jenkins’s parents drank a lot and used drugs every day. Pitts knew these\\n\\nthings because she had daily contact with the family and lived with them off and\\n\\non, and Jenkins lived with her for a time after he ran away from home. No one had\\n\\ncontacted her about testifying at Jenkins’s trial.\\n\\n Jenkins’s second cousin Betty DeLavega testified that her relationship with\\n\\nJenkins was like that of an aunt. She lived with the Jenkins family for five months\\n\\nwhen Jenkins was 10 or 11. She saw Jenkins’s stepfather slap Jenkins “[q]uite a\\n\\nfew times,” and once, he made Jenkins eat his own feces. DeLavega testified that\\n\\nJenkins’s parents used drugs nearly every day and withheld affection from Jenkins\\n\\nand Michael. No one contacted her about Jenkins’s trial; Jenkins’s grandmother\\n\\nhad asked her to testify in the Rule 32 hearing about any abuse she had witnessed.\\n\\n Jenkins’s grandmother Doris Wagoner testified that while Jenkins’s mother\\n\\nwas pregnant with him, her husband was in prison and she drank and took drugs.\\n\\nWhen Jenkins was born, his mother initially put him up for adoption, then changed\\n\\n\\n 15\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 16 of 73\\n\\n\\nher mind weeks later. Jenkins was a lethargic infant and his mother cared more\\n\\nabout partying than about her children. Wagoner testified that she heard of\\n\\nJenkins’s mistreatment as a child but didn’t see it. She could tell from the way that\\n\\nJenkins feared his stepfather that he was being abused, and she knew he slept in a\\n\\nfilthy bed. After Jenkins was arrested for murder, Wagoner spoke with Scofield,\\n\\nwho asked for money relating to his representation of Jenkins. She did not\\n\\nremember whether she talked with Scofield about Jenkins’s background. She could\\n\\nnot remember why she didn’t testify at the trial; she might have been ill, but she\\n\\nwas probably busy.\\n\\n Sherry Seal, the wife of Lonnie Seal, who had testified in the penalty phase\\n\\nof the trial, testified that Jenkins was respectful and generous and that she trusted\\n\\nhim with her children. She related how he supported her family after moving to\\n\\nAlabama, giving them “pretty much” his entire paycheck. She said she would still\\n\\nfeel safe around Jenkins. Two jailers also testified that Jenkins was a respectful,\\n\\ncourteous, model inmate who never complained or caused any problems.\\n\\n Jenkins also introduced medical, school, and juvenile court records from his\\n\\nchildhood. The medical records document his premature birth and his mother’s\\n\\nintention to give him up for adoption. The school records reflect a struggling\\n\\nstudent who was repeatedly promoted to the next grade because of his age, despite\\n\\nserious academic difficulties, chronic truancy, and frequent school changes. The\\n\\n\\n 16\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 17 of 73\\n\\n\\njuvenile court records document Jenkins’s placement in juvenile hall starting at age\\n\\n14, following repeated acts of grand theft auto and running away.\\n\\n The Rule 32 hearing concluded with the testimony of two psychologists\\n\\nabout Jenkins’s history of psychological trauma stemming from his childhood\\n\\nabuse. The defense expert, Dr. David Lisak, had met with Jenkins three times and\\n\\ninterviewed a dozen of his family members, family friends, and associates. Dr.\\n\\nLisak testified that the pervasive abuse and neglect in Jenkins’s childhood impaired\\n\\nhis ability to cope with trauma as an adult. He opined that Jenkins’s childhood\\n\\nabuse adversely affected his cognitive and emotional development and other\\n\\nabilities. He described how Jenkins had been neglected as an infant and sexually\\n\\nabused by his grandfather at age four, which led to a lifetime of bladder and bowel\\n\\ndysfunction. He also discussed the hypothesis that children who are abused are at\\n\\ngreater risk for perpetrating violence later in life, but he acknowledged there would\\n\\nhave been other causative factors in Jenkins’s case. Dr. Lisak described Jenkins’s\\n\\nintellectual capacity as “somewhat borderline,” though he did not perform any\\n\\ntesting of his own.\\n\\n The State’s expert, Dr. Karl Kirkland, had met with Jenkins for several hours\\n\\nand administered several tests. On the Wechsler Adult Intelligence Scale, Jenkins\\n\\n“scored in the range of borderline intellectual functioning which is between mild\\n\\nmental retardation and low average intellectual functioning” with an overall IQ of\\n\\n\\n 17\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 18 of 73\\n\\n\\n76. Dr. Kirkland agreed that Jenkins had been “reared in a chaotic dysfunctional\\n\\nfamily” and generally agreed with Dr. Lisak’s assessment of the negative effects of\\n\\nchildhood trauma.\\n\\n The court considered all of this evidence before denying Rule 32 relief in\\n\\nfull. As relevant here, it found that some of the evidence about the severity of\\n\\nabuse that Jenkins suffered as a child was not credible. It further found that Jenkins\\n\\nsuffered no prejudice from any deficient performance by his penalty-phase counsel\\n\\nbecause the aggravating circumstances outweighed any additional mitigating\\n\\nevidence Downey might have adduced.\\n\\n The Alabama Court of Criminal Appeals affirmed the denial of Rule 32\\n\\nrelief. Jenkins v. State, 972 So. 2d 111 (Ala. Ct. Crim. App. 2004). In relevant part,\\n\\nthe appeals court rejected Jenkins’s ineffective assistance claim, finding that\\n\\ncounsel reasonably chose to pursue a penalty-phase strategy of residual doubt and\\n\\ngood character. Id. at 147. It explained that the new evidence was self-\\n\\ncontradictory in that it asserted that childhood abuse would have made Jenkins a\\n\\nviolent adult yet also asserted that Jenkins was meek and mild. Id. The appeals\\n\\ncourt also agreed with the lower court’s conclusion that Jenkins suffered no\\n\\nprejudice from any error by his counsel because the aggravating circumstances still\\n\\noutweighed any new mitigating ones. Id. at 148. The court also noted that evidence\\n\\n\\n\\n\\n 18\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 19 of 73\\n\\n\\nabout Jenkins’s good behavior in jail awaiting trial was “minimally mitigating” and\\n\\nwould not have affected the sentence. Id. at 149.\\n\\n Because the Supreme Court’s decision in Atkins v. Virginia6 had issued\\n\\nduring the pendency of the Rule 32 appeal, the Alabama appeals court ordered\\n\\nsupplemental briefing on the possible impact of Atkins upon Jenkins’s sentence.\\n\\nJenkins asked the court to stay the proceedings and conduct an evidentiary hearing\\n\\nafter the Alabama legislature enacted a statute about executing the mentally\\n\\nretarded or, in the alternative, to vacate his sentence and remand after the Alabama\\n\\nlegislature acted. He later asserted that he is mentally retarded and therefore\\n\\nineligible for the death penalty. The appeals court proceeded to reject Jenkins’s\\n\\nAtkins claim on the merits because the record established that his IQ was 76 and\\n\\nthat he maintained relationships and employment. Id. at 155. The Alabama\\n\\nSupreme Court summarily affirmed, in relevant part. Ex parte Jenkins, 972 So. 2d\\n\\n159, 165 (Ala. 2005), remanded to 972 So. 2d 165 (Ala. Ct. Crim. App. 2005)\\n\\n(juror misconduct claim), cert. denied, 552 U.S. 1167 (2008).\\n\\n In 2008, Jenkins filed his first petition for a writ of habeas corpus in federal\\n\\ncourt pursuant to 28 U.S.C. § 2254. The federal proceedings were stayed for\\n\\nseveral years while Jenkins pursued additional, unrelated Rule 32 relief in the state\\n\\n\\n6\\n 536 U.S. 304, 321 (2002) (holding that the execution of mentally retarded offenders violates the\\nEighth Amendment). The rule announced in Atkins is retroactive to cases on collateral review. In\\nre Holladay, 331 F.3d 1169, 1172–73 (11th Cir. 2003).\\n 19\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 20 of 73\\n\\n\\ncourts, ultimately unsuccessfully. In the district court, Jenkins moved for an\\n\\nevidentiary hearing on his Atkins claim. After reviewing the state court record,\\n\\nwith a focus on the Rule 32 testimony of Drs. Lisak and Kirkland, the district court\\n\\ndenied the motion in 2015. Noting that our Circuit has accepted the State of\\n\\nAlabama’s definition of mental retardation following Atkins, the court found that\\n\\nJenkins is not intellectually disabled under that standard. In a 347-page order, the\\n\\ndistrict court also denied relief on all of Jenkins’s claims including ineffective\\n\\nassistance of counsel during the penalty phase and affirmed as reasonable the state\\n\\ncourt’s conclusion that Jenkins failed to establish prejudice from any error by his\\n\\ncounsel. The district court later denied reconsideration and a certificate of\\n\\nappealability.\\n\\n Jenkins now appeals. Our Court granted Jenkins a certificate of\\n\\nappealability, 28 U.S.C. § 2253(c), on two issues:\\n\\n 1) Whether the district court erred in denying Appellant’s claim\\n that his trial counsel rendered ineffective assistance of counsel in the\\n investigation and presentation of mitigating evidence during the\\n penalty phase of Appellant’s 1991 trial; and\\n\\n 2) Whether the district court erred in denying Appellant’s claim\\n that he is intellectually disabled and ineligible for the death penalty\\n under Atkins v. Virginia, 536 U.S. 304 (2002).\\n\\n II. STANDARDS OF REVIEW\\n\\n We review the denial of a petition for a writ of habeas corpus de novo.\\n\\nMorrow v. Warden, Ga. Diagnostic Prison, 886 F.3d 1138, 1146 (11th Cir. 2018).\\n\\n 20\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 21 of 73\\n\\n\\nThe Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) instructs that\\n\\na writ of habeas corpus shall not be granted to a state prisoner unless the state\\n\\nadjudication of a claim “resulted in a decision that was contrary to, or involved an\\n\\nunreasonable application of, clearly established Federal law, as determined by the\\n\\nSupreme Court of the United States.” 28 U.S.C. § 2254(d)(1). “[C]learly\\n\\nestablished Federal law” means “the holdings, as opposed to the dicta, of [the\\n\\nSupreme] Court’s decisions as of the time of the relevant state-court decision.”\\n\\nWilliams v. Taylor, 529 U.S. 362, 412 (2000). “[W]hen the last state court to\\n\\ndecide a prisoner’s federal claim explains its decision on the merits in a reasoned\\n\\nopinion . . . a federal habeas court simply reviews the specific reasons given by the\\n\\nstate court and defers to those reasons if they are reasonable.” Wilson v. Sellers,\\n\\n138 S. Ct. 1188, 1192 (2018). A state court’s decision is reasonable “so long as\\n\\n‘fairminded jurists could disagree’ on the correctness of the state court’s decision.”\\n\\nHarrington v. Richter, 562 U.S. 86, 101 (2011) (quoting Yarborough v. Alvarado,\\n\\n541 U.S. 652, 664 (2004)). “[E]ven a strong case for relief does not mean the state\\n\\ncourt’s contrary conclusion was unreasonable.” Id. at 102. Rather, a prisoner must\\n\\nshow that the state court’s ruling “was so lacking in justification that there was an\\n\\nerror well understood and comprehended in existing law beyond any possibility for\\n\\nfairminded disagreement.” Id. at 103. That ruling must have been “‘objectively\\n\\nunreasonable,’ not merely wrong; even ‘clear error’ will not suffice.” White v.\\n\\n\\n 21\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 22 of 73\\n\\n\\nWoodall, 572 U.S. 415, 419 (2014) (quoting Lockyer v. Andrade, 538 U.S. 63, 75\\n\\n(2003)).\\n\\n In addition, “a determination of a factual issue made by a State court shall be\\n\\npresumed to be correct,” and the prisoner bears “the burden of rebutting the\\n\\npresumption of correctness by clear and convincing evidence.” 28 U.S.C.\\n\\n§ 2254(e)(1). AEDPA further instructs that a writ of habeas corpus may not be\\n\\ngranted unless the state court adjudication “resulted in a decision that was based on\\n\\nan unreasonable determination of the facts in light of the evidence presented in the\\n\\nState court proceeding.” Id. § 2254(d)(2). When a federal habeas court makes a\\n\\nfactual finding as part of its habeas determination, we review for clear error, which\\n\\nis a “highly deferential” standard of review. Thomas v. Allen, 607 F.3d 749, 752\\n\\n(11th Cir. 2010) (quoting Holton v. City of Thomasville Sch. Dist., 425 F.3d 1325,\\n\\n1350 (11th Cir. 2005)). “A finding is ‘clearly erroneous’ when although there is\\n\\nevidence to support it, the reviewing court on the entire evidence is left with the\\n\\ndefinite and firm conviction that a mistake has been committed.” United States v.\\n\\nU.S. Gypsum Co., 333 U.S. 364, 395 (1948).\\n\\n III. DISCUSSION\\n\\n A. Ineffective Assistance of Counsel\\n\\n A prisoner alleging that he received ineffective assistance of counsel in\\n\\nviolation of the Sixth Amendment must establish two elements. Strickland v.\\n\\n\\n 22\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 23 of 73\\n\\n\\nWashington, 466 U.S. 668, 687 (1984). “First, the defendant must show that\\n\\ncounsel’s performance was deficient.” Id. In other words, he must show “that\\n\\ncounsel made errors so serious that counsel was not functioning as the ‘counsel’\\n\\nguaranteed the defendant by the Sixth Amendment.” Id. “Second, the defendant\\n\\nmust show that the deficient performance prejudiced the defense.” Id. We discuss\\n\\neach of these elements in turn.\\n\\n 1. Deficient Performance\\n\\n A “doubly” deferential lens sits before our eyes as we review the\\n\\nperformance of Jenkins’s counsel during the penalty phase. Richter, 562 U.S. at\\n\\n105 (quoting Knowles v. Mirzayance, 556 U.S. 111, 123 (2009)). This redoubled\\n\\ndeference arises out of the tandem application of AEDPA’s “highly deferential”\\n\\nreview of a state court’s decision with Strickland’s “highly deferential” review of\\n\\nan attorney’s performance. Id. (quoting Lindh v. Murphy, 521 U.S. 320, 333 n.7\\n\\n(1997), and Strickland, 466 U.S. at 689). Under the Strickland standard, “counsel\\n\\nis strongly presumed to have rendered adequate assistance and made all significant\\n\\ndecisions in the exercise of reasonable professional judgment.” Strickland, 466\\n\\nU.S. at 690. And under AEPDA, we must deny habeas relief if “there is any\\n\\nreasonable argument that counsel satisfied Strickland’s deferential standard.”\\n\\nRichter, 562 U.S. at 105.\\n\\n\\n\\n\\n 23\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 24 of 73\\n\\n\\n Jenkins argues that his counsel performed deficiently by failing to\\n\\n“investigate compelling mitigating evidence or develop a penalty-phase theory.”\\n\\nThe state court disagreed. It quoted decisions of our Court that hold that “there is\\n\\nno per se rule that evidence of a criminal defendant’s troubled childhood must\\n\\nalways be presented as mitigating evidence.” Jenkins, 972 So. 2d at 144 (quoting\\n\\nMarek v. Singletary, 62 F.3d 1295, 1300 (11th Cir. 1995)). Noting that “Downey\\n\\nwas in charge of the penalty phase” but “we do not have the benefit of Downey’s\\n\\ntestimony as to what occurred and why,” it reviewed the trial record to find that\\n\\n“counsel argued residual doubt and Jenkins’s good character at the penalty phase.”\\n\\nId. at 145–46. It concluded that “Downey’s decision to concentrate on reasonable\\n\\ndoubt and to portray Jenkins as a good person was reasonable under the\\n\\ncircumstances.” Id. at 147.\\n\\n Jenkins insists, however, that counsel’s failure to develop mitigating\\n\\nevidence about his childhood abuse was “not the product of strategy, but rather\\n\\narose from ingnorance and inexeperiance [sic].” The problem for Jenkins is that,\\n\\nlike the state courts, this Court also does not know what Downey did or why. The\\n\\nrecord is silent as to his thoughts and intentions as he prepared for the penalty\\n\\nphase. Given that Jenkins bore the burden of overcoming the strong presumption of\\n\\ncompetence and of proving Downey’s deficient performance, we are perplexed\\n\\nwhy Jenkins’s new counsel did not have Downey testify or submit an affidavit in\\n\\n\\n 24\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 25 of 73\\n\\n\\nthe Rule 32 proceeding. See Strickland, 466 U.S. at 687 (“The defendant must\\n\\nshow that counsel’s performance was deficient.”). Nor did Jenkins testify to the\\n\\nsubstance of his conversations with his counsel. “The reasonableness of a trial\\n\\ncounsel’s acts, including lack of investigation or excluding character witnesses\\n\\nfrom the sentencing phase, depends ‘critically’ upon what information the client\\n\\ncommunicated to counsel.” Chandler v. United States, 218 F.3d 1305, 1324 (11th\\n\\nCir. 2000) (en banc) (quoting Strickland, 466 U.S. at 691). There is simply no\\n\\nrecord evidence to support Jenkins’s central assertion that Downey made no\\n\\ntactical choices in his preparation for the penalty phase.\\n\\n Jenkins instead relies on Scofield’s Rule 32 testimony about how he and\\n\\nDowney prepared for the trial and sentencing. We acknowledge that Scofield\\n\\nrepeatedly testified that he did not really know what he was doing and that he\\n\\nfailed to make strategic choices with respect to mitigating circumstances and\\n\\nsentencing. But Scofield was equally clear that Downey—the more experienced\\n\\nlawyer—was in fact responsible for the penalty phase. And Scofield also\\n\\nrepeatedly testified that he did not know what Downey may or may not have done\\n\\nto prepare. Furthermore, his lack of knowledge about Downey’s preparation was\\n\\nunderstandable. He and Downey were not colleagues who shared an office and\\n\\nmight reasonably be expected to know what the other was doing much of the time.\\n\\nScofield and Downey did not even work in the same city.\\n\\n\\n 25\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 26 of 73\\n\\n\\n This case is unlike those in which we have found deficient performance\\n\\nwhere the record showed that counsel investigated insufficiently because he\\n\\n“waited until the eleventh hour” to begin preparing for the penalty phase while\\n\\nfaced with “overwhelming evidence of guilt.” See, e.g., Johnson v. Sec’y, D.O.C.,\\n\\n643 F.3d 907, 932 (11th Cir. 2011). The limited record that Jenkins developed\\n\\nestablishes that Downey spent at least 71 hours preparing for trial, spread out over\\n\\n17 months. The state court interpreted this sparse record as reflecting a penalty-\\n\\nphase strategy of residual doubt, the logical extension of Downey’s guilt-phase\\n\\ninvestigation of the evidence and Scofield’s pursuit at trial of an acquittal based on\\n\\nreasonable doubt. See Jenkins, 972 So. 2d at 147. In such a situation, “[a] lawyer’s\\n\\ntime and effort in preparing to defend his client in the guilt phase of a capital case\\n\\ncontinues to count at the sentencing phase.” Tarver v. Hopper, 169 F.3d 710, 715\\n\\n(11th Cir. 1999).\\n\\n We decline to dictate how many hours or how many sessions an attorney\\n\\nmust spend interviewing his counsel in order to provide effective assistance. “No\\n\\nabsolute rules dictate what is reasonable performance for lawyers.” Chandler, 218\\n\\nF.3d at 1317 (citing Strickland, 466 U.S. at 688–89). But even this limited record\\n\\nundermines Jenkins’s assertion that Downey prepared inadequately. Downey’s 71\\n\\nhours of preparation included 35 hours of jailhouse conversations with Jenkins\\n\\n\\n\\n\\n 26\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 27 of 73\\n\\n\\nduring 18 separate meetings.7 According to Downey’s notes, Scofield was present\\n\\nfor only five of those meetings (12 hours’ worth), leaving 23 hours over 13\\n\\nmeetings between Jenkins and Downey alone. Downey’s notes reveal little about\\n\\nthe content of those conversations beyond broad notations such as “evidence,\\n\\nwitnesses, general status,” and discussions with Jenkins of the evidence and\\n\\ntestimony against him were surely relevant to Downey’s penalty-phase strategy of\\n\\nresidual doubt.\\n\\n But the record does not reveal that Downey knew or should have known that\\n\\nJenkins’s childhood merited investigation beyond that which he may have done.\\n\\nThe record shows that Downey developed enough of a rapport with Jenkins that,\\n\\naccording to his fee declaration, he “g[ave a] Christmas gift [and] discuss[ed] how\\n\\nmuch to tell his family.” But how much Downey knew about Jenkins’s family\\n\\nbackground can only be guessed from Scofield’s Rule 32 hearing testimony.\\n\\nScofield testified that he discussed Jenkins’s background during his dozen or so\\n\\njailhouse meetings with him, and that Downey was present for some of those\\n\\nmeetings. Scofield—and possibly Downey—learned that Jenkins experienced\\n\\nabuse as a child, that his stepfather was brutal, and that he lived on the streets and\\n\\n\\n\\n\\n7\\n We note that the state court clearly erroneously stated that Downey’s fee declaration\\ndocumented 171 hours of pretrial preparation. See Jenkins, 972 So. 2d at 137. However, the state\\ncourt accurately stated that “he spent over 25 hours talking with Jenkins in more than 10 visits to\\nthe jail.” Id. That statement was not an unreasonable determination of the facts.\\n 27\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 28 of 73\\n\\n\\nin juvenile institutions because of his brutal home life. It is against this backdrop\\n\\nthat Downey’s presumptive “decision not to investigate must be directly assessed\\n\\nfor reasonableness in all the circumstances, applying a heavy measure of deference\\n\\nto counsel’s judgments.” Strickland, 466 U.S. at 691.\\n\\n That assessment depends a great deal on what Jenkins told his counsel about\\n\\nhis background. See id. But we do not know whether Downey was present when\\n\\nScofield learned about Jenkins’s childhood abuse, nor what else Downey may have\\n\\nlearned during his 13 solo discussions with Jenkins. Therefore, “given the absence\\n\\nof evidence in the record, we must assume counsel carried out his professional\\n\\nresponsibility and discussed mitigation with his client.” Chandler, 218 F.3d at\\n\\n1324. We will not “turn that presumption on its head by giving [Jenkins] the\\n\\nbenefit of the doubt where it is unclear what [Downey] did or did not do.” Williams\\n\\nv. Head, 185 F.3d 1223, 1235 (11th Cir. 1999). We will not assume that Jenkins\\n\\ntold Downey more than he told Scofield, and Scofield denied remembering that\\n\\nJenkins told him about the severity or frequency of his stepfather’s physical abuse.\\n\\nHe denied remembering Jenkins telling him about being beaten daily, being locked\\n\\nin his room 24/7, or being forced to eat his feces. And there is no suggestion in the\\n\\nrecord that counsel knew or had reason to investigate the allegation of sexual abuse\\n\\nby his stepgrandfather. The record simply does not reflect that Jenkins told either\\n\\nof his lawyers about the severe abuse that some of the Rule 32 witnesses later\\n\\n\\n 28\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 29 of 73\\n\\n\\ndescribed. We will not fault those lawyers for failing to present as mitigation this\\n\\nlimited evidence that they may have reasonably believed to be unremarkable. See\\n\\nStrickland, 466 U.S. at 691.\\n\\n To be sure, counsel’s failure “to discover all reasonably available mitigating\\n\\nevidence” may constitute deficient performance in some cases. Wiggins v. Smith,\\n\\n539 U.S. 510, 524 (2003) (quoting ABA Guidelines for the Appointment and\\n\\nPerformance of Counsel in Death Penalty Cases § 11.4.1(C) (1989)). Our Court\\n\\nhas identified deficient performance in several instances where counsel had good\\n\\nreason to investigate the defendant’s background but failed to do so. See Morrow,\\n\\n886 F.3d at 1147 (collecting cases). But we assess “a particular decision not to\\n\\ninvestigate . . . for reasonableness in all the circumstances, applying a heavy\\n\\nmeasure of deference to counsel’s judgments.” Strickland, 466 U.S. at 691. Thus,\\n\\nwe have rejected a hard rule that counsel’s “strategic decisions can be considered\\n\\nreasonable only if they are preceded by a ‘thorough investigation.’” Williams, 185\\n\\nF.3d at 1237 (quoting Rogers v. Zant, 13 F.3d 384, 387 (11th Cir. 1994)). Because\\n\\nwe simply do not know why Downey chose not to investigate Jenkins’s childhood\\n\\nmore thoroughly, Downey is entitled to the presumption that this strategic decision\\n\\nwas reasonable.\\n\\n And though we have also rejected “a per se rule of ineffective assistance\\n\\nwhere counsel does not consult family members,” Holladay v. Haley, 209 F.3d\\n\\n\\n 29\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 30 of 73\\n\\n\\n1243, 1252 (11th Cir. 2000), we note that Downey did not fail altogether to consult\\n\\nfamily members. Downey had at least one substantive conversation with Jenkins’s\\n\\ngrandmother, and we know from her Rule 32 hearing testimony that she never\\n\\nwitnessed Jenkins’s childhood abuse. Given the silence of the record and the strong\\n\\npresumption of competence to which Downey is entitled, we decline to fault him\\n\\nfor not doing more when Jenkins and his grandmother did not and could not,\\n\\nrespectively, provide a basis for doing so. Cf. United States v. Pease, 240 F.3d 938,\\n\\n941–92 (11th Cir. 2001) (“we cannot say as a matter of law that reliance on a\\n\\nclient’s statements is per se deficient performance”). Nor did Jenkins’s counsel\\n\\nhave an absolute duty to put on abusive-childhood evidence, even if they knew of\\n\\nit. See Chandler, 218 F.3d at 1319 (“Counsel is not required to present every\\n\\nnonfrivolous defense; nor is counsel required to present all mitigation evidence,\\n\\neven if the additional mitigation evidence would not have been incompatible with\\n\\ncounsel’s strategy.”); Waters v. Thomas, 46 F.3d 1506, 1511 (11th Cir. 1995) (en\\n\\nbanc) (“Our decisions are inconsistent with any notion that counsel must present\\n\\nall available mitigating circumstance evidence.”).\\n\\n Instead, we affirm as a reasonable application of Strickland the state court’s\\n\\nconclusion that Downey’s strategic decision to emphasize residual doubt was\\n\\nreasonable. See Jenkins, 972 So. 2d at 147. As we have noted, “[a] lawyer’s time\\n\\nand effort in preparing to defend his client in the guilt phase of a capital case\\n\\n\\n 30\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 31 of 73\\n\\n\\ncontinues to count at the sentencing phase.” Tarver, 169 F.3d at 715. This\\n\\nreasoning is especially applicable where, as here, counsel pursued the reasonable\\n\\nstrategy of arguing for acquittal at the guilt phase and for life at the sentencing\\n\\nphase by highlighting doubt about the evidence. “[R]esidual doubt is perhaps the\\n\\nmost effective strategy to employ at sentencing.” Chandler, 218 F.3d at 1320 n.28.\\n\\n“[E]specially when—as in this case—the evidence of guilt was not overwhelming,\\n\\nwe expect that petitioners can rarely (if ever) prove a lawyer to be ineffective for\\n\\nrelying on this seemingly reasonable strategy to defend his client.” Id. at 1320\\n\\n(footnote omitted); see also Tarver, 169 F.3d at 715–16 (collecting law review\\n\\narticles on the merits of residual-doubt strategy). Not only can such a strategy be\\n\\nconstitutionally adequate, in some cases it may be “the most effective performance\\n\\nin defense to the death penalty.” Tarver, 169 F.3d at 716.\\n\\n We cannot say that Downey’s manifest strategy of residual doubt was\\n\\nobjectively unreasonable under the circumstances. Jenkins’s case is precisely the\\n\\nsort where reminding the jury of the misgivings it may have held about the State’s\\n\\ncase could have been effective. Scofield’s guilt-phase closing argument from that\\n\\nvery morning, which explicitly emphasized doubt and the weakness of the State’s\\n\\nevidence, would have been fresh in the minds of the jury.8 Although we are not\\n\\n\\n8\\n During the penalty phase of a capital case, “[e]vidence presented at the trial of the case may be\\nconsidered insofar as it is relevant to the aggravating and mitigating circumstances without the\\nnecessity of re-introducing that evidence at the sentence hearing.” Ala. Code § 13A-5-45(c).\\n 31\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 32 of 73\\n\\n\\nquestioning the jury’s guilty verdict or the sufficiency of the evidence, which are\\n\\nbeyond the scope of our habeas review, we acknowledge that the State’s case\\n\\nagainst Jenkins was—as Downey reminded the jury during the penalty phase—\\n\\nentirely circumstantial. Indeed, Scofield maintained post-conviction that the case\\n\\nfor reasonable doubt was “[v]ery strong.” No one saw Jenkins commit either the\\n\\nmurder or the robbery and kidnapping that elevated Hogeland’s murder to a capital\\n\\ncrime. Scofield’s guilt-phase closing argument challenged the inferential leaps the\\n\\nState was asking the jury to make in finding that Jenkins committed these capital-\\n\\npredicate offenses and that he formed the intent to kill. This situation was not one\\n\\nwhere a guilt-phase strategy of acquittal—and therefore a penalty-phase strategy of\\n\\nresidual doubt—was unreasonable in light of the evidence. Cf. Johnson, 643 F.3d\\n\\nat 932–33 (deficient sentencing investigation where defendant had confessed to\\n\\ntwo murders).\\n\\n Thus, although Downey’s penalty-phase closing argument did not\\n\\nspecifically invoke lingering doubt, it clearly reminded the jury of the entirely\\n\\ncircumstantial nature of the State’s guilt-phase case. See Chandler, 218 F.3d at\\n\\n1320 & n.6 (a strategy of “focus[ing] on obtaining an acquittal and then, at\\n\\nsentencing, on lingering doubt” is reasonable, even if counsel does not use the\\n\\nwords “lingering doubt” or “residual doubt”). Downey closed this reminder by\\n\\nexplaining to the jury how a foundational ancient legal system proscribed capital\\n\\n\\n 32\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 33 of 73\\n\\n\\npunishment in the absence of eyewitness testimony—the exact situation here. The\\n\\nstate court found that Downey’s penalty-phase closing argument reflected a\\n\\nstrategic “decision to concentrate on reasonable doubt,” see Jenkins, 972 So. 2d at\\n\\n147, a tactical choice entitled to a “strong presumption of correctness,” Marek, 62\\n\\nF.3d at 1300 & n.3.\\n\\n Jenkins suggests other theories of mitigation that Downey might have more\\n\\neffectively pursued,9 but our deferential review under Strickland does not ask\\n\\nwhether counsel could possibly or ideally have been more effective. “The test for\\n\\nineffectiveness is not whether counsel could have done more.” Waters, 46 F.3d at\\n\\n1518. We do not ask whether an attorney’s representation “deviated from best\\n\\npractices or common custom,” and we should resist the temptation to second-guess\\n\\nan attorney with the benefit of our hindsight. Richter, 562 U.S. at 105. We ask only\\n\\n“whether, in light of all the circumstances, the identified acts or omissions were\\n\\n\\n\\n\\n9\\n In particular, Jenkins argues that his counsel should have presented “model prisoner” evidence.\\nThe state court noted, “Good conduct during pretrial incarceration is not necessarily a mitigating\\ncircumstance,” before finding that Jenkins had not met his burden of establishing that his counsel\\neven had information about his conduct in jail. Jenkins, 972 So. 2d at 149. We agree, and we\\ndecline to impose a requirement of sua sponte investigation of this kind in every capital case. Cf.\\nWilliams, 529 U.S. at 396 (finding deficient performance based in part on not returning a call\\nfrom a prison ministry volunteer who offered to vouch for defendant’s good behavior in prison).\\n“Surely, counsel is not required to call a witness to testify to facts such as lack of violent nature\\nwhen the jury has rejected such an approach and has found that the defendant is guilty of\\nmurder.” Griffin v. Wainwright, 760 F.2d 1505, 1512 (11th Cir. 1985), vacated on other grounds,\\n476 U.S. 1112 (1986), reaff’d, 874 F.2d 1397 (11th Cir. 1989).\\n 33\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 34 of 73\\n\\n\\noutside the range of professionally competent assistance.” Strickland, 466 U.S. at\\n\\n690.\\n\\n We conclude that the state court did not unreasonably determine that Jenkins\\n\\nfailed to establish objectively incompetent performance by his counsel during the\\n\\npenalty phase of his trial. Our inquiry is not “what the best lawyers would have\\n\\ndone. Nor is the test even what most good lawyers would have done. We ask only\\n\\nwhether some reasonable lawyer at the trial could have acted, in the circumstances,\\n\\nas defense counsel acted at trial. . . . We are not interested in grading lawyers’\\n\\nperformances.” White v. Singletary, 972 F.2d 1218, 1220–21 (11th Cir. 1992). And\\n\\n“[i]f fairminded jurists could reasonably disagree” about the reasonableness of\\n\\ncounsel’s performance, “then habeas relief is due to be denied.” Johnson, 643 F.3d\\n\\nat 932.\\n\\n 2. Prejudice\\n\\n Even if Jenkins had demonstrated that his counsel performed as no\\n\\nreasonable lawyer could have, he must also demonstrate a “reasonable probability\\n\\nthat, but for counsel’s unprofessional errors, the result of the proceeding would\\n\\nhave been different.” Strickland, 466 U.S. at 694. Given that the jury here\\n\\nrecommended a sentence of death by the narrowest possible vote, 10 to 2, Jenkins\\n\\nneed establish only “a reasonable probability that at least one juror would have\\n\\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537. “A\\n\\n\\n 34\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 35 of 73\\n\\n\\nreasonable probability is a probability sufficient to undermine confidence in the\\n\\noutcome.” Strickland, 466 U.S. at 694. In measuring that probability here, we will\\n\\n“evaluate the totality of the available mitigation evidence—both that adduced at\\n\\ntrial, and the evidence adduced in the habeas proceeding in reweighing it against\\n\\nthe evidence in aggravation.” Williams, 529 U.S. at 397–98.\\n\\n We begin our analysis by clarifying what exactly Jenkins is arguing about\\n\\nprejudice. We do not know precisely which aggravating and mitigating factors the\\n\\njury may have implicitly found before making its recommendation of death. The\\n\\ntrial court found at sentencing two statutory aggravating factors: the robbery and\\n\\nthe kidnapping of the victim. See Ala. Code § 13A-5-49(4). Because none of the\\n\\nmitigation evidence offered in the Rule 32 hearing bears on those factors, we see\\n\\nno reasonable probability that the aggravating factors would have been determined\\n\\ndifferently by the jury or the court. The court also found at sentencing two\\n\\nstatutory mitigating factors: no significant history of prior criminal activity and\\n\\nage. See id. § 13A-5-51(1), (7). There was also strong evidence of Jenkins’s\\n\\nintoxication, which was a key part of Jenkins’s guilt-phase defense and could have\\n\\nsupported a third mitigating factor, impaired capacity. See id. § 13A-5-51(6).10\\n\\n\\n10\\n The jury had heard evidence about all three of these statutory mitigators during the guilt phase\\nof the trial. Downey’s penalty-phase opening statement reviewed for the jury all seven of the\\npossible statutory mitigating factors, and the court also recited them in its jury instructions,\\nreminding the jury that they could consider guilt-phase evidence as well as non-statutory\\nmitigation. Jenkins has not established that there is a substantial probability that the jury’s\\nrecommendation would have been different had it been more explicitly reminded of Jenkins’s\\n 35\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 36 of 73\\n\\n\\nJenkins does not argue that the jury or the court would have found other statutory\\n\\nmitigators in light of the Rule 32 evidence. Rather, he contends that the jury might\\n\\nhave found non-statutory mitigation based on Jenkins’s abusive childhood, low\\n\\nintelligence, or mental health problems that would have outweighed the\\n\\naggravating factors.11\\n\\n The Rule 32 trial court found that Jenkins had not demonstrated prejudice\\n\\neven assuming that counsel had performed deficiently. See Jenkins, 972 So. 2d at\\n\\n138. It concluded that he had not shown a reasonable probability that the sentencer\\n\\n“would have concluded that a weighing of the aggravating and mitigating\\n\\ncircumstances did not warrant death” upon a reweighing of the evidence. See id. It\\n\\nemphasized the depravity of the crime and the aggravating circumstances and\\n\\nfound that the mitigating witnesses were not credible for various reasons. See id.\\n\\nThe state appellate court affirmed, declining to disturb the credibility\\n\\ndeterminations of the trial court. Id. at 142–44. Independently reweighing the\\n\\n\\nage, criminal history, and intoxication, given that the trial court, the Rule 32 trial court, and the\\nstate appellate court all explicitly considered those mitigators and found them outweighed by the\\naggravators.\\n11\\n Alabama law lists seven statutory mitigating circumstances, but states that mitigation “shall\\n. . . not be limited to” those seven circumstances. Ala. Code § 13A-5-51. Further, “mitigating\\ncircumstances shall include any aspect of a defendant’s character or record and any of the\\ncircumstances of the offense that the defendant offers as a basis for a sentence of life\\nimprisonment without parole instead of death, and any other relevant mitigating circumstances\\nwhich the defendant offers as a basis for a sentence of life imprisonment without parole instead\\nof death.” Id. § 13A-5-52. A jury may recommend a sentence of death only if it finds that “one or\\nmore aggravating circumstances . . . exist and that they outweigh the mitigating circumstances, if\\nany.” Id. § 13A-5-46(e)(3).\\n 36\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 37 of 73\\n\\n\\nevidence, it concluded that, “[g]iven the aggravating circumstances that were\\n\\nproven by the State and the facts surrounding Hogeland’s murder, we, like the\\n\\ncircuit court, are confident that death was the appropriate punishment for Jenkins’s\\n\\nactions.” Id. at 148.\\n\\n Our review of this evidence does not show that it is likely that the jury\\n\\nwould have concluded that death was not the appropriate punishment if it had\\n\\nheard the Rule 32 evidence. Much of the proffered mitigation evidence is, in our\\n\\nview, much less mitigating than Jenkins asserts. In particular, the testimony of\\n\\nJenkins’s older brother Michael was likely to have been a double-edged sword. See\\n\\nEvans v. Sec’y, Dep’t of Corr., 703 F.3d 1316, 1327 (11th Cir. 2013) (en banc) (“it\\n\\nis reasonable to conclude that a defendant was not prejudiced when his mitigation\\n\\nevidence was a two-edged sword” (quoting Ponticelli v. Sec’y, Fla. Dep’t of Corr.,\\n\\n690 F.2d 1271, 1296 (11th Cir. 2012)). Michael was the closest relative to testify at\\n\\nthe Rule 32 hearing and his testimony about the abuse Jenkins experienced should\\n\\nhave been the most reliable, since he experienced much of the same physical abuse\\n\\nhimself. The brothers shared the same filthy bedroom, received beatings from the\\n\\nsame paddle, and were frequently absent from school together. But Michael’s\\n\\ntestimony contained some discrepancies, and he admitted to cognitive and memory\\n\\nproblems. Perhaps Michael’s testimony would have made Jenkins appear more\\n\\nsympathetic in the eyes of the jury—but perhaps not. While Jenkins ran away from\\n\\n\\n 37\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 38 of 73\\n\\n\\nhome and eventually committed murder, Michael responded to the same\\n\\nbackground by running away from home, graduating from high school, and\\n\\nproviding health care for both the terminally ill and people with substance use\\n\\ndisorders. It is possible that the jury could even have found Jenkins to be more\\n\\nculpable for not overcoming his difficult upbringing and to pose more of a danger\\n\\nof future violence.\\n\\n We also find reasonable the state court’s identification of other problems\\n\\nwith the mitigating evidence Jenkins argues should have been presented. Hearing\\n\\nthat evidence firsthand in the Rule 32 proceeding, the trial court found the\\n\\ntestimony of the cousins less credible in light of the lack of physical abuse\\n\\ndocumented in the school health records. See Jenkins, 972 So. 2d at 139 (noting\\n\\nthat school records observed a rash and gingivitis). It faulted Pitts in particular for\\n\\npurporting to have witnessed ongoing, severe abuse but taking only minimal,\\n\\nineffective steps to obtain outside help. See id. It further noted that DeLavega and\\n\\nWagoner actually testified that they never saw Jenkins beaten, and it noted all of\\n\\nthe witnesses’ interest in getting Jenkins off death row. All of these factors would\\n\\nlikely have caused a reasonable jury to discount their testimony. See id. at 139–42.\\n\\nWe must accept the trial court’s views of the Rule 32 testimony. “In the absence of\\n\\nclear and convincing evidence, we have no power on federal habeas review to\\n\\nrevisit the state court’s credibility determinations.” Bishop v. Warden, 726 F.3d\\n\\n\\n 38\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 39 of 73\\n\\n\\n1243, 1259 (11th Cir. 2013) (citing Marshall v. Lonberger, 459 U.S. 422, 434\\n\\n(1983)).\\n\\n The state court also noted that Wagoner testified that she was not available\\n\\nto testify at the trial, and it found that Seal’s testimony would have been\\n\\ncumulative to that of her husband. Thus, counsel’s failure to call those two\\n\\nwitnesses could not have been prejudicial to Jenkins. See id. at 141, 142. Neither of\\n\\nthese conclusions involves an unreasonable determination of the facts. We agree\\n\\nthat the Rule 32 testimony of Sherry Seal was substantially similar to the testimony\\n\\nof her husband Lonnie Seal that the jury heard during the penalty phase. Both\\n\\ntestified about meeting Jenkins in California, Jenkins’s assistance with their move\\n\\nto Alabama, and his residence with them in their Alabama home. Sherry added\\n\\nonly that Jenkins sometimes wet his bed. It was not unreasonable for the state court\\n\\nto conclude that Sherry’s testimony was cumulative of her husband’s, and “a\\n\\npetitioner cannot satisfy the prejudice prong of the Strickland test with evidence\\n\\nthat is merely cumulative of evidence already presented.” Rose v. McNeil, 634 F.3d\\n\\n1224, 1243 (11th Cir. 2011).\\n\\n Further, the school, medical, and juvenile court records that Jenkins\\n\\nintroduced in the Rule 32 proceeding have their own problems. As noted, they do\\n\\nnot document or even corroborate the severe level of abuse alleged by some of the\\n\\nwitnesses. Though these documents tell a sad story, they do not create a substantial\\n\\n\\n 39\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 40 of 73\\n\\n\\nlikelihood that the jury would have recommended life. Although they portray\\n\\nJenkins as a victim of truly unfortunate circumstances, they also show his unsavory\\n\\ndelinquent behavior. That delinquency included a history of stealing cars, and\\n\\ngiven that the jury likely inferred that Jenkins stole the red Mazda, new\\n\\ninformation about his history of theft may not have been especially mitigating.\\n\\nSimilarly, we agree with the state court that evidence of Jenkins’s compliant\\n\\nconduct in pretrial detention is “minimally mitigating” where there had been no\\n\\nevidence of jail misconduct as an aggravator. See Jenkins, 972 So. 2d at 149; cf.\\n\\nSkipper v. South Carolina, 476 U.S. 1, 5 & n.1 (1986) (good jail behavior is\\n\\n“potentially mitigating” especially “[w]here the prosecution specifically relies on a\\n\\nprediction of future dangerousness in asking for the death penalty”).\\n\\n Jenkins argues that allowing the jury to see his experience of mental,\\n\\nemotional, and cognitive difficulties, compounded by years of abuse and neglect\\n\\nand considered alongside his mild character and good jail behavior as an adult,\\n\\nwould have convinced them that Hogeland’s murder was an “aberration” that\\n\\ndeserved a punishment less severe than death. That interpretation of Jenkins’s life\\n\\nstory may be conceivable. But more than mere conceivability is required to\\n\\nestablish prejudice: “The likelihood of a different result must be substantial, not\\n\\njust conceivable.” Richter, 562 U.S. at 112 (emphasis added). We do not find the\\n\\nstate court’s decision that a different result was not substantially likely to be an\\n\\n\\n 40\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 41 of 73\\n\\n\\nunreasonable application of Strickland. As we have noted, we think it equally\\n\\nlikely that the jury could have rejected this alternate narrative out of hand in light\\n\\nof having just convicted him of the violent murder of a young mother. We also\\n\\nthink it possible that the jury could have viewed the murder as just one step in a\\n\\nlogical progression of degeneracy that began with behavioral problems at school,\\n\\ncontinued through the juvenile court system, and escalated past grand theft auto\\n\\ninto kidnapping and murder. Taken separately as well as together, the proffered\\n\\nevidence does not compel us to conclude that a verdict of life would have been\\n\\nsubstantially likely. Because fairminded jurists could disagree about the state\\n\\ncourt’s conclusion about Strickland prejudice, we affirm the denial of habeas relief\\n\\non this ground.\\n\\n B. Intellectual Disability\\n\\n Jenkins argues that the district court erred when it denied his claim that he is\\n\\nintellectually disabled and ineligible for the death penalty under Atkins v. Virginia,\\n\\n536 U.S. 304 (2002). The Supreme Court held in Atkins that executing an offender\\n\\nwho has mental retardation12 violates the Eighth Amendment’s ban on excessive\\n\\n\\n\\n\\n12\\n In 2002, the Atkins Court followed prevailing convention in referring to “mentally retarded”\\noffenders. We will use that now-disfavored term only when quoting or discussing older sources.\\nSince 2014, the Supreme Court has exclusively used the terms “intellectually disabled person”\\nand “person with intellectual disability.” See, e.g., Hall v. Florida, 572 U.S. 701, 704 (2014);\\nMoore v. Texas, 139 S. Ct. 666, 672 (2019).\\n 41\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 42 of 73\\n\\n\\npunishments. Id. at 321. It defined mental retardation by quoting the American\\n\\nAssociation on Mental Retardation:\\n\\n Mental retardation refers to substantial limitations in present\\n functioning. It is characterized by significantly subaverage intellectual\\n functioning, existing concurrently with related limitations in two or\\n more of the following applicable adaptive skill areas: communication,\\n self-care, home living, social skills, community use, self-direction,\\n health and safety, functional academics, leisure, and work. Mental\\n retardation manifests before age 18.\\n\\nId. at 308 n.3. Alabama courts have followed that tripartite definition, adopting\\n\\nwhat they consider the “broadest” definition of mental retardation for Atkins\\n\\npurposes:\\n\\n [A] defendant, to be considered mentally retarded, must have\\n significantly subaverage intellectual functioning (an IQ of 70 or\\n below), and significant or substantial deficits in adaptive behavior.\\n Additionally, these problems must have manifested themselves during\\n the developmental period (i.e., before the defendant reached age 18).\\n\\nEx parte Perkins, 851 So. 2d 453, 456 (Ala. 2002).\\n\\n The clinical evidence adduced during the Rule 32 hearing with respect to\\n\\nJenkins’s intellectual ability was that his IQ was 76, reflecting “borderline\\n\\nintellectual functioning.” The record also included the report of the Lunacy\\n\\nCommission13 which found that Jenkins was competent to stand trial, as well as\\n\\n\\n\\n13\\n Alabama law requires a capital defendant to be “committed to the Department of Mental\\nHealth and Mental Retardation for examination” if there is a question about his capacity to stand\\ntrial or if there will be a mental-disease-or-defect defense. Ala. Code § 15-16-22(a). “The\\nassigned mental health professional(s) shall examine the defendant with respect to determining\\nthe presence of any mental disease or defect which, if determined to be present, would affect the\\n 42\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 43 of 73\\n\\n\\nJenkins’s school and juvenile court records. The court also considered all the\\n\\ntestimony it had heard about Jenkins and reviewed the transcript of his murder trial\\n\\nbefore rejecting Jenkins’s claim of mental retardation.\\n\\n On appeal, after reciting the Perkins standard, the Alabama Court of\\n\\nCriminal Appeals explained that Jenkins had not established mental retardation:\\n\\n Dr. Kirkland testified that he performed psychological tests on Jenkins\\n and that Jenkins’s IQ was 76. There was evidence presented at\\n Jenkins’s trial indicating that Jenkins maintained relationships with\\n other individuals and that he had been employed by P.S. Edwards\\n Landscaping Company, Cotton Lowe 76 Service Station, and\\n Paramount Painting Company. The record fails to show that Jenkins\\n meets the most liberal view of mental retardation adopted by the\\n Alabama Supreme Court in Perkins.\\n\\nJenkins, 972 So. 2d at 155. In this habeas appeal, Jenkins raises arguments about\\n\\nall three components of the intellectual disability showing under Atkins and\\n\\nPerkins, which we address in turn. He also appeals the denial of an evidentiary\\n\\nhearing on this issue.\\n\\n 1. Intellectual Function\\n\\n The first component of intellectual disability under Perkins is “significantly\\n\\nsubaverage intellectual functioning (an IQ of 70 or below).” Perkins, 851 So. 2d at\\n\\n456. Jenkins argues that the use of a strict IQ cutoff of 70 was contrary to or an\\n\\nunreasonable application of Atkins.\\n\\n\\ncapacity of the defendant to proceed or continue to trial or which would affect the defendant’s\\ncriminal responsibility at the time of the commission of the crime.” Id.\\n 43\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 44 of 73\\n\\n\\n We disagree. The state court’s decision that Jenkins failed to establish\\n\\nmental retardation in this respect was a reasonable application of the record and of\\n\\nAtkins. As a threshold matter, we note that the state court did not say it was\\n\\napplying a strict cutoff of 70; it merely stated that Dr. Kirkland had measured\\n\\nJenkins’s IQ as 76. Jenkins, 972 So. 2d at 155. Under our deferential review, the\\n\\nquestion is whether the decision that Jenkins had not established the intellectual\\n\\nfunctioning aspect of intellectual disability was an unreasonable application of\\n\\nAtkins. The Supreme Court explained in Atkins that it was leaving “to the State[s]\\n\\nthe task of developing appropriate ways to enforce the constitutional restriction\\n\\nupon [their] execution of sentences.” Atkins, 536 U.S. at 317 (quoting Ford v.\\n\\nWainwright, 477 U.S. 399, 416–17 (1986)). In line with that invitation, our Circuit\\n\\nhas accepted Alabama’s judicial definition of intellectual disability as laid out in\\n\\nPerkins.14 See Thomas, 607 F.3d at 752–53. Jenkins objects, however, that the\\n\\nPerkins standard, which he characterizes as articulating a “strict cutoff” of 70,\\n\\nconflicts with more recent Supreme Court precedent that instead requires a\\n\\n“clinical” definition that goes beyond a single IQ score.\\n\\n In particular, Jenkins points to Hall v. Florida, in which the Supreme Court\\n\\nheld that when an offender’s IQ is near but greater than 70, courts must take into\\n\\n\\n\\n14\\n Notwithstanding the urging of the Alabama Supreme Court, see Perkins, 851 So. 2d at 455\\nn.1, the Alabama legislature has not enacted a statutory definition of intellectual disability.\\n 44\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 45 of 73\\n\\n\\naccount the IQ test’s standard error of measurement as well as non-IQ evidence of\\n\\nintellectual function. See Hall v. Florida, 572 U.S. 701, 723 (2014) (“This Court\\n\\nagrees with the medical experts that when a defendant’s IQ test score falls within\\n\\nthe test’s acknowledged and inherent margin of error, the defendant must be able\\n\\nto present additional evidence of intellectual disability, including testimony\\n\\nregarding adaptive deficits.”). But, as Jenkins concedes, our Circuit has\\n\\nspecifically held that Hall is not retroactive to cases on collateral review. See In re\\n\\nHenry, 757 F.3d 1151, 1161 (11th Cir. 2014) (explaining that Hall does not supply\\n\\na new substantive rule but rather “merely provides new procedures for ensuring\\n\\nthat States do not execute members of an already protected group”). We also note\\n\\nthat Alabama courts have found the Perkins standard to comport with Hall in that\\n\\nit does not prevent an offender from presenting the standard error of measurement\\n\\nor non-IQ evidence of intellectual disability. See Reeves v. State, 226 So. 3d 711,\\n\\n728–29 (Ala. Ct. Crim. App. 2016). Here, the state court did not unreasonably\\n\\napply Atkins when it analyzed the intellectual functioning component of\\n\\nintellectual disability.\\n\\n Jenkins also argues that the state court’s decision on this component in his\\n\\nRule 32 proceeding resulted from an unreasonable determination of the facts. He\\n\\nasserts that his IQ score of 76, when combined with the test’s standard error of\\n\\nmeasurement and considered alongside the Flynn effect (which we discuss below),\\n\\n\\n 45\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 46 of 73\\n\\n\\ndoes show that he has significantly subaverage intellectual function. For the\\n\\nreasons that follow, we disagree.\\n\\n Most fundamentally, neither of the clinicians who testified about Jenkins’s\\n\\nintellectual ability in the Rule 32 hearing opined that he was mentally retarded.\\n\\nThe closest thing Jenkins can point to is the testimony of Dr. Kirkland that his IQ\\n\\nis two standard deviations below the mean. But his entire testimony, in context, did\\n\\nnot establish intellectual disability. To the contrary: Dr. Kirkland testified that\\n\\nJenkins “scored in the range of borderline intellectual functioning which is\\n\\nbetween mild mental retardation and low average intellectual functioning.”\\n\\nFurthermore, Jenkins’s own expert Dr. Lisak fully reviewed Dr. Kirkland’s clinical\\n\\nfindings. He agreed with those findings, concluding that the IQ results show\\n\\n“borderline intelligence.” It is tremendously significant that no clinical assessment\\n\\nin the entire record of these proceedings has found that Jenkins has mental\\n\\nretardation or intellectual disability.\\n\\n Jenkins nonetheless argues that his IQ score of 76 establishes the intellectual\\n\\ncomponent of intellectual ability when the standard error of measurement and the\\n\\nFlynn effect are considered alongside that score. It was not unreasonable for the\\n\\nstate court to disagree. As we have discussed, Jenkins is not retroactively entitled\\n\\nto have the state courts take into account the margin of error of the IQ test. That\\n\\nstatistical margin of error—5 points in this case—takes into account characteristics\\n\\n\\n 46\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 47 of 73\\n\\n\\nof the test and variations due to chance in order to state how confident the tester is\\n\\nthat the measured result accurately states Jenkins’s true, unknowable IQ. A result\\n\\nof 76 with a standard error of measurement of 5 simply means that the tester is\\n\\n68% confident that Jenkins’s true IQ falls between 71 and 81, and 95% confident\\n\\nthat his true IQ falls between 66 and 86. See David H. Kaye & David A. Freedman,\\n\\nReference Guide on Statistics, in Reference Manual on Scientific Evidence 211,\\n\\n243–44 (3d ed. 2011); see also Hall, 572 U.S. at 738–39 (Alito, J., dissenting). Of\\n\\ncourse, Jenkins would like us to consider only the possibility that the test result\\n\\noverstated his IQ—but the standard error of measurement is a two-way street.\\n\\n“[T]he standard error of measurement is a bi-directional concept that does not\\n\\ncarry with it a presumption that an individual’s IQ falls to the bottom of his IQ\\n\\nrange.” Ledford v. Warden, Ga. Diag. & Clas. Prison, 818 F.3d 600, 641 (11th Cir.\\n\\n2016).\\n\\n Even if Jenkins were somehow entitled to subtract the 5-point margin of\\n\\nerror from his score and thereby lower his “true” IQ to 71, he still would not satisfy\\n\\nthe clinical intellectual component of intellectual disability. Thus, he sought in the\\n\\ndistrict court to introduce evidence about the Flynn effect in order to bring his IQ\\n\\ndown to 70 or below. As our Court has explained, the Flynn effect “recognizes the\\n\\nfact that IQ test scores have been increasing over time”:\\n\\n The Flynn effect acknowledges that as an intelligence test ages, or\\n moves farther from the date on which it was standardized, or normed,\\n 47\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 48 of 73\\n\\n\\n the mean score of the population as a whole on that assessment\\n instrument increases, thereby artificially inflating the IQ scores of\\n individual test subjects. Therefore, the IQ test scores must be\\n recalibrated to keep all test subjects on a level playing field.\\n\\nThomas, 607 F.3d at 753. We have said that “[a]n evaluator may also consider the\\n\\n‘Flynn effect’” in assessing an offender’s possible intellectual disability, id.\\n\\n(emphasis added), but neither we nor the Alabama Supreme Court nor the U.S.\\n\\nSupreme Court have said that a court must consider it in order to reasonably apply\\n\\nAtkins. See, e.g., Reeves, 226 So. 3d at 739 (“This Court has repeatedly held that a\\n\\ncircuit court is not required to accept, consider, or apply the ‘Flynn Effect’ in\\n\\ndetermining intellectual disability.”).\\n\\n More problematically, Jenkins did not raise the Flynn effect issue in the\\n\\nAlabama state courts. Jenkins asserts that he implicitly raised it as part and parcel\\n\\nof his Atkins claim during his Rule 32 appeal, but we do not agree that the\\n\\nAlabama courts had a full and fair opportunity to consider this issue, because it\\n\\nwas not explicitly presented to them. See Kelley v. Sec’y for the Dep’t of Corr., 377\\n\\nF.3d 1317, 1349–50 (11th Cir. 2004) (holding that one ineffective-assistance\\n\\nargument was unexhausted despite the petitioner’s general presentation of a\\n\\nStrickland claim in the state courts). Thus, under AEDPA, Jenkins is procedurally\\n\\nbarred from raising this unexhausted claim in his federal habeas petition. See 28\\n\\nU.S.C. § 2254(b)(1)(A), (c); Mason v. Allen, 605 F.3d 1114, 1119 (11th Cir. 2010).\\n\\nWe will allow a prisoner to overcome a procedural bar if he establishes cause for\\n 48\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 49 of 73\\n\\n\\nhis default and prejudice therefrom. See, e.g., Smith v. Jones, 256 F.3d 1135, 1138\\n\\n(11th Cir. 2001). But Jenkins cannot establish prejudice from being barred from\\n\\nasserting this claim now because the Supreme Court has not “clearly established”\\n\\nthat a court must consider the Flynn effect in assessing intellectual disability. Thus,\\n\\nunder AEDPA, it could not form the basis for a federal habeas claim. See 28\\n\\nU.S.C. § 2254(d)(1). We therefore conclude that the state court did not\\n\\nunreasonably determine the facts or unreasonably apply Atkins with respect to the\\n\\nintellectual component of intellectual disability.\\n\\n 2. Adaptive Function\\n\\n The second component of intellectual disability is “significant or substantial\\n\\ndeficits in adaptive behavior.” Perkins, 851 So. 2d at 456. Jenkins argues that the\\n\\nstate court, which mentioned only Jenkins’s successes in employment and social\\n\\nskills, see Jenkins, 972 So. 2d at 155, failed to consider his deficits. He adds that\\n\\neven the mention of employment success does not tell the whole story. Jenkins\\n\\ninsists that his jobs were menial, his employers took advantage of him, and he held\\n\\nthe jobs for only a few months each.\\n\\n But the state court’s terse discussion of Jenkins’s adaptive behavior did not\\n\\ninvolve an unreasonable determination of the facts or an unreasonable application\\n\\nof Atkins. To the contrary, its decision on this issue was reasonable in light of both\\n\\nthe record and Atkins.\\n\\n\\n 49\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 50 of 73\\n\\n\\n Atkins elaborated that the adaptive-skill component of intellectual disability\\n\\nrequires substantial present limitation in at least two of the following areas:\\n\\n“communication, self-care, home living, social skills, community use, self-\\n\\ndirection, health and safety, functional academics, leisure, and work.” Atkins, 536\\n\\nU.S. at 308 n.3 (quoting definition of American Association on Mental\\n\\nRetardation). The record supports the state court’s judgment that Jenkins does not\\n\\nhave substantial deficits in at least two areas. To be sure, we will assume that\\n\\nJenkins may be substantially limited in functional academics; Dr. Kirkland\\n\\nassessed his reading, spelling, and arithmetic skills at a third-grade level. But the\\n\\nrecord does not compel a finding of substantial present deficits in the areas of\\n\\ncommunication, self-care, home living, and work. Although Jenkins emphasizes\\n\\nhis transient employment, the record shows he was able to find work quickly, to\\n\\nwork hard to support the Seals’ household until Lonnie Seal found a job, and, later,\\n\\nto move into his own home. That record evidence is inconsistent with a substantial\\n\\ndeficit in the area of work.\\n\\n As the State observes, the facts of the crime also fail to show significant\\n\\ndeficits in the areas of communication, self-care, community use, and self-\\n\\ndirection. Jenkins was able to communicate well enough to solicit an alibi and to\\n\\nsell his car, including writing out a bill of sale. He then was able to arrange for his\\n\\nflight from Alabama, including purchasing a bus ticket and hitchhiking across the\\n\\n\\n 50\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 51 of 73\\n\\n\\ncountry. Indeed, throughout his trial and postconviction proceedings, the people\\n\\nwho knew Jenkins as an adult consistently described a man who did not have\\n\\nserious difficulties in communicating, forming relationships, working, and caring\\n\\nfor himself. No witness mentioned these kinds of difficulties in his or her\\n\\ntestimony about Jenkins. There was some evidence that Jenkins was “not\\n\\narticulate” and that his house was messy, but that testimony does not rise to the\\n\\nlevel of overwhelming evidence of substantial functional deficits that would\\n\\nsupport a conclusion that the state court unreasonably determined the facts.\\n\\n Of course, the record also contains evidence of Jenkins’s childhood\\n\\nacademic and social deficits, attributable at least in part to his parents’ neglect, but\\n\\nJenkins’s childhood is not directly relevant to our consideration of his present\\n\\nlimitations. Although we acknowledge that the Supreme Court has noted that such\\n\\n“risk factors” may support further exploration of the possibility of intellectual\\n\\ndisability, see Moore v. Texas, 137 S. Ct. 1039, 1051 (2017), the state court record\\n\\ncontains the results of that exploration. Overall, that record supports the state\\n\\ncourt’s conclusion that Jenkins does not have substantial deficits in adaptive\\n\\nbehaviors.\\n\\n 3. Juvenile Onset\\n\\n The third and final component of intellectual disability is manifestation of\\n\\nthe first and second components “before the defendant reached age 18.” Perkins,\\n\\n\\n 51\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 52 of 73\\n\\n\\n851 So. 2d at 456. Jenkins asserts, the State assumes, and we agree that the state\\n\\ncourt did not rule upon this issue, so there is nothing to which we must defer under\\n\\nAEDPA. See, e.g., Brumfield v. Cain, 135 S. Ct. 2269, 2282 (2015). Thus, we will\\n\\nreview the district court’s legal conclusions on this issue de novo, and its factual\\n\\nfindings for clear error. See, e.g., Grossman v. McDonough, 466 F.3d 1325, 1335\\n\\n(11th Cir. 2006). Jenkins argues that he has satisfied the age-of-onset component\\n\\nby showing “a wide range of evidence of deficits in both intellectual and adaptive\\n\\nfunctioning throughout Jenkins’s childhood.”\\n\\n In fact, the ample record of Jenkins’s childhood, as developed in the Rule 32\\n\\nproceeding, does not point to intellectual disability before age 18. School\\n\\nassessments at age 12 reported an IQ score of 83, an adaptive behavior rating of\\n\\n“low,” and noted overall an intellectual capacity of “average,” a diagnosis of\\n\\ndyslexia, and “poor functional skills.” The psychologist attributed Jenkins’s\\n\\ndifficulties to a combination of “[f]requent moves, changes in schools and family\\n\\ninstability” with “[l]ow ability coupled with a specific learning disability” and\\n\\n“[l]ack of responsibility for self control and poor use of time.” Although Jenkins\\n\\nfailed several grades of school, teachers additionally attributed his difficulties to\\n\\nhis behavioral problems and his high number of absences from school.\\n\\n Similarly, IQ tests ordered by the juvenile court at age 14 gave scores of 81\\n\\nand 86 on different tests, in the “[d]ull–[n]ormal” range and reflecting “a definite\\n\\n\\n 52\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 53 of 73\\n\\n\\nlearning disability.” These records all show a child with serious academic deficits\\n\\nand some intellectual and adaptive deficits, but they do not clearly show an\\n\\nintellectually disabled child. In particular, Jenkins’s childhood IQ scores of 83, 81,\\n\\nand 86 fail to point to intellectual disability or mental retardation. We are not left\\n\\nwith the definite and firm conviction that the district court erred when it found that\\n\\n“Jenkins’s [childhood] intelligence scores and adaptive functioning skills were\\n\\nattributable to his learning disability and other circumstances in his life, rather than\\n\\nmental retardation.”\\n\\n In sum, we agree with the district court that the decision of the Alabama\\n\\nCourt of Criminal Appeals that Jenkins does not have intellectual disability was\\n\\nnot contrary to or an unreasonable application of Atkins v. Virginia.\\n\\n 4. Evidentiary Hearing\\n\\n Finally, Jenkins argues that the district court erred when it denied his request\\n\\nfor an evidentiary hearing on his Atkins claim. If Jenkins diligently attempted to\\n\\npresent this claim in the state court, we review for abuse of discretion. Burgess v.\\n\\nComm’r, Ala. Dep’t of Corr., 723 F.3d 1308, 1320 (11th Cir. 2013). “The district\\n\\ncourt abuses its discretion where ‘such a hearing could enable an applicant to prove\\n\\nthe petition’s factual allegations, which, if true, would entitle the applicant to\\n\\nfederal habeas relief.’” Id. (quoting Schriro v. Landrigan, 550 U.S. 465, 474\\n\\n(2007)). If Jenkins did not diligently attempt to develop the factual basis for this\\n\\n\\n 53\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 54 of 73\\n\\n\\nclaim in the state court, absent certain circumstances, he is not entitled to a hearing\\n\\nand we will affirm the district court’s denial. 28 U.S.C. § 2254(e)(2).\\n\\n The parties strenuously dispute whether Jenkins diligently attempted to\\n\\npresent this issue in the state court. On the one hand, the State argues that Jenkins\\n\\nfailed to allege that he was mentally retarded even after the U.S. Supreme Court\\n\\ndecided Atkins in 2002. On the other hand, Jenkins asserts that he raised the\\n\\npossibility of mental retardation as early as his original Rule 32 petition in 1995.\\n\\n As we have noted, shortly after the Supreme Court decided Atkins, the\\n\\nAlabama Court of Criminal Appeals ordered supplemental briefing on the possible\\n\\nimpact of Atkins on Jenkins. Jenkins’s Rule 32 appeal was pending at that point,\\n\\nhaving been fully briefed and orally argued in the state appellate court. Having\\n\\nreviewed the state court records, we agree with the State that Jenkins did not\\n\\ndiligently attempt to present his Atkins claim in the state courts.\\n\\n “Diligence will require in the usual case that the prisoner, at a minimum,\\n\\nseek an evidentiary hearing in state court.” Williams, 529 U.S. at 437. The record\\n\\nclearly shows that Jenkins did not do so. In the initial supplemental brief on Atkins\\n\\nhe filed in 2002, Jenkins merely requested a future hearing after the Alabama\\n\\nlegislature’s enactment of a statute on executing the mentally retarded. Seventeen\\n\\nyears later, that contingency is still in the future. Only in his reply supplemental\\n\\nbrief did Jenkins then assert, for the first time ever, that he is mentally retarded.\\n\\n\\n 54\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 55 of 73\\n\\n\\n We acknowledge that Jenkins’s initial Rule 32 petition included in its\\n\\nallegations of ineffective assistance of counsel a list of unpresented mitigating\\n\\nevidence including his “developmental[] impair[ment]” and “learning disabilities,\\n\\nlow intelligence, poor comprehension and retarded socialization skills.” We\\n\\nconclude that these tangential Strickland references to his mental and social\\n\\nabilities do not amount to diligence. Thus, because Jenkins did not diligently\\n\\nattempt to develop the factual basis for his Atkins claim in the state courts, we\\n\\naffirm the district court’s denial of his request for an evidentiary hearing.\\n\\n Even if Jenkins had diligently attempted to present this claim in the state\\n\\ncourt, however, we conclude in the alternative that the district court did not abuse\\n\\nits discretion when it denied a hearing. We agree, in light of the ample evidence\\n\\nabout intellectual ability that Jenkins adduced in support of his ineffective-\\n\\nassistance claim, that such a hearing would not have enabled Jenkins to prove that\\n\\nhe is intellectually disabled. See Burgess, 723 F.3d at 1320.\\n\\n IV. CONCLUSION\\n\\n The denial of Jenkins’s petition for a writ of habeas corpus is\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 55\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 56 of 73\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435142/\", \"author_raw\": \"BRANCH, Circuit Judge:\"}, {\"author\": \"WILSON, Circuit Judge, dissenting\", \"type\": \"dissent\", \"text\": \"WILSON, Circuit Judge, dissenting:\\n\\n Mark Jenkins must prove that the Alabama state courts acted unreasonably\\n\\nin denying him post-conviction collateral relief. 28 U.S.C. § 2254(d)(1). Jenkins\\n\\nhas satisfied that burden as to his ineffective assistance of counsel claim. He has\\n\\nalso demonstrated that he is entitled to an evidentiary hearing regarding his Atkins\\n\\nv. Virginia claim. Because the majority came to the opposite conclusions, I\\n\\ndissent.\\n\\n I. Ineffective Assistance of Counsel at Penalty Phase\\n\\n Jenkins first asserts that his trial counsel’s failure to investigate Jenkins’\\n\\nabusive upbringing and to present any mitigating evidence at the penalty phase of\\n\\nhis trial constituted ineffective assistance of counsel. The majority rejects this\\n\\ncontention, concluding that the Alabama state courts did not act unreasonably in\\n\\ndenying Jenkins’ ineffective assistance of counsel claim on collateral review. I\\n\\ndisagree and maintain that Jenkins established both deficient performance and\\n\\nprejudice, entitling him to habeas relief. See Strickland v. Washington, 466 U.S.\\n\\n668, 687 (1984).\\n\\n A. Deficient Performance\\n\\n The Alabama state courts concluded that Jenkins’ counsel, Stan Downey,\\n\\ndid not perform deficiently at the penalty phase. The majority finds that\\n\\nconclusion reasonable primarily because Downey (1) was not obligated to\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 57 of 73\\n\\n\\ninvestigate Jenkins’ abusive childhood and (2) reasonably pursued a residual doubt\\n\\nstrategy at the penalty phase. But further review of the record requires me to part\\n\\ncompany with the majority’s reasoning and, ultimately, persuades me that the\\n\\nAlabama courts’ conclusion was indeed unreasonable. See 28 U.S.C. § 2254(d)(1)\\n\\n(providing that a writ of habeas corpus should not be granted to a state prisoner\\n\\nunless the state court adjudication involved “a decision that was contrary to, or\\n\\ninvolved an unreasonable application of, clearly established Federal law, as\\n\\ndetermined by the Supreme Court of the United States”).\\n\\n Jenkins offers numerous examples of how his counsel performed deficiently\\n\\nat the penalty phase of trial. First, counsel failed to thoroughly investigate Jenkins’\\n\\nabusive childhood—an error that constitutes deficient performance. In Williams v.\\n\\nAllen, we held that counsel’s failure to investigate mitigating evidence at the\\n\\npenalty stage constituted ineffective assistance of counsel because counsel drew\\n\\nupon only three sources of information—a psychologist’s report, a presentence\\n\\ninvestigation report, and an interview with the appellant’s mother. 542 F.3d 1326,\\n\\n1340 (11th Cir. 2008); see also Debruce v. Comm’r, Ala. Dep’t of Corr., 758 F.3d\\n\\n1263, 1270, 1273–74 (11th Cir. 2014) (finding counsel’s performance deficient\\n\\nwhere counsel interviewed the petitioner and his mother, but failed to contact\\n\\nexpert witnesses or other family members). In that case, we concluded that\\n\\n\\n\\n\\n 57\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 58 of 73\\n\\n\\ncounsel’s lack of investigation resulted in “an incomplete and misleading\\n\\nunderstanding of Williams’s life history.” Williams, 542 F.3d at 1340.\\n\\n Jenkins’ counsel did even less than the attorneys in Williams and Debruce.\\n\\nThat is, Jenkins’ counsel conducted no investigation. Downey did not consider\\n\\nJenkins’ poor school records, mention Jenkins’ juvenile history, produce\\n\\ndocumentary evidence that Jenkins suffered from severe depression, or pursue\\n\\nrecords demonstrating Jenkins’ significant deficits in cognitive functions. We have\\n\\npreviously found that a counsel’s failure to investigate these rudimentary examples\\n\\nof mitigating evidence constitutes ineffective assistance of counsel. See Daniel v.\\n\\nComm’r, Ala. Dep’t of Corr., 822 F.3d 1248 (2016) (reversing and remanding\\n\\ndistrict court’s dismissal of the ineffective assistance of counsel claim and noting\\n\\nhow counsel should have more thoroughly investigated Daniel’s history of\\n\\nphysical and sexual abuse, his poor school records, and evidence of a potential\\n\\nintellectual disability).\\n\\n Given the lack of preparation, it is not surprising that Downey failed to call\\n\\nany witnesses to testify at the penalty stage about Jenkins’ childhood, thus failing\\n\\nto give a complete depiction of Jenkins’ “excruciating life history.” Wiggins v.\\n\\nSmith, 539 U.S. 510, 537 (2003) (finding deficient performance where counsel did\\n\\nnot adequately investigate petitioner’s life history for mitigating evidence); see\\n\\nalso Debruce, 758 F.3d at 1270, 1273–74; Johnson v. Sec’y, Dep’t of Corr., 643\\n\\n\\n 58\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 59 of 73\\n\\n\\nF.3d 907, 932 (11th Cir. 2011) (“No reasonable attorney, after being told by his\\n\\nclient that he had an abusive upbringing, would fail to interview members of his\\n\\nfamily who were readily available and could corroborate or refute the allegations\\n\\nof abuse.”).1 If Downey had effectively investigated—or investigated at all—he\\n\\nwould have been aware of several available and informed witnesses. Ultimately,\\n\\nonly one witness testified at the penalty phase: Lonnie Seal.2 Downey met with\\n\\nSeal for only fifteen minutes before the penalty phase began. That fifteen minutes\\n\\nrepresents the extent of Downey’s preparation for the penalty phase—a wholly\\n\\ninadequate amount of time under Supreme Court precedent. See Williams v.\\n\\nTaylor, 529 U.S. 362, 395 (2000) (finding mitigation investigation deficient where\\n\\n“counsel did not begin to prepare for that phase of the proceeding until a week\\n\\nbefore the trial”).\\n\\n Despite ample evidence of deficient performance, the Alabama Court of\\n\\nCriminal Appeals (CCA) concluded that Downey performed reasonably. The\\n\\nCCA’s decision rested on three explanations: (1) because Downey did not testify,\\n\\ncourts must presume he acted reasonably; (2) based on the record, Downey played\\n\\n\\n1\\n The majority goes to great pains to maintain that Downey did not know about Jenkins’ abusive\\nchildhood. This assertion is directly refuted by the record. Douglas Scofield, Jenkins’ lead\\nattorney, testified that Jenkins described his traumatic upbringing at meetings at which Downey\\nwas present.\\n2\\n It is unclear whether Downey contacted Seal and requested that he testify during the penalty\\nphase of trial. Given that Downey urged Scofield to conduct Seal’s direct examination because\\nScofield was better acquainted with Seal, it is likely that Scofield made the arrangements.\\n 59\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 60 of 73\\n\\n\\nan active role investigating and preparing for the penalty phase of trial; and (3)\\n\\nDowney’s decision to pursue a residual doubt strategy was reasonable.3 Because\\n\\neach of these findings constituted an unreasonable determination of the facts or an\\n\\nunreasonable application of clearly established federal law, Jenkins is entitled to\\n\\nhabeas relief. See 28 U.S.C. § 2254(d)(1).\\n\\n The CCA’s first rationale—presuming Downey acted reasonably because\\n\\nDowney did not testify at the Rule 32 hearing—was based on both an unreasonable\\n\\ndetermination of facts and an unreasonable application of clearly established law.\\n\\nThe CCA stated that “the record is virtually silent as to what actions were or were\\n\\nnot taken or what was or was not done by Mr. Downey at trial and why” and\\n\\ntherefore assumed that Downey’s actions “could have been reasonable and\\n\\nstrategic under the circumstances.” The majority likewise assumes Downey acted\\n\\nreasonably. This conclusion is based, in large part, on Downey’s failure to testify\\n\\nat the Rule 32 hearing. According to the majority, “[t]he problem for Jenkins is\\n\\nthat, like the state courts, this Court also does not know what Downey did or why.”\\n\\n But the record clearly illustrates Downey’s failure to investigate and overall\\n\\nlack of preparation. First, Downey’s fee declarations—which the majority\\n\\nconcedes were “detailed”—showed no investigation of mitigating circumstances,\\n\\nand no preparation for the penalty phase. Importantly, Downey’s fee declarations\\n\\n\\n3\\n The majority opinion endorsed the CCA’s first and third explanations.\\n 60\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 61 of 73\\n\\n\\nwere exceptionally detailed, recounting his work in five-minute increments. The\\n\\nsingle reference to the “mitigating” phase of trial in the declarations concerned a\\n\\nmeeting with Scofield that occurred only a week before trial. Other than that\\n\\nvague entry, no notation indicates that Downey did anything to prepare for the\\n\\npenalty phase of trial during the nearly two years he represented Jenkins.\\n\\n Second, and more persuasively, Douglas Scofield, Jenkins’ lead attorney,\\n\\nexplicitly testified that Downey told him that Downey had not investigated\\n\\npotential mitigating evidence. Both the Alabama courts and the majority ignore\\n\\nthat fact and instead focus on Scofield’s testimony that, “[a]part from what he has\\n\\ntold me, I don’t know what he has done.” This statement, to which the majority\\n\\nassigns great weight, is hardly illuminating. Of course Scofield knows only what\\n\\nDowney told him. And what Downey told him was this: Downey did not\\n\\ninvestigate potentially mitigating evidence before the penalty phase.\\n\\n In the face of this direct evidence of deficient performance, the Alabama\\n\\ncourts and the majority pivot to Downey’s failure to testify at the Rule 32 hearing.\\n\\nWithout his testimony, the majority reasons, we must assume that Downey’s\\n\\ndecision not to investigate was strategic and therefore his representation\\n\\nreasonable. I disagree, and I have deep reservations about the practical effects of\\n\\nthe majority’s understanding of the law. The majority suggests (if not holds) that\\n\\nthe absence of testimony from an allegedly deficient attorney per se means that the\\n\\n\\n 61\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 62 of 73\\n\\n\\nattorney’s actions were reasonable. It is true that “counsel is strongly presumed to\\n\\nhave rendered adequate assistance,” Strickland, 466 U.S. at 690. But a petitioner\\n\\nmay overcome that burden with evidence of inadequate assistance. And that\\n\\nevidence may take various forms. To satisfy the burden, it is not necessary for the\\n\\nallegedly deficient counsel to testify and admit wrongdoing. Nor is it necessary for\\n\\nthe record to reflect counsel’s “thoughts and intentions as he prepared for the\\n\\npenalty phase,” as the majority seems to require. When evaluating whether a\\n\\ncounsel’s performance was deficient, we ask whether his performance “fell below\\n\\nan objective standard of reasonableness.” Strickland, 466 U.S. 687–88. We do not\\n\\ncontemplate what counsel thought or intended. We consider whether his\\n\\nperformance was objectively unreasonable. See id. Because the Alabama courts\\n\\ndid not apply the proper standard, their conclusion was an unreasonable application\\n\\nof clearly established law. See id.\\n\\n The CCA’s second reason for denying Jenkins relief was also based on an\\n\\nunreasonable application of clearly established federal law. The CCA concluded\\n\\nthat Downey had adequately investigated and prepared for the penalty phase of\\n\\ntrial, emphasizing that Downey met with Jenkins on several occasions and spoke\\n\\nwith Jenkins’ grandmother. Based on these findings, the CCA determined that\\n\\nDowney provided adequate assistance of counsel during the penalty phase. This is\\n\\nan unreasonable application of clearly established law that “[a]n attorney has a\\n\\n\\n 62\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 63 of 73\\n\\n\\nduty to conduct a reasonable investigation, including an investigation of the\\n\\ndefendant’s background, for possible mitigating evidence.” Porter v. Singletary,\\n\\n14 F.3d 554, 557 (11th Cir. 1994).\\n\\n As previously discussed, it is established in the record that Downey knew of\\n\\nJenkins’ horrible upbringing yet still failed to investigate potentially mitigating\\n\\nevidence. “A reasonable investigation . . . should have included, at a minimum,\\n\\ninterviewing other family members who could corroborate the evidence of abuse\\n\\nand speak to the resulting impact on [Jenkins].” Williams, 542 F.3d at 1340. But\\n\\nDowney did not conduct a single interview—of a family member or otherwise—in\\n\\npreparation of the penalty phase.4 Because Downey did not conduct an\\n\\ninvestigation, let alone a reasonable investigation, the CCA’s conclusion that\\n\\nDowney adequately investigated potentially mitigating evidence was an\\n\\nunreasonable application of clearly established law.\\n\\n The CCA’s third rationale for its decision—that Jenkins’ counsel reasonably\\n\\nand strategically decided to focus on residual doubt—was both an unreasonable\\n\\napplication of clearly established law and an unreasonable determination of the\\n\\nfacts. In Williams v. Taylor, the Supreme Court held that otherwise reasonable\\n\\ntrial strategies cannot be justified unless counsel first conducted a constitutionally\\n\\n\\n4\\n Downey’s fee declaration reflects one telephone call he had with Jenkins’ grandmother, Doris\\nWagoner. But that call occurred 18 months before trial and, according to Wagoner, concerned a\\nrequest for money.\\n 63\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 64 of 73\\n\\n\\nadequate investigation. 529 U.S. at 395–96. And here, Downey conducted no\\n\\nsuch investigation. It cannot be said, therefore, that Downey’s decision—if one\\n\\ncan even call it that—to pursue a residual doubt strategy was strategically\\n\\nreasonable. See id. The CCA’s conclusion to the contrary was an unreasonable\\n\\napplication of clearly established law.\\n\\n The conclusion was likewise an unreasonable determination of the facts. On\\n\\ndirect appeal, the Alabama CCA noted that “[t]he evidence overwhelmingly\\n\\npointed to the guilt of the appellant.” Jenkins v. State, 627 So. 2d 1054 (Ala. Crim.\\n\\nApp. Feb. 27, 1992). Then, on appeal from Jenkins’ Rule 32 petition, the CCA\\n\\nagain acknowledged how “[t]he evidence establishing Jenkins’s guilt was\\n\\noverwhelming.” Jenkins v. State, 972 So. 2d 111, 157 (Ala. Crim. App. Feb. 27,\\n\\n2004); see also id. at 156 (“The amount of evidence incriminating Jenkins . . . was\\n\\noverwhelming.”). Despite the “overwhelming” evidence of Jenkins’ guilt, the\\n\\nCCA concluded that Jenkins’ counsel acted reasonably in presenting a residual\\n\\ndoubt defense during the penalty phase of trial. But no reasonable attorney would\\n\\nemploy a residual doubt strategy in the face of such overwhelming evidence. See\\n\\nJohnson, 643 F.3d at 932–33. The Alabama courts’ contrary conclusion, therefore,\\n\\nwas an unreasonable determination of the facts.\\n\\n\\n\\n\\n 64\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 65 of 73\\n\\n\\n B. Prejudice\\n\\n Jenkins’ Rule 32 petition also pleaded sufficient facts to show that his\\n\\ncounsel’s deficient performance prejudiced his defense, satisfying the second\\n\\nStrickland prong. To establish prejudice, a petitioner must show a “reasonable\\n\\nprobability that, but for counsel’s unprofessional errors, the result of the\\n\\nproceeding would have been different.” Strickland, 466 U.S. at 694. Here, the\\n\\njury voted in favor of the death penalty by the slimmest margin permitted under\\n\\nAlabama law: ten votes in favor and two votes against. Jenkins need only\\n\\nestablish, then, “a reasonable probability that at least one juror would have struck\\n\\na different balance.” Wiggins, 539 U.S. at 537 (emphasis added). In measuring\\n\\nthat probability, we “evaluate the totality of the available mitigation evidence—\\n\\nboth that adduced at trial, and the evidence adduced in the habeas proceeding in\\n\\nreweighing it against the evidence in aggravation.” Williams, 529 U.S. at 397.\\n\\n The testimony elicited at the Rule 32 hearing was horrendous. Four family\\n\\nmembers offered detailed, consistent testimony about Jenkins’ traumatic\\n\\nchildhood. The majority describes that testimony, highlighting some of the most\\n\\nappalling facts elicited at the Rule 32 hearing. For example, Jenkins’ stepfather\\n\\nwould beat Jenkins on a semi-daily basis, would lock him in a room for hours\\n\\nwithout food, and even made Jenkins eat his own feces. In addition to the\\n\\nmitigating facts detailed by the majority, the following came to light during the\\n\\n\\n 65\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 66 of 73\\n\\n\\nRule 32 hearing. Jenkins’ mother had an affair while her husband, Jenkins’\\n\\nstepfather, was in prison. When Jenkins’ stepfather was released, he took out his\\n\\nfrustrations about the affair on Jenkins. Because Jenkins’ biological father was\\n\\nJenkins’ stepfather referred to Jenkins in demeaning names like “Mexican trash,”\\n\\nand “Puerto Rican puke.” The stepfather also called him “piece of shit” and\\n\\n“bastard.”\\n\\n The Rule 32 testimony also detailed the alcohol substance abuse, drug\\n\\naddiction, and incessant domestic violence that surrounded Jenkins’ childhood.\\n\\nAnd, most shockingly, the testimony described how Jenkins’ grandfather sexually\\n\\nabused Jenkins when Jenkins was just four years old. As a result of that sexual\\n\\nabuse, Jenkins, suffered bladder and bowel movement issues. Jenkins’\\n\\nincontinence became a trigger for his stepfather: when Jenkins had an accident, his\\n\\nstepfather would beat him or subject him to severe degradation. For example,\\n\\nJenkins’ stepfather would hang Jenkins’ soiled sheets in the front yard for all the\\n\\nneighbors to see, make Jenkins wear his soiled clothes on his head, and, most\\n\\ndisturbingly, make Jenkins eat his own feces. The only way to escape the abuse\\n\\nwas by running away. So, by age twelve, Jenkins lived on the streets.5\\n\\n\\n\\n5\\n The CCA found parts of the family members’ testimony not credible, though it did not\\nexplicitly discredit testimony describing some of the worst abuse. Given the grotesque nature of\\nthe family members’ testimony, and the fact that their testimony was consistent, there is\\nreasonable probability that at least one more juror would have been compelled to vote against a\\ndeath sentence.\\n 66\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 67 of 73\\n\\n\\n Jenkins also introduced medical, school, and juvenile court records from his\\n\\nchildhood, which shed further light on his difficult childhood and intellectual\\n\\ndeficiencies. Juvenile records show Jenkins as “an emotionally disturbed child”\\n\\nwith “psychological problems” and expressed “grave concerns for the emotional\\n\\nwill being of this adolescent”. Moreover, two psychologists testified about\\n\\nJenkins’ history of psychological trauma stemming from his childhood abuse—\\n\\nweighty evidence that is not even considered in the majority’s discussion. Finally,\\n\\ntwo prison guards testified that Jenkins was the ideal prisoner during pretrial\\n\\ndetention. One of the guards even described Jenkins as the best inmate he had\\n\\nevery supervised.\\n\\n Given this compelling mitigating evidence, it is not difficult to conclude that\\n\\nJenkins established “a reasonable probability that at least one juror would have\\n\\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537; see\\n\\nalso Daniel, 822 F.3d at 1276 (listing several factors that support a prejudice\\n\\nfinding, including (1) evidence of impaired intellectual functioning; (2) evidence of\\n\\nchildhood sexual abuse; and (3) the fact that the jury voted only 10–2 in\\n\\nrecommending a death sentence).\\n\\n I. Evidentiary Hearing\\n\\n Jenkins also argues that he is entitled to habeas relief under Atkins v.\\n\\nVirginia, 563 U.S. 304 (2002). He contends that he satisfies the three requirements\\n\\n\\n 67\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 68 of 73\\n\\n\\nnecessary to establish intellectual disability—(1) subaverage intellectual\\n\\nfunctioning; (2) deficits in adaptive behavior; and (3) onset before the age of 18,\\n\\nsee Ex parte Perkins, 851 So. 2d 453, 456 (Ala. 2002)—thus, exempting him from\\n\\nimplementation of the death penalty. Alternatively, Jenkins argues that we should\\n\\nremand for an evidentiary hearing on the Atkins issue.\\n\\n The majority dismisses each argument. First, it concludes that the CCA did\\n\\nnot act unreasonably in concluding that Jenkins failed to satisfy any of the Perkins\\n\\nprongs. Second, it determines that Jenkins is not entitled to an evidentiary hearing\\n\\nbecause he “did not diligently attempt to develop the factual basis for this claim in\\n\\nthe state court.” I disagree.\\n\\n First, it is not possible to undergo a substantive Atkins analysis based on the\\n\\nrecord before us. For one, the CCA’s evaluation of Jenkins’ Atkins claim was a\\n\\nmere three sentences. And, more significantly, Jenkins never had an opportunity to\\n\\npresent evidence on his Atkins claim. To deny Jenkins’ claim, therefore, the\\n\\nmajority relies on evidence from unrelated parts of the record. Such an approach is\\n\\nimproper because, “[p]rior to Atkins, [the petitioner] could not have been expected\\n\\nto necessarily present evidence sufficient to support an Atkins claim because such\\n\\nevidence constituted a double edged sword.” Burgess v. Comm’r, Ala. Dep’t of\\n\\nCorr., 723 F.3d 1308, 1320 (11th Cir. 2013) (internal citation omitted). The\\n\\nevidence was a double-edged sword, we explained, because prior to Atkins,\\n\\n\\n 68\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 69 of 73\\n\\n\\nevidence of intellectual disability could suggest that a defendant would be\\n\\ndangerous in the future; therefore, “a defendant could reasonably decide not to\\n\\nhighlight his mental retardation.” Id. at 1318. For that reason, courts should not\\n\\nrely on pre-Atkins evidence—evidence that was presented in a wholly distinct\\n\\ncontext—to support a finding that a petitioner is not intellectually disabled. So,\\n\\nhere, instead of attempting to make a medical diagnosis based on an insufficient\\n\\nrecord, we should remand for an evidentiary hearing. The majority refuses to do\\n\\nso, reasoning that “Jenkins did not diligently attempt to present his Atkins claim in\\n\\nthe state courts.” I disagree.\\n\\n “[I]f the petitioner was diligent in developing the record in the state habeas\\n\\nproceedings, ‘a federal court may grant an evidentiary hearing without further\\n\\nregard for the provisions of § 2254(3)(2).” Burgess, 723 F.3d at 1320 (citation\\n\\nomitted). “‘Diligence [for purposes of § 2254(d)] depends on whether the prisoner\\n\\nmade a reasonable attempt, in light of the information available at the time, to\\n\\ninvestigate and pursue claims in state court,’ and ‘will require in the usual case that\\n\\nprisoner, at a minimum, seek an evidentiary hearing in state court.’” Id. (quoting\\n\\nWilliams, 529 U.S. at 435 (emphasis added)).\\n\\n The majority concludes that Jenkins failed to diligently pursue his Atkins\\n\\nclaim in state court. But again, based on binding precedent and the record, I must\\n\\npart company with the majority. In conducting its analysis, the majority does not\\n\\n\\n 69\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 70 of 73\\n\\n\\nadequately consider whether Jenkins “made a reasonable attempt, in light of the\\n\\ncircumstances available at the time, to investigate and pursue claims in state\\n\\ncourt.” Id. (emphasis added) (quoting Williams, 529 U.S. at 435). Specifically, the\\n\\nmajority did not consider the unusual procedural posture of Jenkins’ claim.\\n\\n The Supreme Court issued Atkins v. Virginia while Jenkins’ petition for\\n\\npost-conviction collateral relief was pending before the CCA. The CCA, therefore,\\n\\nrequested supplemental briefing on the effect, if any, Atkins had on Jenkins’\\n\\npetition. At that point, neither the Alabama Supreme Court nor the Alabama\\n\\nlegislature had interpreted the groundbreaking case, making Jenkins’ counsel’s\\n\\nnext step unclear. Jenkins’ counsel nonetheless made a reasonable attempt to\\n\\npursue Jenkins’ Atkins claim—he requested that the CCA “vacate his death\\n\\nsentence and remand the case back to the lower court for further proceedings.”\\n\\nThe request appears even more reasonable when considered alongside procedurally\\n\\nanalogous cases decided around the same time.\\n\\n At the time Jenkins’ counsel was preparing the supplemental briefing, there\\n\\nhad been three other post-conviction petitions for relief that shared Jenkins’\\n\\nunusual procedural posture. See Clemons v. State, 55 So. 3d 314 (Ala. Crim. App.\\n\\nAug. 29, 2003); Wood v. State, 891 So. 2d 398 (Ala. Crim. App. Apr. 25, 2003);\\n\\nTarver v. State, 940 So. 2d 312 (Ala. Crim. App. Feb. 27, 2004). These three cases\\n\\nare similar to Jenkins’ petition in that (1) the petitioners were sentenced to death,\\n\\n\\n 70\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 71 of 73\\n\\n\\n(2) the petitioners filed Rule 32 petitions, (3) their Rule 32 petitions were\\n\\ndismissed at the trial court level, (4) while their appeals before the CCA were\\n\\npending, Atkins was decided, and (5) given the Atkins decision, each petitioner\\n\\nsubmitted supplemental briefing. In their supplemental briefs, Clemons, Wood,\\n\\nand Tarver each argued that he was intellectually disabled under Atkins and thus it\\n\\nwould be unconstitutional to execute him. None of the petitioners asked for an\\n\\nevidentiary hearing on their Atkins claim or requested a remand for further\\n\\nproceedings. Despite the lack of request for a hearing, the CCA remanded for an\\n\\nevidentiary hearing in each case.\\n\\n With these cases in mind, it would have been reasonable for Jenkins to think\\n\\nthat broadly stating his Atkins claim would constitute “diligent” pursuit of his\\n\\nclaim. Nonetheless, Jenkins did more than these other petitioners: he specifically\\n\\nrequested that the CCA “vacate his death sentence and remand the case back to the\\n\\nlower court for further proceedings.” Given “the information available at the\\n\\ntime,” Jenkins’ made a “reasonable attempt” to pursue his claims.\\n\\n I also cannot come to terms with the majority’s ultimate conclusion. The\\n\\nmajority determines that Jenkins did not act diligently in pursuing his Atkins claim\\n\\nbecause “[t]he record clearly shows that Jenkins did not [seek an evidentiary\\n\\nhearing].” But Jenkins did ask the CCA to “remand his case for further\\n\\nproceedings.” And, in Alabama, post-conviction relief petitioners with factually\\n\\n\\n 71\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 72 of 73\\n\\n\\nsufficient claims are entitled to an evidentiary hearing. Ala. R. Crim. P. 32.9. So,\\n\\nif the CCA had granted Jenkins’ request to remand “for further proceedings,”\\n\\nJenkins would have received an evidentiary hearing. Jenkins’ request that the\\n\\nCCA “remand for further proceedings,” then, was effectively equivalent to a\\n\\nrequest for an evidentiary hearing. The majority’s superficial conclusion to the\\n\\ncontrary misses the point.\\n\\n II. Conclusion\\n\\n Jenkins’ Rule 32 evidentiary hearing spanned three days. Four family\\n\\nmembers testified about his horrific upbringing. Two psychiatrists testified about\\n\\nhow that upbringing affected Jenkins as an adult. Two prison guards testified\\n\\nabout Jenkins’ exemplary behavior while in pretrial detention. And one friend\\n\\ntestified about Jenkins’ character. Evidence introduced also documented Jenkins’\\n\\nacademic failings and his juvenile record. In short, Jenkins presented an\\n\\nabundance of compelling mitigating evidence. But the jury never heard any of that\\n\\nevidence. Downey failed to introduce it at the penalty phase. In fact, Downey\\n\\nfailed to investigate any mitigating evidence whatsoever. Downey’s failures\\n\\ncannot be attributed to strategic decisions. Nor can they be called reasonable. His\\n\\nperformance can only be described as deficient. And that deficient performance\\n\\nwas surely prejudicial—there is more than a reasonable probability that the\\n\\nhorrifying and detailed Rule 32 testimony would have persuaded one juror to vote\\n\\n\\n 72\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 73 of 73\\n\\n\\nagainst sentencing Jenkins to death. Jenkins is therefore entitled to habeas relief\\n\\non his ineffective assistance of counsel claim.\\n\\n Jenkins is also entitled to a hearing on his Atkins claim. Though Jenkins\\n\\npresented ample mitigating evidence at the Rule 32 hearing, he never had an\\n\\nopportunity to present evidence concerning (1) his intellectual functioning, (2) his\\n\\nadaptive deficits, or (3) whether his potential intellectual disability manifested\\n\\nbefore Jenkins reached 18—the three requirements necessary to establish\\n\\nintellectual disability and thus exempt an individual from a death sentence. The\\n\\nCCA nonetheless decided to evaluate Jenkins’ Atkins claim, apparently believing\\n\\nthat the record, which included no direct evidence concerning whether Jenkins was\\n\\nintellectually disabled, was sufficient to resolve the complex medical issue. In just\\n\\nthree sentences, the CCA arrived at a diagnosis: not intellectually disabled.\\n\\n Despite the insufficient factual record, the majority declines to remand with\\n\\ninstructions to conduct an evidentiary hearing—what I view as the proper\\n\\nrecourse—because Jenkins did not use the phrase “evidentiary hearing” when\\n\\nrequesting further proceedings on his Atkins claim in the Alabama courts. Such a\\n\\nrigid understanding of the standard should not bar a death row petitioner from\\n\\nbringing a constitutional, and potentially lifesaving, claim. I therefore dissent.\\n\\n\\n\\n\\n 73\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435142/\", \"author_raw\": \"WILSON, Circuit Judge, dissenting\"}]}"},"judge1":{"kind":"string","value":"TJOFLAT"},"judge2":{"kind":"string","value":"WILSON"},"judge3":{"kind":"string","value":"BRANCH"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"TJOFLAT\": \", Circuit\", \"WILSON\": \", Circuit\", \"BRANCH\": \", Circuit\"}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":1,"string":"1"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4657889/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Mark Jenkins v. Commissioner, Alabama Department of Corrections\nCourt: U.S. Court of Appeals for the Eleventh Circuit\nDate decided: 2019-08-30\nOpinion type: majority\nHas dissenting opinion: Yes\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nCase: 17-12524 Date Filed: 08/30/2019 Page: 1 of 73\n\n\n [PUBLISH]\n\n IN THE UNITED STATES COURT OF APPEALS\n\n FOR THE ELEVENTH CIRCUIT\n\n\n No. 17-12524\n\n\n\n D.C. Docket No. 4:08-cv-00869-VEH\n\nMARK ALLEN JENKINS,\n\n Petitioner - Appellant,\n\n versus\n\n\n\nCOMMISSIONER, ALABAMA DEPARTMENT OF CORRECTIONS,\n\n Respondent - Appellee.\n\n\n\n Appeal from the United States District Court\n for the Northern District of Alabama\n\n\n (August 30, 2019)\n\nBefore TJOFLAT, WILSON, and BRANCH, Circuit Judges.\n\nBRANCH, Circuit Judge:\n\n Mark Allen Jenkins, an Alabama prisoner sentenced to death for the 1989\n\nmurder of Tammy Ruth Hogeland, appeals the district court’s denial of his petition\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 2 of 73\n\n\nfor a writ of habeas corpus, 28 U.S.C. § 2254. Before us are Jenkins’s arguments\n\nthat he received ineffective assistance of counsel during the penalty phase of his\n\ntrial and that he is intellectually disabled and therefore ineligible for the death\n\npenalty. After careful consideration, and with the benefit of oral argument, we\n\naffirm the denial of Jenkins’s habeas petition on both grounds.\n\n I. BACKGROUND\n\n A. The Crime and Arrest\n\n We summarize the following background narrative from opinions of the\n\nAlabama Court of Criminal Appeals and our own review of the record. See Jenkins\n\nv. State, 627 So. 2d 1034, 1037–40 (Ala. Ct. Crim. App. 1992); Jenkins v. State,\n\n972 So. 2d 111, 119–20 (Ala. Ct. Crim. App. 2004). The events leading up to the\n\nmurder began the evening of April 17, 1989. Jenkins was at the home of his\n\nacquaintance Christine Nicholas. He had met her several months earlier at the\n\nOmelet Shoppe restaurant where she worked. Jenkins was very intoxicated and\n\ntried to seduce her. When she resisted, Jenkins got “real mad” and asked her\n\nseveral times what she would do if someone came up from behind her and grabbed\n\nher.\n\n At around 1 a.m., Jenkins and Nicholas drove to the Riverchase Omelet\n\nShoppe. Jenkins went inside and talked with one of the waitresses, Frieda Vines.\n\nThe manager, Douglas Thrash, recognized Jenkins as a regular customer who knew\n\n\n 2\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 3 of 73\n\n\nall of the waitresses. Thrash overheard part of Jenkins’s conversation with Vines\n\nand heard mention of the Omelet Shoppe location near the Birmingham Airport.\n\nJenkins and Nicholas then returned to Nicholas’s home. Around 2 a.m., Nicholas’s\n\nmother asked Jenkins to leave. He did so, falling down some steps and ramming\n\nhis car into another vehicle in the process.\n\n Meanwhile, 23-year-old Tammy Hogeland was making her way to work at\n\nthe Omelet Shoppe. Her sister Wendy, along with Tammy’s young son, picked\n\nTammy up from her college classes and drove her to the Riverchase Omelet\n\nShoppe a little after 9:30 p.m. Sometime after 10:00, Tammy was sent to work at\n\nthe airport Omelet Shoppe on Tenth Avenue, since that location was unexpectedly\n\nshorthanded. Tammy was wearing her watch, a necklace with the words “special\n\nsister,” her class ring, and her diamond engagement ring. Tammy was working as a\n\ncook that night and was wearing a blue apron, black pants, a white shirt, and white\n\nshoes. She and Sarah Harris were the only employees working at the airport\n\nOmelet Shoppe that night.\n\n At about 2 a.m. on April 18, Harris saw Jenkins—whom she did not know—\n\ndrive up in a red sports car. His arrival was memorable to Harris because the car\n\nnearly jumped the curb and crashed through the restaurant’s glass wall. Jenkins got\n\nout of the car and came into the restaurant, appearing to be intoxicated. He walked\n\nover to Hogeland and began talking to her. Harris later saw Jenkins and Hogeland\n\n\n 3\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 4 of 73\n\n\ndrive away together in the red sports car. That was the last time anyone who knew\n\nHogeland ever saw her. Hogeland left behind her cigarettes, lighter, purse, and\n\npaycheck. She left without telling anyone, which she had never done before.\n\nAlthough Harris saw them drive off, she could not later say whether Hogeland left\n\nwith Jenkins willingly or was instead abducted.\n\n At around 5 a.m. that morning, Bobby and Geraldine Coe had stopped to\n\nbuy gas at a Chevron station on I-59 northeast of Birmingham when they saw\n\nJenkins drive up in the red sports car. The Coes noticed a female who appeared to\n\nbe “passed out” in the front passenger seat, but they could not say whether she was\n\nalive or dead. While Bobby Coe was pumping gas, Jenkins asked him for some\n\ncigarettes and said, “Looks like it’s been a long night and it looks like it’s going to\n\nbe a long day.” Jenkins then said “God bless you” before asking directions to I-\n\n459. Coe gave him directions and got back in his car. As Coe pulled out onto the\n\ninterstate, he saw Jenkins follow him in the red sports car. Coe then saw the car\n\nflash its lights, slow down, and pull to the side of the road between mile markers\n\n151 and 152.\n\n At 8 a.m., Wendy Hogeland learned that Tammy was not at the restaurant,\n\nand their mother called the police. Meanwhile, Jenkins went to the home of Steve\n\nMusser, who noticed that he was wearing the same clothes as the day before.\n\nJenkins told Musser that his car had been stolen the night before and asked him if\n\n\n 4\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 5 of 73\n\n\nhe would say that he had been with him all night. Musser refused. Christine\n\nNicholas also saw Jenkins in a grocery store that morning. He was looking at a\n\nnewspaper, making a phone call, and attempting to sell his Buick. Nicholas loaned\n\nhim $4 for gas. At around 10 o’clock that morning, Jenkins sold his car to Michael\n\nBrooks, a mechanic at a local Chevron station. Jenkins had explained that he\n\nneeded the money so he could visit his sick mother in California. Another\n\nmechanic at the service station drove Jenkins to the Greyhound bus station later\n\nthat morning.\n\n Jenkins awoke the next day on the Greyhound bus in Houston and was\n\nejected from the bus because his fare was used up. He then hitchhiked from\n\nHouston to Phoenix, then to San Diego and Los Angeles. Jenkins was first\n\nidentified as a suspect in Hogeland’s disappearance by Officer Mike Weems of the\n\nHoover Police Department on April 19. Weems ate dinner at the Riverchase\n\nOmelet Shoppe just about every day and had talked with Hogeland just as often.\n\nHe also knew Jenkins from the restaurant. Weems learned of Hogeland’s\n\ndisappearance from the other waitresses. Remembering how Jenkins often talked to\n\nHogeland and passed her notes,1 he gave Jenkins’s name to the missing persons\n\ninvestigator. The investigator also learned that a red Mazda RX-7 sports car, which\n\n\n\n1\n The jury was not permitted to hear, and the record does not reveal, what exactly Jenkins had\nsaid to Hogeland that made Officer Weems suspect him in her disappearance.\n 5\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 6 of 73\n\n\nhad been reported stolen from the service station where Jenkins worked, had been\n\nrecovered on I-459.\n\n The afternoon of April 21, a truck driver who had happened to stop on I-59\n\nnear mile marker 151 discovered Hogeland’s body off the side of the road. She was\n\nnaked, wearing only a watch, and was so badly decomposed that the body had to\n\nbe identified by dental records. From the fractured hyoid bone, it was determined\n\nthat she was manually strangled to death. Also found at the scene were Hogeland’s\n\napron, shoes, bra, panties, pants, and hair net, as well as some beer cans, a Mazda\n\nRX-7 owner’s manual, and other items later determined to have come from the red\n\nMazda. Her necklace and rings were never found.\n\n Alabama authorities issued a warrant for Jenkins’s arrest and also obtained a\n\nfederal fugitive warrant. An FBI agent arrested Jenkins on May 1 in front of his\n\nuncle’s house in Wilmington, California. Jenkins’s uncle later gave the police three\n\nbags of Jenkins’s personal effects. Fibers recovered from Hogeland’s body and\n\nclothing matched those of the Mazda, and hairs matching Hogeland’s were found\n\nin the Mazda. Similarly, fibers from Jenkins’s clothing placed him in the Mazda,\n\nand fibers of his jeans were found on Hogeland’s apron. A bootprint near\n\nHogeland’s body matched a combat boot from among the belongings Jenkins’s\n\nuncle gave to the police, and Jenkins’s roommate said the boots looked just like\n\nthose that Jenkins wore.\n\n\n 6\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 7 of 73\n\n\n A business card like those that had been in the Mazda’s glove box,\n\nbelonging to an owner of the Mazda, was recovered from Jenkins’s wallet after his\n\narrest in California. Later, Jenkins’s cellmate in the Alabama jail reported that\n\nJenkins told him “he had done the crime” and was worried that the couple from the\n\ngas station would identify him or that the police would find his fingerprint on a\n\nbeer can at the scene. A St. Clair County grand jury indicted Jenkins for capital\n\nmurder.\n\n B. The Trial and Sentencing\n\n Jenkins was represented at trial in 1991 by attorneys Douglas Scofield and\n\nStan Downey. Scofield, the lead attorney, had been practicing criminal defense in\n\nBirmingham since 1984 but had never before tried a capital murder case. Jenkins’s\n\nlandlord and his grandmother had arranged the referral and representation before\n\nthe court appointed him. Scofield agreed to the appointment on the condition that a\n\nSt. Clair County lawyer also be appointed. That lawyer was Downey, who had\n\nmore experience than Scofield but whose capital experience was limited to one\n\ncase that had not gone to trial. The two attorneys had agreed that Downey would\n\nbe responsible for jury selection, but Scofield eventually took over during the voir\n\ndire. Per their agreement, Scofield was responsible for the guilt-phase trial,\n\nalthough Downey participated by interviewing Jenkins several times. Scofield also\n\nmet with Jenkins around a dozen times. Jenkins had told him about his background\n\n\n 7\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 8 of 73\n\n\nand childhood, including the fact that he had had a difficult childhood and had\n\nlived on the streets since age 11.\n\n Throughout the weeklong trial, Scofield pursued an outright acquittal. In\n\nparticular, he challenged Harris’s identification of Jenkins as the man who drove\n\naway with Hogeland, and he challenged the medical examiner’s opinions about the\n\ntime and cause of Hogeland’s death. In his closing argument, Scofield took a wide-\n\nranging approach to undermining the State’s wholly circumstantial case against\n\nJenkins. He attacked the State’s timeline of the crime, questioning whether Jenkins\n\nin his intoxicated state would have had the time and the ability to leave Nicholas’s\n\nhouse, steal the Mazda, and abduct and kill Hogeland. He reminded the jury of the\n\nweakness of Harris’s identification of Jenkins and the Coes’ identification of the\n\nwoman in the car. He pointed out that no physical evidence definitively linked\n\nJenkins to the Mazda, to Hogeland’s body, or to her jewelry, which was never\n\nrecovered. And he argued that the evidence provided no plausible motive for\n\nJenkins to have kidnapped, robbed, or killed Hogeland. In sum, he urged the jury\n\nto find reasonable doubt and acquit Jenkins. At the very least, he argued, they\n\nshould find that the State failed to prove Jenkins’s intent to rob and kidnap\n\nHogeland, which were necessary to convict Jenkins of capital murder. He further\n\nargued that, if the jury thought Jenkins killed Hogeland as a crime of passion, they\n\ncould convict him only of manslaughter. The jury deliberated for three hours and\n\n\n 8\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 9 of 73\n\n\nfifteen minutes before convicting Jenkins of capital murder, Ala. Code § 13A-5-\n\n40(a)(1) & (2), i.e., murder committed in the course of kidnapping and robbery.\n\n The penalty phase took place that same afternoon. A week before trial,\n\nScofield and Downey had discussed the penalty phase, and Scofield understood\n\nthat Downey would be responsible for handling the penalty phase. Nonetheless, the\n\nwitness who was present to testify on Jenkins’s behalf during the penalty phase\n\nwas better acquainted with Scofield, so Downey suggested that afternoon that\n\nScofield conduct the direct examination. Scofield urged Downey to interview the\n\nwitness quickly and prepare to examine him, and Downey did so.\n\n Downey handled the penalty presentation to the jury. The State did not\n\npresent any additional aggravation evidence, resting on the evidence presented\n\nduring the guilt phase. One witness, Lonnie Seal, testified on Jenkins’s behalf in\n\norder, in Downey’s words, to “reveal another side of Mark Jenkins to you that you\n\ndon’t know anything about.” Seal had met Jenkins in Fontana, California, in 1988\n\nwhen they both worked at the same garage. Seal and his wife became friends with\n\nJenkins, and he visited in their home. When the Seals decided to move to Alabama\n\na few months later, Jenkins agreed to help with their move and to live with the\n\nSeals in Alabama for a time. It took Lonnie Seal several weeks to find a job, but\n\nJenkins found a job in two days and contributed toward the family’s rent and\n\n\n\n\n 9\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 10 of 73\n\n\ngroceries. Jenkins continued to visit the Seals after he moved into his own home,\n\nand he would always bring a small gift for and play with the Seals’ infant son.\n\n Downey’s closing argument to the jury requested a sentence of life without\n\nparole. Downey reminded the jurors that they had agreed during voir dire that\n\ndeath was not always the appropriate punishment for homicide. Quoting the Bible,2\n\nhe reminded the jury that the Old Testament permitted capital punishment only\n\nwhen supported by the testimony of two or three eyewitnesses, whereas in this case\n\nthere had been only circumstantial evidence. He asserted from Seal’s testimony\n\nthat “Mark had a side of him that was generous, that was giving, and that was\n\nkind,” and argued that the deterrent and retributive purposes of capital punishment\n\nwould not be served in this case.\n\n The jury deliberated for 50 minutes before returning its nonbinding\n\nrecommendation of death by a vote of 10 to 2. See Ala. Code § 13A-5-46(e) (1981)\n\n(amended 2017 to make jury’s verdict binding). After a separate hearing, the court\n\nfound the two statutory aggravating factors for which the State had argued: that the\n\nmurder was committed during the commission of a robbery, and that it was\n\ncommitted during the commission of a kidnapping. See Ala. Code § 13A-5-49(4).\n\n\n\n\n2\n “If anyone kills a person, the murderer shall be put to death on the evidence of witnesses. But\nno person shall be put to death on the testimony of one witness.” Numbers 35:30 (ESV). Downey\nalso paraphrased Deuteronomy 17:6 as “On the evidence of two witnesses or three witnesses, he\nis [sic] shall be put to death. He shall not be put to death on the evidence of one witness.”\n 10\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 11 of 73\n\n\nThe court also found two statutory mitigating factors: that Jenkins had no\n\nsignificant history of prior criminal activity, and his age at the time of the crime\n\n(21). See id. § 13A-5-51(1), (7). The court considered but rejected the statutory\n\nmitigation that, due to his consumption of alcohol, Jenkins’s capacity to appreciate\n\nthe criminality of his conduct or to conform his conduct to the requirements of law\n\nwas substantially impaired. See id. § 13A-5-51(6). The court found that the\n\naggravating circumstances outweighed the mitigating circumstances and, in\n\naccordance with the jury’s recommendation, imposed a sentence of death.\n\n C. The State and Federal Post-Conviction Proceedings\n\n On direct appeal, the Alabama Court of Criminal Appeals affirmed Jenkins’s\n\nconviction and death sentence. Jenkins v. State, 627 So. 2d 1034 (Ala. Ct. Crim.\n\nApp. 1992), aff’d, 627 So. 2d 1054 (Ala. 1993), cert. denied, 511 U.S. 1012\n\n(1994). In 1995, Jenkins filed his first petition for state postconviction relief, Ala.\n\nR. Crim. P. 32,3 alleging, among many other things, that he received ineffective\n\nassistance of counsel during his penalty phase. The Rule 32 trial court conducted a\n\nthree-day hearing at which evidence was presented about the performance of\n\nJenkins’s attorneys.\n\n\n\n\n3\n “Subject to the limitations of Rule 32.2, any defendant who has been convicted of a criminal\noffense may institute a proceeding in the court of original conviction to secure appropriate relief\non the ground that . . . [t]he constitution of the United States or of the State of Alabama requires\na new trial, a new sentence proceeding, or other relief. . . .” Ala. R. Crim. P. 32.1.\n 11\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 12 of 73\n\n\n Scofield testified about the representation that he and Downey provided. In\n\nrelevant part, Scofield asserted that he and Downey had “decided that [developing\n\nmitigation for the penalty phase] would be something that Stan [Downey] would\n\nhandle.” Downey later told Scofield that he hadn’t done any such investigation, but\n\nScofield admitted that, “[a]part from what he has told me, I don’t know what he\n\nhas done.” Scofield testified that Downey “didn’t appear prepared” for the penalty\n\nphase, based on Downey’s last-minute suggestion that Scofield examine Seal since\n\nhe had talked with him before.\n\n In preparation for the penalty phase, Scofield himself did not interview any\n\nmember of Jenkins’s family, but he did not know whether Downey interviewed\n\nany family members. Scofield did interview Jenkins’s landlord and some other\n\nacquaintances but did not do so specifically for the purpose of developing\n\nmitigation. He had met with Jenkins around a dozen times, and Jenkins had told\n\nhim his life story including his experience of abuse as a child. In hindsight,\n\nScofield said he had a “lack of appreciation of mitigating circumstances.” He\n\nrepeatedly denied in his Rule 32 testimony that his omissions during the penalty\n\nphase were strategic: “I didn’t understand what I was doing. . . . I just didn’t\n\nunderstand what I needed to do.”\n\n For reasons that remain unclear, Downey—who is since deceased—did not\n\ntestify or submit an affidavit in the Rule 32 proceeding. Scofield answered,\n\n\n 12\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 13 of 73\n\n\n“Obviously not,” when asked whether he had personal knowledge of everything\n\nDowney may or may not have done to prepare for the penalty phase. The record of\n\nthe extent of Downey’s pretrial preparation is thus limited to Downey’s detailed\n\nfee declaration, which he submitted to the State of Alabama in documentation of\n\nthe compensation due him as appointed counsel. See Ala. Code § 15-12-21. In it,\n\nDowney reported that he spent 71 hours4 on out-of-court preparation for trial. The\n\nreport details 35 hours of interviews with Jenkins in 18 separate visits to the jail\n\nover the course of 17 months, as well as one substantive conversation with\n\nJenkins’s grandmother.5\n\n The Rule 32 trial court also heard testimony from four family members and\n\none friend of Jenkins’s. The family members presented for the first time a stark\n\ndepiction of Jenkins as an unloved and abused child in California. Jenkins’s older\n\nbrother Michael Jenkins testified that all of his siblings have different fathers and\n\nthat his parents often used drugs and fought. Michael testified that Jenkins was\n\nbeaten frequently as a child by his stepfather, if not every day then every other day\n\n\n\n4\n In Jenkins’s state postconviction appeal, the Alabama Court of Criminal Appeals erroneously\nstated that Downey’s fee declaration reports 171 hours of pretrial preparation. See Jenkins, 972\nSo. 2d at 137. We discuss the effect of this discrepancy infra note 7.\n5\n The state appellate court also noted that Downey had moved for a continuance in October 1989\nso that he could conduct “[f]urther discovery and investigation (including a possible trip to\nCalifornia).” Id. We note that Scofield had orally suggested to the trial court in September 1989\nthat more time might be needed with respect to “evidence about the defendant” and witnesses in\nCalifornia, as well as possible psychological evaluation.\n 13\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 14 of 73\n\n\nor every three days. Jenkins was beaten all over his body with a variety of\n\nimplements. Michael was beaten too. Jenkins’s stepfather would lock Jenkins in\n\nhis bedroom for a couple of hours at a time, sometimes without dinner. Some of\n\nthe beatings were prompted by Jenkins’s bedwetting or soiling. Michael testified\n\nthat Jenkins’s stepfather sometimes made Jenkins wear his soiled underwear on his\n\nhead and even forced him to eat his own feces a few times. Michael testified that\n\nJenkins’s bladder and bowel control problems began after a camping trip he took\n\nwith his stepfather and stepgrandfather.\n\n Michael testified that, as children, he and Jenkins were made to work on car\n\nengines late into the night and to shovel horse manure daily. The family moved\n\nfrequently and the children attended many different schools as a result. Michael\n\ntestified that he was Jenkins’s only friend as a child and that Jenkins did poorly in\n\nschool. Jenkins began running away from home as a young teenager and lived with\n\nvarious other family members or on the streets. Michael said that Jenkins was a\n\ngood brother to him and had asked him to testify at his trial, but he never heard\n\nfrom his attorneys or found out when the trial was. Michael admitted to having\n\ncognitive and memory problems due to a head injury years earlier.\n\n Jenkins’s cousin Tammy Pitts testified that, as an infant, Jenkins’s mother\n\nbeat and neglected him, leaving him dirty and in soiled diapers. His stepfather also\n\nbeat Jenkins daily until he left home at age 13, sometimes bruising him so badly\n\n\n 14\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 15 of 73\n\n\nthat he was laid up in bed for days or weeks. Jenkins had problems with\n\nbedwetting and soiling, and his stepfather made him wear soiled clothes to school.\n\nJenkins would be locked in his filthy bedroom around the clock except to attend\n\nschool and do chores. He was usually not allowed to eat dinner with the rest of the\n\nfamily and subsisted on scraps and dog food. The entire family belittled Jenkins\n\nbecause his biological father was Hispanic and his complexion was therefore\n\ndarker. Jenkins’s parents drank a lot and used drugs every day. Pitts knew these\n\nthings because she had daily contact with the family and lived with them off and\n\non, and Jenkins lived with her for a time after he ran away from home. No one had\n\ncontacted her about testifying at Jenkins’s trial.\n\n Jenkins’s second cousin Betty DeLavega testified that her relationship with\n\nJenkins was like that of an aunt. She lived with the Jenkins family for five months\n\nwhen Jenkins was 10 or 11. She saw Jenkins’s stepfather slap Jenkins “[q]uite a\n\nfew times,” and once, he made Jenkins eat his own feces. DeLavega testified that\n\nJenkins’s parents used drugs nearly every day and withheld affection from Jenkins\n\nand Michael. No one contacted her about Jenkins’s trial; Jenkins’s grandmother\n\nhad asked her to testify in the Rule 32 hearing about any abuse she had witnessed.\n\n Jenkins’s grandmother Doris Wagoner testified that while Jenkins’s mother\n\nwas pregnant with him, her husband was in prison and she drank and took drugs.\n\nWhen Jenkins was born, his mother initially put him up for adoption, then changed\n\n\n 15\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 16 of 73\n\n\nher mind weeks later. Jenkins was a lethargic infant and his mother cared more\n\nabout partying than about her children. Wagoner testified that she heard of\n\nJenkins’s mistreatment as a child but didn’t see it. She could tell from the way that\n\nJenkins feared his stepfather that he was being abused, and she knew he slept in a\n\nfilthy bed. After Jenkins was arrested for murder, Wagoner spoke with Scofield,\n\nwho asked for money relating to his representation of Jenkins. She did not\n\nremember whether she talked with Scofield about Jenkins’s background. She could\n\nnot remember why she didn’t testify at the trial; she might have been ill, but she\n\nwas probably busy.\n\n Sherry Seal, the wife of Lonnie Seal, who had testified in the penalty phase\n\nof the trial, testified that Jenkins was respectful and generous and that she trusted\n\nhim with her children. She related how he supported her family after moving to\n\nAlabama, giving them “pretty much” his entire paycheck. She said she would still\n\nfeel safe around Jenkins. Two jailers also testified that Jenkins was a respectful,\n\ncourteous, model inmate who never complained or caused any problems.\n\n Jenkins also introduced medical, school, and juvenile court records from his\n\nchildhood. The medical records document his premature birth and his mother’s\n\nintention to give him up for adoption. The school records reflect a struggling\n\nstudent who was repeatedly promoted to the next grade because of his age, despite\n\nserious academic difficulties, chronic truancy, and frequent school changes. The\n\n\n 16\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 17 of 73\n\n\njuvenile court records document Jenkins’s placement in juvenile hall starting at age\n\n14, following repeated acts of grand theft auto and running away.\n\n The Rule 32 hearing concluded with the testimony of two psychologists\n\nabout Jenkins’s history of psychological trauma stemming from his childhood\n\nabuse. The defense expert, Dr. David Lisak, had met with Jenkins three times and\n\ninterviewed a dozen of his family members, family friends, and associates. Dr.\n\nLisak testified that the pervasive abuse and neglect in Jenkins’s childhood impaired\n\nhis ability to cope with trauma as an adult. He opined that Jenkins’s childhood\n\nabuse adversely affected his cognitive and emotional development and other\n\nabilities. He described how Jenkins had been neglected as an infant and sexually\n\nabused by his grandfather at age four, which led to a lifetime of bladder and bowel\n\ndysfunction. He also discussed the hypothesis that children who are abused are at\n\ngreater risk for perpetrating violence later in life, but he acknowledged there would\n\nhave been other causative factors in Jenkins’s case. Dr. Lisak described Jenkins’s\n\nintellectual capacity as “somewhat borderline,” though he did not perform any\n\ntesting of his own.\n\n The State’s expert, Dr. Karl Kirkland, had met with Jenkins for several hours\n\nand administered several tests. On the Wechsler Adult Intelligence Scale, Jenkins\n\n“scored in the range of borderline intellectual functioning which is between mild\n\nmental retardation and low average intellectual functioning” with an overall IQ of\n\n\n 17\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 18 of 73\n\n\n76. Dr. Kirkland agreed that Jenkins had been “reared in a chaotic dysfunctional\n\nfamily” and generally agreed with Dr. Lisak’s assessment of the negative effects of\n\nchildhood trauma.\n\n The court considered all of this evidence before denying Rule 32 relief in\n\nfull. As relevant here, it found that some of the evidence about the severity of\n\nabuse that Jenkins suffered as a child was not credible. It further found that Jenkins\n\nsuffered no prejudice from any deficient performance by his penalty-phase counsel\n\nbecause the aggravating circumstances outweighed any additional mitigating\n\nevidence Downey might have adduced.\n\n The Alabama Court of Criminal Appeals affirmed the denial of Rule 32\n\nrelief. Jenkins v. State, 972 So. 2d 111 (Ala. Ct. Crim. App. 2004). In relevant part,\n\nthe appeals court rejected Jenkins’s ineffective assistance claim, finding that\n\ncounsel reasonably chose to pursue a penalty-phase strategy of residual doubt and\n\ngood character. Id. at 147. It explained that the new evidence was self-\n\ncontradictory in that it asserted that childhood abuse would have made Jenkins a\n\nviolent adult yet also asserted that Jenkins was meek and mild. Id. The appeals\n\ncourt also agreed with the lower court’s conclusion that Jenkins suffered no\n\nprejudice from any error by his counsel because the aggravating circumstances still\n\noutweighed any new mitigating ones. Id. at 148. The court also noted that evidence\n\n\n\n\n 18\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 19 of 73\n\n\nabout Jenkins’s good behavior in jail awaiting trial was “minimally mitigating” and\n\nwould not have affected the sentence. Id. at 149.\n\n Because the Supreme Court’s decision in Atkins v. Virginia6 had issued\n\nduring the pendency of the Rule 32 appeal, the Alabama appeals court ordered\n\nsupplemental briefing on the possible impact of Atkins upon Jenkins’s sentence.\n\nJenkins asked the court to stay the proceedings and conduct an evidentiary hearing\n\nafter the Alabama legislature enacted a statute about executing the mentally\n\nretarded or, in the alternative, to vacate his sentence and remand after the Alabama\n\nlegislature acted. He later asserted that he is mentally retarded and therefore\n\nineligible for the death penalty. The appeals court proceeded to reject Jenkins’s\n\nAtkins claim on the merits because the record established that his IQ was 76 and\n\nthat he maintained relationships and employment. Id. at 155. The Alabama\n\nSupreme Court summarily affirmed, in relevant part. Ex parte Jenkins, 972 So. 2d\n\n159, 165 (Ala. 2005), remanded to 972 So. 2d 165 (Ala. Ct. Crim. App. 2005)\n\n(juror misconduct claim), cert. denied, 552 U.S. 1167 (2008).\n\n In 2008, Jenkins filed his first petition for a writ of habeas corpus in federal\n\ncourt pursuant to 28 U.S.C. § 2254. The federal proceedings were stayed for\n\nseveral years while Jenkins pursued additional, unrelated Rule 32 relief in the state\n\n\n6\n 536 U.S. 304, 321 (2002) (holding that the execution of mentally retarded offenders violates the\nEighth Amendment). The rule announced in Atkins is retroactive to cases on collateral review. In\nre Holladay, 331 F.3d 1169, 1172–73 (11th Cir. 2003).\n 19\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 20 of 73\n\n\ncourts, ultimately unsuccessfully. In the district court, Jenkins moved for an\n\nevidentiary hearing on his Atkins claim. After reviewing the state court record,\n\nwith a focus on the Rule 32 testimony of Drs. Lisak and Kirkland, the district court\n\ndenied the motion in 2015. Noting that our Circuit has accepted the State of\n\nAlabama’s definition of mental retardation following Atkins, the court found that\n\nJenkins is not intellectually disabled under that standard. In a 347-page order, the\n\ndistrict court also denied relief on all of Jenkins’s claims including ineffective\n\nassistance of counsel during the penalty phase and affirmed as reasonable the state\n\ncourt’s conclusion that Jenkins failed to establish prejudice from any error by his\n\ncounsel. The district court later denied reconsideration and a certificate of\n\nappealability.\n\n Jenkins now appeals. Our Court granted Jenkins a certificate of\n\nappealability, 28 U.S.C. § 2253(c), on two issues:\n\n 1) Whether the district court erred in denying Appellant’s claim\n that his trial counsel rendered ineffective assistance of counsel in the\n investigation and presentation of mitigating evidence during the\n penalty phase of Appellant’s 1991 trial; and\n\n 2) Whether the district court erred in denying Appellant’s claim\n that he is intellectually disabled and ineligible for the death penalty\n under Atkins v. Virginia, 536 U.S. 304 (2002).\n\n II. STANDARDS OF REVIEW\n\n We review the denial of a petition for a writ of habeas corpus de novo.\n\nMorrow v. Warden, Ga. Diagnostic Prison, 886 F.3d 1138, 1146 (11th Cir. 2018).\n\n 20\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 21 of 73\n\n\nThe Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) instructs that\n\na writ of habeas corpus shall not be granted to a state prisoner unless the state\n\nadjudication of a claim “resulted in a decision that was contrary to, or involved an\n\nunreasonable application of, clearly established Federal law, as determined by the\n\nSupreme Court of the United States.” 28 U.S.C. § 2254(d)(1). “[C]learly\n\nestablished Federal law” means “the holdings, as opposed to the dicta, of [the\n\nSupreme] Court’s decisions as of the time of the relevant state-court decision.”\n\nWilliams v. Taylor, 529 U.S. 362, 412 (2000). “[W]hen the last state court to\n\ndecide a prisoner’s federal claim explains its decision on the merits in a reasoned\n\nopinion . . . a federal habeas court simply reviews the specific reasons given by the\n\nstate court and defers to those reasons if they are reasonable.” Wilson v. Sellers,\n\n138 S. Ct. 1188, 1192 (2018). A state court’s decision is reasonable “so long as\n\n‘fairminded jurists could disagree’ on the correctness of the state court’s decision.”\n\nHarrington v. Richter, 562 U.S. 86, 101 (2011) (quoting Yarborough v. Alvarado,\n\n541 U.S. 652, 664 (2004)). “[E]ven a strong case for relief does not mean the state\n\ncourt’s contrary conclusion was unreasonable.” Id. at 102. Rather, a prisoner must\n\nshow that the state court’s ruling “was so lacking in justification that there was an\n\nerror well understood and comprehended in existing law beyond any possibility for\n\nfairminded disagreement.” Id. at 103. That ruling must have been “‘objectively\n\nunreasonable,’ not merely wrong; even ‘clear error’ will not suffice.” White v.\n\n\n 21\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 22 of 73\n\n\nWoodall, 572 U.S. 415, 419 (2014) (quoting Lockyer v. Andrade, 538 U.S. 63, 75\n\n(2003)).\n\n In addition, “a determination of a factual issue made by a State court shall be\n\npresumed to be correct,” and the prisoner bears “the burden of rebutting the\n\npresumption of correctness by clear and convincing evidence.” 28 U.S.C.\n\n§ 2254(e)(1). AEDPA further instructs that a writ of habeas corpus may not be\n\ngranted unless the state court adjudication “resulted in a decision that was based on\n\nan unreasonable determination of the facts in light of the evidence presented in the\n\nState court proceeding.” Id. § 2254(d)(2). When a federal habeas court makes a\n\nfactual finding as part of its habeas determination, we review for clear error, which\n\nis a “highly deferential” standard of review. Thomas v. Allen, 607 F.3d 749, 752\n\n(11th Cir. 2010) (quoting Holton v. City of Thomasville Sch. Dist., 425 F.3d 1325,\n\n1350 (11th Cir. 2005)). “A finding is ‘clearly erroneous’ when although there is\n\nevidence to support it, the reviewing court on the entire evidence is left with the\n\ndefinite and firm conviction that a mistake has been committed.” United States v.\n\nU.S. Gypsum Co., 333 U.S. 364, 395 (1948).\n\n III. DISCUSSION\n\n A. Ineffective Assistance of Counsel\n\n A prisoner alleging that he received ineffective assistance of counsel in\n\nviolation of the Sixth Amendment must establish two elements. Strickland v.\n\n\n 22\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 23 of 73\n\n\nWashington, 466 U.S. 668, 687 (1984). “First, the defendant must show that\n\ncounsel’s performance was deficient.” Id. In other words, he must show “that\n\ncounsel made errors so serious that counsel was not functioning as the ‘counsel’\n\nguaranteed the defendant by the Sixth Amendment.” Id. “Second, the defendant\n\nmust show that the deficient performance prejudiced the defense.” Id. We discuss\n\neach of these elements in turn.\n\n 1. Deficient Performance\n\n A “doubly” deferential lens sits before our eyes as we review the\n\nperformance of Jenkins’s counsel during the penalty phase. Richter, 562 U.S. at\n\n105 (quoting Knowles v. Mirzayance, 556 U.S. 111, 123 (2009)). This redoubled\n\ndeference arises out of the tandem application of AEDPA’s “highly deferential”\n\nreview of a state court’s decision with Strickland’s “highly deferential” review of\n\nan attorney’s performance. Id. (quoting Lindh v. Murphy, 521 U.S. 320, 333 n.7\n\n(1997), and Strickland, 466 U.S. at 689). Under the Strickland standard, “counsel\n\nis strongly presumed to have rendered adequate assistance and made all significant\n\ndecisions in the exercise of reasonable professional judgment.” Strickland, 466\n\nU.S. at 690. And under AEPDA, we must deny habeas relief if “there is any\n\nreasonable argument that counsel satisfied Strickland’s deferential standard.”\n\nRichter, 562 U.S. at 105.\n\n\n\n\n 23\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 24 of 73\n\n\n Jenkins argues that his counsel performed deficiently by failing to\n\n“investigate compelling mitigating evidence or develop a penalty-phase theory.”\n\nThe state court disagreed. It quoted decisions of our Court that hold that “there is\n\nno per se rule that evidence of a criminal defendant’s troubled childhood must\n\nalways be presented as mitigating evidence.” Jenkins, 972 So. 2d at 144 (quoting\n\nMarek v. Singletary, 62 F.3d 1295, 1300 (11th Cir. 1995)). Noting that “Downey\n\nwas in charge of the penalty phase” but “we do not have the benefit of Downey’s\n\ntestimony as to what occurred and why,” it reviewed the trial record to find that\n\n“counsel argued residual doubt and Jenkins’s good character at the penalty phase.”\n\nId. at 145–46. It concluded that “Downey’s decision to concentrate on reasonable\n\ndoubt and to portray Jenkins as a good person was reasonable under the\n\ncircumstances.” Id. at 147.\n\n Jenkins insists, however, that counsel’s failure to develop mitigating\n\nevidence about his childhood abuse was “not the product of strategy, but rather\n\narose from ingnorance and inexeperiance [sic].” The problem for Jenkins is that,\n\nlike the state courts, this Court also does not know what Downey did or why. The\n\nrecord is silent as to his thoughts and intentions as he prepared for the penalty\n\nphase. Given that Jenkins bore the burden of overcoming the strong presumption of\n\ncompetence and of proving Downey’s deficient performance, we are perplexed\n\nwhy Jenkins’s new counsel did not have Downey testify or submit an affidavit in\n\n\n 24\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 25 of 73\n\n\nthe Rule 32 proceeding. See Strickland, 466 U.S. at 687 (“The defendant must\n\nshow that counsel’s performance was deficient.”). Nor did Jenkins testify to the\n\nsubstance of his conversations with his counsel. “The reasonableness of a trial\n\ncounsel’s acts, including lack of investigation or excluding character witnesses\n\nfrom the sentencing phase, depends ‘critically’ upon what information the client\n\ncommunicated to counsel.” Chandler v. United States, 218 F.3d 1305, 1324 (11th\n\nCir. 2000) (en banc) (quoting Strickland, 466 U.S. at 691). There is simply no\n\nrecord evidence to support Jenkins’s central assertion that Downey made no\n\ntactical choices in his preparation for the penalty phase.\n\n Jenkins instead relies on Scofield’s Rule 32 testimony about how he and\n\nDowney prepared for the trial and sentencing. We acknowledge that Scofield\n\nrepeatedly testified that he did not really know what he was doing and that he\n\nfailed to make strategic choices with respect to mitigating circumstances and\n\nsentencing. But Scofield was equally clear that Downey—the more experienced\n\nlawyer—was in fact responsible for the penalty phase. And Scofield also\n\nrepeatedly testified that he did not know what Downey may or may not have done\n\nto prepare. Furthermore, his lack of knowledge about Downey’s preparation was\n\nunderstandable. He and Downey were not colleagues who shared an office and\n\nmight reasonably be expected to know what the other was doing much of the time.\n\nScofield and Downey did not even work in the same city.\n\n\n 25\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 26 of 73\n\n\n This case is unlike those in which we have found deficient performance\n\nwhere the record showed that counsel investigated insufficiently because he\n\n“waited until the eleventh hour” to begin preparing for the penalty phase while\n\nfaced with “overwhelming evidence of guilt.” See, e.g., Johnson v. Sec’y, D.O.C.,\n\n643 F.3d 907, 932 (11th Cir. 2011). The limited record that Jenkins developed\n\nestablishes that Downey spent at least 71 hours preparing for trial, spread out over\n\n17 months. The state court interpreted this sparse record as reflecting a penalty-\n\nphase strategy of residual doubt, the logical extension of Downey’s guilt-phase\n\ninvestigation of the evidence and Scofield’s pursuit at trial of an acquittal based on\n\nreasonable doubt. See Jenkins, 972 So. 2d at 147. In such a situation, “[a] lawyer’s\n\ntime and effort in preparing to defend his client in the guilt phase of a capital case\n\ncontinues to count at the sentencing phase.” Tarver v. Hopper, 169 F.3d 710, 715\n\n(11th Cir. 1999).\n\n We decline to dictate how many hours or how many sessions an attorney\n\nmust spend interviewing his counsel in order to provide effective assistance. “No\n\nabsolute rules dictate what is reasonable performance for lawyers.” Chandler, 218\n\nF.3d at 1317 (citing Strickland, 466 U.S. at 688–89). But even this limited record\n\nundermines Jenkins’s assertion that Downey prepared inadequately. Downey’s 71\n\nhours of preparation included 35 hours of jailhouse conversations with Jenkins\n\n\n\n\n 26\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 27 of 73\n\n\nduring 18 separate meetings.7 According to Downey’s notes, Scofield was present\n\nfor only five of those meetings (12 hours’ worth), leaving 23 hours over 13\n\nmeetings between Jenkins and Downey alone. Downey’s notes reveal little about\n\nthe content of those conversations beyond broad notations such as “evidence,\n\nwitnesses, general status,” and discussions with Jenkins of the evidence and\n\ntestimony against him were surely relevant to Downey’s penalty-phase strategy of\n\nresidual doubt.\n\n But the record does not reveal that Downey knew or should have known that\n\nJenkins’s childhood merited investigation beyond that which he may have done.\n\nThe record shows that Downey developed enough of a rapport with Jenkins that,\n\naccording to his fee declaration, he “g[ave a] Christmas gift [and] discuss[ed] how\n\nmuch to tell his family.” But how much Downey knew about Jenkins’s family\n\nbackground can only be guessed from Scofield’s Rule 32 hearing testimony.\n\nScofield testified that he discussed Jenkins’s background during his dozen or so\n\njailhouse meetings with him, and that Downey was present for some of those\n\nmeetings. Scofield—and possibly Downey—learned that Jenkins experienced\n\nabuse as a child, that his stepfather was brutal, and that he lived on the streets and\n\n\n\n\n7\n We note that the state court clearly erroneously stated that Downey’s fee declaration\ndocumented 171 hours of pretrial preparation. See Jenkins, 972 So. 2d at 137. However, the state\ncourt accurately stated that “he spent over 25 hours talking with Jenkins in more than 10 visits to\nthe jail.” Id. That statement was not an unreasonable determination of the facts.\n 27\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 28 of 73\n\n\nin juvenile institutions because of his brutal home life. It is against this backdrop\n\nthat Downey’s presumptive “decision not to investigate must be directly assessed\n\nfor reasonableness in all the circumstances, applying a heavy measure of deference\n\nto counsel’s judgments.” Strickland, 466 U.S. at 691.\n\n That assessment depends a great deal on what Jenkins told his counsel about\n\nhis background. See id. But we do not know whether Downey was present when\n\nScofield learned about Jenkins’s childhood abuse, nor what else Downey may have\n\nlearned during his 13 solo discussions with Jenkins. Therefore, “given the absence\n\nof evidence in the record, we must assume counsel carried out his professional\n\nresponsibility and discussed mitigation with his client.” Chandler, 218 F.3d at\n\n1324. We will not “turn that presumption on its head by giving [Jenkins] the\n\nbenefit of the doubt where it is unclear what [Downey] did or did not do.” Williams\n\nv. Head, 185 F.3d 1223, 1235 (11th Cir. 1999). We will not assume that Jenkins\n\ntold Downey more than he told Scofield, and Scofield denied remembering that\n\nJenkins told him about the severity or frequency of his stepfather’s physical abuse.\n\nHe denied remembering Jenkins telling him about being beaten daily, being locked\n\nin his room 24/7, or being forced to eat his feces. And there is no suggestion in the\n\nrecord that counsel knew or had reason to investigate the allegation of sexual abuse\n\nby his stepgrandfather. The record simply does not reflect that Jenkins told either\n\nof his lawyers about the severe abuse that some of the Rule 32 witnesses later\n\n\n 28\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 29 of 73\n\n\ndescribed. We will not fault those lawyers for failing to present as mitigation this\n\nlimited evidence that they may have reasonably believed to be unremarkable. See\n\nStrickland, 466 U.S. at 691.\n\n To be sure, counsel’s failure “to discover all reasonably available mitigating\n\nevidence” may constitute deficient performance in some cases. Wiggins v. Smith,\n\n539 U.S. 510, 524 (2003) (quoting ABA Guidelines for the Appointment and\n\nPerformance of Counsel in Death Penalty Cases § 11.4.1(C) (1989)). Our Court\n\nhas identified deficient performance in several instances where counsel had good\n\nreason to investigate the defendant’s background but failed to do so. See Morrow,\n\n886 F.3d at 1147 (collecting cases). But we assess “a particular decision not to\n\ninvestigate . . . for reasonableness in all the circumstances, applying a heavy\n\nmeasure of deference to counsel’s judgments.” Strickland, 466 U.S. at 691. Thus,\n\nwe have rejected a hard rule that counsel’s “strategic decisions can be considered\n\nreasonable only if they are preceded by a ‘thorough investigation.’” Williams, 185\n\nF.3d at 1237 (quoting Rogers v. Zant, 13 F.3d 384, 387 (11th Cir. 1994)). Because\n\nwe simply do not know why Downey chose not to investigate Jenkins’s childhood\n\nmore thoroughly, Downey is entitled to the presumption that this strategic decision\n\nwas reasonable.\n\n And though we have also rejected “a per se rule of ineffective assistance\n\nwhere counsel does not consult family members,” Holladay v. Haley, 209 F.3d\n\n\n 29\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 30 of 73\n\n\n1243, 1252 (11th Cir. 2000), we note that Downey did not fail altogether to consult\n\nfamily members. Downey had at least one substantive conversation with Jenkins’s\n\ngrandmother, and we know from her Rule 32 hearing testimony that she never\n\nwitnessed Jenkins’s childhood abuse. Given the silence of the record and the strong\n\npresumption of competence to which Downey is entitled, we decline to fault him\n\nfor not doing more when Jenkins and his grandmother did not and could not,\n\nrespectively, provide a basis for doing so. Cf. United States v. Pease, 240 F.3d 938,\n\n941–92 (11th Cir. 2001) (“we cannot say as a matter of law that reliance on a\n\nclient’s statements is per se deficient performance”). Nor did Jenkins’s counsel\n\nhave an absolute duty to put on abusive-childhood evidence, even if they knew of\n\nit. See Chandler, 218 F.3d at 1319 (“Counsel is not required to present every\n\nnonfrivolous defense; nor is counsel required to present all mitigation evidence,\n\neven if the additional mitigation evidence would not have been incompatible with\n\ncounsel’s strategy.”); Waters v. Thomas, 46 F.3d 1506, 1511 (11th Cir. 1995) (en\n\nbanc) (“Our decisions are inconsistent with any notion that counsel must present\n\nall available mitigating circumstance evidence.”).\n\n Instead, we affirm as a reasonable application of Strickland the state court’s\n\nconclusion that Downey’s strategic decision to emphasize residual doubt was\n\nreasonable. See Jenkins, 972 So. 2d at 147. As we have noted, “[a] lawyer’s time\n\nand effort in preparing to defend his client in the guilt phase of a capital case\n\n\n 30\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 31 of 73\n\n\ncontinues to count at the sentencing phase.” Tarver, 169 F.3d at 715. This\n\nreasoning is especially applicable where, as here, counsel pursued the reasonable\n\nstrategy of arguing for acquittal at the guilt phase and for life at the sentencing\n\nphase by highlighting doubt about the evidence. “[R]esidual doubt is perhaps the\n\nmost effective strategy to employ at sentencing.” Chandler, 218 F.3d at 1320 n.28.\n\n“[E]specially when—as in this case—the evidence of guilt was not overwhelming,\n\nwe expect that petitioners can rarely (if ever) prove a lawyer to be ineffective for\n\nrelying on this seemingly reasonable strategy to defend his client.” Id. at 1320\n\n(footnote omitted); see also Tarver, 169 F.3d at 715–16 (collecting law review\n\narticles on the merits of residual-doubt strategy). Not only can such a strategy be\n\nconstitutionally adequate, in some cases it may be “the most effective performance\n\nin defense to the death penalty.” Tarver, 169 F.3d at 716.\n\n We cannot say that Downey’s manifest strategy of residual doubt was\n\nobjectively unreasonable under the circumstances. Jenkins’s case is precisely the\n\nsort where reminding the jury of the misgivings it may have held about the State’s\n\ncase could have been effective. Scofield’s guilt-phase closing argument from that\n\nvery morning, which explicitly emphasized doubt and the weakness of the State’s\n\nevidence, would have been fresh in the minds of the jury.8 Although we are not\n\n\n8\n During the penalty phase of a capital case, “[e]vidence presented at the trial of the case may be\nconsidered insofar as it is relevant to the aggravating and mitigating circumstances without the\nnecessity of re-introducing that evidence at the sentence hearing.” Ala. Code § 13A-5-45(c).\n 31\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 32 of 73\n\n\nquestioning the jury’s guilty verdict or the sufficiency of the evidence, which are\n\nbeyond the scope of our habeas review, we acknowledge that the State’s case\n\nagainst Jenkins was—as Downey reminded the jury during the penalty phase—\n\nentirely circumstantial. Indeed, Scofield maintained post-conviction that the case\n\nfor reasonable doubt was “[v]ery strong.” No one saw Jenkins commit either the\n\nmurder or the robbery and kidnapping that elevated Hogeland’s murder to a capital\n\ncrime. Scofield’s guilt-phase closing argument challenged the inferential leaps the\n\nState was asking the jury to make in finding that Jenkins committed these capital-\n\npredicate offenses and that he formed the intent to kill. This situation was not one\n\nwhere a guilt-phase strategy of acquittal—and therefore a penalty-phase strategy of\n\nresidual doubt—was unreasonable in light of the evidence. Cf. Johnson, 643 F.3d\n\nat 932–33 (deficient sentencing investigation where defendant had confessed to\n\ntwo murders).\n\n Thus, although Downey’s penalty-phase closing argument did not\n\nspecifically invoke lingering doubt, it clearly reminded the jury of the entirely\n\ncircumstantial nature of the State’s guilt-phase case. See Chandler, 218 F.3d at\n\n1320 & n.6 (a strategy of “focus[ing] on obtaining an acquittal and then, at\n\nsentencing, on lingering doubt” is reasonable, even if counsel does not use the\n\nwords “lingering doubt” or “residual doubt”). Downey closed this reminder by\n\nexplaining to the jury how a foundational ancient legal system proscribed capital\n\n\n 32\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 33 of 73\n\n\npunishment in the absence of eyewitness testimony—the exact situation here. The\n\nstate court found that Downey’s penalty-phase closing argument reflected a\n\nstrategic “decision to concentrate on reasonable doubt,” see Jenkins, 972 So. 2d at\n\n147, a tactical choice entitled to a “strong presumption of correctness,” Marek, 62\n\nF.3d at 1300 & n.3.\n\n Jenkins suggests other theories of mitigation that Downey might have more\n\neffectively pursued,9 but our deferential review under Strickland does not ask\n\nwhether counsel could possibly or ideally have been more effective. “The test for\n\nineffectiveness is not whether counsel could have done more.” Waters, 46 F.3d at\n\n1518. We do not ask whether an attorney’s representation “deviated from best\n\npractices or common custom,” and we should resist the temptation to second-guess\n\nan attorney with the benefit of our hindsight. Richter, 562 U.S. at 105. We ask only\n\n“whether, in light of all the circumstances, the identified acts or omissions were\n\n\n\n\n9\n In particular, Jenkins argues that his counsel should have presented “model prisoner” evidence.\nThe state court noted, “Good conduct during pretrial incarceration is not necessarily a mitigating\ncircumstance,” before finding that Jenkins had not met his burden of establishing that his counsel\neven had information about his conduct in jail. Jenkins, 972 So. 2d at 149. We agree, and we\ndecline to impose a requirement of sua sponte investigation of this kind in every capital case. Cf.\nWilliams, 529 U.S. at 396 (finding deficient performance based in part on not returning a call\nfrom a prison ministry volunteer who offered to vouch for defendant’s good behavior in prison).\n“Surely, counsel is not required to call a witness to testify to facts such as lack of violent nature\nwhen the jury has rejected such an approach and has found that the defendant is guilty of\nmurder.” Griffin v. Wainwright, 760 F.2d 1505, 1512 (11th Cir. 1985), vacated on other grounds,\n476 U.S. 1112 (1986), reaff’d, 874 F.2d 1397 (11th Cir. 1989).\n 33\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 34 of 73\n\n\noutside the range of professionally competent assistance.” Strickland, 466 U.S. at\n\n690.\n\n We conclude that the state court did not unreasonably determine that Jenkins\n\nfailed to establish objectively incompetent performance by his counsel during the\n\npenalty phase of his trial. Our inquiry is not “what the best lawyers would have\n\ndone. Nor is the test even what most good lawyers would have done. We ask only\n\nwhether some reasonable lawyer at the trial could have acted, in the circumstances,\n\nas defense counsel acted at trial. . . . We are not interested in grading lawyers’\n\nperformances.” White v. Singletary, 972 F.2d 1218, 1220–21 (11th Cir. 1992). And\n\n“[i]f fairminded jurists could reasonably disagree” about the reasonableness of\n\ncounsel’s performance, “then habeas relief is due to be denied.” Johnson, 643 F.3d\n\nat 932.\n\n 2. Prejudice\n\n Even if Jenkins had demonstrated that his counsel performed as no\n\nreasonable lawyer could have, he must also demonstrate a “reasonable probability\n\nthat, but for counsel’s unprofessional errors, the result of the proceeding would\n\nhave been different.” Strickland, 466 U.S. at 694. Given that the jury here\n\nrecommended a sentence of death by the narrowest possible vote, 10 to 2, Jenkins\n\nneed establish only “a reasonable probability that at least one juror would have\n\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537. “A\n\n\n 34\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 35 of 73\n\n\nreasonable probability is a probability sufficient to undermine confidence in the\n\noutcome.” Strickland, 466 U.S. at 694. In measuring that probability here, we will\n\n“evaluate the totality of the available mitigation evidence—both that adduced at\n\ntrial, and the evidence adduced in the habeas proceeding in reweighing it against\n\nthe evidence in aggravation.” Williams, 529 U.S. at 397–98.\n\n We begin our analysis by clarifying what exactly Jenkins is arguing about\n\nprejudice. We do not know precisely which aggravating and mitigating factors the\n\njury may have implicitly found before making its recommendation of death. The\n\ntrial court found at sentencing two statutory aggravating factors: the robbery and\n\nthe kidnapping of the victim. See Ala. Code § 13A-5-49(4). Because none of the\n\nmitigation evidence offered in the Rule 32 hearing bears on those factors, we see\n\nno reasonable probability that the aggravating factors would have been determined\n\ndifferently by the jury or the court. The court also found at sentencing two\n\nstatutory mitigating factors: no significant history of prior criminal activity and\n\nage. See id. § 13A-5-51(1), (7). There was also strong evidence of Jenkins’s\n\nintoxication, which was a key part of Jenkins’s guilt-phase defense and could have\n\nsupported a third mitigating factor, impaired capacity. See id. § 13A-5-51(6).10\n\n\n10\n The jury had heard evidence about all three of these statutory mitigators during the guilt phase\nof the trial. Downey’s penalty-phase opening statement reviewed for the jury all seven of the\npossible statutory mitigating factors, and the court also recited them in its jury instructions,\nreminding the jury that they could consider guilt-phase evidence as well as non-statutory\nmitigation. Jenkins has not established that there is a substantial probability that the jury’s\nrecommendation would have been different had it been more explicitly reminded of Jenkins’s\n 35\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 36 of 73\n\n\nJenkins does not argue that the jury or the court would have found other statutory\n\nmitigators in light of the Rule 32 evidence. Rather, he contends that the jury might\n\nhave found non-statutory mitigation based on Jenkins’s abusive childhood, low\n\nintelligence, or mental health problems that would have outweighed the\n\naggravating factors.11\n\n The Rule 32 trial court found that Jenkins had not demonstrated prejudice\n\neven assuming that counsel had performed deficiently. See Jenkins, 972 So. 2d at\n\n138. It concluded that he had not shown a reasonable probability that the sentencer\n\n“would have concluded that a weighing of the aggravating and mitigating\n\ncircumstances did not warrant death” upon a reweighing of the evidence. See id. It\n\nemphasized the depravity of the crime and the aggravating circumstances and\n\nfound that the mitigating witnesses were not credible for various reasons. See id.\n\nThe state appellate court affirmed, declining to disturb the credibility\n\ndeterminations of the trial court. Id. at 142–44. Independently reweighing the\n\n\nage, criminal history, and intoxication, given that the trial court, the Rule 32 trial court, and the\nstate appellate court all explicitly considered those mitigators and found them outweighed by the\naggravators.\n11\n Alabama law lists seven statutory mitigating circumstances, but states that mitigation “shall\n. . . not be limited to” those seven circumstances. Ala. Code § 13A-5-51. Further, “mitigating\ncircumstances shall include any aspect of a defendant’s character or record and any of the\ncircumstances of the offense that the defendant offers as a basis for a sentence of life\nimprisonment without parole instead of death, and any other relevant mitigating circumstances\nwhich the defendant offers as a basis for a sentence of life imprisonment without parole instead\nof death.” Id. § 13A-5-52. A jury may recommend a sentence of death only if it finds that “one or\nmore aggravating circumstances . . . exist and that they outweigh the mitigating circumstances, if\nany.” Id. § 13A-5-46(e)(3).\n 36\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 37 of 73\n\n\nevidence, it concluded that, “[g]iven the aggravating circumstances that were\n\nproven by the State and the facts surrounding Hogeland’s murder, we, like the\n\ncircuit court, are confident that death was the appropriate punishment for Jenkins’s\n\nactions.” Id. at 148.\n\n Our review of this evidence does not show that it is likely that the jury\n\nwould have concluded that death was not the appropriate punishment if it had\n\nheard the Rule 32 evidence. Much of the proffered mitigation evidence is, in our\n\nview, much less mitigating than Jenkins asserts. In particular, the testimony of\n\nJenkins’s older brother Michael was likely to have been a double-edged sword. See\n\nEvans v. Sec’y, Dep’t of Corr., 703 F.3d 1316, 1327 (11th Cir. 2013) (en banc) (“it\n\nis reasonable to conclude that a defendant was not prejudiced when his mitigation\n\nevidence was a two-edged sword” (quoting Ponticelli v. Sec’y, Fla. Dep’t of Corr.,\n\n690 F.2d 1271, 1296 (11th Cir. 2012)). Michael was the closest relative to testify at\n\nthe Rule 32 hearing and his testimony about the abuse Jenkins experienced should\n\nhave been the most reliable, since he experienced much of the same physical abuse\n\nhimself. The brothers shared the same filthy bedroom, received beatings from the\n\nsame paddle, and were frequently absent from school together. But Michael’s\n\ntestimony contained some discrepancies, and he admitted to cognitive and memory\n\nproblems. Perhaps Michael’s testimony would have made Jenkins appear more\n\nsympathetic in the eyes of the jury—but perhaps not. While Jenkins ran away from\n\n\n 37\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 38 of 73\n\n\nhome and eventually committed murder, Michael responded to the same\n\nbackground by running away from home, graduating from high school, and\n\nproviding health care for both the terminally ill and people with substance use\n\ndisorders. It is possible that the jury could even have found Jenkins to be more\n\nculpable for not overcoming his difficult upbringing and to pose more of a danger\n\nof future violence.\n\n We also find reasonable the state court’s identification of other problems\n\nwith the mitigating evidence Jenkins argues should have been presented. Hearing\n\nthat evidence firsthand in the Rule 32 proceeding, the trial court found the\n\ntestimony of the cousins less credible in light of the lack of physical abuse\n\ndocumented in the school health records. See Jenkins, 972 So. 2d at 139 (noting\n\nthat school records observed a rash and gingivitis). It faulted Pitts in particular for\n\npurporting to have witnessed ongoing, severe abuse but taking only minimal,\n\nineffective steps to obtain outside help. See id. It further noted that DeLavega and\n\nWagoner actually testified that they never saw Jenkins beaten, and it noted all of\n\nthe witnesses’ interest in getting Jenkins off death row. All of these factors would\n\nlikely have caused a reasonable jury to discount their testimony. See id. at 139–42.\n\nWe must accept the trial court’s views of the Rule 32 testimony. “In the absence of\n\nclear and convincing evidence, we have no power on federal habeas review to\n\nrevisit the state court’s credibility determinations.” Bishop v. Warden, 726 F.3d\n\n\n 38\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 39 of 73\n\n\n1243, 1259 (11th Cir. 2013) (citing Marshall v. Lonberger, 459 U.S. 422, 434\n\n(1983)).\n\n The state court also noted that Wagoner testified that she was not available\n\nto testify at the trial, and it found that Seal’s testimony would have been\n\ncumulative to that of her husband. Thus, counsel’s failure to call those two\n\nwitnesses could not have been prejudicial to Jenkins. See id. at 141, 142. Neither of\n\nthese conclusions involves an unreasonable determination of the facts. We agree\n\nthat the Rule 32 testimony of Sherry Seal was substantially similar to the testimony\n\nof her husband Lonnie Seal that the jury heard during the penalty phase. Both\n\ntestified about meeting Jenkins in California, Jenkins’s assistance with their move\n\nto Alabama, and his residence with them in their Alabama home. Sherry added\n\nonly that Jenkins sometimes wet his bed. It was not unreasonable for the state court\n\nto conclude that Sherry’s testimony was cumulative of her husband’s, and “a\n\npetitioner cannot satisfy the prejudice prong of the Strickland test with evidence\n\nthat is merely cumulative of evidence already presented.” Rose v. McNeil, 634 F.3d\n\n1224, 1243 (11th Cir. 2011).\n\n Further, the school, medical, and juvenile court records that Jenkins\n\nintroduced in the Rule 32 proceeding have their own problems. As noted, they do\n\nnot document or even corroborate the severe level of abuse alleged by some of the\n\nwitnesses. Though these documents tell a sad story, they do not create a substantial\n\n\n 39\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 40 of 73\n\n\nlikelihood that the jury would have recommended life. Although they portray\n\nJenkins as a victim of truly unfortunate circumstances, they also show his unsavory\n\ndelinquent behavior. That delinquency included a history of stealing cars, and\n\ngiven that the jury likely inferred that Jenkins stole the red Mazda, new\n\ninformation about his history of theft may not have been especially mitigating.\n\nSimilarly, we agree with the state court that evidence of Jenkins’s compliant\n\nconduct in pretrial detention is “minimally mitigating” where there had been no\n\nevidence of jail misconduct as an aggravator. See Jenkins, 972 So. 2d at 149; cf.\n\nSkipper v. South Carolina, 476 U.S. 1, 5 & n.1 (1986) (good jail behavior is\n\n“potentially mitigating” especially “[w]here the prosecution specifically relies on a\n\nprediction of future dangerousness in asking for the death penalty”).\n\n Jenkins argues that allowing the jury to see his experience of mental,\n\nemotional, and cognitive difficulties, compounded by years of abuse and neglect\n\nand considered alongside his mild character and good jail behavior as an adult,\n\nwould have convinced them that Hogeland’s murder was an “aberration” that\n\ndeserved a punishment less severe than death. That interpretation of Jenkins’s life\n\nstory may be conceivable. But more than mere conceivability is required to\n\nestablish prejudice: “The likelihood of a different result must be substantial, not\n\njust conceivable.” Richter, 562 U.S. at 112 (emphasis added). We do not find the\n\nstate court’s decision that a different result was not substantially likely to be an\n\n\n 40\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 41 of 73\n\n\nunreasonable application of Strickland. As we have noted, we think it equally\n\nlikely that the jury could have rejected this alternate narrative out of hand in light\n\nof having just convicted him of the violent murder of a young mother. We also\n\nthink it possible that the jury could have viewed the murder as just one step in a\n\nlogical progression of degeneracy that began with behavioral problems at school,\n\ncontinued through the juvenile court system, and escalated past grand theft auto\n\ninto kidnapping and murder. Taken separately as well as together, the proffered\n\nevidence does not compel us to conclude that a verdict of life would have been\n\nsubstantially likely. Because fairminded jurists could disagree about the state\n\ncourt’s conclusion about Strickland prejudice, we affirm the denial of habeas relief\n\non this ground.\n\n B. Intellectual Disability\n\n Jenkins argues that the district court erred when it denied his claim that he is\n\nintellectually disabled and ineligible for the death penalty under Atkins v. Virginia,\n\n536 U.S. 304 (2002). The Supreme Court held in Atkins that executing an offender\n\nwho has mental retardation12 violates the Eighth Amendment’s ban on excessive\n\n\n\n\n12\n In 2002, the Atkins Court followed prevailing convention in referring to “mentally retarded”\noffenders. We will use that now-disfavored term only when quoting or discussing older sources.\nSince 2014, the Supreme Court has exclusively used the terms “intellectually disabled person”\nand “person with intellectual disability.” See, e.g., Hall v. Florida, 572 U.S. 701, 704 (2014);\nMoore v. Texas, 139 S. Ct. 666, 672 (2019).\n 41\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 42 of 73\n\n\npunishments. Id. at 321. It defined mental retardation by quoting the American\n\nAssociation on Mental Retardation:\n\n Mental retardation refers to substantial limitations in present\n functioning. It is characterized by significantly subaverage intellectual\n functioning, existing concurrently with related limitations in two or\n more of the following applicable adaptive skill areas: communication,\n self-care, home living, social skills, community use, self-direction,\n health and safety, functional academics, leisure, and work. Mental\n retardation manifests before age 18.\n\nId. at 308 n.3. Alabama courts have followed that tripartite definition, adopting\n\nwhat they consider the “broadest” definition of mental retardation for Atkins\n\npurposes:\n\n [A] defendant, to be considered mentally retarded, must have\n significantly subaverage intellectual functioning (an IQ of 70 or\n below), and significant or substantial deficits in adaptive behavior.\n Additionally, these problems must have manifested themselves during\n the developmental period (i.e., before the defendant reached age 18).\n\nEx parte Perkins, 851 So. 2d 453, 456 (Ala. 2002).\n\n The clinical evidence adduced during the Rule 32 hearing with respect to\n\nJenkins’s intellectual ability was that his IQ was 76, reflecting “borderline\n\nintellectual functioning.” The record also included the report of the Lunacy\n\nCommission13 which found that Jenkins was competent to stand trial, as well as\n\n\n\n13\n Alabama law requires a capital defendant to be “committed to the Department of Mental\nHealth and Mental Retardation for examination” if there is a question about his capacity to stand\ntrial or if there will be a mental-disease-or-defect defense. Ala. Code § 15-16-22(a). “The\nassigned mental health professional(s) shall examine the defendant with respect to determining\nthe presence of any mental disease or defect which, if determined to be present, would affect the\n 42\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 43 of 73\n\n\nJenkins’s school and juvenile court records. The court also considered all the\n\ntestimony it had heard about Jenkins and reviewed the transcript of his murder trial\n\nbefore rejecting Jenkins’s claim of mental retardation.\n\n On appeal, after reciting the Perkins standard, the Alabama Court of\n\nCriminal Appeals explained that Jenkins had not established mental retardation:\n\n Dr. Kirkland testified that he performed psychological tests on Jenkins\n and that Jenkins’s IQ was 76. There was evidence presented at\n Jenkins’s trial indicating that Jenkins maintained relationships with\n other individuals and that he had been employed by P.S. Edwards\n Landscaping Company, Cotton Lowe 76 Service Station, and\n Paramount Painting Company. The record fails to show that Jenkins\n meets the most liberal view of mental retardation adopted by the\n Alabama Supreme Court in Perkins.\n\nJenkins, 972 So. 2d at 155. In this habeas appeal, Jenkins raises arguments about\n\nall three components of the intellectual disability showing under Atkins and\n\nPerkins, which we address in turn. He also appeals the denial of an evidentiary\n\nhearing on this issue.\n\n 1. Intellectual Function\n\n The first component of intellectual disability under Perkins is “significantly\n\nsubaverage intellectual functioning (an IQ of 70 or below).” Perkins, 851 So. 2d at\n\n456. Jenkins argues that the use of a strict IQ cutoff of 70 was contrary to or an\n\nunreasonable application of Atkins.\n\n\ncapacity of the defendant to proceed or continue to trial or which would affect the defendant’s\ncriminal responsibility at the time of the commission of the crime.” Id.\n 43\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 44 of 73\n\n\n We disagree. The state court’s decision that Jenkins failed to establish\n\nmental retardation in this respect was a reasonable application of the record and of\n\nAtkins. As a threshold matter, we note that the state court did not say it was\n\napplying a strict cutoff of 70; it merely stated that Dr. Kirkland had measured\n\nJenkins’s IQ as 76. Jenkins, 972 So. 2d at 155. Under our deferential review, the\n\nquestion is whether the decision that Jenkins had not established the intellectual\n\nfunctioning aspect of intellectual disability was an unreasonable application of\n\nAtkins. The Supreme Court explained in Atkins that it was leaving “to the State[s]\n\nthe task of developing appropriate ways to enforce the constitutional restriction\n\nupon [their] execution of sentences.” Atkins, 536 U.S. at 317 (quoting Ford v.\n\nWainwright, 477 U.S. 399, 416–17 (1986)). In line with that invitation, our Circuit\n\nhas accepted Alabama’s judicial definition of intellectual disability as laid out in\n\nPerkins.14 See Thomas, 607 F.3d at 752–53. Jenkins objects, however, that the\n\nPerkins standard, which he characterizes as articulating a “strict cutoff” of 70,\n\nconflicts with more recent Supreme Court precedent that instead requires a\n\n“clinical” definition that goes beyond a single IQ score.\n\n In particular, Jenkins points to Hall v. Florida, in which the Supreme Court\n\nheld that when an offender’s IQ is near but greater than 70, courts must take into\n\n\n\n14\n Notwithstanding the urging of the Alabama Supreme Court, see Perkins, 851 So. 2d at 455\nn.1, the Alabama legislature has not enacted a statutory definition of intellectual disability.\n 44\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 45 of 73\n\n\naccount the IQ test’s standard error of measurement as well as non-IQ evidence of\n\nintellectual function. See Hall v. Florida, 572 U.S. 701, 723 (2014) (“This Court\n\nagrees with the medical experts that when a defendant’s IQ test score falls within\n\nthe test’s acknowledged and inherent margin of error, the defendant must be able\n\nto present additional evidence of intellectual disability, including testimony\n\nregarding adaptive deficits.”). But, as Jenkins concedes, our Circuit has\n\nspecifically held that Hall is not retroactive to cases on collateral review. See In re\n\nHenry, 757 F.3d 1151, 1161 (11th Cir. 2014) (explaining that Hall does not supply\n\na new substantive rule but rather “merely provides new procedures for ensuring\n\nthat States do not execute members of an already protected group”). We also note\n\nthat Alabama courts have found the Perkins standard to comport with Hall in that\n\nit does not prevent an offender from presenting the standard error of measurement\n\nor non-IQ evidence of intellectual disability. See Reeves v. State, 226 So. 3d 711,\n\n728–29 (Ala. Ct. Crim. App. 2016). Here, the state court did not unreasonably\n\napply Atkins when it analyzed the intellectual functioning component of\n\nintellectual disability.\n\n Jenkins also argues that the state court’s decision on this component in his\n\nRule 32 proceeding resulted from an unreasonable determination of the facts. He\n\nasserts that his IQ score of 76, when combined with the test’s standard error of\n\nmeasurement and considered alongside the Flynn effect (which we discuss below),\n\n\n 45\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 46 of 73\n\n\ndoes show that he has significantly subaverage intellectual function. For the\n\nreasons that follow, we disagree.\n\n Most fundamentally, neither of the clinicians who testified about Jenkins’s\n\nintellectual ability in the Rule 32 hearing opined that he was mentally retarded.\n\nThe closest thing Jenkins can point to is the testimony of Dr. Kirkland that his IQ\n\nis two standard deviations below the mean. But his entire testimony, in context, did\n\nnot establish intellectual disability. To the contrary: Dr. Kirkland testified that\n\nJenkins “scored in the range of borderline intellectual functioning which is\n\nbetween mild mental retardation and low average intellectual functioning.”\n\nFurthermore, Jenkins’s own expert Dr. Lisak fully reviewed Dr. Kirkland’s clinical\n\nfindings. He agreed with those findings, concluding that the IQ results show\n\n“borderline intelligence.” It is tremendously significant that no clinical assessment\n\nin the entire record of these proceedings has found that Jenkins has mental\n\nretardation or intellectual disability.\n\n Jenkins nonetheless argues that his IQ score of 76 establishes the intellectual\n\ncomponent of intellectual ability when the standard error of measurement and the\n\nFlynn effect are considered alongside that score. It was not unreasonable for the\n\nstate court to disagree. As we have discussed, Jenkins is not retroactively entitled\n\nto have the state courts take into account the margin of error of the IQ test. That\n\nstatistical margin of error—5 points in this case—takes into account characteristics\n\n\n 46\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 47 of 73\n\n\nof the test and variations due to chance in order to state how confident the tester is\n\nthat the measured result accurately states Jenkins’s true, unknowable IQ. A result\n\nof 76 with a standard error of measurement of 5 simply means that the tester is\n\n68% confident that Jenkins’s true IQ falls between 71 and 81, and 95% confident\n\nthat his true IQ falls between 66 and 86. See David H. Kaye & David A. Freedman,\n\nReference Guide on Statistics, in Reference Manual on Scientific Evidence 211,\n\n243–44 (3d ed. 2011); see also Hall, 572 U.S. at 738–39 (Alito, J., dissenting). Of\n\ncourse, Jenkins would like us to consider only the possibility that the test result\n\noverstated his IQ—but the standard error of measurement is a two-way street.\n\n“[T]he standard error of measurement is a bi-directional concept that does not\n\ncarry with it a presumption that an individual’s IQ falls to the bottom of his IQ\n\nrange.” Ledford v. Warden, Ga. Diag. & Clas. Prison, 818 F.3d 600, 641 (11th Cir.\n\n2016).\n\n Even if Jenkins were somehow entitled to subtract the 5-point margin of\n\nerror from his score and thereby lower his “true” IQ to 71, he still would not satisfy\n\nthe clinical intellectual component of intellectual disability. Thus, he sought in the\n\ndistrict court to introduce evidence about the Flynn effect in order to bring his IQ\n\ndown to 70 or below. As our Court has explained, the Flynn effect “recognizes the\n\nfact that IQ test scores have been increasing over time”:\n\n The Flynn effect acknowledges that as an intelligence test ages, or\n moves farther from the date on which it was standardized, or normed,\n 47\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 48 of 73\n\n\n the mean score of the population as a whole on that assessment\n instrument increases, thereby artificially inflating the IQ scores of\n individual test subjects. Therefore, the IQ test scores must be\n recalibrated to keep all test subjects on a level playing field.\n\nThomas, 607 F.3d at 753. We have said that “[a]n evaluator may also consider the\n\n‘Flynn effect’” in assessing an offender’s possible intellectual disability, id.\n\n(emphasis added), but neither we nor the Alabama Supreme Court nor the U.S.\n\nSupreme Court have said that a court must consider it in order to reasonably apply\n\nAtkins. See, e.g., Reeves, 226 So. 3d at 739 (“This Court has repeatedly held that a\n\ncircuit court is not required to accept, consider, or apply the ‘Flynn Effect’ in\n\ndetermining intellectual disability.”).\n\n More problematically, Jenkins did not raise the Flynn effect issue in the\n\nAlabama state courts. Jenkins asserts that he implicitly raised it as part and parcel\n\nof his Atkins claim during his Rule 32 appeal, but we do not agree that the\n\nAlabama courts had a full and fair opportunity to consider this issue, because it\n\nwas not explicitly presented to them. See Kelley v. Sec’y for the Dep’t of Corr., 377\n\nF.3d 1317, 1349–50 (11th Cir. 2004) (holding that one ineffective-assistance\n\nargument was unexhausted despite the petitioner’s general presentation of a\n\nStrickland claim in the state courts). Thus, under AEDPA, Jenkins is procedurally\n\nbarred from raising this unexhausted claim in his federal habeas petition. See 28\n\nU.S.C. § 2254(b)(1)(A), (c); Mason v. Allen, 605 F.3d 1114, 1119 (11th Cir. 2010).\n\nWe will allow a prisoner to overcome a procedural bar if he establishes cause for\n 48\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 49 of 73\n\n\nhis default and prejudice therefrom. See, e.g., Smith v. Jones, 256 F.3d 1135, 1138\n\n(11th Cir. 2001). But Jenkins cannot establish prejudice from being barred from\n\nasserting this claim now because the Supreme Court has not “clearly established”\n\nthat a court must consider the Flynn effect in assessing intellectual disability. Thus,\n\nunder AEDPA, it could not form the basis for a federal habeas claim. See 28\n\nU.S.C. § 2254(d)(1). We therefore conclude that the state court did not\n\nunreasonably determine the facts or unreasonably apply Atkins with respect to the\n\nintellectual component of intellectual disability.\n\n 2. Adaptive Function\n\n The second component of intellectual disability is “significant or substantial\n\ndeficits in adaptive behavior.” Perkins, 851 So. 2d at 456. Jenkins argues that the\n\nstate court, which mentioned only Jenkins’s successes in employment and social\n\nskills, see Jenkins, 972 So. 2d at 155, failed to consider his deficits. He adds that\n\neven the mention of employment success does not tell the whole story. Jenkins\n\ninsists that his jobs were menial, his employers took advantage of him, and he held\n\nthe jobs for only a few months each.\n\n But the state court’s terse discussion of Jenkins’s adaptive behavior did not\n\ninvolve an unreasonable determination of the facts or an unreasonable application\n\nof Atkins. To the contrary, its decision on this issue was reasonable in light of both\n\nthe record and Atkins.\n\n\n 49\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 50 of 73\n\n\n Atkins elaborated that the adaptive-skill component of intellectual disability\n\nrequires substantial present limitation in at least two of the following areas:\n\n“communication, self-care, home living, social skills, community use, self-\n\ndirection, health and safety, functional academics, leisure, and work.” Atkins, 536\n\nU.S. at 308 n.3 (quoting definition of American Association on Mental\n\nRetardation). The record supports the state court’s judgment that Jenkins does not\n\nhave substantial deficits in at least two areas. To be sure, we will assume that\n\nJenkins may be substantially limited in functional academics; Dr. Kirkland\n\nassessed his reading, spelling, and arithmetic skills at a third-grade level. But the\n\nrecord does not compel a finding of substantial present deficits in the areas of\n\ncommunication, self-care, home living, and work. Although Jenkins emphasizes\n\nhis transient employment, the record shows he was able to find work quickly, to\n\nwork hard to support the Seals’ household until Lonnie Seal found a job, and, later,\n\nto move into his own home. That record evidence is inconsistent with a substantial\n\ndeficit in the area of work.\n\n As the State observes, the facts of the crime also fail to show significant\n\ndeficits in the areas of communication, self-care, community use, and self-\n\ndirection. Jenkins was able to communicate well enough to solicit an alibi and to\n\nsell his car, including writing out a bill of sale. He then was able to arrange for his\n\nflight from Alabama, including purchasing a bus ticket and hitchhiking across the\n\n\n 50\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 51 of 73\n\n\ncountry. Indeed, throughout his trial and postconviction proceedings, the people\n\nwho knew Jenkins as an adult consistently described a man who did not have\n\nserious difficulties in communicating, forming relationships, working, and caring\n\nfor himself. No witness mentioned these kinds of difficulties in his or her\n\ntestimony about Jenkins. There was some evidence that Jenkins was “not\n\narticulate” and that his house was messy, but that testimony does not rise to the\n\nlevel of overwhelming evidence of substantial functional deficits that would\n\nsupport a conclusion that the state court unreasonably determined the facts.\n\n Of course, the record also contains evidence of Jenkins’s childhood\n\nacademic and social deficits, attributable at least in part to his parents’ neglect, but\n\nJenkins’s childhood is not directly relevant to our consideration of his present\n\nlimitations. Although we acknowledge that the Supreme Court has noted that such\n\n“risk factors” may support further exploration of the possibility of intellectual\n\ndisability, see Moore v. Texas, 137 S. Ct. 1039, 1051 (2017), the state court record\n\ncontains the results of that exploration. Overall, that record supports the state\n\ncourt’s conclusion that Jenkins does not have substantial deficits in adaptive\n\nbehaviors.\n\n 3. Juvenile Onset\n\n The third and final component of intellectual disability is manifestation of\n\nthe first and second components “before the defendant reached age 18.” Perkins,\n\n\n 51\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 52 of 73\n\n\n851 So. 2d at 456. Jenkins asserts, the State assumes, and we agree that the state\n\ncourt did not rule upon this issue, so there is nothing to which we must defer under\n\nAEDPA. See, e.g., Brumfield v. Cain, 135 S. Ct. 2269, 2282 (2015). Thus, we will\n\nreview the district court’s legal conclusions on this issue de novo, and its factual\n\nfindings for clear error. See, e.g., Grossman v. McDonough, 466 F.3d 1325, 1335\n\n(11th Cir. 2006). Jenkins argues that he has satisfied the age-of-onset component\n\nby showing “a wide range of evidence of deficits in both intellectual and adaptive\n\nfunctioning throughout Jenkins’s childhood.”\n\n In fact, the ample record of Jenkins’s childhood, as developed in the Rule 32\n\nproceeding, does not point to intellectual disability before age 18. School\n\nassessments at age 12 reported an IQ score of 83, an adaptive behavior rating of\n\n“low,” and noted overall an intellectual capacity of “average,” a diagnosis of\n\ndyslexia, and “poor functional skills.” The psychologist attributed Jenkins’s\n\ndifficulties to a combination of “[f]requent moves, changes in schools and family\n\ninstability” with “[l]ow ability coupled with a specific learning disability” and\n\n“[l]ack of responsibility for self control and poor use of time.” Although Jenkins\n\nfailed several grades of school, teachers additionally attributed his difficulties to\n\nhis behavioral problems and his high number of absences from school.\n\n Similarly, IQ tests ordered by the juvenile court at age 14 gave scores of 81\n\nand 86 on different tests, in the “[d]ull–[n]ormal” range and reflecting “a definite\n\n\n 52\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 53 of 73\n\n\nlearning disability.” These records all show a child with serious academic deficits\n\nand some intellectual and adaptive deficits, but they do not clearly show an\n\nintellectually disabled child. In particular, Jenkins’s childhood IQ scores of 83, 81,\n\nand 86 fail to point to intellectual disability or mental retardation. We are not left\n\nwith the definite and firm conviction that the district court erred when it found that\n\n“Jenkins’s [childhood] intelligence scores and adaptive functioning skills were\n\nattributable to his learning disability and other circumstances in his life, rather than\n\nmental retardation.”\n\n In sum, we agree with the district court that the decision of the Alabama\n\nCourt of Criminal Appeals that Jenkins does not have intellectual disability was\n\nnot contrary to or an unreasonable application of Atkins v. Virginia.\n\n 4. Evidentiary Hearing\n\n Finally, Jenkins argues that the district court erred when it denied his request\n\nfor an evidentiary hearing on his Atkins claim. If Jenkins diligently attempted to\n\npresent this claim in the state court, we review for abuse of discretion. Burgess v.\n\nComm’r, Ala. Dep’t of Corr., 723 F.3d 1308, 1320 (11th Cir. 2013). “The district\n\ncourt abuses its discretion where ‘such a hearing could enable an applicant to prove\n\nthe petition’s factual allegations, which, if true, would entitle the applicant to\n\nfederal habeas relief.’” Id. (quoting Schriro v. Landrigan, 550 U.S. 465, 474\n\n(2007)). If Jenkins did not diligently attempt to develop the factual basis for this\n\n\n 53\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 54 of 73\n\n\nclaim in the state court, absent certain circumstances, he is not entitled to a hearing\n\nand we will affirm the district court’s denial. 28 U.S.C. § 2254(e)(2).\n\n The parties strenuously dispute whether Jenkins diligently attempted to\n\npresent this issue in the state court. On the one hand, the State argues that Jenkins\n\nfailed to allege that he was mentally retarded even after the U.S. Supreme Court\n\ndecided Atkins in 2002. On the other hand, Jenkins asserts that he raised the\n\npossibility of mental retardation as early as his original Rule 32 petition in 1995.\n\n As we have noted, shortly after the Supreme Court decided Atkins, the\n\nAlabama Court of Criminal Appeals ordered supplemental briefing on the possible\n\nimpact of Atkins on Jenkins. Jenkins’s Rule 32 appeal was pending at that point,\n\nhaving been fully briefed and orally argued in the state appellate court. Having\n\nreviewed the state court records, we agree with the State that Jenkins did not\n\ndiligently attempt to present his Atkins claim in the state courts.\n\n “Diligence will require in the usual case that the prisoner, at a minimum,\n\nseek an evidentiary hearing in state court.” Williams, 529 U.S. at 437. The record\n\nclearly shows that Jenkins did not do so. In the initial supplemental brief on Atkins\n\nhe filed in 2002, Jenkins merely requested a future hearing after the Alabama\n\nlegislature’s enactment of a statute on executing the mentally retarded. Seventeen\n\nyears later, that contingency is still in the future. Only in his reply supplemental\n\nbrief did Jenkins then assert, for the first time ever, that he is mentally retarded.\n\n\n 54\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 55 of 73\n\n\n We acknowledge that Jenkins’s initial Rule 32 petition included in its\n\nallegations of ineffective assistance of counsel a list of unpresented mitigating\n\nevidence including his “developmental[] impair[ment]” and “learning disabilities,\n\nlow intelligence, poor comprehension and retarded socialization skills.” We\n\nconclude that these tangential Strickland references to his mental and social\n\nabilities do not amount to diligence. Thus, because Jenkins did not diligently\n\nattempt to develop the factual basis for his Atkins claim in the state courts, we\n\naffirm the district court’s denial of his request for an evidentiary hearing.\n\n Even if Jenkins had diligently attempted to present this claim in the state\n\ncourt, however, we conclude in the alternative that the district court did not abuse\n\nits discretion when it denied a hearing. We agree, in light of the ample evidence\n\nabout intellectual ability that Jenkins adduced in support of his ineffective-\n\nassistance claim, that such a hearing would not have enabled Jenkins to prove that\n\nhe is intellectually disabled. See Burgess, 723 F.3d at 1320.\n\n IV. CONCLUSION\n\n The denial of Jenkins’s petition for a writ of habeas corpus is\n\n AFFIRMED.\n\n\n\n\n 55\n\n Case: 17-12524 Date Filed: 08/30/2019 Page: 56 of 73\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Mark Jenkins v. Commissioner, Alabama Department of Corrections\\nCourt: U.S. Court of Appeals for the Eleventh Circuit\\nDate decided: 2019-08-30\\nOpinion type: majority\\nHas dissenting opinion: Yes\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nCase: 17-12524 Date Filed: 08/30/2019 Page: 1 of 73\\n\\n\\n [PUBLISH]\\n\\n IN THE UNITED STATES COURT OF APPEALS\\n\\n FOR THE ELEVENTH CIRCUIT\\n\\n\\n No. 17-12524\\n\\n\\n\\n D.C. Docket No. 4:08-cv-00869-VEH\\n\\nMARK ALLEN JENKINS,\\n\\n Petitioner - Appellant,\\n\\n versus\\n\\n\\n\\nCOMMISSIONER, ALABAMA DEPARTMENT OF CORRECTIONS,\\n\\n Respondent - Appellee.\\n\\n\\n\\n Appeal from the United States District Court\\n for the Northern District of Alabama\\n\\n\\n (August 30, 2019)\\n\\nBefore TJOFLAT, WILSON, and BRANCH, Circuit Judges.\\n\\nBRANCH, Circuit Judge:\\n\\n Mark Allen Jenkins, an Alabama prisoner sentenced to death for the 1989\\n\\nmurder of Tammy Ruth Hogeland, appeals the district court’s denial of his petition\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 2 of 73\\n\\n\\nfor a writ of habeas corpus, 28 U.S.C. § 2254. Before us are Jenkins’s arguments\\n\\nthat he received ineffective assistance of counsel during the penalty phase of his\\n\\ntrial and that he is intellectually disabled and therefore ineligible for the death\\n\\npenalty. After careful consideration, and with the benefit of oral argument, we\\n\\naffirm the denial of Jenkins’s habeas petition on both grounds.\\n\\n I. BACKGROUND\\n\\n A. The Crime and Arrest\\n\\n We summarize the following background narrative from opinions of the\\n\\nAlabama Court of Criminal Appeals and our own review of the record. See Jenkins\\n\\nv. State, 627 So. 2d 1034, 1037–40 (Ala. Ct. Crim. App. 1992); Jenkins v. State,\\n\\n972 So. 2d 111, 119–20 (Ala. Ct. Crim. App. 2004). The events leading up to the\\n\\nmurder began the evening of April 17, 1989. Jenkins was at the home of his\\n\\nacquaintance Christine Nicholas. He had met her several months earlier at the\\n\\nOmelet Shoppe restaurant where she worked. Jenkins was very intoxicated and\\n\\ntried to seduce her. When she resisted, Jenkins got “real mad” and asked her\\n\\nseveral times what she would do if someone came up from behind her and grabbed\\n\\nher.\\n\\n At around 1 a.m., Jenkins and Nicholas drove to the Riverchase Omelet\\n\\nShoppe. Jenkins went inside and talked with one of the waitresses, Frieda Vines.\\n\\nThe manager, Douglas Thrash, recognized Jenkins as a regular customer who knew\\n\\n\\n 2\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 3 of 73\\n\\n\\nall of the waitresses. Thrash overheard part of Jenkins’s conversation with Vines\\n\\nand heard mention of the Omelet Shoppe location near the Birmingham Airport.\\n\\nJenkins and Nicholas then returned to Nicholas’s home. Around 2 a.m., Nicholas’s\\n\\nmother asked Jenkins to leave. He did so, falling down some steps and ramming\\n\\nhis car into another vehicle in the process.\\n\\n Meanwhile, 23-year-old Tammy Hogeland was making her way to work at\\n\\nthe Omelet Shoppe. Her sister Wendy, along with Tammy’s young son, picked\\n\\nTammy up from her college classes and drove her to the Riverchase Omelet\\n\\nShoppe a little after 9:30 p.m. Sometime after 10:00, Tammy was sent to work at\\n\\nthe airport Omelet Shoppe on Tenth Avenue, since that location was unexpectedly\\n\\nshorthanded. Tammy was wearing her watch, a necklace with the words “special\\n\\nsister,” her class ring, and her diamond engagement ring. Tammy was working as a\\n\\ncook that night and was wearing a blue apron, black pants, a white shirt, and white\\n\\nshoes. She and Sarah Harris were the only employees working at the airport\\n\\nOmelet Shoppe that night.\\n\\n At about 2 a.m. on April 18, Harris saw Jenkins—whom she did not know—\\n\\ndrive up in a red sports car. His arrival was memorable to Harris because the car\\n\\nnearly jumped the curb and crashed through the restaurant’s glass wall. Jenkins got\\n\\nout of the car and came into the restaurant, appearing to be intoxicated. He walked\\n\\nover to Hogeland and began talking to her. Harris later saw Jenkins and Hogeland\\n\\n\\n 3\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 4 of 73\\n\\n\\ndrive away together in the red sports car. That was the last time anyone who knew\\n\\nHogeland ever saw her. Hogeland left behind her cigarettes, lighter, purse, and\\n\\npaycheck. She left without telling anyone, which she had never done before.\\n\\nAlthough Harris saw them drive off, she could not later say whether Hogeland left\\n\\nwith Jenkins willingly or was instead abducted.\\n\\n At around 5 a.m. that morning, Bobby and Geraldine Coe had stopped to\\n\\nbuy gas at a Chevron station on I-59 northeast of Birmingham when they saw\\n\\nJenkins drive up in the red sports car. The Coes noticed a female who appeared to\\n\\nbe “passed out” in the front passenger seat, but they could not say whether she was\\n\\nalive or dead. While Bobby Coe was pumping gas, Jenkins asked him for some\\n\\ncigarettes and said, “Looks like it’s been a long night and it looks like it’s going to\\n\\nbe a long day.” Jenkins then said “God bless you” before asking directions to I-\\n\\n459. Coe gave him directions and got back in his car. As Coe pulled out onto the\\n\\ninterstate, he saw Jenkins follow him in the red sports car. Coe then saw the car\\n\\nflash its lights, slow down, and pull to the side of the road between mile markers\\n\\n151 and 152.\\n\\n At 8 a.m., Wendy Hogeland learned that Tammy was not at the restaurant,\\n\\nand their mother called the police. Meanwhile, Jenkins went to the home of Steve\\n\\nMusser, who noticed that he was wearing the same clothes as the day before.\\n\\nJenkins told Musser that his car had been stolen the night before and asked him if\\n\\n\\n 4\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 5 of 73\\n\\n\\nhe would say that he had been with him all night. Musser refused. Christine\\n\\nNicholas also saw Jenkins in a grocery store that morning. He was looking at a\\n\\nnewspaper, making a phone call, and attempting to sell his Buick. Nicholas loaned\\n\\nhim $4 for gas. At around 10 o’clock that morning, Jenkins sold his car to Michael\\n\\nBrooks, a mechanic at a local Chevron station. Jenkins had explained that he\\n\\nneeded the money so he could visit his sick mother in California. Another\\n\\nmechanic at the service station drove Jenkins to the Greyhound bus station later\\n\\nthat morning.\\n\\n Jenkins awoke the next day on the Greyhound bus in Houston and was\\n\\nejected from the bus because his fare was used up. He then hitchhiked from\\n\\nHouston to Phoenix, then to San Diego and Los Angeles. Jenkins was first\\n\\nidentified as a suspect in Hogeland’s disappearance by Officer Mike Weems of the\\n\\nHoover Police Department on April 19. Weems ate dinner at the Riverchase\\n\\nOmelet Shoppe just about every day and had talked with Hogeland just as often.\\n\\nHe also knew Jenkins from the restaurant. Weems learned of Hogeland’s\\n\\ndisappearance from the other waitresses. Remembering how Jenkins often talked to\\n\\nHogeland and passed her notes,1 he gave Jenkins’s name to the missing persons\\n\\ninvestigator. The investigator also learned that a red Mazda RX-7 sports car, which\\n\\n\\n\\n1\\n The jury was not permitted to hear, and the record does not reveal, what exactly Jenkins had\\nsaid to Hogeland that made Officer Weems suspect him in her disappearance.\\n 5\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 6 of 73\\n\\n\\nhad been reported stolen from the service station where Jenkins worked, had been\\n\\nrecovered on I-459.\\n\\n The afternoon of April 21, a truck driver who had happened to stop on I-59\\n\\nnear mile marker 151 discovered Hogeland’s body off the side of the road. She was\\n\\nnaked, wearing only a watch, and was so badly decomposed that the body had to\\n\\nbe identified by dental records. From the fractured hyoid bone, it was determined\\n\\nthat she was manually strangled to death. Also found at the scene were Hogeland’s\\n\\napron, shoes, bra, panties, pants, and hair net, as well as some beer cans, a Mazda\\n\\nRX-7 owner’s manual, and other items later determined to have come from the red\\n\\nMazda. Her necklace and rings were never found.\\n\\n Alabama authorities issued a warrant for Jenkins’s arrest and also obtained a\\n\\nfederal fugitive warrant. An FBI agent arrested Jenkins on May 1 in front of his\\n\\nuncle’s house in Wilmington, California. Jenkins’s uncle later gave the police three\\n\\nbags of Jenkins’s personal effects. Fibers recovered from Hogeland’s body and\\n\\nclothing matched those of the Mazda, and hairs matching Hogeland’s were found\\n\\nin the Mazda. Similarly, fibers from Jenkins’s clothing placed him in the Mazda,\\n\\nand fibers of his jeans were found on Hogeland’s apron. A bootprint near\\n\\nHogeland’s body matched a combat boot from among the belongings Jenkins’s\\n\\nuncle gave to the police, and Jenkins’s roommate said the boots looked just like\\n\\nthose that Jenkins wore.\\n\\n\\n 6\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 7 of 73\\n\\n\\n A business card like those that had been in the Mazda’s glove box,\\n\\nbelonging to an owner of the Mazda, was recovered from Jenkins’s wallet after his\\n\\narrest in California. Later, Jenkins’s cellmate in the Alabama jail reported that\\n\\nJenkins told him “he had done the crime” and was worried that the couple from the\\n\\ngas station would identify him or that the police would find his fingerprint on a\\n\\nbeer can at the scene. A St. Clair County grand jury indicted Jenkins for capital\\n\\nmurder.\\n\\n B. The Trial and Sentencing\\n\\n Jenkins was represented at trial in 1991 by attorneys Douglas Scofield and\\n\\nStan Downey. Scofield, the lead attorney, had been practicing criminal defense in\\n\\nBirmingham since 1984 but had never before tried a capital murder case. Jenkins’s\\n\\nlandlord and his grandmother had arranged the referral and representation before\\n\\nthe court appointed him. Scofield agreed to the appointment on the condition that a\\n\\nSt. Clair County lawyer also be appointed. That lawyer was Downey, who had\\n\\nmore experience than Scofield but whose capital experience was limited to one\\n\\ncase that had not gone to trial. The two attorneys had agreed that Downey would\\n\\nbe responsible for jury selection, but Scofield eventually took over during the voir\\n\\ndire. Per their agreement, Scofield was responsible for the guilt-phase trial,\\n\\nalthough Downey participated by interviewing Jenkins several times. Scofield also\\n\\nmet with Jenkins around a dozen times. Jenkins had told him about his background\\n\\n\\n 7\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 8 of 73\\n\\n\\nand childhood, including the fact that he had had a difficult childhood and had\\n\\nlived on the streets since age 11.\\n\\n Throughout the weeklong trial, Scofield pursued an outright acquittal. In\\n\\nparticular, he challenged Harris’s identification of Jenkins as the man who drove\\n\\naway with Hogeland, and he challenged the medical examiner’s opinions about the\\n\\ntime and cause of Hogeland’s death. In his closing argument, Scofield took a wide-\\n\\nranging approach to undermining the State’s wholly circumstantial case against\\n\\nJenkins. He attacked the State’s timeline of the crime, questioning whether Jenkins\\n\\nin his intoxicated state would have had the time and the ability to leave Nicholas’s\\n\\nhouse, steal the Mazda, and abduct and kill Hogeland. He reminded the jury of the\\n\\nweakness of Harris’s identification of Jenkins and the Coes’ identification of the\\n\\nwoman in the car. He pointed out that no physical evidence definitively linked\\n\\nJenkins to the Mazda, to Hogeland’s body, or to her jewelry, which was never\\n\\nrecovered. And he argued that the evidence provided no plausible motive for\\n\\nJenkins to have kidnapped, robbed, or killed Hogeland. In sum, he urged the jury\\n\\nto find reasonable doubt and acquit Jenkins. At the very least, he argued, they\\n\\nshould find that the State failed to prove Jenkins’s intent to rob and kidnap\\n\\nHogeland, which were necessary to convict Jenkins of capital murder. He further\\n\\nargued that, if the jury thought Jenkins killed Hogeland as a crime of passion, they\\n\\ncould convict him only of manslaughter. The jury deliberated for three hours and\\n\\n\\n 8\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 9 of 73\\n\\n\\nfifteen minutes before convicting Jenkins of capital murder, Ala. Code § 13A-5-\\n\\n40(a)(1) & (2), i.e., murder committed in the course of kidnapping and robbery.\\n\\n The penalty phase took place that same afternoon. A week before trial,\\n\\nScofield and Downey had discussed the penalty phase, and Scofield understood\\n\\nthat Downey would be responsible for handling the penalty phase. Nonetheless, the\\n\\nwitness who was present to testify on Jenkins’s behalf during the penalty phase\\n\\nwas better acquainted with Scofield, so Downey suggested that afternoon that\\n\\nScofield conduct the direct examination. Scofield urged Downey to interview the\\n\\nwitness quickly and prepare to examine him, and Downey did so.\\n\\n Downey handled the penalty presentation to the jury. The State did not\\n\\npresent any additional aggravation evidence, resting on the evidence presented\\n\\nduring the guilt phase. One witness, Lonnie Seal, testified on Jenkins’s behalf in\\n\\norder, in Downey’s words, to “reveal another side of Mark Jenkins to you that you\\n\\ndon’t know anything about.” Seal had met Jenkins in Fontana, California, in 1988\\n\\nwhen they both worked at the same garage. Seal and his wife became friends with\\n\\nJenkins, and he visited in their home. When the Seals decided to move to Alabama\\n\\na few months later, Jenkins agreed to help with their move and to live with the\\n\\nSeals in Alabama for a time. It took Lonnie Seal several weeks to find a job, but\\n\\nJenkins found a job in two days and contributed toward the family’s rent and\\n\\n\\n\\n\\n 9\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 10 of 73\\n\\n\\ngroceries. Jenkins continued to visit the Seals after he moved into his own home,\\n\\nand he would always bring a small gift for and play with the Seals’ infant son.\\n\\n Downey’s closing argument to the jury requested a sentence of life without\\n\\nparole. Downey reminded the jurors that they had agreed during voir dire that\\n\\ndeath was not always the appropriate punishment for homicide. Quoting the Bible,2\\n\\nhe reminded the jury that the Old Testament permitted capital punishment only\\n\\nwhen supported by the testimony of two or three eyewitnesses, whereas in this case\\n\\nthere had been only circumstantial evidence. He asserted from Seal’s testimony\\n\\nthat “Mark had a side of him that was generous, that was giving, and that was\\n\\nkind,” and argued that the deterrent and retributive purposes of capital punishment\\n\\nwould not be served in this case.\\n\\n The jury deliberated for 50 minutes before returning its nonbinding\\n\\nrecommendation of death by a vote of 10 to 2. See Ala. Code § 13A-5-46(e) (1981)\\n\\n(amended 2017 to make jury’s verdict binding). After a separate hearing, the court\\n\\nfound the two statutory aggravating factors for which the State had argued: that the\\n\\nmurder was committed during the commission of a robbery, and that it was\\n\\ncommitted during the commission of a kidnapping. See Ala. Code § 13A-5-49(4).\\n\\n\\n\\n\\n2\\n “If anyone kills a person, the murderer shall be put to death on the evidence of witnesses. But\\nno person shall be put to death on the testimony of one witness.” Numbers 35:30 (ESV). Downey\\nalso paraphrased Deuteronomy 17:6 as “On the evidence of two witnesses or three witnesses, he\\nis [sic] shall be put to death. He shall not be put to death on the evidence of one witness.”\\n 10\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 11 of 73\\n\\n\\nThe court also found two statutory mitigating factors: that Jenkins had no\\n\\nsignificant history of prior criminal activity, and his age at the time of the crime\\n\\n(21). See id. § 13A-5-51(1), (7). The court considered but rejected the statutory\\n\\nmitigation that, due to his consumption of alcohol, Jenkins’s capacity to appreciate\\n\\nthe criminality of his conduct or to conform his conduct to the requirements of law\\n\\nwas substantially impaired. See id. § 13A-5-51(6). The court found that the\\n\\naggravating circumstances outweighed the mitigating circumstances and, in\\n\\naccordance with the jury’s recommendation, imposed a sentence of death.\\n\\n C. The State and Federal Post-Conviction Proceedings\\n\\n On direct appeal, the Alabama Court of Criminal Appeals affirmed Jenkins’s\\n\\nconviction and death sentence. Jenkins v. State, 627 So. 2d 1034 (Ala. Ct. Crim.\\n\\nApp. 1992), aff’d, 627 So. 2d 1054 (Ala. 1993), cert. denied, 511 U.S. 1012\\n\\n(1994). In 1995, Jenkins filed his first petition for state postconviction relief, Ala.\\n\\nR. Crim. P. 32,3 alleging, among many other things, that he received ineffective\\n\\nassistance of counsel during his penalty phase. The Rule 32 trial court conducted a\\n\\nthree-day hearing at which evidence was presented about the performance of\\n\\nJenkins’s attorneys.\\n\\n\\n\\n\\n3\\n “Subject to the limitations of Rule 32.2, any defendant who has been convicted of a criminal\\noffense may institute a proceeding in the court of original conviction to secure appropriate relief\\non the ground that . . . [t]he constitution of the United States or of the State of Alabama requires\\na new trial, a new sentence proceeding, or other relief. . . .” Ala. R. Crim. P. 32.1.\\n 11\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 12 of 73\\n\\n\\n Scofield testified about the representation that he and Downey provided. In\\n\\nrelevant part, Scofield asserted that he and Downey had “decided that [developing\\n\\nmitigation for the penalty phase] would be something that Stan [Downey] would\\n\\nhandle.” Downey later told Scofield that he hadn’t done any such investigation, but\\n\\nScofield admitted that, “[a]part from what he has told me, I don’t know what he\\n\\nhas done.” Scofield testified that Downey “didn’t appear prepared” for the penalty\\n\\nphase, based on Downey’s last-minute suggestion that Scofield examine Seal since\\n\\nhe had talked with him before.\\n\\n In preparation for the penalty phase, Scofield himself did not interview any\\n\\nmember of Jenkins’s family, but he did not know whether Downey interviewed\\n\\nany family members. Scofield did interview Jenkins’s landlord and some other\\n\\nacquaintances but did not do so specifically for the purpose of developing\\n\\nmitigation. He had met with Jenkins around a dozen times, and Jenkins had told\\n\\nhim his life story including his experience of abuse as a child. In hindsight,\\n\\nScofield said he had a “lack of appreciation of mitigating circumstances.” He\\n\\nrepeatedly denied in his Rule 32 testimony that his omissions during the penalty\\n\\nphase were strategic: “I didn’t understand what I was doing. . . . I just didn’t\\n\\nunderstand what I needed to do.”\\n\\n For reasons that remain unclear, Downey—who is since deceased—did not\\n\\ntestify or submit an affidavit in the Rule 32 proceeding. Scofield answered,\\n\\n\\n 12\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 13 of 73\\n\\n\\n“Obviously not,” when asked whether he had personal knowledge of everything\\n\\nDowney may or may not have done to prepare for the penalty phase. The record of\\n\\nthe extent of Downey’s pretrial preparation is thus limited to Downey’s detailed\\n\\nfee declaration, which he submitted to the State of Alabama in documentation of\\n\\nthe compensation due him as appointed counsel. See Ala. Code § 15-12-21. In it,\\n\\nDowney reported that he spent 71 hours4 on out-of-court preparation for trial. The\\n\\nreport details 35 hours of interviews with Jenkins in 18 separate visits to the jail\\n\\nover the course of 17 months, as well as one substantive conversation with\\n\\nJenkins’s grandmother.5\\n\\n The Rule 32 trial court also heard testimony from four family members and\\n\\none friend of Jenkins’s. The family members presented for the first time a stark\\n\\ndepiction of Jenkins as an unloved and abused child in California. Jenkins’s older\\n\\nbrother Michael Jenkins testified that all of his siblings have different fathers and\\n\\nthat his parents often used drugs and fought. Michael testified that Jenkins was\\n\\nbeaten frequently as a child by his stepfather, if not every day then every other day\\n\\n\\n\\n4\\n In Jenkins’s state postconviction appeal, the Alabama Court of Criminal Appeals erroneously\\nstated that Downey’s fee declaration reports 171 hours of pretrial preparation. See Jenkins, 972\\nSo. 2d at 137. We discuss the effect of this discrepancy infra note 7.\\n5\\n The state appellate court also noted that Downey had moved for a continuance in October 1989\\nso that he could conduct “[f]urther discovery and investigation (including a possible trip to\\nCalifornia).” Id. We note that Scofield had orally suggested to the trial court in September 1989\\nthat more time might be needed with respect to “evidence about the defendant” and witnesses in\\nCalifornia, as well as possible psychological evaluation.\\n 13\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 14 of 73\\n\\n\\nor every three days. Jenkins was beaten all over his body with a variety of\\n\\nimplements. Michael was beaten too. Jenkins’s stepfather would lock Jenkins in\\n\\nhis bedroom for a couple of hours at a time, sometimes without dinner. Some of\\n\\nthe beatings were prompted by Jenkins’s bedwetting or soiling. Michael testified\\n\\nthat Jenkins’s stepfather sometimes made Jenkins wear his soiled underwear on his\\n\\nhead and even forced him to eat his own feces a few times. Michael testified that\\n\\nJenkins’s bladder and bowel control problems began after a camping trip he took\\n\\nwith his stepfather and stepgrandfather.\\n\\n Michael testified that, as children, he and Jenkins were made to work on car\\n\\nengines late into the night and to shovel horse manure daily. The family moved\\n\\nfrequently and the children attended many different schools as a result. Michael\\n\\ntestified that he was Jenkins’s only friend as a child and that Jenkins did poorly in\\n\\nschool. Jenkins began running away from home as a young teenager and lived with\\n\\nvarious other family members or on the streets. Michael said that Jenkins was a\\n\\ngood brother to him and had asked him to testify at his trial, but he never heard\\n\\nfrom his attorneys or found out when the trial was. Michael admitted to having\\n\\ncognitive and memory problems due to a head injury years earlier.\\n\\n Jenkins’s cousin Tammy Pitts testified that, as an infant, Jenkins’s mother\\n\\nbeat and neglected him, leaving him dirty and in soiled diapers. His stepfather also\\n\\nbeat Jenkins daily until he left home at age 13, sometimes bruising him so badly\\n\\n\\n 14\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 15 of 73\\n\\n\\nthat he was laid up in bed for days or weeks. Jenkins had problems with\\n\\nbedwetting and soiling, and his stepfather made him wear soiled clothes to school.\\n\\nJenkins would be locked in his filthy bedroom around the clock except to attend\\n\\nschool and do chores. He was usually not allowed to eat dinner with the rest of the\\n\\nfamily and subsisted on scraps and dog food. The entire family belittled Jenkins\\n\\nbecause his biological father was Hispanic and his complexion was therefore\\n\\ndarker. Jenkins’s parents drank a lot and used drugs every day. Pitts knew these\\n\\nthings because she had daily contact with the family and lived with them off and\\n\\non, and Jenkins lived with her for a time after he ran away from home. No one had\\n\\ncontacted her about testifying at Jenkins’s trial.\\n\\n Jenkins’s second cousin Betty DeLavega testified that her relationship with\\n\\nJenkins was like that of an aunt. She lived with the Jenkins family for five months\\n\\nwhen Jenkins was 10 or 11. She saw Jenkins’s stepfather slap Jenkins “[q]uite a\\n\\nfew times,” and once, he made Jenkins eat his own feces. DeLavega testified that\\n\\nJenkins’s parents used drugs nearly every day and withheld affection from Jenkins\\n\\nand Michael. No one contacted her about Jenkins’s trial; Jenkins’s grandmother\\n\\nhad asked her to testify in the Rule 32 hearing about any abuse she had witnessed.\\n\\n Jenkins’s grandmother Doris Wagoner testified that while Jenkins’s mother\\n\\nwas pregnant with him, her husband was in prison and she drank and took drugs.\\n\\nWhen Jenkins was born, his mother initially put him up for adoption, then changed\\n\\n\\n 15\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 16 of 73\\n\\n\\nher mind weeks later. Jenkins was a lethargic infant and his mother cared more\\n\\nabout partying than about her children. Wagoner testified that she heard of\\n\\nJenkins’s mistreatment as a child but didn’t see it. She could tell from the way that\\n\\nJenkins feared his stepfather that he was being abused, and she knew he slept in a\\n\\nfilthy bed. After Jenkins was arrested for murder, Wagoner spoke with Scofield,\\n\\nwho asked for money relating to his representation of Jenkins. She did not\\n\\nremember whether she talked with Scofield about Jenkins’s background. She could\\n\\nnot remember why she didn’t testify at the trial; she might have been ill, but she\\n\\nwas probably busy.\\n\\n Sherry Seal, the wife of Lonnie Seal, who had testified in the penalty phase\\n\\nof the trial, testified that Jenkins was respectful and generous and that she trusted\\n\\nhim with her children. She related how he supported her family after moving to\\n\\nAlabama, giving them “pretty much” his entire paycheck. She said she would still\\n\\nfeel safe around Jenkins. Two jailers also testified that Jenkins was a respectful,\\n\\ncourteous, model inmate who never complained or caused any problems.\\n\\n Jenkins also introduced medical, school, and juvenile court records from his\\n\\nchildhood. The medical records document his premature birth and his mother’s\\n\\nintention to give him up for adoption. The school records reflect a struggling\\n\\nstudent who was repeatedly promoted to the next grade because of his age, despite\\n\\nserious academic difficulties, chronic truancy, and frequent school changes. The\\n\\n\\n 16\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 17 of 73\\n\\n\\njuvenile court records document Jenkins’s placement in juvenile hall starting at age\\n\\n14, following repeated acts of grand theft auto and running away.\\n\\n The Rule 32 hearing concluded with the testimony of two psychologists\\n\\nabout Jenkins’s history of psychological trauma stemming from his childhood\\n\\nabuse. The defense expert, Dr. David Lisak, had met with Jenkins three times and\\n\\ninterviewed a dozen of his family members, family friends, and associates. Dr.\\n\\nLisak testified that the pervasive abuse and neglect in Jenkins’s childhood impaired\\n\\nhis ability to cope with trauma as an adult. He opined that Jenkins’s childhood\\n\\nabuse adversely affected his cognitive and emotional development and other\\n\\nabilities. He described how Jenkins had been neglected as an infant and sexually\\n\\nabused by his grandfather at age four, which led to a lifetime of bladder and bowel\\n\\ndysfunction. He also discussed the hypothesis that children who are abused are at\\n\\ngreater risk for perpetrating violence later in life, but he acknowledged there would\\n\\nhave been other causative factors in Jenkins’s case. Dr. Lisak described Jenkins’s\\n\\nintellectual capacity as “somewhat borderline,” though he did not perform any\\n\\ntesting of his own.\\n\\n The State’s expert, Dr. Karl Kirkland, had met with Jenkins for several hours\\n\\nand administered several tests. On the Wechsler Adult Intelligence Scale, Jenkins\\n\\n“scored in the range of borderline intellectual functioning which is between mild\\n\\nmental retardation and low average intellectual functioning” with an overall IQ of\\n\\n\\n 17\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 18 of 73\\n\\n\\n76. Dr. Kirkland agreed that Jenkins had been “reared in a chaotic dysfunctional\\n\\nfamily” and generally agreed with Dr. Lisak’s assessment of the negative effects of\\n\\nchildhood trauma.\\n\\n The court considered all of this evidence before denying Rule 32 relief in\\n\\nfull. As relevant here, it found that some of the evidence about the severity of\\n\\nabuse that Jenkins suffered as a child was not credible. It further found that Jenkins\\n\\nsuffered no prejudice from any deficient performance by his penalty-phase counsel\\n\\nbecause the aggravating circumstances outweighed any additional mitigating\\n\\nevidence Downey might have adduced.\\n\\n The Alabama Court of Criminal Appeals affirmed the denial of Rule 32\\n\\nrelief. Jenkins v. State, 972 So. 2d 111 (Ala. Ct. Crim. App. 2004). In relevant part,\\n\\nthe appeals court rejected Jenkins’s ineffective assistance claim, finding that\\n\\ncounsel reasonably chose to pursue a penalty-phase strategy of residual doubt and\\n\\ngood character. Id. at 147. It explained that the new evidence was self-\\n\\ncontradictory in that it asserted that childhood abuse would have made Jenkins a\\n\\nviolent adult yet also asserted that Jenkins was meek and mild. Id. The appeals\\n\\ncourt also agreed with the lower court’s conclusion that Jenkins suffered no\\n\\nprejudice from any error by his counsel because the aggravating circumstances still\\n\\noutweighed any new mitigating ones. Id. at 148. The court also noted that evidence\\n\\n\\n\\n\\n 18\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 19 of 73\\n\\n\\nabout Jenkins’s good behavior in jail awaiting trial was “minimally mitigating” and\\n\\nwould not have affected the sentence. Id. at 149.\\n\\n Because the Supreme Court’s decision in Atkins v. Virginia6 had issued\\n\\nduring the pendency of the Rule 32 appeal, the Alabama appeals court ordered\\n\\nsupplemental briefing on the possible impact of Atkins upon Jenkins’s sentence.\\n\\nJenkins asked the court to stay the proceedings and conduct an evidentiary hearing\\n\\nafter the Alabama legislature enacted a statute about executing the mentally\\n\\nretarded or, in the alternative, to vacate his sentence and remand after the Alabama\\n\\nlegislature acted. He later asserted that he is mentally retarded and therefore\\n\\nineligible for the death penalty. The appeals court proceeded to reject Jenkins’s\\n\\nAtkins claim on the merits because the record established that his IQ was 76 and\\n\\nthat he maintained relationships and employment. Id. at 155. The Alabama\\n\\nSupreme Court summarily affirmed, in relevant part. Ex parte Jenkins, 972 So. 2d\\n\\n159, 165 (Ala. 2005), remanded to 972 So. 2d 165 (Ala. Ct. Crim. App. 2005)\\n\\n(juror misconduct claim), cert. denied, 552 U.S. 1167 (2008).\\n\\n In 2008, Jenkins filed his first petition for a writ of habeas corpus in federal\\n\\ncourt pursuant to 28 U.S.C. § 2254. The federal proceedings were stayed for\\n\\nseveral years while Jenkins pursued additional, unrelated Rule 32 relief in the state\\n\\n\\n6\\n 536 U.S. 304, 321 (2002) (holding that the execution of mentally retarded offenders violates the\\nEighth Amendment). The rule announced in Atkins is retroactive to cases on collateral review. In\\nre Holladay, 331 F.3d 1169, 1172–73 (11th Cir. 2003).\\n 19\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 20 of 73\\n\\n\\ncourts, ultimately unsuccessfully. In the district court, Jenkins moved for an\\n\\nevidentiary hearing on his Atkins claim. After reviewing the state court record,\\n\\nwith a focus on the Rule 32 testimony of Drs. Lisak and Kirkland, the district court\\n\\ndenied the motion in 2015. Noting that our Circuit has accepted the State of\\n\\nAlabama’s definition of mental retardation following Atkins, the court found that\\n\\nJenkins is not intellectually disabled under that standard. In a 347-page order, the\\n\\ndistrict court also denied relief on all of Jenkins’s claims including ineffective\\n\\nassistance of counsel during the penalty phase and affirmed as reasonable the state\\n\\ncourt’s conclusion that Jenkins failed to establish prejudice from any error by his\\n\\ncounsel. The district court later denied reconsideration and a certificate of\\n\\nappealability.\\n\\n Jenkins now appeals. Our Court granted Jenkins a certificate of\\n\\nappealability, 28 U.S.C. § 2253(c), on two issues:\\n\\n 1) Whether the district court erred in denying Appellant’s claim\\n that his trial counsel rendered ineffective assistance of counsel in the\\n investigation and presentation of mitigating evidence during the\\n penalty phase of Appellant’s 1991 trial; and\\n\\n 2) Whether the district court erred in denying Appellant’s claim\\n that he is intellectually disabled and ineligible for the death penalty\\n under Atkins v. Virginia, 536 U.S. 304 (2002).\\n\\n II. STANDARDS OF REVIEW\\n\\n We review the denial of a petition for a writ of habeas corpus de novo.\\n\\nMorrow v. Warden, Ga. Diagnostic Prison, 886 F.3d 1138, 1146 (11th Cir. 2018).\\n\\n 20\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 21 of 73\\n\\n\\nThe Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) instructs that\\n\\na writ of habeas corpus shall not be granted to a state prisoner unless the state\\n\\nadjudication of a claim “resulted in a decision that was contrary to, or involved an\\n\\nunreasonable application of, clearly established Federal law, as determined by the\\n\\nSupreme Court of the United States.” 28 U.S.C. § 2254(d)(1). “[C]learly\\n\\nestablished Federal law” means “the holdings, as opposed to the dicta, of [the\\n\\nSupreme] Court’s decisions as of the time of the relevant state-court decision.”\\n\\nWilliams v. Taylor, 529 U.S. 362, 412 (2000). “[W]hen the last state court to\\n\\ndecide a prisoner’s federal claim explains its decision on the merits in a reasoned\\n\\nopinion . . . a federal habeas court simply reviews the specific reasons given by the\\n\\nstate court and defers to those reasons if they are reasonable.” Wilson v. Sellers,\\n\\n138 S. Ct. 1188, 1192 (2018). A state court’s decision is reasonable “so long as\\n\\n‘fairminded jurists could disagree’ on the correctness of the state court’s decision.”\\n\\nHarrington v. Richter, 562 U.S. 86, 101 (2011) (quoting Yarborough v. Alvarado,\\n\\n541 U.S. 652, 664 (2004)). “[E]ven a strong case for relief does not mean the state\\n\\ncourt’s contrary conclusion was unreasonable.” Id. at 102. Rather, a prisoner must\\n\\nshow that the state court’s ruling “was so lacking in justification that there was an\\n\\nerror well understood and comprehended in existing law beyond any possibility for\\n\\nfairminded disagreement.” Id. at 103. That ruling must have been “‘objectively\\n\\nunreasonable,’ not merely wrong; even ‘clear error’ will not suffice.” White v.\\n\\n\\n 21\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 22 of 73\\n\\n\\nWoodall, 572 U.S. 415, 419 (2014) (quoting Lockyer v. Andrade, 538 U.S. 63, 75\\n\\n(2003)).\\n\\n In addition, “a determination of a factual issue made by a State court shall be\\n\\npresumed to be correct,” and the prisoner bears “the burden of rebutting the\\n\\npresumption of correctness by clear and convincing evidence.” 28 U.S.C.\\n\\n§ 2254(e)(1). AEDPA further instructs that a writ of habeas corpus may not be\\n\\ngranted unless the state court adjudication “resulted in a decision that was based on\\n\\nan unreasonable determination of the facts in light of the evidence presented in the\\n\\nState court proceeding.” Id. § 2254(d)(2). When a federal habeas court makes a\\n\\nfactual finding as part of its habeas determination, we review for clear error, which\\n\\nis a “highly deferential” standard of review. Thomas v. Allen, 607 F.3d 749, 752\\n\\n(11th Cir. 2010) (quoting Holton v. City of Thomasville Sch. Dist., 425 F.3d 1325,\\n\\n1350 (11th Cir. 2005)). “A finding is ‘clearly erroneous’ when although there is\\n\\nevidence to support it, the reviewing court on the entire evidence is left with the\\n\\ndefinite and firm conviction that a mistake has been committed.” United States v.\\n\\nU.S. Gypsum Co., 333 U.S. 364, 395 (1948).\\n\\n III. DISCUSSION\\n\\n A. Ineffective Assistance of Counsel\\n\\n A prisoner alleging that he received ineffective assistance of counsel in\\n\\nviolation of the Sixth Amendment must establish two elements. Strickland v.\\n\\n\\n 22\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 23 of 73\\n\\n\\nWashington, 466 U.S. 668, 687 (1984). “First, the defendant must show that\\n\\ncounsel’s performance was deficient.” Id. In other words, he must show “that\\n\\ncounsel made errors so serious that counsel was not functioning as the ‘counsel’\\n\\nguaranteed the defendant by the Sixth Amendment.” Id. “Second, the defendant\\n\\nmust show that the deficient performance prejudiced the defense.” Id. We discuss\\n\\neach of these elements in turn.\\n\\n 1. Deficient Performance\\n\\n A “doubly” deferential lens sits before our eyes as we review the\\n\\nperformance of Jenkins’s counsel during the penalty phase. Richter, 562 U.S. at\\n\\n105 (quoting Knowles v. Mirzayance, 556 U.S. 111, 123 (2009)). This redoubled\\n\\ndeference arises out of the tandem application of AEDPA’s “highly deferential”\\n\\nreview of a state court’s decision with Strickland’s “highly deferential” review of\\n\\nan attorney’s performance. Id. (quoting Lindh v. Murphy, 521 U.S. 320, 333 n.7\\n\\n(1997), and Strickland, 466 U.S. at 689). Under the Strickland standard, “counsel\\n\\nis strongly presumed to have rendered adequate assistance and made all significant\\n\\ndecisions in the exercise of reasonable professional judgment.” Strickland, 466\\n\\nU.S. at 690. And under AEPDA, we must deny habeas relief if “there is any\\n\\nreasonable argument that counsel satisfied Strickland’s deferential standard.”\\n\\nRichter, 562 U.S. at 105.\\n\\n\\n\\n\\n 23\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 24 of 73\\n\\n\\n Jenkins argues that his counsel performed deficiently by failing to\\n\\n“investigate compelling mitigating evidence or develop a penalty-phase theory.”\\n\\nThe state court disagreed. It quoted decisions of our Court that hold that “there is\\n\\nno per se rule that evidence of a criminal defendant’s troubled childhood must\\n\\nalways be presented as mitigating evidence.” Jenkins, 972 So. 2d at 144 (quoting\\n\\nMarek v. Singletary, 62 F.3d 1295, 1300 (11th Cir. 1995)). Noting that “Downey\\n\\nwas in charge of the penalty phase” but “we do not have the benefit of Downey’s\\n\\ntestimony as to what occurred and why,” it reviewed the trial record to find that\\n\\n“counsel argued residual doubt and Jenkins’s good character at the penalty phase.”\\n\\nId. at 145–46. It concluded that “Downey’s decision to concentrate on reasonable\\n\\ndoubt and to portray Jenkins as a good person was reasonable under the\\n\\ncircumstances.” Id. at 147.\\n\\n Jenkins insists, however, that counsel’s failure to develop mitigating\\n\\nevidence about his childhood abuse was “not the product of strategy, but rather\\n\\narose from ingnorance and inexeperiance [sic].” The problem for Jenkins is that,\\n\\nlike the state courts, this Court also does not know what Downey did or why. The\\n\\nrecord is silent as to his thoughts and intentions as he prepared for the penalty\\n\\nphase. Given that Jenkins bore the burden of overcoming the strong presumption of\\n\\ncompetence and of proving Downey’s deficient performance, we are perplexed\\n\\nwhy Jenkins’s new counsel did not have Downey testify or submit an affidavit in\\n\\n\\n 24\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 25 of 73\\n\\n\\nthe Rule 32 proceeding. See Strickland, 466 U.S. at 687 (“The defendant must\\n\\nshow that counsel’s performance was deficient.”). Nor did Jenkins testify to the\\n\\nsubstance of his conversations with his counsel. “The reasonableness of a trial\\n\\ncounsel’s acts, including lack of investigation or excluding character witnesses\\n\\nfrom the sentencing phase, depends ‘critically’ upon what information the client\\n\\ncommunicated to counsel.” Chandler v. United States, 218 F.3d 1305, 1324 (11th\\n\\nCir. 2000) (en banc) (quoting Strickland, 466 U.S. at 691). There is simply no\\n\\nrecord evidence to support Jenkins’s central assertion that Downey made no\\n\\ntactical choices in his preparation for the penalty phase.\\n\\n Jenkins instead relies on Scofield’s Rule 32 testimony about how he and\\n\\nDowney prepared for the trial and sentencing. We acknowledge that Scofield\\n\\nrepeatedly testified that he did not really know what he was doing and that he\\n\\nfailed to make strategic choices with respect to mitigating circumstances and\\n\\nsentencing. But Scofield was equally clear that Downey—the more experienced\\n\\nlawyer—was in fact responsible for the penalty phase. And Scofield also\\n\\nrepeatedly testified that he did not know what Downey may or may not have done\\n\\nto prepare. Furthermore, his lack of knowledge about Downey’s preparation was\\n\\nunderstandable. He and Downey were not colleagues who shared an office and\\n\\nmight reasonably be expected to know what the other was doing much of the time.\\n\\nScofield and Downey did not even work in the same city.\\n\\n\\n 25\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 26 of 73\\n\\n\\n This case is unlike those in which we have found deficient performance\\n\\nwhere the record showed that counsel investigated insufficiently because he\\n\\n“waited until the eleventh hour” to begin preparing for the penalty phase while\\n\\nfaced with “overwhelming evidence of guilt.” See, e.g., Johnson v. Sec’y, D.O.C.,\\n\\n643 F.3d 907, 932 (11th Cir. 2011). The limited record that Jenkins developed\\n\\nestablishes that Downey spent at least 71 hours preparing for trial, spread out over\\n\\n17 months. The state court interpreted this sparse record as reflecting a penalty-\\n\\nphase strategy of residual doubt, the logical extension of Downey’s guilt-phase\\n\\ninvestigation of the evidence and Scofield’s pursuit at trial of an acquittal based on\\n\\nreasonable doubt. See Jenkins, 972 So. 2d at 147. In such a situation, “[a] lawyer’s\\n\\ntime and effort in preparing to defend his client in the guilt phase of a capital case\\n\\ncontinues to count at the sentencing phase.” Tarver v. Hopper, 169 F.3d 710, 715\\n\\n(11th Cir. 1999).\\n\\n We decline to dictate how many hours or how many sessions an attorney\\n\\nmust spend interviewing his counsel in order to provide effective assistance. “No\\n\\nabsolute rules dictate what is reasonable performance for lawyers.” Chandler, 218\\n\\nF.3d at 1317 (citing Strickland, 466 U.S. at 688–89). But even this limited record\\n\\nundermines Jenkins’s assertion that Downey prepared inadequately. Downey’s 71\\n\\nhours of preparation included 35 hours of jailhouse conversations with Jenkins\\n\\n\\n\\n\\n 26\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 27 of 73\\n\\n\\nduring 18 separate meetings.7 According to Downey’s notes, Scofield was present\\n\\nfor only five of those meetings (12 hours’ worth), leaving 23 hours over 13\\n\\nmeetings between Jenkins and Downey alone. Downey’s notes reveal little about\\n\\nthe content of those conversations beyond broad notations such as “evidence,\\n\\nwitnesses, general status,” and discussions with Jenkins of the evidence and\\n\\ntestimony against him were surely relevant to Downey’s penalty-phase strategy of\\n\\nresidual doubt.\\n\\n But the record does not reveal that Downey knew or should have known that\\n\\nJenkins’s childhood merited investigation beyond that which he may have done.\\n\\nThe record shows that Downey developed enough of a rapport with Jenkins that,\\n\\naccording to his fee declaration, he “g[ave a] Christmas gift [and] discuss[ed] how\\n\\nmuch to tell his family.” But how much Downey knew about Jenkins’s family\\n\\nbackground can only be guessed from Scofield’s Rule 32 hearing testimony.\\n\\nScofield testified that he discussed Jenkins’s background during his dozen or so\\n\\njailhouse meetings with him, and that Downey was present for some of those\\n\\nmeetings. Scofield—and possibly Downey—learned that Jenkins experienced\\n\\nabuse as a child, that his stepfather was brutal, and that he lived on the streets and\\n\\n\\n\\n\\n7\\n We note that the state court clearly erroneously stated that Downey’s fee declaration\\ndocumented 171 hours of pretrial preparation. See Jenkins, 972 So. 2d at 137. However, the state\\ncourt accurately stated that “he spent over 25 hours talking with Jenkins in more than 10 visits to\\nthe jail.” Id. That statement was not an unreasonable determination of the facts.\\n 27\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 28 of 73\\n\\n\\nin juvenile institutions because of his brutal home life. It is against this backdrop\\n\\nthat Downey’s presumptive “decision not to investigate must be directly assessed\\n\\nfor reasonableness in all the circumstances, applying a heavy measure of deference\\n\\nto counsel’s judgments.” Strickland, 466 U.S. at 691.\\n\\n That assessment depends a great deal on what Jenkins told his counsel about\\n\\nhis background. See id. But we do not know whether Downey was present when\\n\\nScofield learned about Jenkins’s childhood abuse, nor what else Downey may have\\n\\nlearned during his 13 solo discussions with Jenkins. Therefore, “given the absence\\n\\nof evidence in the record, we must assume counsel carried out his professional\\n\\nresponsibility and discussed mitigation with his client.” Chandler, 218 F.3d at\\n\\n1324. We will not “turn that presumption on its head by giving [Jenkins] the\\n\\nbenefit of the doubt where it is unclear what [Downey] did or did not do.” Williams\\n\\nv. Head, 185 F.3d 1223, 1235 (11th Cir. 1999). We will not assume that Jenkins\\n\\ntold Downey more than he told Scofield, and Scofield denied remembering that\\n\\nJenkins told him about the severity or frequency of his stepfather’s physical abuse.\\n\\nHe denied remembering Jenkins telling him about being beaten daily, being locked\\n\\nin his room 24/7, or being forced to eat his feces. And there is no suggestion in the\\n\\nrecord that counsel knew or had reason to investigate the allegation of sexual abuse\\n\\nby his stepgrandfather. The record simply does not reflect that Jenkins told either\\n\\nof his lawyers about the severe abuse that some of the Rule 32 witnesses later\\n\\n\\n 28\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 29 of 73\\n\\n\\ndescribed. We will not fault those lawyers for failing to present as mitigation this\\n\\nlimited evidence that they may have reasonably believed to be unremarkable. See\\n\\nStrickland, 466 U.S. at 691.\\n\\n To be sure, counsel’s failure “to discover all reasonably available mitigating\\n\\nevidence” may constitute deficient performance in some cases. Wiggins v. Smith,\\n\\n539 U.S. 510, 524 (2003) (quoting ABA Guidelines for the Appointment and\\n\\nPerformance of Counsel in Death Penalty Cases § 11.4.1(C) (1989)). Our Court\\n\\nhas identified deficient performance in several instances where counsel had good\\n\\nreason to investigate the defendant’s background but failed to do so. See Morrow,\\n\\n886 F.3d at 1147 (collecting cases). But we assess “a particular decision not to\\n\\ninvestigate . . . for reasonableness in all the circumstances, applying a heavy\\n\\nmeasure of deference to counsel’s judgments.” Strickland, 466 U.S. at 691. Thus,\\n\\nwe have rejected a hard rule that counsel’s “strategic decisions can be considered\\n\\nreasonable only if they are preceded by a ‘thorough investigation.’” Williams, 185\\n\\nF.3d at 1237 (quoting Rogers v. Zant, 13 F.3d 384, 387 (11th Cir. 1994)). Because\\n\\nwe simply do not know why Downey chose not to investigate Jenkins’s childhood\\n\\nmore thoroughly, Downey is entitled to the presumption that this strategic decision\\n\\nwas reasonable.\\n\\n And though we have also rejected “a per se rule of ineffective assistance\\n\\nwhere counsel does not consult family members,” Holladay v. Haley, 209 F.3d\\n\\n\\n 29\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 30 of 73\\n\\n\\n1243, 1252 (11th Cir. 2000), we note that Downey did not fail altogether to consult\\n\\nfamily members. Downey had at least one substantive conversation with Jenkins’s\\n\\ngrandmother, and we know from her Rule 32 hearing testimony that she never\\n\\nwitnessed Jenkins’s childhood abuse. Given the silence of the record and the strong\\n\\npresumption of competence to which Downey is entitled, we decline to fault him\\n\\nfor not doing more when Jenkins and his grandmother did not and could not,\\n\\nrespectively, provide a basis for doing so. Cf. United States v. Pease, 240 F.3d 938,\\n\\n941–92 (11th Cir. 2001) (“we cannot say as a matter of law that reliance on a\\n\\nclient’s statements is per se deficient performance”). Nor did Jenkins’s counsel\\n\\nhave an absolute duty to put on abusive-childhood evidence, even if they knew of\\n\\nit. See Chandler, 218 F.3d at 1319 (“Counsel is not required to present every\\n\\nnonfrivolous defense; nor is counsel required to present all mitigation evidence,\\n\\neven if the additional mitigation evidence would not have been incompatible with\\n\\ncounsel’s strategy.”); Waters v. Thomas, 46 F.3d 1506, 1511 (11th Cir. 1995) (en\\n\\nbanc) (“Our decisions are inconsistent with any notion that counsel must present\\n\\nall available mitigating circumstance evidence.”).\\n\\n Instead, we affirm as a reasonable application of Strickland the state court’s\\n\\nconclusion that Downey’s strategic decision to emphasize residual doubt was\\n\\nreasonable. See Jenkins, 972 So. 2d at 147. As we have noted, “[a] lawyer’s time\\n\\nand effort in preparing to defend his client in the guilt phase of a capital case\\n\\n\\n 30\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 31 of 73\\n\\n\\ncontinues to count at the sentencing phase.” Tarver, 169 F.3d at 715. This\\n\\nreasoning is especially applicable where, as here, counsel pursued the reasonable\\n\\nstrategy of arguing for acquittal at the guilt phase and for life at the sentencing\\n\\nphase by highlighting doubt about the evidence. “[R]esidual doubt is perhaps the\\n\\nmost effective strategy to employ at sentencing.” Chandler, 218 F.3d at 1320 n.28.\\n\\n“[E]specially when—as in this case—the evidence of guilt was not overwhelming,\\n\\nwe expect that petitioners can rarely (if ever) prove a lawyer to be ineffective for\\n\\nrelying on this seemingly reasonable strategy to defend his client.” Id. at 1320\\n\\n(footnote omitted); see also Tarver, 169 F.3d at 715–16 (collecting law review\\n\\narticles on the merits of residual-doubt strategy). Not only can such a strategy be\\n\\nconstitutionally adequate, in some cases it may be “the most effective performance\\n\\nin defense to the death penalty.” Tarver, 169 F.3d at 716.\\n\\n We cannot say that Downey’s manifest strategy of residual doubt was\\n\\nobjectively unreasonable under the circumstances. Jenkins’s case is precisely the\\n\\nsort where reminding the jury of the misgivings it may have held about the State’s\\n\\ncase could have been effective. Scofield’s guilt-phase closing argument from that\\n\\nvery morning, which explicitly emphasized doubt and the weakness of the State’s\\n\\nevidence, would have been fresh in the minds of the jury.8 Although we are not\\n\\n\\n8\\n During the penalty phase of a capital case, “[e]vidence presented at the trial of the case may be\\nconsidered insofar as it is relevant to the aggravating and mitigating circumstances without the\\nnecessity of re-introducing that evidence at the sentence hearing.” Ala. Code § 13A-5-45(c).\\n 31\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 32 of 73\\n\\n\\nquestioning the jury’s guilty verdict or the sufficiency of the evidence, which are\\n\\nbeyond the scope of our habeas review, we acknowledge that the State’s case\\n\\nagainst Jenkins was—as Downey reminded the jury during the penalty phase—\\n\\nentirely circumstantial. Indeed, Scofield maintained post-conviction that the case\\n\\nfor reasonable doubt was “[v]ery strong.” No one saw Jenkins commit either the\\n\\nmurder or the robbery and kidnapping that elevated Hogeland’s murder to a capital\\n\\ncrime. Scofield’s guilt-phase closing argument challenged the inferential leaps the\\n\\nState was asking the jury to make in finding that Jenkins committed these capital-\\n\\npredicate offenses and that he formed the intent to kill. This situation was not one\\n\\nwhere a guilt-phase strategy of acquittal—and therefore a penalty-phase strategy of\\n\\nresidual doubt—was unreasonable in light of the evidence. Cf. Johnson, 643 F.3d\\n\\nat 932–33 (deficient sentencing investigation where defendant had confessed to\\n\\ntwo murders).\\n\\n Thus, although Downey’s penalty-phase closing argument did not\\n\\nspecifically invoke lingering doubt, it clearly reminded the jury of the entirely\\n\\ncircumstantial nature of the State’s guilt-phase case. See Chandler, 218 F.3d at\\n\\n1320 & n.6 (a strategy of “focus[ing] on obtaining an acquittal and then, at\\n\\nsentencing, on lingering doubt” is reasonable, even if counsel does not use the\\n\\nwords “lingering doubt” or “residual doubt”). Downey closed this reminder by\\n\\nexplaining to the jury how a foundational ancient legal system proscribed capital\\n\\n\\n 32\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 33 of 73\\n\\n\\npunishment in the absence of eyewitness testimony—the exact situation here. The\\n\\nstate court found that Downey’s penalty-phase closing argument reflected a\\n\\nstrategic “decision to concentrate on reasonable doubt,” see Jenkins, 972 So. 2d at\\n\\n147, a tactical choice entitled to a “strong presumption of correctness,” Marek, 62\\n\\nF.3d at 1300 & n.3.\\n\\n Jenkins suggests other theories of mitigation that Downey might have more\\n\\neffectively pursued,9 but our deferential review under Strickland does not ask\\n\\nwhether counsel could possibly or ideally have been more effective. “The test for\\n\\nineffectiveness is not whether counsel could have done more.” Waters, 46 F.3d at\\n\\n1518. We do not ask whether an attorney’s representation “deviated from best\\n\\npractices or common custom,” and we should resist the temptation to second-guess\\n\\nan attorney with the benefit of our hindsight. Richter, 562 U.S. at 105. We ask only\\n\\n“whether, in light of all the circumstances, the identified acts or omissions were\\n\\n\\n\\n\\n9\\n In particular, Jenkins argues that his counsel should have presented “model prisoner” evidence.\\nThe state court noted, “Good conduct during pretrial incarceration is not necessarily a mitigating\\ncircumstance,” before finding that Jenkins had not met his burden of establishing that his counsel\\neven had information about his conduct in jail. Jenkins, 972 So. 2d at 149. We agree, and we\\ndecline to impose a requirement of sua sponte investigation of this kind in every capital case. Cf.\\nWilliams, 529 U.S. at 396 (finding deficient performance based in part on not returning a call\\nfrom a prison ministry volunteer who offered to vouch for defendant’s good behavior in prison).\\n“Surely, counsel is not required to call a witness to testify to facts such as lack of violent nature\\nwhen the jury has rejected such an approach and has found that the defendant is guilty of\\nmurder.” Griffin v. Wainwright, 760 F.2d 1505, 1512 (11th Cir. 1985), vacated on other grounds,\\n476 U.S. 1112 (1986), reaff’d, 874 F.2d 1397 (11th Cir. 1989).\\n 33\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 34 of 73\\n\\n\\noutside the range of professionally competent assistance.” Strickland, 466 U.S. at\\n\\n690.\\n\\n We conclude that the state court did not unreasonably determine that Jenkins\\n\\nfailed to establish objectively incompetent performance by his counsel during the\\n\\npenalty phase of his trial. Our inquiry is not “what the best lawyers would have\\n\\ndone. Nor is the test even what most good lawyers would have done. We ask only\\n\\nwhether some reasonable lawyer at the trial could have acted, in the circumstances,\\n\\nas defense counsel acted at trial. . . . We are not interested in grading lawyers’\\n\\nperformances.” White v. Singletary, 972 F.2d 1218, 1220–21 (11th Cir. 1992). And\\n\\n“[i]f fairminded jurists could reasonably disagree” about the reasonableness of\\n\\ncounsel’s performance, “then habeas relief is due to be denied.” Johnson, 643 F.3d\\n\\nat 932.\\n\\n 2. Prejudice\\n\\n Even if Jenkins had demonstrated that his counsel performed as no\\n\\nreasonable lawyer could have, he must also demonstrate a “reasonable probability\\n\\nthat, but for counsel’s unprofessional errors, the result of the proceeding would\\n\\nhave been different.” Strickland, 466 U.S. at 694. Given that the jury here\\n\\nrecommended a sentence of death by the narrowest possible vote, 10 to 2, Jenkins\\n\\nneed establish only “a reasonable probability that at least one juror would have\\n\\nstruck a different balance” between life and death. Wiggins, 539 U.S. at 537. “A\\n\\n\\n 34\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 35 of 73\\n\\n\\nreasonable probability is a probability sufficient to undermine confidence in the\\n\\noutcome.” Strickland, 466 U.S. at 694. In measuring that probability here, we will\\n\\n“evaluate the totality of the available mitigation evidence—both that adduced at\\n\\ntrial, and the evidence adduced in the habeas proceeding in reweighing it against\\n\\nthe evidence in aggravation.” Williams, 529 U.S. at 397–98.\\n\\n We begin our analysis by clarifying what exactly Jenkins is arguing about\\n\\nprejudice. We do not know precisely which aggravating and mitigating factors the\\n\\njury may have implicitly found before making its recommendation of death. The\\n\\ntrial court found at sentencing two statutory aggravating factors: the robbery and\\n\\nthe kidnapping of the victim. See Ala. Code § 13A-5-49(4). Because none of the\\n\\nmitigation evidence offered in the Rule 32 hearing bears on those factors, we see\\n\\nno reasonable probability that the aggravating factors would have been determined\\n\\ndifferently by the jury or the court. The court also found at sentencing two\\n\\nstatutory mitigating factors: no significant history of prior criminal activity and\\n\\nage. See id. § 13A-5-51(1), (7). There was also strong evidence of Jenkins’s\\n\\nintoxication, which was a key part of Jenkins’s guilt-phase defense and could have\\n\\nsupported a third mitigating factor, impaired capacity. See id. § 13A-5-51(6).10\\n\\n\\n10\\n The jury had heard evidence about all three of these statutory mitigators during the guilt phase\\nof the trial. Downey’s penalty-phase opening statement reviewed for the jury all seven of the\\npossible statutory mitigating factors, and the court also recited them in its jury instructions,\\nreminding the jury that they could consider guilt-phase evidence as well as non-statutory\\nmitigation. Jenkins has not established that there is a substantial probability that the jury’s\\nrecommendation would have been different had it been more explicitly reminded of Jenkins’s\\n 35\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 36 of 73\\n\\n\\nJenkins does not argue that the jury or the court would have found other statutory\\n\\nmitigators in light of the Rule 32 evidence. Rather, he contends that the jury might\\n\\nhave found non-statutory mitigation based on Jenkins’s abusive childhood, low\\n\\nintelligence, or mental health problems that would have outweighed the\\n\\naggravating factors.11\\n\\n The Rule 32 trial court found that Jenkins had not demonstrated prejudice\\n\\neven assuming that counsel had performed deficiently. See Jenkins, 972 So. 2d at\\n\\n138. It concluded that he had not shown a reasonable probability that the sentencer\\n\\n“would have concluded that a weighing of the aggravating and mitigating\\n\\ncircumstances did not warrant death” upon a reweighing of the evidence. See id. It\\n\\nemphasized the depravity of the crime and the aggravating circumstances and\\n\\nfound that the mitigating witnesses were not credible for various reasons. See id.\\n\\nThe state appellate court affirmed, declining to disturb the credibility\\n\\ndeterminations of the trial court. Id. at 142–44. Independently reweighing the\\n\\n\\nage, criminal history, and intoxication, given that the trial court, the Rule 32 trial court, and the\\nstate appellate court all explicitly considered those mitigators and found them outweighed by the\\naggravators.\\n11\\n Alabama law lists seven statutory mitigating circumstances, but states that mitigation “shall\\n. . . not be limited to” those seven circumstances. Ala. Code § 13A-5-51. Further, “mitigating\\ncircumstances shall include any aspect of a defendant’s character or record and any of the\\ncircumstances of the offense that the defendant offers as a basis for a sentence of life\\nimprisonment without parole instead of death, and any other relevant mitigating circumstances\\nwhich the defendant offers as a basis for a sentence of life imprisonment without parole instead\\nof death.” Id. § 13A-5-52. A jury may recommend a sentence of death only if it finds that “one or\\nmore aggravating circumstances . . . exist and that they outweigh the mitigating circumstances, if\\nany.” Id. § 13A-5-46(e)(3).\\n 36\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 37 of 73\\n\\n\\nevidence, it concluded that, “[g]iven the aggravating circumstances that were\\n\\nproven by the State and the facts surrounding Hogeland’s murder, we, like the\\n\\ncircuit court, are confident that death was the appropriate punishment for Jenkins’s\\n\\nactions.” Id. at 148.\\n\\n Our review of this evidence does not show that it is likely that the jury\\n\\nwould have concluded that death was not the appropriate punishment if it had\\n\\nheard the Rule 32 evidence. Much of the proffered mitigation evidence is, in our\\n\\nview, much less mitigating than Jenkins asserts. In particular, the testimony of\\n\\nJenkins’s older brother Michael was likely to have been a double-edged sword. See\\n\\nEvans v. Sec’y, Dep’t of Corr., 703 F.3d 1316, 1327 (11th Cir. 2013) (en banc) (“it\\n\\nis reasonable to conclude that a defendant was not prejudiced when his mitigation\\n\\nevidence was a two-edged sword” (quoting Ponticelli v. Sec’y, Fla. Dep’t of Corr.,\\n\\n690 F.2d 1271, 1296 (11th Cir. 2012)). Michael was the closest relative to testify at\\n\\nthe Rule 32 hearing and his testimony about the abuse Jenkins experienced should\\n\\nhave been the most reliable, since he experienced much of the same physical abuse\\n\\nhimself. The brothers shared the same filthy bedroom, received beatings from the\\n\\nsame paddle, and were frequently absent from school together. But Michael’s\\n\\ntestimony contained some discrepancies, and he admitted to cognitive and memory\\n\\nproblems. Perhaps Michael’s testimony would have made Jenkins appear more\\n\\nsympathetic in the eyes of the jury—but perhaps not. While Jenkins ran away from\\n\\n\\n 37\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 38 of 73\\n\\n\\nhome and eventually committed murder, Michael responded to the same\\n\\nbackground by running away from home, graduating from high school, and\\n\\nproviding health care for both the terminally ill and people with substance use\\n\\ndisorders. It is possible that the jury could even have found Jenkins to be more\\n\\nculpable for not overcoming his difficult upbringing and to pose more of a danger\\n\\nof future violence.\\n\\n We also find reasonable the state court’s identification of other problems\\n\\nwith the mitigating evidence Jenkins argues should have been presented. Hearing\\n\\nthat evidence firsthand in the Rule 32 proceeding, the trial court found the\\n\\ntestimony of the cousins less credible in light of the lack of physical abuse\\n\\ndocumented in the school health records. See Jenkins, 972 So. 2d at 139 (noting\\n\\nthat school records observed a rash and gingivitis). It faulted Pitts in particular for\\n\\npurporting to have witnessed ongoing, severe abuse but taking only minimal,\\n\\nineffective steps to obtain outside help. See id. It further noted that DeLavega and\\n\\nWagoner actually testified that they never saw Jenkins beaten, and it noted all of\\n\\nthe witnesses’ interest in getting Jenkins off death row. All of these factors would\\n\\nlikely have caused a reasonable jury to discount their testimony. See id. at 139–42.\\n\\nWe must accept the trial court’s views of the Rule 32 testimony. “In the absence of\\n\\nclear and convincing evidence, we have no power on federal habeas review to\\n\\nrevisit the state court’s credibility determinations.” Bishop v. Warden, 726 F.3d\\n\\n\\n 38\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 39 of 73\\n\\n\\n1243, 1259 (11th Cir. 2013) (citing Marshall v. Lonberger, 459 U.S. 422, 434\\n\\n(1983)).\\n\\n The state court also noted that Wagoner testified that she was not available\\n\\nto testify at the trial, and it found that Seal’s testimony would have been\\n\\ncumulative to that of her husband. Thus, counsel’s failure to call those two\\n\\nwitnesses could not have been prejudicial to Jenkins. See id. at 141, 142. Neither of\\n\\nthese conclusions involves an unreasonable determination of the facts. We agree\\n\\nthat the Rule 32 testimony of Sherry Seal was substantially similar to the testimony\\n\\nof her husband Lonnie Seal that the jury heard during the penalty phase. Both\\n\\ntestified about meeting Jenkins in California, Jenkins’s assistance with their move\\n\\nto Alabama, and his residence with them in their Alabama home. Sherry added\\n\\nonly that Jenkins sometimes wet his bed. It was not unreasonable for the state court\\n\\nto conclude that Sherry’s testimony was cumulative of her husband’s, and “a\\n\\npetitioner cannot satisfy the prejudice prong of the Strickland test with evidence\\n\\nthat is merely cumulative of evidence already presented.” Rose v. McNeil, 634 F.3d\\n\\n1224, 1243 (11th Cir. 2011).\\n\\n Further, the school, medical, and juvenile court records that Jenkins\\n\\nintroduced in the Rule 32 proceeding have their own problems. As noted, they do\\n\\nnot document or even corroborate the severe level of abuse alleged by some of the\\n\\nwitnesses. Though these documents tell a sad story, they do not create a substantial\\n\\n\\n 39\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 40 of 73\\n\\n\\nlikelihood that the jury would have recommended life. Although they portray\\n\\nJenkins as a victim of truly unfortunate circumstances, they also show his unsavory\\n\\ndelinquent behavior. That delinquency included a history of stealing cars, and\\n\\ngiven that the jury likely inferred that Jenkins stole the red Mazda, new\\n\\ninformation about his history of theft may not have been especially mitigating.\\n\\nSimilarly, we agree with the state court that evidence of Jenkins’s compliant\\n\\nconduct in pretrial detention is “minimally mitigating” where there had been no\\n\\nevidence of jail misconduct as an aggravator. See Jenkins, 972 So. 2d at 149; cf.\\n\\nSkipper v. South Carolina, 476 U.S. 1, 5 & n.1 (1986) (good jail behavior is\\n\\n“potentially mitigating” especially “[w]here the prosecution specifically relies on a\\n\\nprediction of future dangerousness in asking for the death penalty”).\\n\\n Jenkins argues that allowing the jury to see his experience of mental,\\n\\nemotional, and cognitive difficulties, compounded by years of abuse and neglect\\n\\nand considered alongside his mild character and good jail behavior as an adult,\\n\\nwould have convinced them that Hogeland’s murder was an “aberration” that\\n\\ndeserved a punishment less severe than death. That interpretation of Jenkins’s life\\n\\nstory may be conceivable. But more than mere conceivability is required to\\n\\nestablish prejudice: “The likelihood of a different result must be substantial, not\\n\\njust conceivable.” Richter, 562 U.S. at 112 (emphasis added). We do not find the\\n\\nstate court’s decision that a different result was not substantially likely to be an\\n\\n\\n 40\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 41 of 73\\n\\n\\nunreasonable application of Strickland. As we have noted, we think it equally\\n\\nlikely that the jury could have rejected this alternate narrative out of hand in light\\n\\nof having just convicted him of the violent murder of a young mother. We also\\n\\nthink it possible that the jury could have viewed the murder as just one step in a\\n\\nlogical progression of degeneracy that began with behavioral problems at school,\\n\\ncontinued through the juvenile court system, and escalated past grand theft auto\\n\\ninto kidnapping and murder. Taken separately as well as together, the proffered\\n\\nevidence does not compel us to conclude that a verdict of life would have been\\n\\nsubstantially likely. Because fairminded jurists could disagree about the state\\n\\ncourt’s conclusion about Strickland prejudice, we affirm the denial of habeas relief\\n\\non this ground.\\n\\n B. Intellectual Disability\\n\\n Jenkins argues that the district court erred when it denied his claim that he is\\n\\nintellectually disabled and ineligible for the death penalty under Atkins v. Virginia,\\n\\n536 U.S. 304 (2002). The Supreme Court held in Atkins that executing an offender\\n\\nwho has mental retardation12 violates the Eighth Amendment’s ban on excessive\\n\\n\\n\\n\\n12\\n In 2002, the Atkins Court followed prevailing convention in referring to “mentally retarded”\\noffenders. We will use that now-disfavored term only when quoting or discussing older sources.\\nSince 2014, the Supreme Court has exclusively used the terms “intellectually disabled person”\\nand “person with intellectual disability.” See, e.g., Hall v. Florida, 572 U.S. 701, 704 (2014);\\nMoore v. Texas, 139 S. Ct. 666, 672 (2019).\\n 41\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 42 of 73\\n\\n\\npunishments. Id. at 321. It defined mental retardation by quoting the American\\n\\nAssociation on Mental Retardation:\\n\\n Mental retardation refers to substantial limitations in present\\n functioning. It is characterized by significantly subaverage intellectual\\n functioning, existing concurrently with related limitations in two or\\n more of the following applicable adaptive skill areas: communication,\\n self-care, home living, social skills, community use, self-direction,\\n health and safety, functional academics, leisure, and work. Mental\\n retardation manifests before age 18.\\n\\nId. at 308 n.3. Alabama courts have followed that tripartite definition, adopting\\n\\nwhat they consider the “broadest” definition of mental retardation for Atkins\\n\\npurposes:\\n\\n [A] defendant, to be considered mentally retarded, must have\\n significantly subaverage intellectual functioning (an IQ of 70 or\\n below), and significant or substantial deficits in adaptive behavior.\\n Additionally, these problems must have manifested themselves during\\n the developmental period (i.e., before the defendant reached age 18).\\n\\nEx parte Perkins, 851 So. 2d 453, 456 (Ala. 2002).\\n\\n The clinical evidence adduced during the Rule 32 hearing with respect to\\n\\nJenkins’s intellectual ability was that his IQ was 76, reflecting “borderline\\n\\nintellectual functioning.” The record also included the report of the Lunacy\\n\\nCommission13 which found that Jenkins was competent to stand trial, as well as\\n\\n\\n\\n13\\n Alabama law requires a capital defendant to be “committed to the Department of Mental\\nHealth and Mental Retardation for examination” if there is a question about his capacity to stand\\ntrial or if there will be a mental-disease-or-defect defense. Ala. Code § 15-16-22(a). “The\\nassigned mental health professional(s) shall examine the defendant with respect to determining\\nthe presence of any mental disease or defect which, if determined to be present, would affect the\\n 42\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 43 of 73\\n\\n\\nJenkins’s school and juvenile court records. The court also considered all the\\n\\ntestimony it had heard about Jenkins and reviewed the transcript of his murder trial\\n\\nbefore rejecting Jenkins’s claim of mental retardation.\\n\\n On appeal, after reciting the Perkins standard, the Alabama Court of\\n\\nCriminal Appeals explained that Jenkins had not established mental retardation:\\n\\n Dr. Kirkland testified that he performed psychological tests on Jenkins\\n and that Jenkins’s IQ was 76. There was evidence presented at\\n Jenkins’s trial indicating that Jenkins maintained relationships with\\n other individuals and that he had been employed by P.S. Edwards\\n Landscaping Company, Cotton Lowe 76 Service Station, and\\n Paramount Painting Company. The record fails to show that Jenkins\\n meets the most liberal view of mental retardation adopted by the\\n Alabama Supreme Court in Perkins.\\n\\nJenkins, 972 So. 2d at 155. In this habeas appeal, Jenkins raises arguments about\\n\\nall three components of the intellectual disability showing under Atkins and\\n\\nPerkins, which we address in turn. He also appeals the denial of an evidentiary\\n\\nhearing on this issue.\\n\\n 1. Intellectual Function\\n\\n The first component of intellectual disability under Perkins is “significantly\\n\\nsubaverage intellectual functioning (an IQ of 70 or below).” Perkins, 851 So. 2d at\\n\\n456. Jenkins argues that the use of a strict IQ cutoff of 70 was contrary to or an\\n\\nunreasonable application of Atkins.\\n\\n\\ncapacity of the defendant to proceed or continue to trial or which would affect the defendant’s\\ncriminal responsibility at the time of the commission of the crime.” Id.\\n 43\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 44 of 73\\n\\n\\n We disagree. The state court’s decision that Jenkins failed to establish\\n\\nmental retardation in this respect was a reasonable application of the record and of\\n\\nAtkins. As a threshold matter, we note that the state court did not say it was\\n\\napplying a strict cutoff of 70; it merely stated that Dr. Kirkland had measured\\n\\nJenkins’s IQ as 76. Jenkins, 972 So. 2d at 155. Under our deferential review, the\\n\\nquestion is whether the decision that Jenkins had not established the intellectual\\n\\nfunctioning aspect of intellectual disability was an unreasonable application of\\n\\nAtkins. The Supreme Court explained in Atkins that it was leaving “to the State[s]\\n\\nthe task of developing appropriate ways to enforce the constitutional restriction\\n\\nupon [their] execution of sentences.” Atkins, 536 U.S. at 317 (quoting Ford v.\\n\\nWainwright, 477 U.S. 399, 416–17 (1986)). In line with that invitation, our Circuit\\n\\nhas accepted Alabama’s judicial definition of intellectual disability as laid out in\\n\\nPerkins.14 See Thomas, 607 F.3d at 752–53. Jenkins objects, however, that the\\n\\nPerkins standard, which he characterizes as articulating a “strict cutoff” of 70,\\n\\nconflicts with more recent Supreme Court precedent that instead requires a\\n\\n“clinical” definition that goes beyond a single IQ score.\\n\\n In particular, Jenkins points to Hall v. Florida, in which the Supreme Court\\n\\nheld that when an offender’s IQ is near but greater than 70, courts must take into\\n\\n\\n\\n14\\n Notwithstanding the urging of the Alabama Supreme Court, see Perkins, 851 So. 2d at 455\\nn.1, the Alabama legislature has not enacted a statutory definition of intellectual disability.\\n 44\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 45 of 73\\n\\n\\naccount the IQ test’s standard error of measurement as well as non-IQ evidence of\\n\\nintellectual function. See Hall v. Florida, 572 U.S. 701, 723 (2014) (“This Court\\n\\nagrees with the medical experts that when a defendant’s IQ test score falls within\\n\\nthe test’s acknowledged and inherent margin of error, the defendant must be able\\n\\nto present additional evidence of intellectual disability, including testimony\\n\\nregarding adaptive deficits.”). But, as Jenkins concedes, our Circuit has\\n\\nspecifically held that Hall is not retroactive to cases on collateral review. See In re\\n\\nHenry, 757 F.3d 1151, 1161 (11th Cir. 2014) (explaining that Hall does not supply\\n\\na new substantive rule but rather “merely provides new procedures for ensuring\\n\\nthat States do not execute members of an already protected group”). We also note\\n\\nthat Alabama courts have found the Perkins standard to comport with Hall in that\\n\\nit does not prevent an offender from presenting the standard error of measurement\\n\\nor non-IQ evidence of intellectual disability. See Reeves v. State, 226 So. 3d 711,\\n\\n728–29 (Ala. Ct. Crim. App. 2016). Here, the state court did not unreasonably\\n\\napply Atkins when it analyzed the intellectual functioning component of\\n\\nintellectual disability.\\n\\n Jenkins also argues that the state court’s decision on this component in his\\n\\nRule 32 proceeding resulted from an unreasonable determination of the facts. He\\n\\nasserts that his IQ score of 76, when combined with the test’s standard error of\\n\\nmeasurement and considered alongside the Flynn effect (which we discuss below),\\n\\n\\n 45\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 46 of 73\\n\\n\\ndoes show that he has significantly subaverage intellectual function. For the\\n\\nreasons that follow, we disagree.\\n\\n Most fundamentally, neither of the clinicians who testified about Jenkins’s\\n\\nintellectual ability in the Rule 32 hearing opined that he was mentally retarded.\\n\\nThe closest thing Jenkins can point to is the testimony of Dr. Kirkland that his IQ\\n\\nis two standard deviations below the mean. But his entire testimony, in context, did\\n\\nnot establish intellectual disability. To the contrary: Dr. Kirkland testified that\\n\\nJenkins “scored in the range of borderline intellectual functioning which is\\n\\nbetween mild mental retardation and low average intellectual functioning.”\\n\\nFurthermore, Jenkins’s own expert Dr. Lisak fully reviewed Dr. Kirkland’s clinical\\n\\nfindings. He agreed with those findings, concluding that the IQ results show\\n\\n“borderline intelligence.” It is tremendously significant that no clinical assessment\\n\\nin the entire record of these proceedings has found that Jenkins has mental\\n\\nretardation or intellectual disability.\\n\\n Jenkins nonetheless argues that his IQ score of 76 establishes the intellectual\\n\\ncomponent of intellectual ability when the standard error of measurement and the\\n\\nFlynn effect are considered alongside that score. It was not unreasonable for the\\n\\nstate court to disagree. As we have discussed, Jenkins is not retroactively entitled\\n\\nto have the state courts take into account the margin of error of the IQ test. That\\n\\nstatistical margin of error—5 points in this case—takes into account characteristics\\n\\n\\n 46\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 47 of 73\\n\\n\\nof the test and variations due to chance in order to state how confident the tester is\\n\\nthat the measured result accurately states Jenkins’s true, unknowable IQ. A result\\n\\nof 76 with a standard error of measurement of 5 simply means that the tester is\\n\\n68% confident that Jenkins’s true IQ falls between 71 and 81, and 95% confident\\n\\nthat his true IQ falls between 66 and 86. See David H. Kaye & David A. Freedman,\\n\\nReference Guide on Statistics, in Reference Manual on Scientific Evidence 211,\\n\\n243–44 (3d ed. 2011); see also Hall, 572 U.S. at 738–39 (Alito, J., dissenting). Of\\n\\ncourse, Jenkins would like us to consider only the possibility that the test result\\n\\noverstated his IQ—but the standard error of measurement is a two-way street.\\n\\n“[T]he standard error of measurement is a bi-directional concept that does not\\n\\ncarry with it a presumption that an individual’s IQ falls to the bottom of his IQ\\n\\nrange.” Ledford v. Warden, Ga. Diag. & Clas. Prison, 818 F.3d 600, 641 (11th Cir.\\n\\n2016).\\n\\n Even if Jenkins were somehow entitled to subtract the 5-point margin of\\n\\nerror from his score and thereby lower his “true” IQ to 71, he still would not satisfy\\n\\nthe clinical intellectual component of intellectual disability. Thus, he sought in the\\n\\ndistrict court to introduce evidence about the Flynn effect in order to bring his IQ\\n\\ndown to 70 or below. As our Court has explained, the Flynn effect “recognizes the\\n\\nfact that IQ test scores have been increasing over time”:\\n\\n The Flynn effect acknowledges that as an intelligence test ages, or\\n moves farther from the date on which it was standardized, or normed,\\n 47\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 48 of 73\\n\\n\\n the mean score of the population as a whole on that assessment\\n instrument increases, thereby artificially inflating the IQ scores of\\n individual test subjects. Therefore, the IQ test scores must be\\n recalibrated to keep all test subjects on a level playing field.\\n\\nThomas, 607 F.3d at 753. We have said that “[a]n evaluator may also consider the\\n\\n‘Flynn effect’” in assessing an offender’s possible intellectual disability, id.\\n\\n(emphasis added), but neither we nor the Alabama Supreme Court nor the U.S.\\n\\nSupreme Court have said that a court must consider it in order to reasonably apply\\n\\nAtkins. See, e.g., Reeves, 226 So. 3d at 739 (“This Court has repeatedly held that a\\n\\ncircuit court is not required to accept, consider, or apply the ‘Flynn Effect’ in\\n\\ndetermining intellectual disability.”).\\n\\n More problematically, Jenkins did not raise the Flynn effect issue in the\\n\\nAlabama state courts. Jenkins asserts that he implicitly raised it as part and parcel\\n\\nof his Atkins claim during his Rule 32 appeal, but we do not agree that the\\n\\nAlabama courts had a full and fair opportunity to consider this issue, because it\\n\\nwas not explicitly presented to them. See Kelley v. Sec’y for the Dep’t of Corr., 377\\n\\nF.3d 1317, 1349–50 (11th Cir. 2004) (holding that one ineffective-assistance\\n\\nargument was unexhausted despite the petitioner’s general presentation of a\\n\\nStrickland claim in the state courts). Thus, under AEDPA, Jenkins is procedurally\\n\\nbarred from raising this unexhausted claim in his federal habeas petition. See 28\\n\\nU.S.C. § 2254(b)(1)(A), (c); Mason v. Allen, 605 F.3d 1114, 1119 (11th Cir. 2010).\\n\\nWe will allow a prisoner to overcome a procedural bar if he establishes cause for\\n 48\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 49 of 73\\n\\n\\nhis default and prejudice therefrom. See, e.g., Smith v. Jones, 256 F.3d 1135, 1138\\n\\n(11th Cir. 2001). But Jenkins cannot establish prejudice from being barred from\\n\\nasserting this claim now because the Supreme Court has not “clearly established”\\n\\nthat a court must consider the Flynn effect in assessing intellectual disability. Thus,\\n\\nunder AEDPA, it could not form the basis for a federal habeas claim. See 28\\n\\nU.S.C. § 2254(d)(1). We therefore conclude that the state court did not\\n\\nunreasonably determine the facts or unreasonably apply Atkins with respect to the\\n\\nintellectual component of intellectual disability.\\n\\n 2. Adaptive Function\\n\\n The second component of intellectual disability is “significant or substantial\\n\\ndeficits in adaptive behavior.” Perkins, 851 So. 2d at 456. Jenkins argues that the\\n\\nstate court, which mentioned only Jenkins’s successes in employment and social\\n\\nskills, see Jenkins, 972 So. 2d at 155, failed to consider his deficits. He adds that\\n\\neven the mention of employment success does not tell the whole story. Jenkins\\n\\ninsists that his jobs were menial, his employers took advantage of him, and he held\\n\\nthe jobs for only a few months each.\\n\\n But the state court’s terse discussion of Jenkins’s adaptive behavior did not\\n\\ninvolve an unreasonable determination of the facts or an unreasonable application\\n\\nof Atkins. To the contrary, its decision on this issue was reasonable in light of both\\n\\nthe record and Atkins.\\n\\n\\n 49\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 50 of 73\\n\\n\\n Atkins elaborated that the adaptive-skill component of intellectual disability\\n\\nrequires substantial present limitation in at least two of the following areas:\\n\\n“communication, self-care, home living, social skills, community use, self-\\n\\ndirection, health and safety, functional academics, leisure, and work.” Atkins, 536\\n\\nU.S. at 308 n.3 (quoting definition of American Association on Mental\\n\\nRetardation). The record supports the state court’s judgment that Jenkins does not\\n\\nhave substantial deficits in at least two areas. To be sure, we will assume that\\n\\nJenkins may be substantially limited in functional academics; Dr. Kirkland\\n\\nassessed his reading, spelling, and arithmetic skills at a third-grade level. But the\\n\\nrecord does not compel a finding of substantial present deficits in the areas of\\n\\ncommunication, self-care, home living, and work. Although Jenkins emphasizes\\n\\nhis transient employment, the record shows he was able to find work quickly, to\\n\\nwork hard to support the Seals’ household until Lonnie Seal found a job, and, later,\\n\\nto move into his own home. That record evidence is inconsistent with a substantial\\n\\ndeficit in the area of work.\\n\\n As the State observes, the facts of the crime also fail to show significant\\n\\ndeficits in the areas of communication, self-care, community use, and self-\\n\\ndirection. Jenkins was able to communicate well enough to solicit an alibi and to\\n\\nsell his car, including writing out a bill of sale. He then was able to arrange for his\\n\\nflight from Alabama, including purchasing a bus ticket and hitchhiking across the\\n\\n\\n 50\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 51 of 73\\n\\n\\ncountry. Indeed, throughout his trial and postconviction proceedings, the people\\n\\nwho knew Jenkins as an adult consistently described a man who did not have\\n\\nserious difficulties in communicating, forming relationships, working, and caring\\n\\nfor himself. No witness mentioned these kinds of difficulties in his or her\\n\\ntestimony about Jenkins. There was some evidence that Jenkins was “not\\n\\narticulate” and that his house was messy, but that testimony does not rise to the\\n\\nlevel of overwhelming evidence of substantial functional deficits that would\\n\\nsupport a conclusion that the state court unreasonably determined the facts.\\n\\n Of course, the record also contains evidence of Jenkins’s childhood\\n\\nacademic and social deficits, attributable at least in part to his parents’ neglect, but\\n\\nJenkins’s childhood is not directly relevant to our consideration of his present\\n\\nlimitations. Although we acknowledge that the Supreme Court has noted that such\\n\\n“risk factors” may support further exploration of the possibility of intellectual\\n\\ndisability, see Moore v. Texas, 137 S. Ct. 1039, 1051 (2017), the state court record\\n\\ncontains the results of that exploration. Overall, that record supports the state\\n\\ncourt’s conclusion that Jenkins does not have substantial deficits in adaptive\\n\\nbehaviors.\\n\\n 3. Juvenile Onset\\n\\n The third and final component of intellectual disability is manifestation of\\n\\nthe first and second components “before the defendant reached age 18.” Perkins,\\n\\n\\n 51\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 52 of 73\\n\\n\\n851 So. 2d at 456. Jenkins asserts, the State assumes, and we agree that the state\\n\\ncourt did not rule upon this issue, so there is nothing to which we must defer under\\n\\nAEDPA. See, e.g., Brumfield v. Cain, 135 S. Ct. 2269, 2282 (2015). Thus, we will\\n\\nreview the district court’s legal conclusions on this issue de novo, and its factual\\n\\nfindings for clear error. See, e.g., Grossman v. McDonough, 466 F.3d 1325, 1335\\n\\n(11th Cir. 2006). Jenkins argues that he has satisfied the age-of-onset component\\n\\nby showing “a wide range of evidence of deficits in both intellectual and adaptive\\n\\nfunctioning throughout Jenkins’s childhood.”\\n\\n In fact, the ample record of Jenkins’s childhood, as developed in the Rule 32\\n\\nproceeding, does not point to intellectual disability before age 18. School\\n\\nassessments at age 12 reported an IQ score of 83, an adaptive behavior rating of\\n\\n“low,” and noted overall an intellectual capacity of “average,” a diagnosis of\\n\\ndyslexia, and “poor functional skills.” The psychologist attributed Jenkins’s\\n\\ndifficulties to a combination of “[f]requent moves, changes in schools and family\\n\\ninstability” with “[l]ow ability coupled with a specific learning disability” and\\n\\n“[l]ack of responsibility for self control and poor use of time.” Although Jenkins\\n\\nfailed several grades of school, teachers additionally attributed his difficulties to\\n\\nhis behavioral problems and his high number of absences from school.\\n\\n Similarly, IQ tests ordered by the juvenile court at age 14 gave scores of 81\\n\\nand 86 on different tests, in the “[d]ull–[n]ormal” range and reflecting “a definite\\n\\n\\n 52\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 53 of 73\\n\\n\\nlearning disability.” These records all show a child with serious academic deficits\\n\\nand some intellectual and adaptive deficits, but they do not clearly show an\\n\\nintellectually disabled child. In particular, Jenkins’s childhood IQ scores of 83, 81,\\n\\nand 86 fail to point to intellectual disability or mental retardation. We are not left\\n\\nwith the definite and firm conviction that the district court erred when it found that\\n\\n“Jenkins’s [childhood] intelligence scores and adaptive functioning skills were\\n\\nattributable to his learning disability and other circumstances in his life, rather than\\n\\nmental retardation.”\\n\\n In sum, we agree with the district court that the decision of the Alabama\\n\\nCourt of Criminal Appeals that Jenkins does not have intellectual disability was\\n\\nnot contrary to or an unreasonable application of Atkins v. Virginia.\\n\\n 4. Evidentiary Hearing\\n\\n Finally, Jenkins argues that the district court erred when it denied his request\\n\\nfor an evidentiary hearing on his Atkins claim. If Jenkins diligently attempted to\\n\\npresent this claim in the state court, we review for abuse of discretion. Burgess v.\\n\\nComm’r, Ala. Dep’t of Corr., 723 F.3d 1308, 1320 (11th Cir. 2013). “The district\\n\\ncourt abuses its discretion where ‘such a hearing could enable an applicant to prove\\n\\nthe petition’s factual allegations, which, if true, would entitle the applicant to\\n\\nfederal habeas relief.’” Id. (quoting Schriro v. Landrigan, 550 U.S. 465, 474\\n\\n(2007)). If Jenkins did not diligently attempt to develop the factual basis for this\\n\\n\\n 53\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 54 of 73\\n\\n\\nclaim in the state court, absent certain circumstances, he is not entitled to a hearing\\n\\nand we will affirm the district court’s denial. 28 U.S.C. § 2254(e)(2).\\n\\n The parties strenuously dispute whether Jenkins diligently attempted to\\n\\npresent this issue in the state court. On the one hand, the State argues that Jenkins\\n\\nfailed to allege that he was mentally retarded even after the U.S. Supreme Court\\n\\ndecided Atkins in 2002. On the other hand, Jenkins asserts that he raised the\\n\\npossibility of mental retardation as early as his original Rule 32 petition in 1995.\\n\\n As we have noted, shortly after the Supreme Court decided Atkins, the\\n\\nAlabama Court of Criminal Appeals ordered supplemental briefing on the possible\\n\\nimpact of Atkins on Jenkins. Jenkins’s Rule 32 appeal was pending at that point,\\n\\nhaving been fully briefed and orally argued in the state appellate court. Having\\n\\nreviewed the state court records, we agree with the State that Jenkins did not\\n\\ndiligently attempt to present his Atkins claim in the state courts.\\n\\n “Diligence will require in the usual case that the prisoner, at a minimum,\\n\\nseek an evidentiary hearing in state court.” Williams, 529 U.S. at 437. The record\\n\\nclearly shows that Jenkins did not do so. In the initial supplemental brief on Atkins\\n\\nhe filed in 2002, Jenkins merely requested a future hearing after the Alabama\\n\\nlegislature’s enactment of a statute on executing the mentally retarded. Seventeen\\n\\nyears later, that contingency is still in the future. Only in his reply supplemental\\n\\nbrief did Jenkins then assert, for the first time ever, that he is mentally retarded.\\n\\n\\n 54\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 55 of 73\\n\\n\\n We acknowledge that Jenkins’s initial Rule 32 petition included in its\\n\\nallegations of ineffective assistance of counsel a list of unpresented mitigating\\n\\nevidence including his “developmental[] impair[ment]” and “learning disabilities,\\n\\nlow intelligence, poor comprehension and retarded socialization skills.” We\\n\\nconclude that these tangential Strickland references to his mental and social\\n\\nabilities do not amount to diligence. Thus, because Jenkins did not diligently\\n\\nattempt to develop the factual basis for his Atkins claim in the state courts, we\\n\\naffirm the district court’s denial of his request for an evidentiary hearing.\\n\\n Even if Jenkins had diligently attempted to present this claim in the state\\n\\ncourt, however, we conclude in the alternative that the district court did not abuse\\n\\nits discretion when it denied a hearing. We agree, in light of the ample evidence\\n\\nabout intellectual ability that Jenkins adduced in support of his ineffective-\\n\\nassistance claim, that such a hearing would not have enabled Jenkins to prove that\\n\\nhe is intellectually disabled. See Burgess, 723 F.3d at 1320.\\n\\n IV. CONCLUSION\\n\\n The denial of Jenkins’s petition for a writ of habeas corpus is\\n\\n AFFIRMED.\\n\\n\\n\\n\\n 55\\n\\n Case: 17-12524 Date Filed: 08/30/2019 Page: 56 of 73\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4167,"cells":{"id":{"kind":"number","value":4645248,"string":"4,645,248"},"name":{"kind":"string","value":"UNITED STATES of America, Appellee v. Mir ISLAM, Appellant"},"name_abbreviation":{"kind":"string","value":"United States v. Mir Islam"},"decision_date":{"kind":"string","value":"2019-08-02"},"docket_number":{"kind":"string","value":"18-3003"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Katsas, Silberman, Williams\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge KATSAS\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued November 19, 2018 Decided August 2, 2019\\n\\n No. 18-3003\\n\\n UNITED STATES OF AMERICA,\\n APPELLEE\\n\\n v.\\n\\n MIR ISLAM,\\n APPELLANT\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:15-cr-00067-1)\\n\\n\\n Jonathan Zucker, appointed by the court, argued the cause\\nand filed the briefs for appellant.\\n\\n Peter S. Smith, Assistant U.S. Attorney, argued the cause\\nfor appellee. With him on the brief were Jessie K. Liu, U.S.\\nAttorney, and Elizabeth Trosman, John P. Mannarino, and\\nJonathan P. Hooks, Assistant U.S. Attorneys.\\n\\n Before: KATSAS, Circuit Judge, and SILBERMAN and\\nWILLIAMS, Senior Circuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge KATSAS.\\n\\f 2\\n KATSAS, Circuit Judge: This appeal addresses the legal\\nconsequences of a criminal defendant’s failure to object to a\\nmagistrate judge’s adverse report and recommendation. We\\nalso consider claims that defense counsel provided ineffective\\nassistance during proceedings to revoke a term of supervised\\nrelease.\\n\\n I\\n\\n Mir Islam received two federal sentences for various\\noffenses. In the Southern District of New York, Islam pleaded\\nguilty to crimes involving credit-card fraud, identity theft, and\\ncomputer hacking. An SDNY judge sentenced him to one day\\nof imprisonment followed by three years of supervised release.\\nIn the District of Columbia, Islam pleaded guilty to crimes\\ninvolving the theft and online publication of personal\\ninformation, conveying false information about the use of\\nexplosives, and cyber-stalking. A DDC judge sentenced him\\nto two years of imprisonment followed by three years of\\nsupervised release. The respective terms of supervised release\\nwere subject to substantially similar conditions.\\n\\n After Islam served both prison sentences, he began\\nconcurrently serving the terms of supervised release. Because\\nIslam was living in New York, the SDNY probation office\\nconducted supervision for both courts.\\n\\n On January 18, 2017, Islam was arrested in New York for\\nviolating the conditions of his supervised release. On January\\n19, the SDNY probation office filed with the SDNY a petition\\nto revoke the supervised release. On April 11, the SDNY\\njudge revoked Islam’s original term of supervised release and\\nimposed two years of imprisonment followed by one year of\\nsupervised release subject to the same conditions previously\\nimposed. She recommended that the imprisonment and\\nsupervised release run concurrently with any further\\n\\f 3\\npunishment that the DDC judge might impose. With credit for\\ntime served, Islam was released from SDNY custody on\\nAugust 17, 2017.\\n\\n In the meantime, the DDC probation office filed its own\\npetition to revoke with the DDC. On February 1, 2017, the\\nDDC judge issued a warrant for Islam’s arrest, which was\\nlodged as a detainer to be executed upon Islam’s release from\\nSDNY custody. On August 14, Islam moved to dismiss the\\ndetainer and to transfer the DDC case to the SDNY. On the\\nsame day, the DDC judge denied the motion in a minute order.\\nUpon his release from SDNY custody, Islam was held under\\nthe detainer and then transferred to the District of Columbia.\\nOn September 6, Islam arrived in the District, was arrested\\nunder the warrant, and appeared before a magistrate judge\\nassigned to his case.\\n\\n The magistrate judge scheduled a revocation hearing for\\nSeptember 15, but Islam sought and received two continuances.\\nOn October 27, Islam moved to dismiss the petition for\\nrevocation. Among other things, he argued that the delay\\nbetween his arrest in New York and revocation proceedings in\\nthe DDC violated both the Due Process Clause of the Fifth\\nAmendment and Federal Rule of Criminal Procedure\\n32.1(a)(1). Islam also sought to transfer the matter to the\\nSDNY. On November 8, the magistrate judge held a hearing\\non these issues.\\n\\n On December 4, the magistrate judge issued a thirty-page\\nreport and recommendation rejecting Islam’s various\\narguments. The magistrate judge proposed finding that Islam\\nhad engaged in unauthorized travel, failed to identify\\ncomputers and other electronic devices to which he had access,\\nfailed to allow monitoring of those devices, failed to attend\\nmental-health counseling, and failed to provide requested\\n\\f 4\\nfinancial information—all in violation of his supervised-\\nrelease conditions. The magistrate judge recommended that\\nthe district court impose four months of imprisonment. She\\nrecommended no further supervision because the SDNY\\nprobation office would be supervising compliance with the\\nconditions imposed in the SDNY, which were “nearly\\nidentical” to those imposed in the DDC. App. 98. The report\\nand recommendation stated that “any party who objects to the\\nproposed findings or recommendations herein must file written\\nobjections within fourteen days” of service. App. 99. It\\nfurther warned that the parties “may waive their right of appeal\\nfrom an order of the District Court adopting such findings and\\nrecommendations” if they “fail to file timely objections.” Id.\\nIslam did not file any objections.\\n\\n The district court held a revocation hearing on December\\n19, 2017. At the outset, the court asked the parties if they\\nobjected to the report and recommendation. Through counsel,\\nIslam replied that he did not. Still, the court went over each\\nviolation and confirmed that Islam had no objections. The\\ncourt adopted the proposed findings that Islam had violated his\\nsupervised-release conditions in five different respects,\\nrevoked his supervised release, and imposed nine months of\\nimprisonment followed by 24 months of supervised release.\\nThe court rejected the magistrate judge’s recommendation of a\\nfour-month prison term with no further supervised release, as\\nwell as the SDNY judge’s recommendation of a concurrent\\nprison term. Regarding the relationship between the SDNY\\nand DDC proceedings, the district court explained: “[T]his\\nwas a separate matter and there was a separate criminal\\nproceeding here. This is a violation of the terms with respect\\nto that separate criminal proceeding.” App. 110.\\n\\f 5\\n II\\n\\n On appeal, Islam seeks to raise the same delay argument\\nthat he pressed unsuccessfully before the magistrate judge.\\nHe contends that the lag between his January 2017 arrest and\\nhis December 2017 revocation hearing violated both due\\nprocess and Rule 32.1(a)(1). The government responds that\\nIslam forfeited his right to appeal the district court’s decision\\nrejecting these claims by failing to object to the magistrate\\njudge’s adverse recommendation. We agree.\\n\\n A\\n\\n The Federal Magistrates Act permits a district court to\\n“designate a magistrate judge to conduct hearings” on various\\ncivil and criminal matters, 28 U.S.C. § 636(b)(1)(B), including\\npetitions “to modify, revoke, or terminate supervised release,”\\n18 U.S.C. § 3401(i). For such petitions, the magistrate judge\\nmust make “proposed findings of fact and recommendations,”\\nid., as she must for all other dispositive matters referred to her,\\nsee 28 U.S.C. § 636(b)(1)(B); Fed. R. Crim. P. 59(b)(1); Fed.\\nR. Civ. P. 72(b)(1). Within 14 days of the recommended\\ndisposition of a criminal case, any party “may” file objections,\\nFed. R. Crim. P. 59(b)(2), in which case the district court “must\\nconsider” the objection “de novo,” id. 59(b)(3). “[F]ailure to\\nobject in accordance with this rule waives a party’s right to\\nreview.” Id. 59(b)(2). Failure to object also “waives a\\nparty’s right to review” of any non-dispositive matter referred\\nto a magistrate judge in a criminal case. Id. 59(a).\\n\\n The Supreme Court explained the rationale for this waiver\\nrule in Thomas v. Arn, 474 U.S. 140 (1985), which held that\\nthe courts of appeals could make and enforce such a rule\\nthrough their supervisory powers. See id. at 155. In this\\ncontext as in others, contemporaneous objections help to\\n“focus attention” on any genuinely contested issues. Id. at\\n\\f 6\\n147. Moreover, without the waiver rule, litigants could\\nsandbag a district court “by failing to object and then\\nappealing,” which would either “force the court of appeals to\\nconsider claims that were never reviewed by the district court,\\nor force the district court to review every issue in every case.”\\nId. at 148. Many courts of appeals had adopted the waiver rule\\napproved in Thomas, see id. at 146 n.4, and the Supreme Court,\\nin 1983, codified it for non-dispositive matters in civil cases,\\nsee Fed. R. Civ. P. 72(a). After the Ninth Circuit later refused\\nto apply the waiver rule to criminal cases, United States v.\\nAbonce-Barrera, 257 F.3d 959, 967–69 (9th Cir. 2001), the\\nSupreme Court, in 2005, codified it for criminal cases as well,\\nsee Fed. R. Crim. P. 59(a), (b)(2).\\n\\n The parties briefed this case as if Thomas still controls,\\nwhich would require us to determine whether to adopt its\\nwaiver rule as an exercise of supervisory power. But as we\\nhave explained, Rule 59(b)(2) now governs, and it provides\\nthat the “[f]ailure to object” to a magistrate judge’s proposed\\nfindings and recommendations on any dispositive matter in a\\ncriminal case “waives a party’s right to review.” Because\\nIslam did not timely object to the report and recommendation\\nin this case, he cannot appeal the district court’s decision\\nadopting it.\\n\\n B\\n\\n Because Rule 59(b)(2) is not cast in jurisdictional terms,\\nthe courts have discretion to excuse a waiver under the rule.\\nIslam argues that the district court excused his waiver here and\\nthat we should do likewise. We disagree.\\n\\n As for the district court, Islam highlights one passing\\ncomment, made orally at the revocation hearing, that the court\\nhad reviewed and agreed with the “analysis and conclusions”\\nof the magistrate judge. App. 105. That hardly constitutes a\\n\\f 7\\ndecision to excuse the waiver. To the contrary, in its very next\\nbreath, the court “note[d] for the record as well that the parties\\nhave not objected to [the] report and recommendation.” Id.\\nAnd again, the court later noted that “Mr. Islam is not at this\\npoint contesting [the] report and recommendation.” App. 106.\\n\\n In this Court, Islam presses his claims of unlawful delay\\ndespite failing to preserve them below. We review such\\nunpreserved claims only for plain error. Puckett v. United\\nStates, 556 U.S. 129, 135 (2009). Here, there was no error.\\n\\n Islam’s due-process argument is meritless. To violate\\ndue process, a delay in revocation proceedings must be both\\nunreasonable and prejudicial. See Sutherland v. McCall, 709\\nF.2d 730, 732 (D.C. Cir. 1983). Islam complains about the\\neleven-month delay between his January 2017 arrest in New\\nYork and his final revocation hearing in December 2017. But\\nthe DDC arrest warrant was not executed until September 6,\\n2017. And it is the “execution of the warrant and custody\\nunder that warrant” that serves as “the operative event\\ntriggering any loss of liberty” for purposes of due process.\\nMoody v. Daggett, 429 U.S. 78, 87 (1976). So, the due-\\nprocess clock for the DDC revocation hearing did not begin to\\nrun until Islam was arrested on the DDC charges in September.\\n\\n Islam contends that he was arrested in January 2017 for\\nviolating both the conditions imposed by the SDNY and those\\nimposed by the DDC. He argues that the arrest was made by\\nthe probation officer based on probable cause that Islam\\nviolated both sets of conditions. But the SDNY record\\nindicates that the arrest was made on a warrant issued by the\\nSDNY judge for violations of the SDNY conditions. United\\nStates v. Islam, No. 12-cr-810 (S.D.N.Y.), ECF Docs. 36, 39,\\n46. Thus, the magistrate judge permissibly found that Islam\\nwas not arrested on the DDC charges until September, and the\\n\\f 8\\ndistrict court permissibly adopted the finding. Given a\\nSeptember arrest date, the December revocation hearing—\\npushed back twice by continuances requested by Islam—was\\nnot unreasonably delayed. See Howard v. Caufield, 765 F.3d\\n1, 3, 13 (D.C. Cir. 2014) (“minor delay” of four months\\nbetween arrest and revocation hearing was not unreasonable);\\nUnited States v. Madden, 515 F.3d 601, 607 (6th Cir. 2008)\\n(delay “caused by [the defendant’s] own conduct” is not\\nunreasonable).\\n\\n Islam’s due-process argument also fails for lack of any\\nprejudice. The only harm he alleges is losing the ability to\\nseek a prison term in the DDC that would run concurrently with\\nthe one imposed by the SDNY. But the district court did\\nconsider—and expressly reject—the possibility of a concurrent\\npunishment. The court explained that the SDNY and DDC\\nproceedings involved separate offenses, that the totality of\\nIslam’s misconduct was concerning, and that additional\\nincarceration was therefore appropriate. Islam provides no\\nreason to think that the district court might have assessed these\\nissues differently had it imposed punishment sooner rather than\\nlater.\\n\\n Islam further argues that the delay violated Rule\\n32.1(a)(1), which provides that a person arrested for violating\\nconditions of supervised release must be brought before a\\nmagistrate judge “without unnecessary delay.” Nothing in\\nthat sparse text expands due-process protections in this area.\\nTo the contrary, the rule simply codifies these protections.\\nSee, e.g., United States v. Ruby, 706 F.3d 1221, 1226 (10th Cir.\\n2013). Islam’s Rule 32.1 argument thus fares no better than\\nhis due-process argument.\\n\\f 9\\n III\\n\\n In the alternative, Islam seeks a remand for an evidentiary\\nhearing on two claims alleging ineffective assistance of counsel\\nduring the revocation proceedings. Although we remand\\ncolorable claims of ineffective assistance, we have “never held\\nthat any claim of ineffective assistance of counsel …\\nautomatically entitles a party to an evidentiary remand.”\\nUnited States v. Sitzmann, 893 F.3d 811, 831 (D.C. Cir. 2018)\\n(per curiam) (quotation marks omitted). When the record\\n“clearly shows” that the claim is meritless, or when no further\\nfactual development is needed, we may dispose of the claim\\nwithout remanding. Id. at 831–32.\\n\\n As an initial matter, it is unclear whether Islam had any\\nright to effective assistance during his revocation proceedings.\\nThe Sixth Amendment right to counsel does not attach to\\nrevocation hearings. Baker v. Sard, 486 F.2d 415, 423 (D.C.\\nCir. 1972). The Due Process Clause does attach, but it\\nguarantees counsel only in rare cases; as the Supreme Court has\\nexplained, the participation of counsel “will probably be both\\nundesirable and constitutionally unnecessary in most\\nrevocation hearings.” Gagnon v. Scarpelli, 411 U.S. 778, 790\\n(1973). Islam did have a statutory right to counsel, 18 U.S.C.\\n§ 3006A(a)(1)(E), but it is unclear whether that encompasses a\\nright to effective assistance, see United States v. Eskridge, 445\\nF.3d 930, 933 (7th Cir. 2006). Because the government is\\nwilling to assume that ineffective-assistance principles apply to\\nthis case, and that they parallel ones that govern under the Sixth\\nAmendment, we do likewise.\\n\\n To establish ineffective assistance, a defendant must show\\nboth that “counsel’s performance was deficient” and that “the\\ndeficient performance prejudiced the defense.” Strickland v.\\nWashington, 466 U.S. 668, 687 (1984). Deficient\\n\\f 10\\nperformance must fall “below an objective standard of\\nreasonableness.” Id. at 688. Prejudice requires a “reasonable\\nprobability” that, but for the deficient performance, “the result\\nof the proceeding would have been different.” Id. at 694.\\nThe failure to raise a meritless objection is not deficient\\nperformance. See Sitzmann, 893 F.3d at 833.\\n\\n Islam raises two allegations of ineffective assistance.\\nFirst, he claims that his counsel was ineffective in failing to\\nobject to the magistrate judge’s recommendation to reject the\\nclaim of unreasonable delay. As explained above, however,\\nthe unreasonable-delay claim lacks merit, and the delay caused\\nno prejudice. The waiver of this claim does not amount to\\nineffective assistance.\\n\\n Next, Islam complains that his counsel waited to file a\\ntransfer motion until October 2017, when the SDNY\\nrevocation proceedings were complete and the DDC\\nproceedings were already well underway. Islam contends that\\nif counsel had filed the motion in April 2017, when he first\\nlearned of the DDC charges, the motion likely would have been\\ngranted, and Islam then would have received concurrent rather\\nthan consecutive prison terms. But a transfer decision is left\\nto the district court’s discretion, 18 U.S.C. § 3605, and there is\\nno reasonable probability that the court here would have\\nordered a transfer. Again, Islam cannot show prejudice.\\n\\n Islam’s contrary argument focuses on the magistrate\\njudge’s recommendation to reject the October transfer motion\\nbecause she had already invested substantial time in the case.\\nIslam overlooks the district court’s prior decision rejecting out-\\nof-hand the August transfer motion, which was filed before the\\nmagistrate judge had begun working on the case. Islam also\\noverlooks the district court’s subsequent revocation decision,\\nwhich rejected the magistrate judge’s recommendation to cede\\n\\f 11\\nfuture supervision to the SDNY. Moreover, the supervised-\\nrelease violations charged in the DDC were broader than those\\ncharged in the SDNY, as only the former involved financial-\\ndisclosure obligations. And the underlying offenses in the\\nDDC, which originally produced a prison sentence of two\\nyears, were far more serious than those in the SDNY, which\\noriginally produced a prison sentence of one day. Under these\\ncircumstances, the likelihood of the district court’s granting an\\nearlier motion to transfer would have been slim at best.\\nWithout any prejudice, Islam’s second ineffective-assistance\\nclaim fares no better than his first.\\n\\n Affirmed.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4422501/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge KATSAS\"}]}"},"judge1":{"kind":"string","value":"KATSAS"},"judge2":{"kind":"string","value":"SILBERMAN"},"judge3":{"kind":"string","value":"WILLIAMS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4645248/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: UNITED STATES of America, Appellee v. Mir ISLAM, Appellant\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-02\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued November 19, 2018 Decided August 2, 2019\n\n No. 18-3003\n\n UNITED STATES OF AMERICA,\n APPELLEE\n\n v.\n\n MIR ISLAM,\n APPELLANT\n\n\n Appeal from the United States District Court\n for the District of Columbia\n (No. 1:15-cr-00067-1)\n\n\n Jonathan Zucker, appointed by the court, argued the cause\nand filed the briefs for appellant.\n\n Peter S. Smith, Assistant U.S. Attorney, argued the cause\nfor appellee. With him on the brief were Jessie K. Liu, U.S.\nAttorney, and Elizabeth Trosman, John P. Mannarino, and\nJonathan P. Hooks, Assistant U.S. Attorneys.\n\n Before: KATSAS, Circuit Judge, and SILBERMAN and\nWILLIAMS, Senior Circuit Judges.\n\n Opinion for the Court filed by Circuit Judge KATSAS.\n\f 2\n KATSAS, Circuit Judge: This appeal addresses the legal\nconsequences of a criminal defendant’s failure to object to a\nmagistrate judge’s adverse report and recommendation. We\nalso consider claims that defense counsel provided ineffective\nassistance during proceedings to revoke a term of supervised\nrelease.\n\n I\n\n Mir Islam received two federal sentences for various\noffenses. In the Southern District of New York, Islam pleaded\nguilty to crimes involving credit-card fraud, identity theft, and\ncomputer hacking. An SDNY judge sentenced him to one day\nof imprisonment followed by three years of supervised release.\nIn the District of Columbia, Islam pleaded guilty to crimes\ninvolving the theft and online publication of personal\ninformation, conveying false information about the use of\nexplosives, and cyber-stalking. A DDC judge sentenced him\nto two years of imprisonment followed by three years of\nsupervised release. The respective terms of supervised release\nwere subject to substantially similar conditions.\n\n After Islam served both prison sentences, he began\nconcurrently serving the terms of supervised release. Because\nIslam was living in New York, the SDNY probation office\nconducted supervision for both courts.\n\n On January 18, 2017, Islam was arrested in New York for\nviolating the conditions of his supervised release. On January\n19, the SDNY probation office filed with the SDNY a petition\nto revoke the supervised release. On April 11, the SDNY\njudge revoked Islam’s original term of supervised release and\nimposed two years of imprisonment followed by one year of\nsupervised release subject to the same conditions previously\nimposed. She recommended that the imprisonment and\nsupervised release run concurrently with any further\n\f 3\npunishment that the DDC judge might impose. With credit for\ntime served, Islam was released from SDNY custody on\nAugust 17, 2017.\n\n In the meantime, the DDC probation office filed its own\npetition to revoke with the DDC. On February 1, 2017, the\nDDC judge issued a warrant for Islam’s arrest, which was\nlodged as a detainer to be executed upon Islam’s release from\nSDNY custody. On August 14, Islam moved to dismiss the\ndetainer and to transfer the DDC case to the SDNY. On the\nsame day, the DDC judge denied the motion in a minute order.\nUpon his release from SDNY custody, Islam was held under\nthe detainer and then transferred to the District of Columbia.\nOn September 6, Islam arrived in the District, was arrested\nunder the warrant, and appeared before a magistrate judge\nassigned to his case.\n\n The magistrate judge scheduled a revocation hearing for\nSeptember 15, but Islam sought and received two continuances.\nOn October 27, Islam moved to dismiss the petition for\nrevocation. Among other things, he argued that the delay\nbetween his arrest in New York and revocation proceedings in\nthe DDC violated both the Due Process Clause of the Fifth\nAmendment and Federal Rule of Criminal Procedure\n32.1(a)(1). Islam also sought to transfer the matter to the\nSDNY. On November 8, the magistrate judge held a hearing\non these issues.\n\n On December 4, the magistrate judge issued a thirty-page\nreport and recommendation rejecting Islam’s various\narguments. The magistrate judge proposed finding that Islam\nhad engaged in unauthorized travel, failed to identify\ncomputers and other electronic devices to which he had access,\nfailed to allow monitoring of those devices, failed to attend\nmental-health counseling, and failed to provide requested\n\f 4\nfinancial information—all in violation of his supervised-\nrelease conditions. The magistrate judge recommended that\nthe district court impose four months of imprisonment. She\nrecommended no further supervision because the SDNY\nprobation office would be supervising compliance with the\nconditions imposed in the SDNY, which were “nearly\nidentical” to those imposed in the DDC. App. 98. The report\nand recommendation stated that “any party who objects to the\nproposed findings or recommendations herein must file written\nobjections within fourteen days” of service. App. 99. It\nfurther warned that the parties “may waive their right of appeal\nfrom an order of the District Court adopting such findings and\nrecommendations” if they “fail to file timely objections.” Id.\nIslam did not file any objections.\n\n The district court held a revocation hearing on December\n19, 2017. At the outset, the court asked the parties if they\nobjected to the report and recommendation. Through counsel,\nIslam replied that he did not. Still, the court went over each\nviolation and confirmed that Islam had no objections. The\ncourt adopted the proposed findings that Islam had violated his\nsupervised-release conditions in five different respects,\nrevoked his supervised release, and imposed nine months of\nimprisonment followed by 24 months of supervised release.\nThe court rejected the magistrate judge’s recommendation of a\nfour-month prison term with no further supervised release, as\nwell as the SDNY judge’s recommendation of a concurrent\nprison term. Regarding the relationship between the SDNY\nand DDC proceedings, the district court explained: “[T]his\nwas a separate matter and there was a separate criminal\nproceeding here. This is a violation of the terms with respect\nto that separate criminal proceeding.” App. 110.\n\f 5\n II\n\n On appeal, Islam seeks to raise the same delay argument\nthat he pressed unsuccessfully before the magistrate judge.\nHe contends that the lag between his January 2017 arrest and\nhis December 2017 revocation hearing violated both due\nprocess and Rule 32.1(a)(1). The government responds that\nIslam forfeited his right to appeal the district court’s decision\nrejecting these claims by failing to object to the magistrate\njudge’s adverse recommendation. We agree.\n\n A\n\n The Federal Magistrates Act permits a district court to\n“designate a magistrate judge to conduct hearings” on various\ncivil and criminal matters, 28 U.S.C. § 636(b)(1)(B), including\npetitions “to modify, revoke, or terminate supervised release,”\n18 U.S.C. § 3401(i). For such petitions, the magistrate judge\nmust make “proposed findings of fact and recommendations,”\nid., as she must for all other dispositive matters referred to her,\nsee 28 U.S.C. § 636(b)(1)(B); Fed. R. Crim. P. 59(b)(1); Fed.\nR. Civ. P. 72(b)(1). Within 14 days of the recommended\ndisposition of a criminal case, any party “may” file objections,\nFed. R. Crim. P. 59(b)(2), in which case the district court “must\nconsider” the objection “de novo,” id. 59(b)(3). “[F]ailure to\nobject in accordance with this rule waives a party’s right to\nreview.” Id. 59(b)(2). Failure to object also “waives a\nparty’s right to review” of any non-dispositive matter referred\nto a magistrate judge in a criminal case. Id. 59(a).\n\n The Supreme Court explained the rationale for this waiver\nrule in Thomas v. Arn, 474 U.S. 140 (1985), which held that\nthe courts of appeals could make and enforce such a rule\nthrough their supervisory powers. See id. at 155. In this\ncontext as in others, contemporaneous objections help to\n“focus attention” on any genuinely contested issues. Id. at\n\f 6\n147. Moreover, without the waiver rule, litigants could\nsandbag a district court “by failing to object and then\nappealing,” which would either “force the court of appeals to\nconsider claims that were never reviewed by the district court,\nor force the district court to review every issue in every case.”\nId. at 148. Many courts of appeals had adopted the waiver rule\napproved in Thomas, see id. at 146 n.4, and the Supreme Court,\nin 1983, codified it for non-dispositive matters in civil cases,\nsee Fed. R. Civ. P. 72(a). After the Ninth Circuit later refused\nto apply the waiver rule to criminal cases, United States v.\nAbonce-Barrera, 257 F.3d 959, 967–69 (9th Cir. 2001), the\nSupreme Court, in 2005, codified it for criminal cases as well,\nsee Fed. R. Crim. P. 59(a), (b)(2).\n\n The parties briefed this case as if Thomas still controls,\nwhich would require us to determine whether to adopt its\nwaiver rule as an exercise of supervisory power. But as we\nhave explained, Rule 59(b)(2) now governs, and it provides\nthat the “[f]ailure to object” to a magistrate judge’s proposed\nfindings and recommendations on any dispositive matter in a\ncriminal case “waives a party’s right to review.” Because\nIslam did not timely object to the report and recommendation\nin this case, he cannot appeal the district court’s decision\nadopting it.\n\n B\n\n Because Rule 59(b)(2) is not cast in jurisdictional terms,\nthe courts have discretion to excuse a waiver under the rule.\nIslam argues that the district court excused his waiver here and\nthat we should do likewise. We disagree.\n\n As for the district court, Islam highlights one passing\ncomment, made orally at the revocation hearing, that the court\nhad reviewed and agreed with the “analysis and conclusions”\nof the magistrate judge. App. 105. That hardly constitutes a\n\f 7\ndecision to excuse the waiver. To the contrary, in its very next\nbreath, the court “note[d] for the record as well that the parties\nhave not objected to [the] report and recommendation.” Id.\nAnd again, the court later noted that “Mr. Islam is not at this\npoint contesting [the] report and recommendation.” App. 106.\n\n In this Court, Islam presses his claims of unlawful delay\ndespite failing to preserve them below. We review such\nunpreserved claims only for plain error. Puckett v. United\nStates, 556 U.S. 129, 135 (2009). Here, there was no error.\n\n Islam’s due-process argument is meritless. To violate\ndue process, a delay in revocation proceedings must be both\nunreasonable and prejudicial. See Sutherland v. McCall, 709\nF.2d 730, 732 (D.C. Cir. 1983). Islam complains about the\neleven-month delay between his January 2017 arrest in New\nYork and his final revocation hearing in December 2017. But\nthe DDC arrest warrant was not executed until September 6,\n2017. And it is the “execution of the warrant and custody\nunder that warrant” that serves as “the operative event\ntriggering any loss of liberty” for purposes of due process.\nMoody v. Daggett, 429 U.S. 78, 87 (1976). So, the due-\nprocess clock for the DDC revocation hearing did not begin to\nrun until Islam was arrested on the DDC charges in September.\n\n Islam contends that he was arrested in January 2017 for\nviolating both the conditions imposed by the SDNY and those\nimposed by the DDC. He argues that the arrest was made by\nthe probation officer based on probable cause that Islam\nviolated both sets of conditions. But the SDNY record\nindicates that the arrest was made on a warrant issued by the\nSDNY judge for violations of the SDNY conditions. United\nStates v. Islam, No. 12-cr-810 (S.D.N.Y.), ECF Docs. 36, 39,\n46. Thus, the magistrate judge permissibly found that Islam\nwas not arrested on the DDC charges until September, and the\n\f 8\ndistrict court permissibly adopted the finding. Given a\nSeptember arrest date, the December revocation hearing—\npushed back twice by continuances requested by Islam—was\nnot unreasonably delayed. See Howard v. Caufield, 765 F.3d\n1, 3, 13 (D.C. Cir. 2014) (“minor delay” of four months\nbetween arrest and revocation hearing was not unreasonable);\nUnited States v. Madden, 515 F.3d 601, 607 (6th Cir. 2008)\n(delay “caused by [the defendant’s] own conduct” is not\nunreasonable).\n\n Islam’s due-process argument also fails for lack of any\nprejudice. The only harm he alleges is losing the ability to\nseek a prison term in the DDC that would run concurrently with\nthe one imposed by the SDNY. But the district court did\nconsider—and expressly reject—the possibility of a concurrent\npunishment. The court explained that the SDNY and DDC\nproceedings involved separate offenses, that the totality of\nIslam’s misconduct was concerning, and that additional\nincarceration was therefore appropriate. Islam provides no\nreason to think that the district court might have assessed these\nissues differently had it imposed punishment sooner rather than\nlater.\n\n Islam further argues that the delay violated Rule\n32.1(a)(1), which provides that a person arrested for violating\nconditions of supervised release must be brought before a\nmagistrate judge “without unnecessary delay.” Nothing in\nthat sparse text expands due-process protections in this area.\nTo the contrary, the rule simply codifies these protections.\nSee, e.g., United States v. Ruby, 706 F.3d 1221, 1226 (10th Cir.\n2013). Islam’s Rule 32.1 argument thus fares no better than\nhis due-process argument.\n\f 9\n III\n\n In the alternative, Islam seeks a remand for an evidentiary\nhearing on two claims alleging ineffective assistance of counsel\nduring the revocation proceedings. Although we remand\ncolorable claims of ineffective assistance, we have “never held\nthat any claim of ineffective assistance of counsel …\nautomatically entitles a party to an evidentiary remand.”\nUnited States v. Sitzmann, 893 F.3d 811, 831 (D.C. Cir. 2018)\n(per curiam) (quotation marks omitted). When the record\n“clearly shows” that the claim is meritless, or when no further\nfactual development is needed, we may dispose of the claim\nwithout remanding. Id. at 831–32.\n\n As an initial matter, it is unclear whether Islam had any\nright to effective assistance during his revocation proceedings.\nThe Sixth Amendment right to counsel does not attach to\nrevocation hearings. Baker v. Sard, 486 F.2d 415, 423 (D.C.\nCir. 1972). The Due Process Clause does attach, but it\nguarantees counsel only in rare cases; as the Supreme Court has\nexplained, the participation of counsel “will probably be both\nundesirable and constitutionally unnecessary in most\nrevocation hearings.” Gagnon v. Scarpelli, 411 U.S. 778, 790\n(1973). Islam did have a statutory right to counsel, 18 U.S.C.\n§ 3006A(a)(1)(E), but it is unclear whether that encompasses a\nright to effective assistance, see United States v. Eskridge, 445\nF.3d 930, 933 (7th Cir. 2006). Because the government is\nwilling to assume that ineffective-assistance principles apply to\nthis case, and that they parallel ones that govern under the Sixth\nAmendment, we do likewise.\n\n To establish ineffective assistance, a defendant must show\nboth that “counsel’s performance was deficient” and that “the\ndeficient performance prejudiced the defense.” Strickland v.\nWashington, 466 U.S. 668, 687 (1984). Deficient\n\f 10\nperformance must fall “below an objective standard of\nreasonableness.” Id. at 688. Prejudice requires a “reasonable\nprobability” that, but for the deficient performance, “the result\nof the proceeding would have been different.” Id. at 694.\nThe failure to raise a meritless objection is not deficient\nperformance. See Sitzmann, 893 F.3d at 833.\n\n Islam raises two allegations of ineffective assistance.\nFirst, he claims that his counsel was ineffective in failing to\nobject to the magistrate judge’s recommendation to reject the\nclaim of unreasonable delay. As explained above, however,\nthe unreasonable-delay claim lacks merit, and the delay caused\nno prejudice. The waiver of this claim does not amount to\nineffective assistance.\n\n Next, Islam complains that his counsel waited to file a\ntransfer motion until October 2017, when the SDNY\nrevocation proceedings were complete and the DDC\nproceedings were already well underway. Islam contends that\nif counsel had filed the motion in April 2017, when he first\nlearned of the DDC charges, the motion likely would have been\ngranted, and Islam then would have received concurrent rather\nthan consecutive prison terms. But a transfer decision is left\nto the district court’s discretion, 18 U.S.C. § 3605, and there is\nno reasonable probability that the court here would have\nordered a transfer. Again, Islam cannot show prejudice.\n\n Islam’s contrary argument focuses on the magistrate\njudge’s recommendation to reject the October transfer motion\nbecause she had already invested substantial time in the case.\nIslam overlooks the district court’s prior decision rejecting out-\nof-hand the August transfer motion, which was filed before the\nmagistrate judge had begun working on the case. Islam also\noverlooks the district court’s subsequent revocation decision,\nwhich rejected the magistrate judge’s recommendation to cede\n\f 11\nfuture supervision to the SDNY. Moreover, the supervised-\nrelease violations charged in the DDC were broader than those\ncharged in the SDNY, as only the former involved financial-\ndisclosure obligations. And the underlying offenses in the\nDDC, which originally produced a prison sentence of two\nyears, were far more serious than those in the SDNY, which\noriginally produced a prison sentence of one day. Under these\ncircumstances, the likelihood of the district court’s granting an\nearlier motion to transfer would have been slim at best.\nWithout any prejudice, Islam’s second ineffective-assistance\nclaim fares no better than his first.\n\n Affirmed.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: UNITED STATES of America, Appellee v. Mir ISLAM, Appellant\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-02\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued November 19, 2018 Decided August 2, 2019\\n\\n No. 18-3003\\n\\n UNITED STATES OF AMERICA,\\n APPELLEE\\n\\n v.\\n\\n MIR ISLAM,\\n APPELLANT\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:15-cr-00067-1)\\n\\n\\n Jonathan Zucker, appointed by the court, argued the cause\\nand filed the briefs for appellant.\\n\\n Peter S. Smith, Assistant U.S. Attorney, argued the cause\\nfor appellee. With him on the brief were Jessie K. Liu, U.S.\\nAttorney, and Elizabeth Trosman, John P. Mannarino, and\\nJonathan P. Hooks, Assistant U.S. Attorneys.\\n\\n Before: KATSAS, Circuit Judge, and SILBERMAN and\\nWILLIAMS, Senior Circuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge KATSAS.\\n\\f 2\\n KATSAS, Circuit Judge: This appeal addresses the legal\\nconsequences of a criminal defendant’s failure to object to a\\nmagistrate judge’s adverse report and recommendation. We\\nalso consider claims that defense counsel provided ineffective\\nassistance during proceedings to revoke a term of supervised\\nrelease.\\n\\n I\\n\\n Mir Islam received two federal sentences for various\\noffenses. In the Southern District of New York, Islam pleaded\\nguilty to crimes involving credit-card fraud, identity theft, and\\ncomputer hacking. An SDNY judge sentenced him to one day\\nof imprisonment followed by three years of supervised release.\\nIn the District of Columbia, Islam pleaded guilty to crimes\\ninvolving the theft and online publication of personal\\ninformation, conveying false information about the use of\\nexplosives, and cyber-stalking. A DDC judge sentenced him\\nto two years of imprisonment followed by three years of\\nsupervised release. The respective terms of supervised release\\nwere subject to substantially similar conditions.\\n\\n After Islam served both prison sentences, he began\\nconcurrently serving the terms of supervised release. Because\\nIslam was living in New York, the SDNY probation office\\nconducted supervision for both courts.\\n\\n On January 18, 2017, Islam was arrested in New York for\\nviolating the conditions of his supervised release. On January\\n19, the SDNY probation office filed with the SDNY a petition\\nto revoke the supervised release. On April 11, the SDNY\\njudge revoked Islam’s original term of supervised release and\\nimposed two years of imprisonment followed by one year of\\nsupervised release subject to the same conditions previously\\nimposed. She recommended that the imprisonment and\\nsupervised release run concurrently with any further\\n\\f 3\\npunishment that the DDC judge might impose. With credit for\\ntime served, Islam was released from SDNY custody on\\nAugust 17, 2017.\\n\\n In the meantime, the DDC probation office filed its own\\npetition to revoke with the DDC. On February 1, 2017, the\\nDDC judge issued a warrant for Islam’s arrest, which was\\nlodged as a detainer to be executed upon Islam’s release from\\nSDNY custody. On August 14, Islam moved to dismiss the\\ndetainer and to transfer the DDC case to the SDNY. On the\\nsame day, the DDC judge denied the motion in a minute order.\\nUpon his release from SDNY custody, Islam was held under\\nthe detainer and then transferred to the District of Columbia.\\nOn September 6, Islam arrived in the District, was arrested\\nunder the warrant, and appeared before a magistrate judge\\nassigned to his case.\\n\\n The magistrate judge scheduled a revocation hearing for\\nSeptember 15, but Islam sought and received two continuances.\\nOn October 27, Islam moved to dismiss the petition for\\nrevocation. Among other things, he argued that the delay\\nbetween his arrest in New York and revocation proceedings in\\nthe DDC violated both the Due Process Clause of the Fifth\\nAmendment and Federal Rule of Criminal Procedure\\n32.1(a)(1). Islam also sought to transfer the matter to the\\nSDNY. On November 8, the magistrate judge held a hearing\\non these issues.\\n\\n On December 4, the magistrate judge issued a thirty-page\\nreport and recommendation rejecting Islam’s various\\narguments. The magistrate judge proposed finding that Islam\\nhad engaged in unauthorized travel, failed to identify\\ncomputers and other electronic devices to which he had access,\\nfailed to allow monitoring of those devices, failed to attend\\nmental-health counseling, and failed to provide requested\\n\\f 4\\nfinancial information—all in violation of his supervised-\\nrelease conditions. The magistrate judge recommended that\\nthe district court impose four months of imprisonment. She\\nrecommended no further supervision because the SDNY\\nprobation office would be supervising compliance with the\\nconditions imposed in the SDNY, which were “nearly\\nidentical” to those imposed in the DDC. App. 98. The report\\nand recommendation stated that “any party who objects to the\\nproposed findings or recommendations herein must file written\\nobjections within fourteen days” of service. App. 99. It\\nfurther warned that the parties “may waive their right of appeal\\nfrom an order of the District Court adopting such findings and\\nrecommendations” if they “fail to file timely objections.” Id.\\nIslam did not file any objections.\\n\\n The district court held a revocation hearing on December\\n19, 2017. At the outset, the court asked the parties if they\\nobjected to the report and recommendation. Through counsel,\\nIslam replied that he did not. Still, the court went over each\\nviolation and confirmed that Islam had no objections. The\\ncourt adopted the proposed findings that Islam had violated his\\nsupervised-release conditions in five different respects,\\nrevoked his supervised release, and imposed nine months of\\nimprisonment followed by 24 months of supervised release.\\nThe court rejected the magistrate judge’s recommendation of a\\nfour-month prison term with no further supervised release, as\\nwell as the SDNY judge’s recommendation of a concurrent\\nprison term. Regarding the relationship between the SDNY\\nand DDC proceedings, the district court explained: “[T]his\\nwas a separate matter and there was a separate criminal\\nproceeding here. This is a violation of the terms with respect\\nto that separate criminal proceeding.” App. 110.\\n\\f 5\\n II\\n\\n On appeal, Islam seeks to raise the same delay argument\\nthat he pressed unsuccessfully before the magistrate judge.\\nHe contends that the lag between his January 2017 arrest and\\nhis December 2017 revocation hearing violated both due\\nprocess and Rule 32.1(a)(1). The government responds that\\nIslam forfeited his right to appeal the district court’s decision\\nrejecting these claims by failing to object to the magistrate\\njudge’s adverse recommendation. We agree.\\n\\n A\\n\\n The Federal Magistrates Act permits a district court to\\n“designate a magistrate judge to conduct hearings” on various\\ncivil and criminal matters, 28 U.S.C. § 636(b)(1)(B), including\\npetitions “to modify, revoke, or terminate supervised release,”\\n18 U.S.C. § 3401(i). For such petitions, the magistrate judge\\nmust make “proposed findings of fact and recommendations,”\\nid., as she must for all other dispositive matters referred to her,\\nsee 28 U.S.C. § 636(b)(1)(B); Fed. R. Crim. P. 59(b)(1); Fed.\\nR. Civ. P. 72(b)(1). Within 14 days of the recommended\\ndisposition of a criminal case, any party “may” file objections,\\nFed. R. Crim. P. 59(b)(2), in which case the district court “must\\nconsider” the objection “de novo,” id. 59(b)(3). “[F]ailure to\\nobject in accordance with this rule waives a party’s right to\\nreview.” Id. 59(b)(2). Failure to object also “waives a\\nparty’s right to review” of any non-dispositive matter referred\\nto a magistrate judge in a criminal case. Id. 59(a).\\n\\n The Supreme Court explained the rationale for this waiver\\nrule in Thomas v. Arn, 474 U.S. 140 (1985), which held that\\nthe courts of appeals could make and enforce such a rule\\nthrough their supervisory powers. See id. at 155. In this\\ncontext as in others, contemporaneous objections help to\\n“focus attention” on any genuinely contested issues. Id. at\\n\\f 6\\n147. Moreover, without the waiver rule, litigants could\\nsandbag a district court “by failing to object and then\\nappealing,” which would either “force the court of appeals to\\nconsider claims that were never reviewed by the district court,\\nor force the district court to review every issue in every case.”\\nId. at 148. Many courts of appeals had adopted the waiver rule\\napproved in Thomas, see id. at 146 n.4, and the Supreme Court,\\nin 1983, codified it for non-dispositive matters in civil cases,\\nsee Fed. R. Civ. P. 72(a). After the Ninth Circuit later refused\\nto apply the waiver rule to criminal cases, United States v.\\nAbonce-Barrera, 257 F.3d 959, 967–69 (9th Cir. 2001), the\\nSupreme Court, in 2005, codified it for criminal cases as well,\\nsee Fed. R. Crim. P. 59(a), (b)(2).\\n\\n The parties briefed this case as if Thomas still controls,\\nwhich would require us to determine whether to adopt its\\nwaiver rule as an exercise of supervisory power. But as we\\nhave explained, Rule 59(b)(2) now governs, and it provides\\nthat the “[f]ailure to object” to a magistrate judge’s proposed\\nfindings and recommendations on any dispositive matter in a\\ncriminal case “waives a party’s right to review.” Because\\nIslam did not timely object to the report and recommendation\\nin this case, he cannot appeal the district court’s decision\\nadopting it.\\n\\n B\\n\\n Because Rule 59(b)(2) is not cast in jurisdictional terms,\\nthe courts have discretion to excuse a waiver under the rule.\\nIslam argues that the district court excused his waiver here and\\nthat we should do likewise. We disagree.\\n\\n As for the district court, Islam highlights one passing\\ncomment, made orally at the revocation hearing, that the court\\nhad reviewed and agreed with the “analysis and conclusions”\\nof the magistrate judge. App. 105. That hardly constitutes a\\n\\f 7\\ndecision to excuse the waiver. To the contrary, in its very next\\nbreath, the court “note[d] for the record as well that the parties\\nhave not objected to [the] report and recommendation.” Id.\\nAnd again, the court later noted that “Mr. Islam is not at this\\npoint contesting [the] report and recommendation.” App. 106.\\n\\n In this Court, Islam presses his claims of unlawful delay\\ndespite failing to preserve them below. We review such\\nunpreserved claims only for plain error. Puckett v. United\\nStates, 556 U.S. 129, 135 (2009). Here, there was no error.\\n\\n Islam’s due-process argument is meritless. To violate\\ndue process, a delay in revocation proceedings must be both\\nunreasonable and prejudicial. See Sutherland v. McCall, 709\\nF.2d 730, 732 (D.C. Cir. 1983). Islam complains about the\\neleven-month delay between his January 2017 arrest in New\\nYork and his final revocation hearing in December 2017. But\\nthe DDC arrest warrant was not executed until September 6,\\n2017. And it is the “execution of the warrant and custody\\nunder that warrant” that serves as “the operative event\\ntriggering any loss of liberty” for purposes of due process.\\nMoody v. Daggett, 429 U.S. 78, 87 (1976). So, the due-\\nprocess clock for the DDC revocation hearing did not begin to\\nrun until Islam was arrested on the DDC charges in September.\\n\\n Islam contends that he was arrested in January 2017 for\\nviolating both the conditions imposed by the SDNY and those\\nimposed by the DDC. He argues that the arrest was made by\\nthe probation officer based on probable cause that Islam\\nviolated both sets of conditions. But the SDNY record\\nindicates that the arrest was made on a warrant issued by the\\nSDNY judge for violations of the SDNY conditions. United\\nStates v. Islam, No. 12-cr-810 (S.D.N.Y.), ECF Docs. 36, 39,\\n46. Thus, the magistrate judge permissibly found that Islam\\nwas not arrested on the DDC charges until September, and the\\n\\f 8\\ndistrict court permissibly adopted the finding. Given a\\nSeptember arrest date, the December revocation hearing—\\npushed back twice by continuances requested by Islam—was\\nnot unreasonably delayed. See Howard v. Caufield, 765 F.3d\\n1, 3, 13 (D.C. Cir. 2014) (“minor delay” of four months\\nbetween arrest and revocation hearing was not unreasonable);\\nUnited States v. Madden, 515 F.3d 601, 607 (6th Cir. 2008)\\n(delay “caused by [the defendant’s] own conduct” is not\\nunreasonable).\\n\\n Islam’s due-process argument also fails for lack of any\\nprejudice. The only harm he alleges is losing the ability to\\nseek a prison term in the DDC that would run concurrently with\\nthe one imposed by the SDNY. But the district court did\\nconsider—and expressly reject—the possibility of a concurrent\\npunishment. The court explained that the SDNY and DDC\\nproceedings involved separate offenses, that the totality of\\nIslam’s misconduct was concerning, and that additional\\nincarceration was therefore appropriate. Islam provides no\\nreason to think that the district court might have assessed these\\nissues differently had it imposed punishment sooner rather than\\nlater.\\n\\n Islam further argues that the delay violated Rule\\n32.1(a)(1), which provides that a person arrested for violating\\nconditions of supervised release must be brought before a\\nmagistrate judge “without unnecessary delay.” Nothing in\\nthat sparse text expands due-process protections in this area.\\nTo the contrary, the rule simply codifies these protections.\\nSee, e.g., United States v. Ruby, 706 F.3d 1221, 1226 (10th Cir.\\n2013). Islam’s Rule 32.1 argument thus fares no better than\\nhis due-process argument.\\n\\f 9\\n III\\n\\n In the alternative, Islam seeks a remand for an evidentiary\\nhearing on two claims alleging ineffective assistance of counsel\\nduring the revocation proceedings. Although we remand\\ncolorable claims of ineffective assistance, we have “never held\\nthat any claim of ineffective assistance of counsel …\\nautomatically entitles a party to an evidentiary remand.”\\nUnited States v. Sitzmann, 893 F.3d 811, 831 (D.C. Cir. 2018)\\n(per curiam) (quotation marks omitted). When the record\\n“clearly shows” that the claim is meritless, or when no further\\nfactual development is needed, we may dispose of the claim\\nwithout remanding. Id. at 831–32.\\n\\n As an initial matter, it is unclear whether Islam had any\\nright to effective assistance during his revocation proceedings.\\nThe Sixth Amendment right to counsel does not attach to\\nrevocation hearings. Baker v. Sard, 486 F.2d 415, 423 (D.C.\\nCir. 1972). The Due Process Clause does attach, but it\\nguarantees counsel only in rare cases; as the Supreme Court has\\nexplained, the participation of counsel “will probably be both\\nundesirable and constitutionally unnecessary in most\\nrevocation hearings.” Gagnon v. Scarpelli, 411 U.S. 778, 790\\n(1973). Islam did have a statutory right to counsel, 18 U.S.C.\\n§ 3006A(a)(1)(E), but it is unclear whether that encompasses a\\nright to effective assistance, see United States v. Eskridge, 445\\nF.3d 930, 933 (7th Cir. 2006). Because the government is\\nwilling to assume that ineffective-assistance principles apply to\\nthis case, and that they parallel ones that govern under the Sixth\\nAmendment, we do likewise.\\n\\n To establish ineffective assistance, a defendant must show\\nboth that “counsel’s performance was deficient” and that “the\\ndeficient performance prejudiced the defense.” Strickland v.\\nWashington, 466 U.S. 668, 687 (1984). Deficient\\n\\f 10\\nperformance must fall “below an objective standard of\\nreasonableness.” Id. at 688. Prejudice requires a “reasonable\\nprobability” that, but for the deficient performance, “the result\\nof the proceeding would have been different.” Id. at 694.\\nThe failure to raise a meritless objection is not deficient\\nperformance. See Sitzmann, 893 F.3d at 833.\\n\\n Islam raises two allegations of ineffective assistance.\\nFirst, he claims that his counsel was ineffective in failing to\\nobject to the magistrate judge’s recommendation to reject the\\nclaim of unreasonable delay. As explained above, however,\\nthe unreasonable-delay claim lacks merit, and the delay caused\\nno prejudice. The waiver of this claim does not amount to\\nineffective assistance.\\n\\n Next, Islam complains that his counsel waited to file a\\ntransfer motion until October 2017, when the SDNY\\nrevocation proceedings were complete and the DDC\\nproceedings were already well underway. Islam contends that\\nif counsel had filed the motion in April 2017, when he first\\nlearned of the DDC charges, the motion likely would have been\\ngranted, and Islam then would have received concurrent rather\\nthan consecutive prison terms. But a transfer decision is left\\nto the district court’s discretion, 18 U.S.C. § 3605, and there is\\nno reasonable probability that the court here would have\\nordered a transfer. Again, Islam cannot show prejudice.\\n\\n Islam’s contrary argument focuses on the magistrate\\njudge’s recommendation to reject the October transfer motion\\nbecause she had already invested substantial time in the case.\\nIslam overlooks the district court’s prior decision rejecting out-\\nof-hand the August transfer motion, which was filed before the\\nmagistrate judge had begun working on the case. Islam also\\noverlooks the district court’s subsequent revocation decision,\\nwhich rejected the magistrate judge’s recommendation to cede\\n\\f 11\\nfuture supervision to the SDNY. Moreover, the supervised-\\nrelease violations charged in the DDC were broader than those\\ncharged in the SDNY, as only the former involved financial-\\ndisclosure obligations. And the underlying offenses in the\\nDDC, which originally produced a prison sentence of two\\nyears, were far more serious than those in the SDNY, which\\noriginally produced a prison sentence of one day. Under these\\ncircumstances, the likelihood of the district court’s granting an\\nearlier motion to transfer would have been slim at best.\\nWithout any prejudice, Islam’s second ineffective-assistance\\nclaim fares no better than his first.\\n\\n Affirmed.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4168,"cells":{"id":{"kind":"number","value":4645249,"string":"4,645,249"},"name":{"kind":"string","value":"EXHAUSTLESS INC., Petitioner v. FEDERAL AVIATION ADMINISTRATION, Respondent"},"name_abbreviation":{"kind":"string","value":"Exhaustless Inc. v. Fed. Aviation Admin."},"decision_date":{"kind":"string","value":"2019-08-02"},"docket_number":{"kind":"string","value":"18-1303; C/w 18-1304"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Henderson, Srinivasan, Pillard\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge SRINIVASAN\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued May 13, 2019 Decided August 2, 2019\\n\\n No. 18-1303\\n\\n EXHAUSTLESS INC.,\\n PETITIONER\\n\\n v.\\n\\n FEDERAL AVIATION ADMINISTRATION,\\n RESPONDENT\\n\\n\\n Consolidated with 18-1304\\n\\n\\n On Petitions for Review of Orders\\n of the Federal Aviation Administration\\n\\n\\n Kevin M. Blair argued the cause and filed the briefs for\\npetitioner.\\n\\n Benjamin M. Shultz, Attorney, U.S. Department of Justice,\\nargued the cause for respondent. With him on the brief were\\nMichael S. Raab, Attorney, U.S. Department of Justice, and\\nSteven G. Bradbury, General Counsel, Paul M. Geier,\\nAssistant General Counsel, Joy K. Park, Senior Trial Attorney,\\nand Arjun Garg, Chief Counsel, Federal Aviation\\nAdministration.\\n\\f 2\\n Before: HENDERSON, SRINIVASAN, and PILLARD, Circuit\\nJudges.\\n\\n Opinion for the Court filed by Circuit Judge SRINIVASAN.\\n\\n SRINIVASAN, Circuit Judge: In 1968, the Federal\\nAviation Administration began limiting the number of flights\\nserving LaGuardia and John F. Kennedy Airports in New York\\nin order to reduce flight delays. In 2000, Congress mandated\\nthe repeal of the relevant regulations based on concerns about\\ntheir anticompetitive effects. The phase-out process, however,\\ncaused flight delays to skyrocket at LaGuardia and JFK\\nAirports. The FAA then issued interim orders again limiting\\nthe number of flights serving those airports. The FAA has\\nsince extended the interim orders many times as efforts to\\nestablish a permanent solution have failed.\\n\\n Exhaustless, Inc., brings two petitions for review of the\\nlatest interim extension orders. Exhaustless would like\\nLaGuardia and JFK Airports to implement the company’s\\npatent-pending product, Aviation 2.0 Operating System, to\\nmanage the allocation of takeoff and landing “slots” to airlines.\\n\\n We dismiss Exhaustless’s petitions for lack of standing.\\nThe company fails to demonstrate that vacating the interim\\nFAA orders would redress its injury—i.e., a lack of market\\nopportunity for its product. Vacating the interim orders would\\nleave takeoffs and landings at the airports unregulated,\\neliminating the need for the company’s product at the federal\\nlevel. To the extent Exhaustless argues that the local airport\\nauthority could employ Aviation 2.0 if there were no federal\\nregulation, we find any such possibility too speculative to\\nsupport the company’s standing to bring these petitions.\\n\\f 3\\n I.\\n\\n The Federal Aviation Act calls for the FAA to “assign by\\nregulation or order the use of the airspace necessary to ensure\\nthe safety of aircraft and the efficient use of airspace.” 49\\nU.S.C. § 40103(b)(1). Navigable airspace includes the\\n“airspace needed to ensure safety in the takeoff and landing of\\naircraft.” Id. § 40102(a)(32).\\n\\n Since 1968, the FAA has restricted the number of takeoffs\\nand landings at certain highly congested airports in order to\\nreduce inefficient flight delays. The restrictions were codified\\nin a series of regulations known as the High Density Rule. As\\nof 2000, the rule placed numerical limits on the hourly takeoffs\\nand landings at five highly congested airports: Newark\\nLiberty, LaGuardia, JFK, O’Hare, and Ronald Reagan\\nWashington National. 14 C.F.R. § 93.123 (2000).\\n\\n By then, Congress had grown concerned with the High\\nDensity Rule’s collateral effects on airport access for carriers\\nand competition among carriers. Acting on those concerns in\\n2000, Congress prohibited the use of the High Density Rule at\\nLaGuardia or JFK Airports after January 1, 2007. 49 U.S.C.\\n§ 41715(a). For the period leading up to that date, Congress\\ndirected the FAA to grant slot exemptions for carriers servicing\\nsmaller airports and carriers with little or no existing service at\\nthe airports. Id. § 41716.\\n\\n Congress’s action led to an immediate increase in airport\\ncongestion at LaGuardia. As the FAA began granting slot\\nexemptions, “the number of scheduled flight operations at\\nLaGuardia began to far exceed the airport’s capacity even\\nunder optimal operating conditions.” 71 Fed. Reg. 54,331,\\n54,331 (Sept. 14, 2006). The average minutes of delay for\\narriving flights increased 144% between March and September\\n\\f 4\\nof 2000. Id. at 54,332. By September 2000, flight delays at\\nLaGuardia accounted for 25% of the delays nationwide. Id.\\nThe FAA responded by limiting the number of slot exemptions.\\nFrom late 2000 until the end of 2006, the High Density Rule,\\nwith the exemption cap, governed the number of slots at\\nLaGuardia. Id. at 54,332 & n.9.\\n\\n Because the High Density Rule was set to expire by 2007,\\nthe FAA, in August 2006, proposed a new permanent\\ncongestion management rule for LaGuardia and requested\\ncomments. 71 Fed. Reg. 51,360 (Aug. 29, 2006). A few weeks\\nlater, the agency explained that the permanent rule would not\\nbe finalized by the end of the year and that it was necessary to\\nimplement an interim rule to avert crippling delays. 71 Fed.\\nReg. 54,331 (Sept. 14, 2006).\\n\\n The FAA issued an interim order in December 2006. 71\\nFed. Reg. 77,854 (Dec. 27, 2006). The rule made clear that it\\nwas a temporary measure and reiterated the agency’s “need to\\ncomplete the rulemaking, because the final decision in that\\nproceeding should establish a more rational basis for the\\nregulation of flight operations at LaGuardia.” Id. at 77,856.\\nThe interim rule resembled the High Density Rule and\\ngenerally grandfathered the slots held by airlines under the\\nprevious regime. Id. at 77,859–61.\\n\\n Regulatory efforts concerning JFK Airport followed a\\nsomewhat different path but ended in much the same place.\\nWith respect to JFK, the FAA allowed the High Density Rule\\nto expire in 2007 without a replacement. Unsurprisingly, the\\nnumber of flights at JFK spiked, and with more planes came\\nmore delays. In 2007, the average daily operations at JFK\\nincreased 21% over the prior year and on-time arrival rates\\ndeclined from 68.5% to 62.2%. 73 Fed. Reg. 3510, 3511 (Jan.\\n18, 2008).\\n\\f 5\\n In 2008, the FAA published an interim order limiting the\\nnumber of takeoffs and landings at JFK. Id. at 3516–42. Like\\nthe LaGuardia order, the JFK order stressed its “short-term\\nnature,” stating that it was “not intended to create a long-term\\nsolution to congestion.” Id. at 3513–14.\\n\\n The FAA’s first attempt at a permanent solution for\\nLaGuardia and JFK Airports came via rules promulgated in\\nOctober 2008. 73 Fed. Reg. 60,574 (Oct. 10, 2008); 73 Fed.\\nReg. 60,544 (Oct. 10, 2008). Under those rules, a portion of\\nthe slots would be allocated using an auction. 73 Fed. Reg. at\\n60,577; 73 Fed. Reg. at 60,547. A number of airlines and trade\\ngroups, along with the local airport authority, promptly\\nchallenged the rules in our court. The challengers moved for\\nan immediate stay, contending that the FAA lacked statutory\\nauthority to conduct slot auctions. We granted the motion.\\nOrder, Port Auth. of N.Y. & N.J. v. FAA, No. 08-1329 (D.C.\\nCir. Dec. 8, 2008), J.A. 365. The FAA then rescinded the rules.\\n74 Fed. Reg. 52,134 (Oct. 9, 2009); 74 Fed. Reg. 52,132 (Oct.\\n9, 2009).\\n\\n The agency extended the interim orders for both\\nLaGuardia and JFK Airports until October 2011, noting that a\\npermanent solution would require more time. 74 Fed. Reg.\\n51,653 (Oct. 7, 2009); 74 Fed. Reg. 51,650 (Oct. 7, 2009). A\\nseries of additional extensions followed. In April 2011, the\\nFAA extended the orders until October 2013. 76 Fed. Reg.\\n18,620 (Apr. 4, 2011); 76 Fed. Reg. 18,616 (Apr. 4, 2011). In\\nMay 2013, the FAA extended the orders to October 2014. 78\\nFed. Reg. 28,278 (May 14, 2013); 78 Fed. Reg. 28,276 (May\\n14, 2013). And in March 2014, the FAA again extended the\\norders, to October 2016. 79 Fed. Reg. 17,222 (Mar. 27, 2014);\\n79 Fed. Reg. 16,854 (Mar. 26, 2014).\\n\\f 6\\n In January 2015, the agency proposed a final rule for New\\nYork–area airports that included a secondary market for the\\npurchase, sale, lease, or trade of slots between airlines. 80 Fed.\\nReg. 1274 (Jan. 8, 2015). But in May 2016, after receiving\\ncomments, the FAA withdrew the proposed rule. 81 Fed. Reg.\\n30,218 (May 16, 2016). That led the agency to extend the\\ninterim orders yet again, until October 2018. 81 Fed. Reg.\\n33,126 (May 25, 2016); 81 Fed. Reg. 32,636 (May 24, 2016).\\n\\n In September 2018, the FAA once more extended the\\ninterim orders for LaGuardia and JFK Airports, this time until\\nOctober 2020. 83 Fed. Reg. 47,065 (Sept. 18, 2018); 83 Fed.\\nReg. 46,865 (Sept. 17, 2018). Those latest extensions are at\\nissue here. While the orders largely match the prior extensions\\nin substance, they are less committal about a permanent rule,\\nstating only that the agency “will continue to consider potential\\nrulemaking in the future to codify the slot management policies\\nat [LaGuardia], and also at John F. Kennedy International\\nAirport (JFK).” 83 Fed. Reg. at 47,065.\\n\\n Petitioner Exhaustless, Inc., as noted, has developed a\\npatent-pending product called Aviation 2.0 Operating Standard\\nfor allocating airline slots at airports. Using Aviation 2.0,\\ncarriers would compete in semi-annual auctions to purchase\\nslots for a six-month period, with the total number of slots\\ndetermined by Exhaustless using its proprietary technology.\\nPassengers would then pay demand-calibrated congestion\\npremiums (on top of their airfare) when purchasing tickets.\\nBoth the congestion premiums and the auction proceeds would\\ngo to Exhaustless.\\n\\n II.\\n\\n Exhaustless asserts several challenges to the latest interim\\nextension orders in its petitions for review, including\\n\\f 7\\narguments that the FAA exceeded its statutory authority and\\nviolated the Administrative Procedure Act. We cannot address\\nthe merits of those claims unless Exhaustless has constitutional\\nstanding. See Steel Co. v. Citizens for a Better Env’t, 523 U.S.\\n83, 101–02 (1998). To establish standing, Exhaustless must\\ndemonstrate that: (i) it has suffered an injury-in-fact that is\\n“concrete and particularized” and “actual or imminent”; (ii) the\\ninjury is “fairly traceable to the challenged action” of the\\nrespondent; and (iii) it is “likely, as opposed to merely\\nspeculative, that the injury will be redressed by a favorable\\ndecision.” Lujan v. Defs. of Wildlife, 504 U.S. 555, 560–61\\n(1992) (formatting modified).\\n\\n The latter two elements, traceability and redressability,\\n“overlap as two sides of a causation coin.” Dynalantic Corp.\\nv. Dep’t of Def., 115 F.3d 1012, 1017 (D.C. Cir. 1997). When\\na petitioner itself is the object of the challenged agency action,\\nthere usually is little doubt of causation. See Lujan, 504 U.S.\\nat 561–62. But when a petitioner’s injury arises from an\\nagency’s “unlawful regulation (or lack of regulation) of\\nsomeone else,” causation often is “substantially more difficult”\\nto establish because the petitioner must demonstrate that the\\ninjury does not result from “the independent action of some\\nthird party not before the court.” Id. at 560, 562.\\n\\n Exhaustless fails to demonstrate redressability. The\\ncompany contends that it “is being deprived of the opportunity\\nto compete in the market with its patent-pending proposed\\nsolution” Aviation 2.0 so long as the FAA’s interim orders\\nremain in place. Exhaustless Br. 25; id. at 27. Exhaustless\\nenvisions that, if the interim orders—including their limitations\\non slots at LaGuardia and JFK—were withdrawn, the agency\\nwould then “transfer the management of service (slot volumes),\\nfor congestion-prevention purposes, to Exhaustless.”\\nExhaustless Inc., Aviation 2.0—Explained, FAA Add. A39.\\n\\f 8\\n Vacatur of the interim orders, however, would not get the\\ncompany closer to that goal. Without the orders, there would\\nbe no federally mandated number of takeoff and landing slots\\nat LaGuardia and JFK Airports, no scarce resource for the FAA\\nto auction, and hence no market for Exhaustless’s product, at\\nleast as concerns the federal government. The relief sought by\\nthe company in its petitions for review—i.e., vacatur of the\\ninterim orders—thus would not redress its claimed injury.\\n\\n The appropriate administrative channel for Exhaustless to\\npursue instead is a petition for rulemaking with the agency to\\nemploy the company’s technology. And because the interim\\nrules are revocable at will, the orders challenged by\\nExhaustless in this proceeding do not stand in the way of the\\ncompany’s attempting to persuade the FAA to adopt its\\ntechnology via a rulemaking. Indeed, Exhaustless has already\\nfiled such a petition for rulemaking. Petition for Rulemaking—\\nFAA-2007-0001 (filed May 21, 2018), J.A. 499. The petition\\nremains pending with the FAA, and, should the agency reject\\nit, Exhaustless would have standing to seek judicial review.\\n\\n At oral argument, Exhaustless submitted that vacatur of\\nthe FAA’s interim rules would create a different market\\nopportunity: vacatur in theory would result in transfer of\\ncontrol over flight schedules at LaGuardia and JFK Airports\\nfrom federal to local authority, i.e., the Port Authority of New\\nYork and New Jersey, which could then elect to use Aviation\\n2.0. That argument for standing fares no better.\\n\\n While the loss of an opportunity to compete for business\\ncan constitute Article III injury, there must be a “realistic\\npossibility” of winning the eventual competition. Ranger\\nCellular v. FCC, 348 F.3d 1044, 1050 (D.C. Cir. 2003)\\n(quoting Albuquerque Indian Rights v. Lujan, 930 F.2d 49, 56\\n(D.C. Cir. 1991)). Here, the hurdles to the Port Authority’s\\n\\f 9\\nadoption of Aviation 2.0 to manage flights at the airports are\\ntoo significant and too numerous for us to find it “likely, as\\nopposed to merely speculative,” that vacatur of the interim\\norders would redress Exhaustless’s injury. Lujan, 504 U.S. at\\n561 (internal quotation marks omitted).\\n\\n First, the FAA operates under a duty “to ensure . . . the\\nefficient use of airspace.” 49 U.S.C. § 40103(b)(1). Since\\n1968, the Administration has fulfilled that responsibility by\\nlimiting the number of takeoffs and landings at LaGuardia and\\nJFK because the airports cannot accommodate the number of\\nflights airlines would like to operate there without causing\\nundue congestion. See 33 Fed. Reg. 17,896 (Dec. 3, 1968).\\nThe Chief Operating Officer of the FAA’s Air Traffic\\nAdministration testified that, absent the interim orders, she\\nwould expect “demand for additional flights . . . to far exceed\\nthe runway capacity resulting in extensive localized and\\nsystemic delays and flight cancellations.” Bristol Decl. ¶ 4,\\nFAA Add. A28. Recall that, in 2000, merely allowing\\nexemptions from slot limitations caused LaGuardia to account\\nfor 25% of flight delays nationwide. 71 Fed. Reg. at 54,332.\\nExhaustless’s suggestion that our vacating the interim orders\\nwould lead the FAA to delegate authority over flight schedules\\nat LaGuardia and JFK Airports to the Port Authority defies\\nhistory and blinks reality.\\n\\n Second, both LaGuardia and JFK Airports have accepted\\nfederal grants for airport development under the Airport\\nImprovement Program. Federal Aviation Administration, FY\\n2019 Primary Entitlements (May 10, 2019),\\nhttps://www.faa.gov/airports/aip/grantapportion_data/media/F\\nY-2019-Primary-Entitlements.pdf. As a condition of accessing\\nthose funds, the airports must pledge that they will be\\n“available for public use on reasonable conditions and without\\nunjust discrimination.” 49 U.S.C. § 47107(a)(1). That\\n\\f 10\\nassurance prohibits airlines from assessing unreasonable fees.\\nSee Air Transp. Ass’n of Am. v. DOT, 613 F.3d 206, 210 (D.C.\\nCir. 2010). And Congress has assigned the Secretary of\\nTransportation primary responsibility for determining whether\\nairport fees are reasonable. See 49 U.S.C. § 47129. Under\\ncurrent regulations, airports may charge landing fees so long as\\nthey do not exceed the historical costs captured by an airport’s\\n“rate base.” 78 Fed. Reg. 55,330, 55,333–35 (Sept. 10, 2013);\\nAir Transp., 613 F.3d at 211. But Exhaustless’s technology\\nrelies on charging carriers a market-clearing auction price\\nrather than a cost-based landing fee, and in doing so runs into\\nconflict with JFK’s and LaGuardia’s grant assurances.\\n\\n Third, there are substantial obstacles to charging\\npassengers a “dedicated Congestion-Prevention Premium,” as\\nExhaustless contemplates. Exhaustless Inc., Aviation 2.0—\\nExplained, FAA Add. A39. To the extent the company\\nenvisions that the local airport authority would assess the\\npremium, it fails to account for the Anti-Head Tax Act, which\\nprovides that any “political subdivision of a State . . . may not\\nlevy or collect a tax, fee, head charge, or other charge on . . . an\\nindividual traveling in air commerce.” 49 U.S.C. § 40116(b).\\nTo the extent the company contemplates collection of the fee\\nby the airlines, the FAA’s Director of the Office of Aviation\\nAnalysis explained that Exhaustless’s “proposal would require\\nthe carriers to substantially redesign their technology to\\nintegrate a dynamic third-party fee” and that it was “unlikely\\nthat carriers would do this voluntarily.” Homan Decl. ¶ 4, FAA\\nAdd. A56. And if the companies decline to collect the fee\\nvoluntarily, Exhaustless does not explain the airports’ authority\\nto compel airlines to assess the charge.\\n\\n Fourth, adopting Exhaustless’s proposal could jeopardize\\nthe United States’ compliance with international agreements on\\ncommercial air travel. For instance, an agreement between the\\n\\f 11\\nUnited States and Canada guarantees Canadian airlines a\\nminimum of 42 slots at LaGuardia. See Air Transport\\nAgreement Between the Government of the United States and\\nthe Government of Canada, T.I.A.S. No. 07-312, Ann. II § 1\\n(Mar. 12, 2007). Yet Exhaustless evidently seeks to auction\\noff all the slots at LaGuardia without regard to a carrier’s\\nnationality.\\n\\n Finally, even if Exhaustless were able to overcome each of\\nthose hurdles, Aviation 2.0 remains an unproven product. The\\nproduct has yet to be adopted by any airport in the nation even\\nthough the vast majority of them are not subject to FAA slot\\nregulation and thus could adopt it today. We find it doubtful\\nthat two of the busiest airports in the nation would volunteer to\\nact as the test sites for Aviation 2.0, even assuming the agency\\nwould permit them to do so. In view of all of those legal and\\npractical obstacles, the notion that vacating the interim orders\\nwould create a business opportunity for Exhaustless amounts\\nto mere conjecture.\\n\\n Our conclusion in that regard does not mean that the\\nchallenged interim orders are entirely insulated from review.\\nStanding principles under Article III exist to ensure that a\\nlitigant alleges “such a personal stake in the outcome of the\\ncontroversy as to warrant his invocation of federal-court\\njurisdiction.” Summers v. Earth Island Inst., 555 U.S. 488, 493\\n(2009) (internal quotation marks omitted). An airline or airport\\nauthority likely would have standing to petition for review of\\nthe orders. But no member of the regulated community has\\njoined Exhaustless’s challenge or (as far as we know) filed its\\nown petition for review, suggesting a form of acceptance\\namong the parties having the most direct stake. Article III\\ndenies us any license to disrupt that evident acceptance today.\\n\\f 12\\n * * * * *\\n\\n For the foregoing reasons, we dismiss Exhaustless’s\\npetitions for review for lack of jurisdiction.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4422502/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge SRINIVASAN\"}]}"},"judge1":{"kind":"string","value":"HENDERSON"},"judge2":{"kind":"string","value":"SRINIVASAN"},"judge3":{"kind":"string","value":"PILLARD"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4645249/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: EXHAUSTLESS INC., Petitioner v. FEDERAL AVIATION ADMINISTRATION, Respondent\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-02\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued May 13, 2019 Decided August 2, 2019\n\n No. 18-1303\n\n EXHAUSTLESS INC.,\n PETITIONER\n\n v.\n\n FEDERAL AVIATION ADMINISTRATION,\n RESPONDENT\n\n\n Consolidated with 18-1304\n\n\n On Petitions for Review of Orders\n of the Federal Aviation Administration\n\n\n Kevin M. Blair argued the cause and filed the briefs for\npetitioner.\n\n Benjamin M. Shultz, Attorney, U.S. Department of Justice,\nargued the cause for respondent. With him on the brief were\nMichael S. Raab, Attorney, U.S. Department of Justice, and\nSteven G. Bradbury, General Counsel, Paul M. Geier,\nAssistant General Counsel, Joy K. Park, Senior Trial Attorney,\nand Arjun Garg, Chief Counsel, Federal Aviation\nAdministration.\n\f 2\n Before: HENDERSON, SRINIVASAN, and PILLARD, Circuit\nJudges.\n\n Opinion for the Court filed by Circuit Judge SRINIVASAN.\n\n SRINIVASAN, Circuit Judge: In 1968, the Federal\nAviation Administration began limiting the number of flights\nserving LaGuardia and John F. Kennedy Airports in New York\nin order to reduce flight delays. In 2000, Congress mandated\nthe repeal of the relevant regulations based on concerns about\ntheir anticompetitive effects. The phase-out process, however,\ncaused flight delays to skyrocket at LaGuardia and JFK\nAirports. The FAA then issued interim orders again limiting\nthe number of flights serving those airports. The FAA has\nsince extended the interim orders many times as efforts to\nestablish a permanent solution have failed.\n\n Exhaustless, Inc., brings two petitions for review of the\nlatest interim extension orders. Exhaustless would like\nLaGuardia and JFK Airports to implement the company’s\npatent-pending product, Aviation 2.0 Operating System, to\nmanage the allocation of takeoff and landing “slots” to airlines.\n\n We dismiss Exhaustless’s petitions for lack of standing.\nThe company fails to demonstrate that vacating the interim\nFAA orders would redress its injury—i.e., a lack of market\nopportunity for its product. Vacating the interim orders would\nleave takeoffs and landings at the airports unregulated,\neliminating the need for the company’s product at the federal\nlevel. To the extent Exhaustless argues that the local airport\nauthority could employ Aviation 2.0 if there were no federal\nregulation, we find any such possibility too speculative to\nsupport the company’s standing to bring these petitions.\n\f 3\n I.\n\n The Federal Aviation Act calls for the FAA to “assign by\nregulation or order the use of the airspace necessary to ensure\nthe safety of aircraft and the efficient use of airspace.” 49\nU.S.C. § 40103(b)(1). Navigable airspace includes the\n“airspace needed to ensure safety in the takeoff and landing of\naircraft.” Id. § 40102(a)(32).\n\n Since 1968, the FAA has restricted the number of takeoffs\nand landings at certain highly congested airports in order to\nreduce inefficient flight delays. The restrictions were codified\nin a series of regulations known as the High Density Rule. As\nof 2000, the rule placed numerical limits on the hourly takeoffs\nand landings at five highly congested airports: Newark\nLiberty, LaGuardia, JFK, O’Hare, and Ronald Reagan\nWashington National. 14 C.F.R. § 93.123 (2000).\n\n By then, Congress had grown concerned with the High\nDensity Rule’s collateral effects on airport access for carriers\nand competition among carriers. Acting on those concerns in\n2000, Congress prohibited the use of the High Density Rule at\nLaGuardia or JFK Airports after January 1, 2007. 49 U.S.C.\n§ 41715(a). For the period leading up to that date, Congress\ndirected the FAA to grant slot exemptions for carriers servicing\nsmaller airports and carriers with little or no existing service at\nthe airports. Id. § 41716.\n\n Congress’s action led to an immediate increase in airport\ncongestion at LaGuardia. As the FAA began granting slot\nexemptions, “the number of scheduled flight operations at\nLaGuardia began to far exceed the airport’s capacity even\nunder optimal operating conditions.” 71 Fed. Reg. 54,331,\n54,331 (Sept. 14, 2006). The average minutes of delay for\narriving flights increased 144% between March and September\n\f 4\nof 2000. Id. at 54,332. By September 2000, flight delays at\nLaGuardia accounted for 25% of the delays nationwide. Id.\nThe FAA responded by limiting the number of slot exemptions.\nFrom late 2000 until the end of 2006, the High Density Rule,\nwith the exemption cap, governed the number of slots at\nLaGuardia. Id. at 54,332 & n.9.\n\n Because the High Density Rule was set to expire by 2007,\nthe FAA, in August 2006, proposed a new permanent\ncongestion management rule for LaGuardia and requested\ncomments. 71 Fed. Reg. 51,360 (Aug. 29, 2006). A few weeks\nlater, the agency explained that the permanent rule would not\nbe finalized by the end of the year and that it was necessary to\nimplement an interim rule to avert crippling delays. 71 Fed.\nReg. 54,331 (Sept. 14, 2006).\n\n The FAA issued an interim order in December 2006. 71\nFed. Reg. 77,854 (Dec. 27, 2006). The rule made clear that it\nwas a temporary measure and reiterated the agency’s “need to\ncomplete the rulemaking, because the final decision in that\nproceeding should establish a more rational basis for the\nregulation of flight operations at LaGuardia.” Id. at 77,856.\nThe interim rule resembled the High Density Rule and\ngenerally grandfathered the slots held by airlines under the\nprevious regime. Id. at 77,859–61.\n\n Regulatory efforts concerning JFK Airport followed a\nsomewhat different path but ended in much the same place.\nWith respect to JFK, the FAA allowed the High Density Rule\nto expire in 2007 without a replacement. Unsurprisingly, the\nnumber of flights at JFK spiked, and with more planes came\nmore delays. In 2007, the average daily operations at JFK\nincreased 21% over the prior year and on-time arrival rates\ndeclined from 68.5% to 62.2%. 73 Fed. Reg. 3510, 3511 (Jan.\n18, 2008).\n\f 5\n In 2008, the FAA published an interim order limiting the\nnumber of takeoffs and landings at JFK. Id. at 3516–42. Like\nthe LaGuardia order, the JFK order stressed its “short-term\nnature,” stating that it was “not intended to create a long-term\nsolution to congestion.” Id. at 3513–14.\n\n The FAA’s first attempt at a permanent solution for\nLaGuardia and JFK Airports came via rules promulgated in\nOctober 2008. 73 Fed. Reg. 60,574 (Oct. 10, 2008); 73 Fed.\nReg. 60,544 (Oct. 10, 2008). Under those rules, a portion of\nthe slots would be allocated using an auction. 73 Fed. Reg. at\n60,577; 73 Fed. Reg. at 60,547. A number of airlines and trade\ngroups, along with the local airport authority, promptly\nchallenged the rules in our court. The challengers moved for\nan immediate stay, contending that the FAA lacked statutory\nauthority to conduct slot auctions. We granted the motion.\nOrder, Port Auth. of N.Y. & N.J. v. FAA, No. 08-1329 (D.C.\nCir. Dec. 8, 2008), J.A. 365. The FAA then rescinded the rules.\n74 Fed. Reg. 52,134 (Oct. 9, 2009); 74 Fed. Reg. 52,132 (Oct.\n9, 2009).\n\n The agency extended the interim orders for both\nLaGuardia and JFK Airports until October 2011, noting that a\npermanent solution would require more time. 74 Fed. Reg.\n51,653 (Oct. 7, 2009); 74 Fed. Reg. 51,650 (Oct. 7, 2009). A\nseries of additional extensions followed. In April 2011, the\nFAA extended the orders until October 2013. 76 Fed. Reg.\n18,620 (Apr. 4, 2011); 76 Fed. Reg. 18,616 (Apr. 4, 2011). In\nMay 2013, the FAA extended the orders to October 2014. 78\nFed. Reg. 28,278 (May 14, 2013); 78 Fed. Reg. 28,276 (May\n14, 2013). And in March 2014, the FAA again extended the\norders, to October 2016. 79 Fed. Reg. 17,222 (Mar. 27, 2014);\n79 Fed. Reg. 16,854 (Mar. 26, 2014).\n\f 6\n In January 2015, the agency proposed a final rule for New\nYork–area airports that included a secondary market for the\npurchase, sale, lease, or trade of slots between airlines. 80 Fed.\nReg. 1274 (Jan. 8, 2015). But in May 2016, after receiving\ncomments, the FAA withdrew the proposed rule. 81 Fed. Reg.\n30,218 (May 16, 2016). That led the agency to extend the\ninterim orders yet again, until October 2018. 81 Fed. Reg.\n33,126 (May 25, 2016); 81 Fed. Reg. 32,636 (May 24, 2016).\n\n In September 2018, the FAA once more extended the\ninterim orders for LaGuardia and JFK Airports, this time until\nOctober 2020. 83 Fed. Reg. 47,065 (Sept. 18, 2018); 83 Fed.\nReg. 46,865 (Sept. 17, 2018). Those latest extensions are at\nissue here. While the orders largely match the prior extensions\nin substance, they are less committal about a permanent rule,\nstating only that the agency “will continue to consider potential\nrulemaking in the future to codify the slot management policies\nat [LaGuardia], and also at John F. Kennedy International\nAirport (JFK).” 83 Fed. Reg. at 47,065.\n\n Petitioner Exhaustless, Inc., as noted, has developed a\npatent-pending product called Aviation 2.0 Operating Standard\nfor allocating airline slots at airports. Using Aviation 2.0,\ncarriers would compete in semi-annual auctions to purchase\nslots for a six-month period, with the total number of slots\ndetermined by Exhaustless using its proprietary technology.\nPassengers would then pay demand-calibrated congestion\npremiums (on top of their airfare) when purchasing tickets.\nBoth the congestion premiums and the auction proceeds would\ngo to Exhaustless.\n\n II.\n\n Exhaustless asserts several challenges to the latest interim\nextension orders in its petitions for review, including\n\f 7\narguments that the FAA exceeded its statutory authority and\nviolated the Administrative Procedure Act. We cannot address\nthe merits of those claims unless Exhaustless has constitutional\nstanding. See Steel Co. v. Citizens for a Better Env’t, 523 U.S.\n83, 101–02 (1998). To establish standing, Exhaustless must\ndemonstrate that: (i) it has suffered an injury-in-fact that is\n“concrete and particularized” and “actual or imminent”; (ii) the\ninjury is “fairly traceable to the challenged action” of the\nrespondent; and (iii) it is “likely, as opposed to merely\nspeculative, that the injury will be redressed by a favorable\ndecision.” Lujan v. Defs. of Wildlife, 504 U.S. 555, 560–61\n(1992) (formatting modified).\n\n The latter two elements, traceability and redressability,\n“overlap as two sides of a causation coin.” Dynalantic Corp.\nv. Dep’t of Def., 115 F.3d 1012, 1017 (D.C. Cir. 1997). When\na petitioner itself is the object of the challenged agency action,\nthere usually is little doubt of causation. See Lujan, 504 U.S.\nat 561–62. But when a petitioner’s injury arises from an\nagency’s “unlawful regulation (or lack of regulation) of\nsomeone else,” causation often is “substantially more difficult”\nto establish because the petitioner must demonstrate that the\ninjury does not result from “the independent action of some\nthird party not before the court.” Id. at 560, 562.\n\n Exhaustless fails to demonstrate redressability. The\ncompany contends that it “is being deprived of the opportunity\nto compete in the market with its patent-pending proposed\nsolution” Aviation 2.0 so long as the FAA’s interim orders\nremain in place. Exhaustless Br. 25; id. at 27. Exhaustless\nenvisions that, if the interim orders—including their limitations\non slots at LaGuardia and JFK—were withdrawn, the agency\nwould then “transfer the management of service (slot volumes),\nfor congestion-prevention purposes, to Exhaustless.”\nExhaustless Inc., Aviation 2.0—Explained, FAA Add. A39.\n\f 8\n Vacatur of the interim orders, however, would not get the\ncompany closer to that goal. Without the orders, there would\nbe no federally mandated number of takeoff and landing slots\nat LaGuardia and JFK Airports, no scarce resource for the FAA\nto auction, and hence no market for Exhaustless’s product, at\nleast as concerns the federal government. The relief sought by\nthe company in its petitions for review—i.e., vacatur of the\ninterim orders—thus would not redress its claimed injury.\n\n The appropriate administrative channel for Exhaustless to\npursue instead is a petition for rulemaking with the agency to\nemploy the company’s technology. And because the interim\nrules are revocable at will, the orders challenged by\nExhaustless in this proceeding do not stand in the way of the\ncompany’s attempting to persuade the FAA to adopt its\ntechnology via a rulemaking. Indeed, Exhaustless has already\nfiled such a petition for rulemaking. Petition for Rulemaking—\nFAA-2007-0001 (filed May 21, 2018), J.A. 499. The petition\nremains pending with the FAA, and, should the agency reject\nit, Exhaustless would have standing to seek judicial review.\n\n At oral argument, Exhaustless submitted that vacatur of\nthe FAA’s interim rules would create a different market\nopportunity: vacatur in theory would result in transfer of\ncontrol over flight schedules at LaGuardia and JFK Airports\nfrom federal to local authority, i.e., the Port Authority of New\nYork and New Jersey, which could then elect to use Aviation\n2.0. That argument for standing fares no better.\n\n While the loss of an opportunity to compete for business\ncan constitute Article III injury, there must be a “realistic\npossibility” of winning the eventual competition. Ranger\nCellular v. FCC, 348 F.3d 1044, 1050 (D.C. Cir. 2003)\n(quoting Albuquerque Indian Rights v. Lujan, 930 F.2d 49, 56\n(D.C. Cir. 1991)). Here, the hurdles to the Port Authority’s\n\f 9\nadoption of Aviation 2.0 to manage flights at the airports are\ntoo significant and too numerous for us to find it “likely, as\nopposed to merely speculative,” that vacatur of the interim\norders would redress Exhaustless’s injury. Lujan, 504 U.S. at\n561 (internal quotation marks omitted).\n\n First, the FAA operates under a duty “to ensure . . . the\nefficient use of airspace.” 49 U.S.C. § 40103(b)(1). Since\n1968, the Administration has fulfilled that responsibility by\nlimiting the number of takeoffs and landings at LaGuardia and\nJFK because the airports cannot accommodate the number of\nflights airlines would like to operate there without causing\nundue congestion. See 33 Fed. Reg. 17,896 (Dec. 3, 1968).\nThe Chief Operating Officer of the FAA’s Air Traffic\nAdministration testified that, absent the interim orders, she\nwould expect “demand for additional flights . . . to far exceed\nthe runway capacity resulting in extensive localized and\nsystemic delays and flight cancellations.” Bristol Decl. ¶ 4,\nFAA Add. A28. Recall that, in 2000, merely allowing\nexemptions from slot limitations caused LaGuardia to account\nfor 25% of flight delays nationwide. 71 Fed. Reg. at 54,332.\nExhaustless’s suggestion that our vacating the interim orders\nwould lead the FAA to delegate authority over flight schedules\nat LaGuardia and JFK Airports to the Port Authority defies\nhistory and blinks reality.\n\n Second, both LaGuardia and JFK Airports have accepted\nfederal grants for airport development under the Airport\nImprovement Program. Federal Aviation Administration, FY\n2019 Primary Entitlements (May 10, 2019),\nhttps://www.faa.gov/airports/aip/grantapportion_data/media/F\nY-2019-Primary-Entitlements.pdf. As a condition of accessing\nthose funds, the airports must pledge that they will be\n“available for public use on reasonable conditions and without\nunjust discrimination.” 49 U.S.C. § 47107(a)(1). That\n\f 10\nassurance prohibits airlines from assessing unreasonable fees.\nSee Air Transp. Ass’n of Am. v. DOT, 613 F.3d 206, 210 (D.C.\nCir. 2010). And Congress has assigned the Secretary of\nTransportation primary responsibility for determining whether\nairport fees are reasonable. See 49 U.S.C. § 47129. Under\ncurrent regulations, airports may charge landing fees so long as\nthey do not exceed the historical costs captured by an airport’s\n“rate base.” 78 Fed. Reg. 55,330, 55,333–35 (Sept. 10, 2013);\nAir Transp., 613 F.3d at 211. But Exhaustless’s technology\nrelies on charging carriers a market-clearing auction price\nrather than a cost-based landing fee, and in doing so runs into\nconflict with JFK’s and LaGuardia’s grant assurances.\n\n Third, there are substantial obstacles to charging\npassengers a “dedicated Congestion-Prevention Premium,” as\nExhaustless contemplates. Exhaustless Inc., Aviation 2.0—\nExplained, FAA Add. A39. To the extent the company\nenvisions that the local airport authority would assess the\npremium, it fails to account for the Anti-Head Tax Act, which\nprovides that any “political subdivision of a State . . . may not\nlevy or collect a tax, fee, head charge, or other charge on . . . an\nindividual traveling in air commerce.” 49 U.S.C. § 40116(b).\nTo the extent the company contemplates collection of the fee\nby the airlines, the FAA’s Director of the Office of Aviation\nAnalysis explained that Exhaustless’s “proposal would require\nthe carriers to substantially redesign their technology to\nintegrate a dynamic third-party fee” and that it was “unlikely\nthat carriers would do this voluntarily.” Homan Decl. ¶ 4, FAA\nAdd. A56. And if the companies decline to collect the fee\nvoluntarily, Exhaustless does not explain the airports’ authority\nto compel airlines to assess the charge.\n\n Fourth, adopting Exhaustless’s proposal could jeopardize\nthe United States’ compliance with international agreements on\ncommercial air travel. For instance, an agreement between the\n\f 11\nUnited States and Canada guarantees Canadian airlines a\nminimum of 42 slots at LaGuardia. See Air Transport\nAgreement Between the Government of the United States and\nthe Government of Canada, T.I.A.S. No. 07-312, Ann. II § 1\n(Mar. 12, 2007). Yet Exhaustless evidently seeks to auction\noff all the slots at LaGuardia without regard to a carrier’s\nnationality.\n\n Finally, even if Exhaustless were able to overcome each of\nthose hurdles, Aviation 2.0 remains an unproven product. The\nproduct has yet to be adopted by any airport in the nation even\nthough the vast majority of them are not subject to FAA slot\nregulation and thus could adopt it today. We find it doubtful\nthat two of the busiest airports in the nation would volunteer to\nact as the test sites for Aviation 2.0, even assuming the agency\nwould permit them to do so. In view of all of those legal and\npractical obstacles, the notion that vacating the interim orders\nwould create a business opportunity for Exhaustless amounts\nto mere conjecture.\n\n Our conclusion in that regard does not mean that the\nchallenged interim orders are entirely insulated from review.\nStanding principles under Article III exist to ensure that a\nlitigant alleges “such a personal stake in the outcome of the\ncontroversy as to warrant his invocation of federal-court\njurisdiction.” Summers v. Earth Island Inst., 555 U.S. 488, 493\n(2009) (internal quotation marks omitted). An airline or airport\nauthority likely would have standing to petition for review of\nthe orders. But no member of the regulated community has\njoined Exhaustless’s challenge or (as far as we know) filed its\nown petition for review, suggesting a form of acceptance\namong the parties having the most direct stake. Article III\ndenies us any license to disrupt that evident acceptance today.\n\f 12\n * * * * *\n\n For the foregoing reasons, we dismiss Exhaustless’s\npetitions for review for lack of jurisdiction.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: EXHAUSTLESS INC., Petitioner v. FEDERAL AVIATION ADMINISTRATION, Respondent\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-02\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued May 13, 2019 Decided August 2, 2019\\n\\n No. 18-1303\\n\\n EXHAUSTLESS INC.,\\n PETITIONER\\n\\n v.\\n\\n FEDERAL AVIATION ADMINISTRATION,\\n RESPONDENT\\n\\n\\n Consolidated with 18-1304\\n\\n\\n On Petitions for Review of Orders\\n of the Federal Aviation Administration\\n\\n\\n Kevin M. Blair argued the cause and filed the briefs for\\npetitioner.\\n\\n Benjamin M. Shultz, Attorney, U.S. Department of Justice,\\nargued the cause for respondent. With him on the brief were\\nMichael S. Raab, Attorney, U.S. Department of Justice, and\\nSteven G. Bradbury, General Counsel, Paul M. Geier,\\nAssistant General Counsel, Joy K. Park, Senior Trial Attorney,\\nand Arjun Garg, Chief Counsel, Federal Aviation\\nAdministration.\\n\\f 2\\n Before: HENDERSON, SRINIVASAN, and PILLARD, Circuit\\nJudges.\\n\\n Opinion for the Court filed by Circuit Judge SRINIVASAN.\\n\\n SRINIVASAN, Circuit Judge: In 1968, the Federal\\nAviation Administration began limiting the number of flights\\nserving LaGuardia and John F. Kennedy Airports in New York\\nin order to reduce flight delays. In 2000, Congress mandated\\nthe repeal of the relevant regulations based on concerns about\\ntheir anticompetitive effects. The phase-out process, however,\\ncaused flight delays to skyrocket at LaGuardia and JFK\\nAirports. The FAA then issued interim orders again limiting\\nthe number of flights serving those airports. The FAA has\\nsince extended the interim orders many times as efforts to\\nestablish a permanent solution have failed.\\n\\n Exhaustless, Inc., brings two petitions for review of the\\nlatest interim extension orders. Exhaustless would like\\nLaGuardia and JFK Airports to implement the company’s\\npatent-pending product, Aviation 2.0 Operating System, to\\nmanage the allocation of takeoff and landing “slots” to airlines.\\n\\n We dismiss Exhaustless’s petitions for lack of standing.\\nThe company fails to demonstrate that vacating the interim\\nFAA orders would redress its injury—i.e., a lack of market\\nopportunity for its product. Vacating the interim orders would\\nleave takeoffs and landings at the airports unregulated,\\neliminating the need for the company’s product at the federal\\nlevel. To the extent Exhaustless argues that the local airport\\nauthority could employ Aviation 2.0 if there were no federal\\nregulation, we find any such possibility too speculative to\\nsupport the company’s standing to bring these petitions.\\n\\f 3\\n I.\\n\\n The Federal Aviation Act calls for the FAA to “assign by\\nregulation or order the use of the airspace necessary to ensure\\nthe safety of aircraft and the efficient use of airspace.” 49\\nU.S.C. § 40103(b)(1). Navigable airspace includes the\\n“airspace needed to ensure safety in the takeoff and landing of\\naircraft.” Id. § 40102(a)(32).\\n\\n Since 1968, the FAA has restricted the number of takeoffs\\nand landings at certain highly congested airports in order to\\nreduce inefficient flight delays. The restrictions were codified\\nin a series of regulations known as the High Density Rule. As\\nof 2000, the rule placed numerical limits on the hourly takeoffs\\nand landings at five highly congested airports: Newark\\nLiberty, LaGuardia, JFK, O’Hare, and Ronald Reagan\\nWashington National. 14 C.F.R. § 93.123 (2000).\\n\\n By then, Congress had grown concerned with the High\\nDensity Rule’s collateral effects on airport access for carriers\\nand competition among carriers. Acting on those concerns in\\n2000, Congress prohibited the use of the High Density Rule at\\nLaGuardia or JFK Airports after January 1, 2007. 49 U.S.C.\\n§ 41715(a). For the period leading up to that date, Congress\\ndirected the FAA to grant slot exemptions for carriers servicing\\nsmaller airports and carriers with little or no existing service at\\nthe airports. Id. § 41716.\\n\\n Congress’s action led to an immediate increase in airport\\ncongestion at LaGuardia. As the FAA began granting slot\\nexemptions, “the number of scheduled flight operations at\\nLaGuardia began to far exceed the airport’s capacity even\\nunder optimal operating conditions.” 71 Fed. Reg. 54,331,\\n54,331 (Sept. 14, 2006). The average minutes of delay for\\narriving flights increased 144% between March and September\\n\\f 4\\nof 2000. Id. at 54,332. By September 2000, flight delays at\\nLaGuardia accounted for 25% of the delays nationwide. Id.\\nThe FAA responded by limiting the number of slot exemptions.\\nFrom late 2000 until the end of 2006, the High Density Rule,\\nwith the exemption cap, governed the number of slots at\\nLaGuardia. Id. at 54,332 & n.9.\\n\\n Because the High Density Rule was set to expire by 2007,\\nthe FAA, in August 2006, proposed a new permanent\\ncongestion management rule for LaGuardia and requested\\ncomments. 71 Fed. Reg. 51,360 (Aug. 29, 2006). A few weeks\\nlater, the agency explained that the permanent rule would not\\nbe finalized by the end of the year and that it was necessary to\\nimplement an interim rule to avert crippling delays. 71 Fed.\\nReg. 54,331 (Sept. 14, 2006).\\n\\n The FAA issued an interim order in December 2006. 71\\nFed. Reg. 77,854 (Dec. 27, 2006). The rule made clear that it\\nwas a temporary measure and reiterated the agency’s “need to\\ncomplete the rulemaking, because the final decision in that\\nproceeding should establish a more rational basis for the\\nregulation of flight operations at LaGuardia.” Id. at 77,856.\\nThe interim rule resembled the High Density Rule and\\ngenerally grandfathered the slots held by airlines under the\\nprevious regime. Id. at 77,859–61.\\n\\n Regulatory efforts concerning JFK Airport followed a\\nsomewhat different path but ended in much the same place.\\nWith respect to JFK, the FAA allowed the High Density Rule\\nto expire in 2007 without a replacement. Unsurprisingly, the\\nnumber of flights at JFK spiked, and with more planes came\\nmore delays. In 2007, the average daily operations at JFK\\nincreased 21% over the prior year and on-time arrival rates\\ndeclined from 68.5% to 62.2%. 73 Fed. Reg. 3510, 3511 (Jan.\\n18, 2008).\\n\\f 5\\n In 2008, the FAA published an interim order limiting the\\nnumber of takeoffs and landings at JFK. Id. at 3516–42. Like\\nthe LaGuardia order, the JFK order stressed its “short-term\\nnature,” stating that it was “not intended to create a long-term\\nsolution to congestion.” Id. at 3513–14.\\n\\n The FAA’s first attempt at a permanent solution for\\nLaGuardia and JFK Airports came via rules promulgated in\\nOctober 2008. 73 Fed. Reg. 60,574 (Oct. 10, 2008); 73 Fed.\\nReg. 60,544 (Oct. 10, 2008). Under those rules, a portion of\\nthe slots would be allocated using an auction. 73 Fed. Reg. at\\n60,577; 73 Fed. Reg. at 60,547. A number of airlines and trade\\ngroups, along with the local airport authority, promptly\\nchallenged the rules in our court. The challengers moved for\\nan immediate stay, contending that the FAA lacked statutory\\nauthority to conduct slot auctions. We granted the motion.\\nOrder, Port Auth. of N.Y. & N.J. v. FAA, No. 08-1329 (D.C.\\nCir. Dec. 8, 2008), J.A. 365. The FAA then rescinded the rules.\\n74 Fed. Reg. 52,134 (Oct. 9, 2009); 74 Fed. Reg. 52,132 (Oct.\\n9, 2009).\\n\\n The agency extended the interim orders for both\\nLaGuardia and JFK Airports until October 2011, noting that a\\npermanent solution would require more time. 74 Fed. Reg.\\n51,653 (Oct. 7, 2009); 74 Fed. Reg. 51,650 (Oct. 7, 2009). A\\nseries of additional extensions followed. In April 2011, the\\nFAA extended the orders until October 2013. 76 Fed. Reg.\\n18,620 (Apr. 4, 2011); 76 Fed. Reg. 18,616 (Apr. 4, 2011). In\\nMay 2013, the FAA extended the orders to October 2014. 78\\nFed. Reg. 28,278 (May 14, 2013); 78 Fed. Reg. 28,276 (May\\n14, 2013). And in March 2014, the FAA again extended the\\norders, to October 2016. 79 Fed. Reg. 17,222 (Mar. 27, 2014);\\n79 Fed. Reg. 16,854 (Mar. 26, 2014).\\n\\f 6\\n In January 2015, the agency proposed a final rule for New\\nYork–area airports that included a secondary market for the\\npurchase, sale, lease, or trade of slots between airlines. 80 Fed.\\nReg. 1274 (Jan. 8, 2015). But in May 2016, after receiving\\ncomments, the FAA withdrew the proposed rule. 81 Fed. Reg.\\n30,218 (May 16, 2016). That led the agency to extend the\\ninterim orders yet again, until October 2018. 81 Fed. Reg.\\n33,126 (May 25, 2016); 81 Fed. Reg. 32,636 (May 24, 2016).\\n\\n In September 2018, the FAA once more extended the\\ninterim orders for LaGuardia and JFK Airports, this time until\\nOctober 2020. 83 Fed. Reg. 47,065 (Sept. 18, 2018); 83 Fed.\\nReg. 46,865 (Sept. 17, 2018). Those latest extensions are at\\nissue here. While the orders largely match the prior extensions\\nin substance, they are less committal about a permanent rule,\\nstating only that the agency “will continue to consider potential\\nrulemaking in the future to codify the slot management policies\\nat [LaGuardia], and also at John F. Kennedy International\\nAirport (JFK).” 83 Fed. Reg. at 47,065.\\n\\n Petitioner Exhaustless, Inc., as noted, has developed a\\npatent-pending product called Aviation 2.0 Operating Standard\\nfor allocating airline slots at airports. Using Aviation 2.0,\\ncarriers would compete in semi-annual auctions to purchase\\nslots for a six-month period, with the total number of slots\\ndetermined by Exhaustless using its proprietary technology.\\nPassengers would then pay demand-calibrated congestion\\npremiums (on top of their airfare) when purchasing tickets.\\nBoth the congestion premiums and the auction proceeds would\\ngo to Exhaustless.\\n\\n II.\\n\\n Exhaustless asserts several challenges to the latest interim\\nextension orders in its petitions for review, including\\n\\f 7\\narguments that the FAA exceeded its statutory authority and\\nviolated the Administrative Procedure Act. We cannot address\\nthe merits of those claims unless Exhaustless has constitutional\\nstanding. See Steel Co. v. Citizens for a Better Env’t, 523 U.S.\\n83, 101–02 (1998). To establish standing, Exhaustless must\\ndemonstrate that: (i) it has suffered an injury-in-fact that is\\n“concrete and particularized” and “actual or imminent”; (ii) the\\ninjury is “fairly traceable to the challenged action” of the\\nrespondent; and (iii) it is “likely, as opposed to merely\\nspeculative, that the injury will be redressed by a favorable\\ndecision.” Lujan v. Defs. of Wildlife, 504 U.S. 555, 560–61\\n(1992) (formatting modified).\\n\\n The latter two elements, traceability and redressability,\\n“overlap as two sides of a causation coin.” Dynalantic Corp.\\nv. Dep’t of Def., 115 F.3d 1012, 1017 (D.C. Cir. 1997). When\\na petitioner itself is the object of the challenged agency action,\\nthere usually is little doubt of causation. See Lujan, 504 U.S.\\nat 561–62. But when a petitioner’s injury arises from an\\nagency’s “unlawful regulation (or lack of regulation) of\\nsomeone else,” causation often is “substantially more difficult”\\nto establish because the petitioner must demonstrate that the\\ninjury does not result from “the independent action of some\\nthird party not before the court.” Id. at 560, 562.\\n\\n Exhaustless fails to demonstrate redressability. The\\ncompany contends that it “is being deprived of the opportunity\\nto compete in the market with its patent-pending proposed\\nsolution” Aviation 2.0 so long as the FAA’s interim orders\\nremain in place. Exhaustless Br. 25; id. at 27. Exhaustless\\nenvisions that, if the interim orders—including their limitations\\non slots at LaGuardia and JFK—were withdrawn, the agency\\nwould then “transfer the management of service (slot volumes),\\nfor congestion-prevention purposes, to Exhaustless.”\\nExhaustless Inc., Aviation 2.0—Explained, FAA Add. A39.\\n\\f 8\\n Vacatur of the interim orders, however, would not get the\\ncompany closer to that goal. Without the orders, there would\\nbe no federally mandated number of takeoff and landing slots\\nat LaGuardia and JFK Airports, no scarce resource for the FAA\\nto auction, and hence no market for Exhaustless’s product, at\\nleast as concerns the federal government. The relief sought by\\nthe company in its petitions for review—i.e., vacatur of the\\ninterim orders—thus would not redress its claimed injury.\\n\\n The appropriate administrative channel for Exhaustless to\\npursue instead is a petition for rulemaking with the agency to\\nemploy the company’s technology. And because the interim\\nrules are revocable at will, the orders challenged by\\nExhaustless in this proceeding do not stand in the way of the\\ncompany’s attempting to persuade the FAA to adopt its\\ntechnology via a rulemaking. Indeed, Exhaustless has already\\nfiled such a petition for rulemaking. Petition for Rulemaking—\\nFAA-2007-0001 (filed May 21, 2018), J.A. 499. The petition\\nremains pending with the FAA, and, should the agency reject\\nit, Exhaustless would have standing to seek judicial review.\\n\\n At oral argument, Exhaustless submitted that vacatur of\\nthe FAA’s interim rules would create a different market\\nopportunity: vacatur in theory would result in transfer of\\ncontrol over flight schedules at LaGuardia and JFK Airports\\nfrom federal to local authority, i.e., the Port Authority of New\\nYork and New Jersey, which could then elect to use Aviation\\n2.0. That argument for standing fares no better.\\n\\n While the loss of an opportunity to compete for business\\ncan constitute Article III injury, there must be a “realistic\\npossibility” of winning the eventual competition. Ranger\\nCellular v. FCC, 348 F.3d 1044, 1050 (D.C. Cir. 2003)\\n(quoting Albuquerque Indian Rights v. Lujan, 930 F.2d 49, 56\\n(D.C. Cir. 1991)). Here, the hurdles to the Port Authority’s\\n\\f 9\\nadoption of Aviation 2.0 to manage flights at the airports are\\ntoo significant and too numerous for us to find it “likely, as\\nopposed to merely speculative,” that vacatur of the interim\\norders would redress Exhaustless’s injury. Lujan, 504 U.S. at\\n561 (internal quotation marks omitted).\\n\\n First, the FAA operates under a duty “to ensure . . . the\\nefficient use of airspace.” 49 U.S.C. § 40103(b)(1). Since\\n1968, the Administration has fulfilled that responsibility by\\nlimiting the number of takeoffs and landings at LaGuardia and\\nJFK because the airports cannot accommodate the number of\\nflights airlines would like to operate there without causing\\nundue congestion. See 33 Fed. Reg. 17,896 (Dec. 3, 1968).\\nThe Chief Operating Officer of the FAA’s Air Traffic\\nAdministration testified that, absent the interim orders, she\\nwould expect “demand for additional flights . . . to far exceed\\nthe runway capacity resulting in extensive localized and\\nsystemic delays and flight cancellations.” Bristol Decl. ¶ 4,\\nFAA Add. A28. Recall that, in 2000, merely allowing\\nexemptions from slot limitations caused LaGuardia to account\\nfor 25% of flight delays nationwide. 71 Fed. Reg. at 54,332.\\nExhaustless’s suggestion that our vacating the interim orders\\nwould lead the FAA to delegate authority over flight schedules\\nat LaGuardia and JFK Airports to the Port Authority defies\\nhistory and blinks reality.\\n\\n Second, both LaGuardia and JFK Airports have accepted\\nfederal grants for airport development under the Airport\\nImprovement Program. Federal Aviation Administration, FY\\n2019 Primary Entitlements (May 10, 2019),\\nhttps://www.faa.gov/airports/aip/grantapportion_data/media/F\\nY-2019-Primary-Entitlements.pdf. As a condition of accessing\\nthose funds, the airports must pledge that they will be\\n“available for public use on reasonable conditions and without\\nunjust discrimination.” 49 U.S.C. § 47107(a)(1). That\\n\\f 10\\nassurance prohibits airlines from assessing unreasonable fees.\\nSee Air Transp. Ass’n of Am. v. DOT, 613 F.3d 206, 210 (D.C.\\nCir. 2010). And Congress has assigned the Secretary of\\nTransportation primary responsibility for determining whether\\nairport fees are reasonable. See 49 U.S.C. § 47129. Under\\ncurrent regulations, airports may charge landing fees so long as\\nthey do not exceed the historical costs captured by an airport’s\\n“rate base.” 78 Fed. Reg. 55,330, 55,333–35 (Sept. 10, 2013);\\nAir Transp., 613 F.3d at 211. But Exhaustless’s technology\\nrelies on charging carriers a market-clearing auction price\\nrather than a cost-based landing fee, and in doing so runs into\\nconflict with JFK’s and LaGuardia’s grant assurances.\\n\\n Third, there are substantial obstacles to charging\\npassengers a “dedicated Congestion-Prevention Premium,” as\\nExhaustless contemplates. Exhaustless Inc., Aviation 2.0—\\nExplained, FAA Add. A39. To the extent the company\\nenvisions that the local airport authority would assess the\\npremium, it fails to account for the Anti-Head Tax Act, which\\nprovides that any “political subdivision of a State . . . may not\\nlevy or collect a tax, fee, head charge, or other charge on . . . an\\nindividual traveling in air commerce.” 49 U.S.C. § 40116(b).\\nTo the extent the company contemplates collection of the fee\\nby the airlines, the FAA’s Director of the Office of Aviation\\nAnalysis explained that Exhaustless’s “proposal would require\\nthe carriers to substantially redesign their technology to\\nintegrate a dynamic third-party fee” and that it was “unlikely\\nthat carriers would do this voluntarily.” Homan Decl. ¶ 4, FAA\\nAdd. A56. And if the companies decline to collect the fee\\nvoluntarily, Exhaustless does not explain the airports’ authority\\nto compel airlines to assess the charge.\\n\\n Fourth, adopting Exhaustless’s proposal could jeopardize\\nthe United States’ compliance with international agreements on\\ncommercial air travel. For instance, an agreement between the\\n\\f 11\\nUnited States and Canada guarantees Canadian airlines a\\nminimum of 42 slots at LaGuardia. See Air Transport\\nAgreement Between the Government of the United States and\\nthe Government of Canada, T.I.A.S. No. 07-312, Ann. II § 1\\n(Mar. 12, 2007). Yet Exhaustless evidently seeks to auction\\noff all the slots at LaGuardia without regard to a carrier’s\\nnationality.\\n\\n Finally, even if Exhaustless were able to overcome each of\\nthose hurdles, Aviation 2.0 remains an unproven product. The\\nproduct has yet to be adopted by any airport in the nation even\\nthough the vast majority of them are not subject to FAA slot\\nregulation and thus could adopt it today. We find it doubtful\\nthat two of the busiest airports in the nation would volunteer to\\nact as the test sites for Aviation 2.0, even assuming the agency\\nwould permit them to do so. In view of all of those legal and\\npractical obstacles, the notion that vacating the interim orders\\nwould create a business opportunity for Exhaustless amounts\\nto mere conjecture.\\n\\n Our conclusion in that regard does not mean that the\\nchallenged interim orders are entirely insulated from review.\\nStanding principles under Article III exist to ensure that a\\nlitigant alleges “such a personal stake in the outcome of the\\ncontroversy as to warrant his invocation of federal-court\\njurisdiction.” Summers v. Earth Island Inst., 555 U.S. 488, 493\\n(2009) (internal quotation marks omitted). An airline or airport\\nauthority likely would have standing to petition for review of\\nthe orders. But no member of the regulated community has\\njoined Exhaustless’s challenge or (as far as we know) filed its\\nown petition for review, suggesting a form of acceptance\\namong the parties having the most direct stake. Article III\\ndenies us any license to disrupt that evident acceptance today.\\n\\f 12\\n * * * * *\\n\\n For the foregoing reasons, we dismiss Exhaustless’s\\npetitions for review for lack of jurisdiction.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4169,"cells":{"id":{"kind":"number","value":4645250,"string":"4,645,250"},"name":{"kind":"string","value":"ALLEGHENY DEFENSE PROJECT, Et Al., Petitioners v. FEDERAL ENERGY REGULATORY COMMISSION, Respondent Anadarko Energy Services Company, Et Al., Intervenors"},"name_abbreviation":{"kind":"string","value":"Allegheny Def. Project v. Fed. Energy Regulatory Comm'n"},"decision_date":{"kind":"string","value":"2019-08-02"},"docket_number":{"kind":"string","value":"17-1098; C/w 17-1128, 17-1263, 18-1030"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Garland, Tatel, Millett\", \"parties\": \"\", \"opinions\": [{\"author\": \"\", \"type\": \"020lead\", \"text\": \"\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/9889046/\", \"author_raw\": \"\"}, {\"author\": \"\", \"type\": \"030concurrence\", \"text\": \"\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/9889047/\", \"author_raw\": \"\"}, {\"author\": \"PER CURIAM\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued December 7, 2018 Decided August 2, 2019\\n\\n No. 17-1098\\n\\n ALLEGHENY DEFENSE PROJECT, ET AL.,\\n PETITIONERS\\n\\n v.\\n\\n FEDERAL ENERGY REGULATORY COMMISSION,\\n RESPONDENT\\n\\n ANADARKO ENERGY SERVICES COMPANY, ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 17-1128, 17-1263, 18-1030\\n\\n\\n On Petitions for Review of Orders of the\\n Federal Energy Regulatory Commission\\n\\n\\n Elizabeth F. Benson argued the cause for petitioners\\nAllegheny Defense Project, et al. Siobhan K. Cole argued the\\ncause for petitioners Hilltop Hollow Limited Partnership, et al.\\nWith them on the joint briefs was Benjamin A. Luckett.\\nMichael N. Onufrak and Derek O. Teaney entered appearances.\\n\\n Beth G. Pacella, Deputy Solicitor, Federal Energy\\nRegulatory Commission, argued the cause for respondent.\\nWith her on the brief were James P. Danly, General Counsel,\\n\\n 2\\nand Robert H. Solomon, Solicitor. Anand Viswanathan,\\nAttorney, entered an appearance.\\n\\n John F. Stoviak argued the cause for intervenors. With\\nhim on the joint brief were Pamela S. Goodwin, Elizabeth U.\\nWitmer, Patrick F. Nugent, Kevin M. Sweeney, Jesse Stuart\\nUnkenholz, and Scott Borden Grover.\\n\\n Before: GARLAND, Chief Judge, and TATEL and MILLETT,\\nCircuit Judges.\\n\\n Opinion for the court filed PER CURIAM.\\n\\n Concurring opinion filed by Circuit Judge MILLETT.\\n\\n PER CURIAM: In February 2017, the Federal Energy\\nRegulatory Commission approved a natural gas pipeline\\nexpansion called the Atlantic Sunrise Project, which stretches\\nfrom northern Pennsylvania, across the Carolinas, and into\\nAlabama. Environmental Associations whose members live\\nand work in the areas affected by the Project and individual\\nHomeowners whose property was used for the Project seek\\nreview of the Commission’s orders permitting the\\nTranscontinental Gas Pipe Line Company to move forward\\nwith the pipeline expansion. Because the challenges to the\\nCommission’s decision cannot surmount the deferential\\nstandards of agency review and binding circuit precedent, we\\ndeny the petitions.\\n\\n I\\n\\n A\\n\\n Under the Natural Gas Act, 15 U.S.C. § 717 et seq., a\\ncompany wishing to construct a natural gas pipeline must first\\n\\n 3\\nobtain a certificate of “public convenience and necessity” from\\nthe Federal Energy Regulatory Commission. See id.\\n§ 717f(c); Myersville Citizens for a Rural Community, Inc. v.\\nFERC, 783 F.3d 1301, 1307 (D.C. Cir. 2015). The\\nCommission “shall * * * issue[]” the certificate if it finds that\\nthe proposed project “is or will be required by the present or\\nfuture public convenience and necessity.” 15 U.S.C.\\n§ 717f(e).\\n\\n When a pipeline company files a certificate application,\\nthe Commission reviews it under criteria set forth in its\\nCertificate Policy Statement, 88 FERC ¶ 61,227 (1999),\\nclarified, 90 FERC ¶ 61,128, further clarified, 92 FERC\\n¶ 61,094 (2000). The Certificate Policy Statement directs the\\nCommission to consider whether the project meets a market\\nneed and whether the public benefits of the project outweigh\\nthe harms. See Sierra Club v. FERC, 867 F.3d 1357, 1379\\n(D.C. Cir. 2017). If market need and public benefit are both\\nestablished, the Commission will issue a certificate authorizing\\nthe pipeline’s construction. Id. Once that certificate is\\ngranted, the Natural Gas Act empowers the private certificate\\nholder to exercise eminent domain authority if it “cannot\\nacquire by contract, or is unable to agree with the owner of\\nproperty to the compensation to be paid for, the necessary\\nright-of-way to construct, operate, and maintain a pipeline[.]”\\n15 U.S.C. § 717f(h).\\n\\n A party wishing to challenge the Commission’s issuance\\nof a certificate of public convenience and necessity must file a\\npetition for rehearing with the Commission. 15 U.S.C.\\n§ 717r(a). Until the Commission disposes of that rehearing\\npetition, the agency action is not final for purposes of judicial\\nreview. See id. § 717r(a)–(b); Clifton Power Corp. v. FERC,\\n294 F.3d 108, 110–111 (D.C. Cir. 2002). The filing and\\ndisposition of such a rehearing petition is thus a mandatory\\n\\n 4\\nprerequisite to obtaining judicial review of the Commission’s\\naction. See Delaware Riverkeeper Network v. FERC, 857\\nF.3d 388, 399 (D.C. Cir. 2017); Clifton Power Corp., 294 F.3d\\nat 110–111. Congress directed that petitions for rehearing\\nmay be “deemed to have been denied” if the Commission has\\nnot “act[ed] upon the application for rehearing within thirty\\ndays after it is filed[.]” 15 U.S.C. § 717r(a).\\n\\n The National Environmental Policy Act (“NEPA”), 42\\nU.S.C. § 4321 et seq., “require[s] the Commission to consider\\nand disclose the environmental effects of the actions it\\ncertifies.” Delaware Riverkeeper Network, 857 F.3d at 394\\n(internal quotation marks omitted). The Commission may\\nfulfill this requirement by compiling an Environmental Impact\\nStatement, which must consider, among other things, the\\nproposed project’s “indirect” environmental effects. See 42\\nU.S.C. § 4332(C); 40 C.F.R. §§ 1508.8, 1508.25(c).\\n\\n B\\n\\n In March 2015, the Transcontinental Gas Pipe Line\\nCompany (“Transco”) applied for a certificate of public\\nconvenience and necessity to build the Atlantic Sunrise Project.\\nAfter notice and public comment, the Commission issued a\\nfinal Environmental Impact Statement in December 2016. In\\nthat Statement, the Commission concluded that “neither\\nconstruction nor operation of the Project would significantly\\ncontribute to [greenhouse-gas] cumulative effects or climate\\nchange.” J.A. 323.\\n\\n On February 3, 2017, the Commission granted Transco its\\nrequested certificate of public convenience and necessity. 158\\nFERC ¶ 61,125 (2017) (“Certificate Order”). Environmental\\nAssociations and the private Homeowners whose land would\\nbecome subject to an easement for the Project both filed\\n\\n 5\\npetitions for rehearing with the Commission, along with\\nmotions for a stay of construction pending disposition of their\\npetitions. Before Congress’s 30-day deadline for action on the\\nrehearing petitions expired, the Commission “granted”\\nrehearing, but only “for the limited purpose of further\\nconsideration.” J.A. 600 (“Certificate Tolling Order”). The\\nCommission took no action on the stay motions for more than\\nfive months, and then denied them.\\n\\n In late August, a Pennsylvania federal district court\\npresiding over Transco’s eminent domain action entered an\\norder that declared Transco’s “right to immediate possession of\\nthe properties in question,” based on the presumed validity of\\nFERC’s Certificate Order. Transcontinental Gas Pipe Line\\nCo. v. Permanent Easements for 2.14 Acres & Temp.\\nEasements for 3.59 Acres in Conestoga Township, Lancaster\\nCounty, Pa., Tax Parcel No. 1201606900000, 2017 WL\\n3624250, at *1, *3 (E.D. Pa. Aug. 23, 2017) (rejecting the\\nHomeowners’ objections as “attacks on the FERC order itself,”\\nwhich “can only be challenged in front of FERC, and then in\\nthe United States Court of Appeals for the District of Columbia\\nCircuit”), aff’d, 907 F.3d 725 (3d Cir. 2018).\\n\\n On September 15, 2017, while the Environmental\\nAssociations’ and Homeowners’ petitions for rehearing were\\nstill pending, the Commission issued an order authorizing\\nTransco to begin construction of the Project. J.A. 616\\n(“Construction Order”). Transco broke ground that same day.\\n\\n The Environmental Associations promptly moved for\\nrehearing of the Construction Order and an order halting\\nconstruction. As its 30-day statutory deadline for action on\\nthe Construction Order rehearing petition approached, the\\nCommission again issued an order granting rehearing “for the\\n\\n 6\\nlimited purpose of further consideration.” J.A. 815\\n(“Construction Tolling Order”).\\n\\n The Commission eventually denied the Homeowners’ and\\nEnvironmental Associations’ petitions for rehearing in\\nDecember 2017, more than nine months after rehearing was\\nsought and three months after construction began. 161 FERC\\n¶ 61,250 (2017) (“Certificate Rehearing Order”). Three\\nmonths after that—and nearly six months after construction\\ncommenced—the Commission denied rehearing of the\\nConstruction Order. 162 FERC ¶ 61,192 (2018)\\n(“Construction Rehearing Order”).\\n\\n II\\n\\n This consolidated case arises out of four petitions for\\nreview collectively challenging the Certificate Order, the\\nCertificate Tolling Order, the Construction Order, the\\nConstruction Tolling Order, and the Certificate Rehearing\\nOrder. Because the Homeowners and Environmental\\nAssociations both properly sought rehearing of the Certificate\\nRehearing Order, which encompasses all of their claims for our\\nreview and is the final agency decision greenlighting the\\nProject, this court has jurisdiction. See 15 U.S.C. § 717r. 1\\n\\n The Commission’s factual findings are conclusive so long\\nas they are supported by substantial evidence. See 15 U.S.C.\\n§ 717(b). Where the Commission “has examined the relevant\\nconsiderations and articulated a satisfactory explanation for its\\naction, including a rational connection between the facts found\\nand the choice made,” we must uphold its decision. FERC v.\\n\\n 1\\n We therefore deny as moot the motions to dismiss the\\npetitions filed before the Certificate Rehearing Order issued, Nos.\\n17-1098 and 17-1128.\\n\\n 7\\nElectric Power Supply Ass’n, 136 S. Ct. 760, 782 (2016)\\n(formatting altered).\\n\\n The Homeowners and Environmental Associations argue\\nthat the Commission’s Certificate Order suffers from fatal\\nsubstantive and procedural flaws. Specifically, they argue\\nthat the Commission improperly conducted its environmental\\nassessment under NEPA, failed to substantiate market need for\\nthe Project as required by the Natural Gas Act, and denied them\\ndue process by authorizing construction to commence before\\nthe issuance of the Certificate Order could be judicially\\nreviewed. None of those arguments succeed.\\n\\n A\\n\\n The administrative record forecloses the Homeowners’\\nand Environmental Associations’ three NEPA arguments.\\n\\n First, the Homeowners and Environmental Associations\\nargue that the Commission did not factor downstream\\ngreenhouse-gas emissions into its evaluation of the Project’s\\nenvironmental impacts.\\n\\n The Homeowners and Environmental Associations are\\ncorrect that customers’ burning of the natural gas that the\\nProject transports will produce greenhouse-gas emissions.\\nSee Sierra Club, 867 F.3d at 1374. They are also correct that\\nNEPA required the Commission to consider both the direct and\\nindirect environmental effects of the Project, and that, despite\\nwhat the Commission argues, the downstream greenhouse-gas\\nemissions are just such an indirect effect. See id.; 40 C.F.R.\\n§ 1502.16(b).\\n\\n But that is as far as the argument gets them, because the\\nCommission already took the steps the Homeowners and\\n\\n 8\\nEnvironmental Associations request. The Commission\\naddressed downstream emissions in the Environmental Impact\\nStatement, the Certificate Order, and the Certificate Rehearing\\nOrder. In particular, the Commission estimated the amount of\\nCO2 emissions resulting from the gas that the Project would\\ntransport and predicted that those emissions would be partially\\noffset by reductions in higher carbon-emitting fuel that the\\nProject’s natural gas would replace. Neither the Homeowners\\nnor the Environmental Associations have identified what more\\nthe Commission should have said. That failure is fatal.\\nUnsubstantiated objections are not enough to stop an agency’s\\naction.\\n\\n Second, the Homeowners and Environmental Associations\\nassert that the Commission impermissibly segmented its\\nenvironmental review by failing to consider the synergistic\\neffect of the Project on emissions associated with a different\\npipeline—the Southeast Market Pipeline. Improper\\nsegmentation occurs when FERC creates separate\\nEnvironmental Impact Statements for interconnected pipelines\\nthat should have been evaluated as a single project. See City\\nof Boston Delegation v. FERC, 897 F.3d 241, 251–252 (D.C.\\nCir. 2018).\\n\\n But NEPA requires more to make out an improper\\nsegmentation claim than the Homeowners and Environmental\\nAssociations have offered. Specifically, the Commission has\\nno duty to consider the environmental effects of a separate\\nproject if the project in question has “substantial independent\\nutility.” City of Boston Delegation, 897 F.3d at 252 (internal\\nquotation marks omitted); see also Coalition on Sensible\\nTransp., Inc. v. Dole, 826 F.2d 60, 69 (D.C. Cir. 1987) (asking\\n“whether one project will serve a significant purpose even if a\\nsecond related project is not built”).\\n\\n 9\\n The record supports the Commission’s determination that\\nthe Southeast Market Pipeline and the Project are\\nindependently justified and that each has its own distinct utility.\\nCity of Boston Delegation, 897 F.3d at 252. That is because,\\neven if the Project were never built, the Southeast Market\\nPipeline still would be connected to enough natural gas supply\\nto exceed its capacity. J.A. 843. On this record, that\\ndisproves any claim of improper segmentation.\\n\\n Third, the Homeowners and Environmental Associations\\ncontend that the Commission was deficient in its analysis of an\\nalternative site for the Project called the Conestoga Route.\\nUnder NEPA, agencies must “identify the reasonable\\nalternatives to the contemplated action,” Minisink Residents for\\nEnvtl. Pres. & Safety v. FERC, 762 F.3d 97, 102 (D.C. Cir.\\n2014), and “discuss[] * * * the relevant issues and opposing\\nviewpoints,” Myersville, 783 F.3d at 1324–1325 (quoting\\nNevada v. Department of Energy, 457 F.3d 78, 93 (D.C. Cir.\\n2006)). This requirement helps ensure that agency actions are\\n“fully informed and well-considered.” Id. at 1325 (internal\\nquotation marks omitted).\\n\\n The Commission’s consideration of the Conestoga Route\\npasses NEPA muster. The Commission “look[ed] hard at the\\nenvironmental effects of its decision” by comparing the\\nproposed route with the Conestoga Route across highly\\ndetailed criteria. Minisink, 762 F.3d at 102 (formatting\\naltered). 2 In doing so, the Commission openly grappled with\\nthe factors favoring the Conestoga Route and reasonably\\nexplained why the proposed route was nonetheless superior.\\n\\n 2\\n The “Waterbodies crossed” criterion, for example, quantifies\\nwaterbodies for “Trout Stocking,” “High Quality Cold Water\\nFishes,” “Cold Water Fishes,” “High Quality Warm Water Fishes,”\\nand “Warm Water Fishes.” J.A. 318.\\n\\n 10\\nIt acknowledged that the Conestoga Route was shorter and\\nwould follow existing rights-of-way for more of its length than\\nthe proposed route would, but it placed more weight on the\\nproposed route’s virtues, including that it would cross fewer\\nrecreational areas and would avoid potentially dangerous\\nelevation changes. That suffices for NEPA. See Myersville,\\n783 F.3d at 1308; Minisink, 762 F.3d at 112.\\n\\n B\\n\\n The Homeowners and Environmental Associations\\nseparately claim that the Commission’s market-need\\ndetermination violated the Natural Gas Act. A showing of\\nmarket need is a precondition for the Commission to issue a\\ncertificate of public convenience and necessity. See Sierra\\nClub, 867 F.3d at 1379; Myersville, 783 F.3d at 1309; 90 FERC\\nat ¶ 61,392; 88 FERC at ¶ 61,746. The certificate applicant\\ncan make its required showing of market need for the pipeline\\n“by presenting evidence of ‘preconstruction contracts’ for gas\\ntransportation service.” Sierra Club, 867 F.3d at 1379.\\n\\n The Commission held Transco to this obligation. Its\\nfinding of market need rested on the existence of contracts with\\nshippers for 100% of the Project’s capacity. That alone is\\nenough. See Sierra Club, 867 F.3d at 1379; Myersville, 783\\nF.3d at 1311; Minisink, 762 F.3d at 111 n.10. And the\\nCommission did not stop there. It also relied on comments by\\ntwo shippers and one end-user, as well as a study submitted by\\none of the Environmental Associations, all of which reinforced\\nthe demand for the natural gas shipments.\\n\\n 11\\n C\\n\\n Finally, the Environmental Associations and Homeowners\\nargue that the Commission’s authorization for construction to\\ngo forward while their rehearing petitions were still pending—\\nand thus before the Commission’s decision was final and\\njudicially reviewable—denied them due process. Circuit\\nprecedent forecloses their claims.\\n\\n A due-process claim turns on two essential inquiries.\\nFirst, is there a “liberty or property interest of which a person\\nhas been deprived”? Swarthout v. Cooke, 562 U.S. 216, 219\\n(2011). Second, were the “procedures followed” by the\\ngovernment in encroaching on those interests “constitutionally\\nsufficient”? Id.\\n\\n The Environmental Associations run into trouble on the\\nfirst prong of that test. They assert that liberty and property\\ninterests were bestowed upon them by the Pennsylvania\\nConstitution’s Environmental Rights Amendment, Pa. Const.\\nArt. 1, § 27, and the Natural Gas Act’s review procedures, 15\\nU.S.C. §§ 717f(c)(1)(B), 717r(a)–(b). This court has already\\nrejected those claims, and those decisions control here. See\\nDelaware Riverkeeper Network v. FERC, 895 F.3d 102, 110\\n(D.C. Cir. 2018) (Pennsylvania Constitution); Griffith v.\\nFederal Labor Relations Auth., 842 F.2d 487, 495 (D.C. Cir.\\n1988) (Natural Gas Act review procedures); see generally\\nLaShawn A. v. Barry, 87 F.3d 1389, 1395 (D.C. Cir. 1996)\\n(“One three-judge panel * * * does not have the authority to\\noverrule another three-judge panel of the court.”).\\n\\n The Homeowners, for their part, lose on the second prong\\nof the due-process analysis. They argue that the\\nCommission’s delay in acting on their rehearing petitions while\\nauthorizing construction to start “denied [them] the right to be\\n\\n 12\\nheard on whether Transco’s taking of their property actually\\nsatisfies the public use requirement of the Fifth Amendment.”\\nPetitioners’ Opening Br. 47.\\n\\n But, again, circuit precedent says otherwise. We have\\nheld that, as long as FERC’s public-convenience-and-necessity\\ndetermination is not legally deficient, it necessarily satisfies the\\nFifth Amendment’s public-use requirement. See Midcoast\\nInterstate Transmission, Inc. v. FERC, 198 F.3d 960, 973 (D.C.\\nCir. 2000). The Homeowners make no claim that they were\\ndeprived of a meaningful opportunity to be heard as part of the\\nCommission’s proceedings leading up to its issuance of the\\nCertificate Order, and they make no effort to distinguish (or\\neven acknowledge) our holding in Midcoast.\\n\\n * * *\\n\\n For the foregoing reasons, the petitions for review are\\ndenied.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4422503/\", \"author_raw\": \"PER CURIAM\"}, {\"author\": \"MILLETT, Circuit Judge, concurring\", \"type\": \"concurrence\", \"text\": \"MILLETT, Circuit Judge, concurring: As for the\\nHomeowners’ due-process claim, I recognize that circuit\\nprecedent ties my hands. But the Commission has twisted our\\nprecedent into a Kafkaesque regime. Under it, the\\nCommission can keep homeowners in seemingly endless\\nadministrative limbo while energy companies plow ahead\\nseizing land and constructing the very pipeline that the\\nprocedurally handcuffed homeowners seek to stop. The\\nCommission does so by casting aside the time limit on\\nrehearing that Congress ordered—treating its decision as final-\\nenough for the pipeline companies to go forward with their\\nconstruction plans, but not final for the injured landowners to\\nobtain judicial review. This case starkly illustrates why that is\\nnot right.\\n\\n I\\n\\n My concern is not one of outcomes. The law and\\nadministrative record dictate who should win in this case, as in\\nall Commission cases. My concern is about fair process and,\\nin particular, the ability of those who are directly injured—the\\nindividuals whose property is taken in whole or in part by\\nCommission order—to have their day in court before it is too\\nlate.\\n\\n The Homeowners in this case are the Erb and Hoffman\\nfamilies. Their “much beloved properties,” J.A. 581, are\\nlocated in Southeastern Pennsylvania, see Transcript of\\nEvidentiary Hearing 9, 53, ECF No. 27 (E.D. Pa. July 20, 2017)\\n(“Hearing Tr.”). That was where the Erbs built their “dream\\nhome” and planned for their three sons to settle one day. Id.\\nat 10–11. The Hoffmans’ house is tucked among “rolling\\nhills” on their property—a home designed to be so private that\\nit could not be seen from the road. Id. at 52–53. They built\\ntheir lives there, among “lots of wildlife,” including the scores\\nof deer and turkeys they fed each day. Id. at 54. Both\\n\\n 2\\nfamilies cherished the quiet, secluded nature of the places\\nwhere they chose to live.\\n\\n That was until the Commission allowed the\\nTranscontinental Gas Pipeline Company (“Transco”) to move\\nin. In October 2015, the Commission notified the Erbs and the\\nHoffmans that a pipeline under consideration might cut right\\nthrough their land. That would mean “removing topsoil, trees,\\nshrubs, brush, roots, and large rocks, and then removing or\\nblasting additional soil and bedrock to create a trench for the\\npipeline,” and giving Transco a permanent right-of-way\\nthrough their yards. J.A. 581. The Erbs were “deathly afraid\\nof the pipeline” and did not “want to be anywhere near it.”\\nHearing Tr. 45. The Hoffmans found the idea “unacceptable”\\nand “disturbing,” because Transco’s right-of-way in the middle\\nof their property would “totally take[] [their] privacy away.”\\nId. at 59.\\n\\n The Erbs and the Hoffmans fought hard before the\\nCommission to preserve their land. After the Certificate\\nOrder issued, the Homeowners filed their request for rehearing\\nas well as a motion to stay the Certificate Order pending the\\nagency’s final rehearing decision. During the 30-day time\\nperiod allotted by Congress for agency action on the\\nHomeowners’ rehearing requests, the Commission ignored the\\napplication for a stay. And before the 30th day passed, the\\nCommission issued a so-called “tolling order.” J.A. 600\\n(“Certificate Tolling Order”). Offering no explanation\\nwhatsoever for its delay, the Commission just declared that\\nrehearing was “granted” for the sole purpose of buying the\\nCommission more time. Id. (“[R]ehearing of the\\nCommission’s order is hereby granted for the limited purpose\\nof further consideration[.]”).\\n\\n 3\\n The upshot of the Commission’s self-help was that its\\ncontinued inaction on rehearing—the non-finality of the\\nCertificate Order—jurisdictionally locked the Homeowners\\nout of federal court. See Delaware Riverkeeper Network v.\\nFERC, 857 F.3d 388, 393 (D.C. Cir. 2017); Clifton Power\\nCorp. v. FERC, 294 F.3d 108, 110–111 (D.C. Cir. 2002).\\nIndeed, when the Homeowners petitioned this court for review\\nof both the Certificate Order and the Certificate Tolling Order,\\nboth the Commission and Transco were quick to seek dismissal\\nof the petitions as “incurably premature” because the rehearing\\nrequests had not yet been resolved. See Motion of Movant-\\nIntervenor Transcontinental Gas Pipe Line Company, LLC to\\nDismiss the Petitions for Review 15, No. 17-1128 (June 30,\\n2017); see also Motion of Federal Energy Regulatory\\nCommission to Apply Disposition of the Motion to Dismiss\\nFiled in Docket No. 17-1098 to the Instant Petitions, No. 17-\\n1128 (June 30, 2017); Motion to Dismiss for Lack of\\nJurisdiction 5–6, No. 17-1098 (April 28, 2017).\\n\\n While non-final for the Erbs and Hoffmans, the\\nCommission’s order was still final enough for Transco to\\nprevail in an eminent domain action in a Pennsylvania federal\\ndistrict court and to acquire the needed easements over the\\nErbs’ and Hoffmans’ land. Transco told the district court that,\\nin deciding whether eminent domain is appropriate, it “must\\nconsider that the [Certificate Order] is final” and that its finding\\nof public convenience and necessity controlled the question of\\nwhether the land was being taken for a public use. Hearing Tr.\\n138–139 (emphasis added). On that basis, the district court\\ngranted Transco the “right to immediate possession of the\\nproperties[.]” Transcontinental Gas Pipe Line Co. v.\\nPermanent Easements for 2.14 Acres & Temp. Easements for\\n3.59 Acres in Conestoga Township, Lancaster County, Pa., Tax\\nParcel No. 1201606900000, 2017 WL 3624250, at *1 (E.D. Pa.\\nAug. 23, 2017), aff’d, 907 F.3d 725 (3d Cir. 2018); see also id.\\n\\n 4\\nat *3–4 (rejecting the Homeowners’ claims as “attacks on the\\nFERC order itself,” which the district court did not have\\njurisdiction to consider).\\n\\n On August 31, 2017—eight days after Transco prevailed\\nin its eminent domain action and more than six months after the\\nHomeowners asked the Commission for a stay of the\\nCertificate Order—the Commission denied the Homeowners’\\nrequest for a stay. The Commission reasoned that the\\nHomeowners’ objections to Transco bulldozing and blasting its\\npipeline into their homesteads were nothing more than\\n“generalized claims of environmental harm [that] do not\\nconstitute sufficient evidence of irreparable harm that would\\njustify a stay.” 160 FERC ¶ 61,042, 2017 WL 3835932, at *2\\n(Feb. 9, 2017).\\n\\n Two weeks later, apparently still too busy to act on the\\nHomeowners’ rehearing petition, the Commission nonetheless\\nfound the time to issue a Construction Order authorizing\\nTransco to start construction on the Homeowners’ land. J.A.\\n616 (“Construction Order”). Which Transco promptly did.\\nJ.A. 616; Oral Arg. Tr. 19. And when the 30-day time limit\\nfor action on a request for rehearing of the Construction Order\\napproached, guess what? The Commission once again issued\\nan order that did nothing but give the Commission more time\\nto decide. See J.A. 815. Meanwhile, Transco’s construction\\ncontinued apace. And the Homeowners remained trapped\\nbefore the agency.\\n\\n The Commission did not issue a final, appealable\\ncertificate decision until December 6, 2017, months after\\nTransco had started construction.\\n\\n 5\\n II\\n\\n Circuit precedent gave the Commission the tools it has\\nused to create this administrative quagmire for those who seek\\nto challenge its decisions. In my view, we should put an end\\nto it. A scheme that walls homeowners off from timely\\njudicial review of the Commission’s public-use determination,\\nwhile allowing eminent domain and functionally irreversible\\nconstruction to go forward, is in substantial tension with\\nstatutory text and runs roughshod over basic principles of fair\\nprocess.\\n\\n A\\n\\n 1\\n\\n This much is clear: The Natural Gas Act makes the filing\\nof a rehearing petition a jurisdictional precondition to obtaining\\njudicial review of Commission decisions. 15 U.S.C.\\n§ 717r(a)–(b). The Act spells out what the Commission can\\ndo once an application for rehearing is filed. It “shall have\\npower to grant or deny rehearing or to abrogate or modify its\\norder[.]” Id. § 717r(a). Unless the Commission so “acts\\nupon the application for rehearing within thirty days,”\\nrehearing “may be deemed to have been denied.” Id.\\n\\n The most natural reading of that language is that Congress\\ntold the Commission to “act[] upon the application for\\nrehearing” and to do so “within thirty days.” 15 U.S.C.\\n§ 717r(a).\\n\\n The Commission reasons that it has “acted” because it\\ngave itself more time. But the Natural Gas Act requires the\\nCommission not just to act, but to “act[] upon the application”\\nitself. 15 U.S.C. § 717r(a). And Congress specified which\\n\\n 6\\nCommission actions count as “acts upon the application”: The\\nCommission could “grant” rehearing, “deny” it, “modify” the\\nunderlying order, or “abrogate” it. Id. Casting aside\\nCongress’s time limit is not on that menu.\\n\\n The Natural Gas Act answers the question of what should\\nhappen if the Commission finds itself unable to act within the\\nallotted 30 days. Rehearing “may be deemed denied” by the\\naggrieved party, who may then obtain judicial review. 15\\nU.S.C. § 717r(a); see id. § 717r(b); cf. 28 U.S.C. § 2675(a)\\n(Under the Federal Tort Claims Act, “[t]he failure of\\nan agency to make final disposition of a claim within six\\nmonths after it is filed shall, at the option of the claimant any\\ntime thereafter, be deemed a final denial of the claim for\\npurposes of this section.”); 42 U.S.C. § 2000e-16(c) (Under\\nTitle VII of the Civil Rights Act of 1964, agency inaction after\\na specified time period for agency decision permits the\\ncomplainant to proceed to court.). There is no point to that\\nstatutory time limit if the Commission can ignore it for no\\nreason at all and with no consequence at all.\\n\\n Congress, in other words, gave the Commission 30 days to\\nfish or cut bait. Trapping an aggrieved party in administrative\\nlimbo while the Commission spends several months thinking\\nabout whether to go fishing is not an option.\\n\\n 2\\n\\n But circuit precedent says otherwise. We have held that\\nthe Commission’s tolling orders qualify under the statute as an\\naction upon the rehearing request. See Delaware Riverkeeper\\nNetwork v. FERC, 895 F.3d 102, 113 (D.C. Cir. 2018); see also\\nMoreau v. FERC, 982 F.2d 556, 564 (D.C. Cir. 1993) (FERC\\nhad “yet to rule on the merits” after forty-four days).\\n\\n 7\\n Delaware Riverkeeper and Moreau faithfully followed the\\nfountainhead of circuit precedent upholding the Commission’s\\ntolling orders: California Company v. Federal Power\\nCommission, 411 F.2d 720 (D.C. Cir. 1969) (per curiam). But\\nthat decision long predates modern statutory construction\\njurisprudence, and did not claim to turn on the best reading of\\nthe statutory text. To the contrary, California Company\\ndescribed the Commission’s reading of Section 717r(a) as “far\\nfrom self-evident.” Id. at 722. Nonetheless, the court\\ndeferred to the Commission because its approach “avoids * * *\\nadministrative and judicial problems.” Id.\\n\\n The court in Delaware Riverkeeper found itself bound by\\nCalifornia Company, and thus yielded to concerns about the\\nworkability of a scheme that ties the Commission’s often\\ncomplicated decisionmaking to a tight 30-day timeframe. See\\n895 F.3d at 113. But, alas, “[i]f you give a mouse a cookie\\n* * *.” 1 The Commission has taken this court’s patience and\\nturned it into a license to routinely blow past Congress’s\\ndeadline, granting itself as much time as it desires to act on\\nrehearing requests. One recent study showed that, between\\n2009 and 2017, the Commission issued tolling orders in\\nresponse to 99% of requests for rehearing of pipeline\\ncertification decisions. Petition for an Extraordinary Writ, In\\nre Appalachian Voices, et al., No. 18-1006 at Exhibit G (Jan.\\n8, 2018) (“Exhibit G”) (cataloguing tolling orders issued in 74\\nout of 75 pipeline certifications between 2009 and 2017); see\\nalso Berkley v. Mountain Valley Pipeline, LLC, 896 F.3d 624,\\n631 n.4 (4th Cir. 2018) (noting that FERC does not dispute that\\nit “regularly” issues tolling orders). FERC has issued a\\nboilerplate tolling order in response to every motion for\\nrehearing of a pipeline certification decision since 2017 too.\\n\\n 1\\n See LAURA NUMEROFF, IF YOU GIVE A MOUSE A COOKIE\\n(1985).\\n\\n 8\\n\\n It also bears noting that this court’s acceptance of tolling\\norders started in a case that involved disputes over money, not\\nproperty. See California Co., 411 F.2d at 720 (rate dispute).\\nThe same is true of all of the other cases cited in Delaware\\nRiverkeeper. See 895 F.3d at 113 (citing Kokajko v. FERC,\\n837 F.2d 524, 526 (1st Cir. 1988) (fees); General American Oil\\nCo. of Tex. v. Federal Power Comm’n, 409 F.2d 597, 599 (5th\\nCir. 1969) (rates)). Because disputes over monetary payments\\ncan be fixed later, the consequences of Commission delay were\\ntemporary and remediable. One side or the other would have\\nto bear the financial cost while administrative and judicial\\nlitigation went forward. The tolling order just assigned that\\nburden to the party that lost before the agency.\\n\\n But allowing the Commission to take its time while private\\nproperty is being destroyed is another thing altogether. Under\\nthe Natural Gas Act, “any holder of a certificate * * * may\\nacquire the [property] by the exercise of the right of eminent\\ndomain.” 15 U.S.C. § 717f(h). On top of that, courts\\ninvolved in eminent domain proceedings—including in this\\ncase—routinely treat the Commission’s non-final certificate\\norder as final and conclusive evidence that the taking serves a\\npublic purpose. See Transcontinental Gas Pipe Line Co.,\\n2017 WL 3624250, at *4 (collecting cases).\\n\\n Congress further directed that the filing of an application\\nfor rehearing “shall not, unless specifically ordered by the\\nCommission, operate as a stay of the Commission’s\\n[certificate] order.” 15 U.S.C. § 717r(c). Of course, in doing\\nso, Congress presumably expected that rehearing decisions\\nwould be resolved within 30 days, as the statute says. See id.\\n§ 717r(a). In other words, as Congress designed the rehearing\\nsystem, eminent domain proceedings would likely not\\nconclude before the Commission acted on rehearing and\\n\\n 9\\nafforded the applicant an opportunity for judicial review of the\\npublic-use determination.\\n\\n The Commission’s use of tolling orders upends that\\nbalanced framework. Now the Commission can, in effect,\\nsplit the atom of finality knowing that its certificate orders will\\nbe treated as conclusive in eminent domain proceedings, while\\nshielding those same orders from judicial scrutiny as non-final.\\nThis imbalance is what allowed Transco’s lawyer in this case\\nto tell the Pennsylvania district court that, “as to this process,\\nthe eminent domain process, the [certificate] order is final,”\\nHearing Tr. 80, while it and the Commission told this court that\\nthe very same certificate order does not constitute final agency\\naction for purposes of the Homeowners’ effort to obtain\\njudicial review, see Motion of Movant-Intervenor at 15\\n(arguing that the Homeowners’ petition for review of the\\nCertificate Order must be dismissed as “‘incurably premature’\\nbecause a Tolling Order was timely issued and the Commission\\nhas not yet ruled on the merits of the Requests for Rehearing”);\\nsee also Commission’s Br. 8 (asserting the same argument).\\n\\n Making that bad situation worse, the Commission that says\\nit is too busy to act on rehearing applications nevertheless\\nconsistently manages to find the time to grant orders\\nauthorizing construction to go forward while rehearing is still\\npending. See Construction Order; Exhibit G (cataloguing\\ncases where FERC authorized construction during the tolling\\nperiod).\\n\\n The result is that the Commission can toll until the cows\\ncome home and thereby forestall judicial review while people’s\\nhomesteads are being destroyed. And the Commission knows\\nexactly what it is doing with its repeated issuance of cookie-\\ncutter tolling orders. Commissioner Glick has openly\\nacknowledged that the Commission’s use of tolling orders to\\n\\n 10\\ndelay rehearing orders causes “landowners, communities, and\\nthe environment” to “suffer needless and avoidable harm”\\nbecause, “while the parties await their opportunity to challenge\\nthe Commission’s certificate decision in court,” the developer\\ngoes ahead and “begin[s] construction on the new pipeline\\nfacility[.]” PennEast Pipeline Co., 163 FERC ¶ 61,159, 2018\\nWL 2453596, at *4 (May 30, 2018) (Glick, Comm’r,\\nconcurring).\\n\\n Of course, the Commission advises the companies that\\nthey proceed at the risk that the agency may have a change of\\nheart on rehearing. Commission’s Br. 24. Perhaps such\\nwords work when the certificate orders just involve monetary\\npayments for fees or rates. But they ring hollow in eminent\\ndomain and construction cases like this. Once the property is\\nin the company’s possession, the land is cleared, trees are cut\\ndown, and the pipeline is cemented into a family’s backyard, it\\nis difficult if not impossible to unshuffle the deck. The\\ndamage to property rights, property values, and the\\nenvironment is done. That creates rather than “avoids * * *\\nadministrative and judicial problems.” California Co., 411\\nF.2d at 722.\\n\\n In my view, the better course is to assume that Congress\\n“says in a statute what it means and means in a statute what it\\nsays there.” Connecticut Nat’l Bank v. Germain, 503 U.S.\\n249, 254 (1992). So when Congress said that rehearing\\npetitions may be deemed denied and agency action final after\\n30 days, that is what it meant.\\n\\n III\\n\\n Circuit precedent has already rejected a due-process\\nchallenge to the Commission’s tolling orders. See Delaware\\nRiverkeeper, 895 F.3d at 112–113. Yet, in my view, the\\n\\n 11\\nauthority our precedent affords the Commission to issue tolling\\norders while simultaneously allowing eminent domain\\nproceedings and construction to proceed—all before the\\naffected landowner can obtain judicial review of the\\nCommission’s public-use determination—skates on thin\\nconstitutional ice. And courts must construe statutes to avoid,\\nrather than to create, constitutional problems. Zadvydas v.\\nDavis, 533 U.S. 678, 689 (2001) (describing this “cardinal\\nprinciple” of statutory construction).\\n\\n A\\n\\n The Due Process Clause of the Fifth Amendment\\nguarantees that “[n]o person shall * * * be deprived of life,\\nliberty, or property, without due process of law[.]” U.S.\\nConst. Amend. V. Due process “calls for such procedural\\nprotections as the particular situation demands” to\\nmeaningfully protect the constitutional right at stake.\\nMorrissey v. Brewer, 408 U.S. 471, 481 (1972). At a\\nminimum, due process requires an “opportunity to be heard.”\\nGrannis v. Ordean, 234 U.S. 385, 394 (1914). And that\\nopportunity “must be granted at a meaningful time and in a\\nmeaningful manner.” Armstrong v. Manzo, 380 U.S. 545, 552\\n(1965). This requirement not only “ensure[s] abstract fair\\nplay to the individual.” Fuentes v. Shevin, 407 U.S. 67, 80–81\\n(1972). It also “protect[s] his use and possession of property\\nfrom arbitrary encroachment” by “minimiz[ing] substantively\\nunfair or mistaken deprivations of property[.]” Id. at 81.\\nThat “danger” is “especially great” when the government\\n“seizes goods simply upon the application of and for the benefit\\nof a private party” such as Transco. Id.\\n\\n Landowners like the Erb and Hoffman families\\n“[u]ndoubtedly * * * have a constitutionally protected property\\ninterest in their home.” Freeman v. F.D.I.C., 56 F.3d 1394,\\n\\n 12\\n1403 (D.C. Cir. 1995). Determining whether the\\nCommission’s procedures for allowing private third parties to\\nbuild on private land pass constitutional muster therefore\\nrequires weighing (i) “the private interest that will be affected\\nby the official action”; (ii) “the risk of an erroneous deprivation\\nof such interest through the procedures used”; and (iii) “the\\nGovernment’s interest, including the function involved and the\\nfiscal and administrative burdens that the additional or\\nsubstitute procedural requirement would entail.” Mathews v.\\nEldridge, 424 U.S. 319, 335 (1976). That balance weighs\\nagainst the Commission’s current process of saddling\\nlandowners with all the harm and risk of delay in its own\\ndecisionmaking. And it weighs heavily in favor of reading the\\nstatute’s 30-day time limit as the plain text prescribes to avoid\\nthis constitutional question.\\n\\n 1\\n\\n The right of the Erb and Hoffman families “to maintain\\ncontrol over [their] home[s], and to be free from governmental\\ninterference, is a private interest of historic and continuing\\nimportance.” United States v. James Daniel Good Real\\nProperty, 510 U.S. 43, 53–54 (1993); id. at 49 (“The\\nGovernment does not, and could not, dispute that the seizure of\\nGood’s home and 4-acre parcel deprived him of property\\ninterests protected by the Due Process Clause.”).\\n\\n And the Commission’s decisions authorizing Transco’s\\n“physical invasion” of the families’ homesteads, Loretto v.\\nTeleprompter Manhattan CATV Corp., 458 U.S. 419, 426\\n(1982), constitute a “government intrusion of an unusually\\nserious character,” id. at 433. Homeowners “suffer[] a special\\nkind of injury when a stranger directly invades and occupies\\nthe owner’s property.” Id. at 436 (emphasis omitted); see also\\nJames Daniel Good Real Property, 510 U.S. at 61 (An\\n\\n 13\\n“essential principle” is that “[i]ndividual freedom finds\\ntangible expression in property rights,” particularly where the\\n“privacy of the home and those who take shelter within it” is at\\nstake.); Hendler v. United States, 952 F.2d 1364, 1374 (Fed.\\nCir. 1991) (“In the bundle of rights we call property, one of the\\nmost valued is the right to sole and exclusive possession—the\\nright to exclude strangers[.]”) (emphasis omitted); cf. United\\nChurch of the Med. Ctr. v. Medical Ctr. Comm’n, 689 F.2d 693,\\n701 (7th Cir. 1982) (“It is settled beyond the need for citation\\n* * * that a given piece of property is considered to be unique,\\nand its loss is always an irreparable injury.”).\\n\\n All the more so because building the pipeline would cause\\n“permanent, irreparable environmental harm” to the Erb and\\nHoffman families’ lands. J.A. 581. “Environmental injury,\\nby its nature, can seldom be adequately remedied by money\\ndamages and is often permanent or at least of long duration,\\ni.e., irreparable.” Amoco Prod. Co. v. Village of Gambell, 480\\nU.S. 531, 545 (1987); see also National Wildlife Fed’n v.\\nBurford, 835 F.2d 305, 323–325 (D.C. Cir. 1987)\\n(“[D]estroying wildlife habitat, air and water quality, natural\\nbeauty, and other environmental and aesthetic values and\\ninterests” constitutes irreparable harm.). After all,\\nconstructing a gas pipeline is not a tidy intrusion. It requires\\ncutting down the families’ trees, digging up their soil, blasting\\ntheir bedrock, displacing wildlife, and polluting the air.\\n\\n Once Commission-authorized construction starts, then\\n“meaningful backward-looking relief to rectify an[]\\nunconstitutional deprivation” is rarely a viable option.\\nMcKesson Corp. v. Division of Alcoholic Beverages &\\nTobacco, 496 U.S. 18, 31 (1990) (emphasis added); see also\\nAmoco, 480 U.S. at 545; RoDa Drilling Co. v. Siegal, 552 F.3d\\n1203, 1210 (10th Cir. 2009) (“[C]ourts have held that, when\\ninterests involving real property are at stake, preliminary\\n\\n 14\\ninjunctive relief can be particularly appropriate because of the\\nunique nature of the property interest.”) (formatting altered)\\n(collecting additional cases).\\n\\n 2\\n\\n The Commission’s process leaves landowners like the\\nErbs and the Hoffmans to bear a material “risk of an erroneous\\ndeprivation.” See Mathews, 424 U.S. at 335. Every time a\\ncourt reverses, vacates, or remands a Commission pipeline\\ncertification decision—or when the Commission grants a\\nlandowner’s rehearing motion on the merits—it means that the\\npeople who had owned and organized their lives on the\\npipeline-occupied land were subject to an erroneous\\ndeprivation by the initial decision. See, e.g., Sierra Club v.\\nFERC, 867 F.3d 1357, 1379 (D.C. Cir. 2017); Delaware\\nRiverkeeper Network v. FERC, 753 F.3d 1304, 1320 (D.C. Cir.\\n2014).\\n\\n The Commission’s prolonged tolling while construction\\nproceeds apace compounds the problem. The Supreme Court\\nhas recently recognized that a “property owner has suffered a\\nviolation of his Fifth Amendment rights when the government\\ntakes his property without just compensation,” regardless of\\nwhether damages might later ensue. Knick v. Township of\\nScott, Pa., 139 S. Ct. 2162, 2168 (2019) (Takings Clause case\\ninvolving regulation by state government); see also id. at 2170.\\nPrompt access to federal court review of the lawfulness of the\\ntaking, including the public use determination, is part of the\\nprotection the Fifth Amendment affords. That access is\\nnecessary to avoid “hand[ing] authority over federal takings\\nclaims to state courts[,]” id., or—as here—to the same federal\\nagency that authorized eminent domain in the first place, see\\nCity of Cincinnati v. Vester, 281 U.S. 439, 446 (1930) (“[T]he\\nquestion what is a public use is a judicial one.”).\\n\\n 15\\n\\n Unsurprisingly, Transco supports the current state of\\naffairs. It reasons that homeowners can always seek\\nmandamus relief under the All Writs Act, 28 U.S.C. § 1651,\\nwhile the pipeline company’s bulldozers keep bulldozing. We\\nagreed in Delaware Riverkeeper, 895 F.3d at 113, and I\\nrecognize that decision binds this panel.\\n\\n But the due-process question is whether the Commission’s\\nprocess protects against erroneous deprivations. Mandamus\\nonly protects against the most extreme outlier—the clearest,\\nmost obvious, and wholly irreparable agency errors, and even\\nthen only if a court chooses to use its discretion to grant that\\nextraordinary remedy. See Cheney v. United States Dist.\\nCourt for D.C., 542 U.S. 367, 380–381 (2004) (Mandamus\\nrequires a showing of “exceptional circumstances,” the absence\\nof “other adequate means” to obtain relief, and a “clear and\\nindisputable” right to the writ, and even still “the issuing court,\\nin the exercise of its discretion, must be satisfied that the writ\\nis appropriate under the circumstances.”) (formatting altered).\\n\\n The proof is in the pudding. The Commission routinely\\nfends off mandamus actions while its tolling orders remain in\\neffect. See, e.g., In re Appalachian Voices, No. 18-1006 (D.C.\\nCir. Feb. 2, 2018) (denying property owners’ petition for a stay\\nof pipeline construction under the All Writs Act); In re\\nAppalachian Voices, No. 18-1271 (4th Cir. March 21, 2018)\\n(same); Coalition to Reroute Nexus v. FERC, No. 17-4302 (6th\\nCir. March 15, 2018) (same). Neither the Commission nor\\nTransco has cited a single instance in which a petitioner\\nopposing pipeline construction has succeeded by invoking the\\nAll Writs Act.\\n\\n The question then becomes “whether ‘the risk of an\\nerroneous deprivation’ would be reduced” by “procedural\\n\\n 16\\nsafeguards” that would prevent pipeline construction from\\nbeginning until the Commission acts on the merits of rehearing\\nrequests. UDC Chairs Chapter, American Ass’n of Univ.\\nProfessors v. Board of Trustees, 56 F.3d 1469, 1474 (D.C. Cir.\\n1995) (quoting Mathews, 424 U.S. at 335).\\n\\n The answer is plain as day. Requiring the Commission to\\nact in a timely manner on applications for rehearing would\\nallow landowners to obtain both final agency and federal court\\nreview before construction starts. Or the Commission could\\nachieve the same result just by declining to issue construction\\norders until it resolves certificate rehearing requests on the\\nmerits.\\n\\n 3\\n\\n The Commission has no legitimate interest that outweighs\\nthe unfairness and risk of harm imposed by the current regime,\\nespecially given how easy it would be for the Commission to\\nfix the problem.\\n\\n The Commission insists that tolling orders are necessary\\nbecause it takes more than 30 days to resolve the issues raised\\nin applications for rehearing. Oral Arg. Tr. 33. The short\\nanswer is that the Commission should raise that argument with\\nCongress, which prescribed the 30-day timeframe for decision.\\nAn equally short answer is that the Commission could try\\nworking with rehearing applicants to obtain more time by, for\\nexample, agreeing to hold its hand on construction orders or\\nstaying initial certificate orders to forestall eminent domain\\nproceedings before it takes final action.\\n\\n If that is too administratively burdensome, then the\\nCommission could try the easiest path of all: take absolutely\\nno action on the rehearing application. That would have the\\n\\n 17\\neffect of denying the request as a matter of law. See 15 U.S.C.\\n§ 717r(a). And that approach would have opened the\\ncourthouse doors to the Homeowners four months before the\\neminent domain decision and five months before construction\\nstarted. See id. § 717r(b).\\n\\n Lastly, the Commission appeals to the “public need” for\\nthe pipeline itself. Oral Arg. Tr. 35. That is pure question-\\nbegging. The public need for the Atlantic Sunrise Project—\\nwhether the Commission’s finding of public convenience and\\nnecessity was lawful—is precisely the question for which the\\nErb and Hoffman families seek judicial review.\\n\\n Anyhow, it is well-settled that “the right to procedural due\\nprocess is ‘absolute’ in the sense that it does not depend upon\\nthe merits of a claimant’s substantive assertions[.]” Carey v.\\nPiphus, 435 U.S. 247, 266 (1978); see also Fuentes, 407 U.S.\\nat 87 (“The right to be heard does not depend upon an advance\\nshowing that one will surely prevail at the hearing.”). In other\\nwords, “[t]o one who protests against the taking of his property\\nwithout due process of law, it is no answer to say that in his\\nparticular case due process of law would have led to the same\\nresult[.]” Coe v. Armour Fertilizer Works, 237 U.S. 413, 424\\n(1915). The question is not whether the Erbs and Hoffmans\\nwill win; it is only whether they have a right to be fairly and\\ntimely heard before their “federal claim dies aborning” in the\\nCommission’s process. Knick, 139 S. Ct. at 2167.\\n\\n * * * * *\\n\\n In cases involving private property rights, the Commission\\nhas transformed this court’s decisions upholding its tolling\\norders into a bureaucratic purgatory that only Dante could love.\\nWhile I acknowledge that circuit precedent currently forecloses\\nthe Homeowners’ constitutional challenge to the tolling orders,\\n\\n 18\\nthis case starkly illustrates why a second look by us or by the\\nCommission is overdue.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4422503/\", \"author_raw\": \"MILLETT, Circuit Judge, concurring\"}]}"},"judge1":{"kind":"string","value":"GARLAND"},"judge2":{"kind":"string","value":"TATEL"},"judge3":{"kind":"string","value":"MILLETT"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":4,"string":"4"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":2,"string":"2"},"other_count":{"kind":"number","value":2,"string":"2"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4645250/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":1,"string":"1"},"030concurrence_count":{"kind":"number","value":1,"string":"1"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: ALLEGHENY DEFENSE PROJECT, Et Al., Petitioners v. FEDERAL ENERGY REGULATORY COMMISSION, Respondent Anadarko Energy Services Company, Et Al., Intervenors\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-02\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: Yes\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued December 7, 2018 Decided August 2, 2019\n\n No. 17-1098\n\n ALLEGHENY DEFENSE PROJECT, ET AL.,\n PETITIONERS\n\n v.\n\n FEDERAL ENERGY REGULATORY COMMISSION,\n RESPONDENT\n\n ANADARKO ENERGY SERVICES COMPANY, ET AL.,\n INTERVENORS\n\n\n Consolidated with 17-1128, 17-1263, 18-1030\n\n\n On Petitions for Review of Orders of the\n Federal Energy Regulatory Commission\n\n\n Elizabeth F. Benson argued the cause for petitioners\nAllegheny Defense Project, et al. Siobhan K. Cole argued the\ncause for petitioners Hilltop Hollow Limited Partnership, et al.\nWith them on the joint briefs was Benjamin A. Luckett.\nMichael N. Onufrak and Derek O. Teaney entered appearances.\n\n Beth G. Pacella, Deputy Solicitor, Federal Energy\nRegulatory Commission, argued the cause for respondent.\nWith her on the brief were James P. Danly, General Counsel,\n\n 2\nand Robert H. Solomon, Solicitor. Anand Viswanathan,\nAttorney, entered an appearance.\n\n John F. Stoviak argued the cause for intervenors. With\nhim on the joint brief were Pamela S. Goodwin, Elizabeth U.\nWitmer, Patrick F. Nugent, Kevin M. Sweeney, Jesse Stuart\nUnkenholz, and Scott Borden Grover.\n\n Before: GARLAND, Chief Judge, and TATEL and MILLETT,\nCircuit Judges.\n\n Opinion for the court filed PER CURIAM.\n\n Concurring opinion filed by Circuit Judge MILLETT.\n\n PER CURIAM: In February 2017, the Federal Energy\nRegulatory Commission approved a natural gas pipeline\nexpansion called the Atlantic Sunrise Project, which stretches\nfrom northern Pennsylvania, across the Carolinas, and into\nAlabama. Environmental Associations whose members live\nand work in the areas affected by the Project and individual\nHomeowners whose property was used for the Project seek\nreview of the Commission’s orders permitting the\nTranscontinental Gas Pipe Line Company to move forward\nwith the pipeline expansion. Because the challenges to the\nCommission’s decision cannot surmount the deferential\nstandards of agency review and binding circuit precedent, we\ndeny the petitions.\n\n I\n\n A\n\n Under the Natural Gas Act, 15 U.S.C. § 717 et seq., a\ncompany wishing to construct a natural gas pipeline must first\n\n 3\nobtain a certificate of “public convenience and necessity” from\nthe Federal Energy Regulatory Commission. See id.\n§ 717f(c); Myersville Citizens for a Rural Community, Inc. v.\nFERC, 783 F.3d 1301, 1307 (D.C. Cir. 2015). The\nCommission “shall * * * issue[]” the certificate if it finds that\nthe proposed project “is or will be required by the present or\nfuture public convenience and necessity.” 15 U.S.C.\n§ 717f(e).\n\n When a pipeline company files a certificate application,\nthe Commission reviews it under criteria set forth in its\nCertificate Policy Statement, 88 FERC ¶ 61,227 (1999),\nclarified, 90 FERC ¶ 61,128, further clarified, 92 FERC\n¶ 61,094 (2000). The Certificate Policy Statement directs the\nCommission to consider whether the project meets a market\nneed and whether the public benefits of the project outweigh\nthe harms. See Sierra Club v. FERC, 867 F.3d 1357, 1379\n(D.C. Cir. 2017). If market need and public benefit are both\nestablished, the Commission will issue a certificate authorizing\nthe pipeline’s construction. Id. Once that certificate is\ngranted, the Natural Gas Act empowers the private certificate\nholder to exercise eminent domain authority if it “cannot\nacquire by contract, or is unable to agree with the owner of\nproperty to the compensation to be paid for, the necessary\nright-of-way to construct, operate, and maintain a pipeline[.]”\n15 U.S.C. § 717f(h).\n\n A party wishing to challenge the Commission’s issuance\nof a certificate of public convenience and necessity must file a\npetition for rehearing with the Commission. 15 U.S.C.\n§ 717r(a). Until the Commission disposes of that rehearing\npetition, the agency action is not final for purposes of judicial\nreview. See id. § 717r(a)–(b); Clifton Power Corp. v. FERC,\n294 F.3d 108, 110–111 (D.C. Cir. 2002). The filing and\ndisposition of such a rehearing petition is thus a mandatory\n\n 4\nprerequisite to obtaining judicial review of the Commission’s\naction. See Delaware Riverkeeper Network v. FERC, 857\nF.3d 388, 399 (D.C. Cir. 2017); Clifton Power Corp., 294 F.3d\nat 110–111. Congress directed that petitions for rehearing\nmay be “deemed to have been denied” if the Commission has\nnot “act[ed] upon the application for rehearing within thirty\ndays after it is filed[.]” 15 U.S.C. § 717r(a).\n\n The National Environmental Policy Act (“NEPA”), 42\nU.S.C. § 4321 et seq., “require[s] the Commission to consider\nand disclose the environmental effects of the actions it\ncertifies.” Delaware Riverkeeper Network, 857 F.3d at 394\n(internal quotation marks omitted). The Commission may\nfulfill this requirement by compiling an Environmental Impact\nStatement, which must consider, among other things, the\nproposed project’s “indirect” environmental effects. See 42\nU.S.C. § 4332(C); 40 C.F.R. §§ 1508.8, 1508.25(c).\n\n B\n\n In March 2015, the Transcontinental Gas Pipe Line\nCompany (“Transco”) applied for a certificate of public\nconvenience and necessity to build the Atlantic Sunrise Project.\nAfter notice and public comment, the Commission issued a\nfinal Environmental Impact Statement in December 2016. In\nthat Statement, the Commission concluded that “neither\nconstruction nor operation of the Project would significantly\ncontribute to [greenhouse-gas] cumulative effects or climate\nchange.” J.A. 323.\n\n On February 3, 2017, the Commission granted Transco its\nrequested certificate of public convenience and necessity. 158\nFERC ¶ 61,125 (2017) (“Certificate Order”). Environmental\nAssociations and the private Homeowners whose land would\nbecome subject to an easement for the Project both filed\n\n 5\npetitions for rehearing with the Commission, along with\nmotions for a stay of construction pending disposition of their\npetitions. Before Congress’s 30-day deadline for action on the\nrehearing petitions expired, the Commission “granted”\nrehearing, but only “for the limited purpose of further\nconsideration.” J.A. 600 (“Certificate Tolling Order”). The\nCommission took no action on the stay motions for more than\nfive months, and then denied them.\n\n In late August, a Pennsylvania federal district court\npresiding over Transco’s eminent domain action entered an\norder that declared Transco’s “right to immediate possession of\nthe properties in question,” based on the presumed validity of\nFERC’s Certificate Order. Transcontinental Gas Pipe Line\nCo. v. Permanent Easements for 2.14 Acres & Temp.\nEasements for 3.59 Acres in Conestoga Township, Lancaster\nCounty, Pa., Tax Parcel No. 1201606900000, 2017 WL\n3624250, at *1, *3 (E.D. Pa. Aug. 23, 2017) (rejecting the\nHomeowners’ objections as “attacks on the FERC order itself,”\nwhich “can only be challenged in front of FERC, and then in\nthe United States Court of Appeals for the District of Columbia\nCircuit”), aff’d, 907 F.3d 725 (3d Cir. 2018).\n\n On September 15, 2017, while the Environmental\nAssociations’ and Homeowners’ petitions for rehearing were\nstill pending, the Commission issued an order authorizing\nTransco to begin construction of the Project. J.A. 616\n(“Construction Order”). Transco broke ground that same day.\n\n The Environmental Associations promptly moved for\nrehearing of the Construction Order and an order halting\nconstruction. As its 30-day statutory deadline for action on\nthe Construction Order rehearing petition approached, the\nCommission again issued an order granting rehearing “for the\n\n 6\nlimited purpose of further consideration.” J.A. 815\n(“Construction Tolling Order”).\n\n The Commission eventually denied the Homeowners’ and\nEnvironmental Associations’ petitions for rehearing in\nDecember 2017, more than nine months after rehearing was\nsought and three months after construction began. 161 FERC\n¶ 61,250 (2017) (“Certificate Rehearing Order”). Three\nmonths after that—and nearly six months after construction\ncommenced—the Commission denied rehearing of the\nConstruction Order. 162 FERC ¶ 61,192 (2018)\n(“Construction Rehearing Order”).\n\n II\n\n This consolidated case arises out of four petitions for\nreview collectively challenging the Certificate Order, the\nCertificate Tolling Order, the Construction Order, the\nConstruction Tolling Order, and the Certificate Rehearing\nOrder. Because the Homeowners and Environmental\nAssociations both properly sought rehearing of the Certificate\nRehearing Order, which encompasses all of their claims for our\nreview and is the final agency decision greenlighting the\nProject, this court has jurisdiction. See 15 U.S.C. § 717r. 1\n\n The Commission’s factual findings are conclusive so long\nas they are supported by substantial evidence. See 15 U.S.C.\n§ 717(b). Where the Commission “has examined the relevant\nconsiderations and articulated a satisfactory explanation for its\naction, including a rational connection between the facts found\nand the choice made,” we must uphold its decision. FERC v.\n\n 1\n We therefore deny as moot the motions to dismiss the\npetitions filed before the Certificate Rehearing Order issued, Nos.\n17-1098 and 17-1128.\n\n 7\nElectric Power Supply Ass’n, 136 S. Ct. 760, 782 (2016)\n(formatting altered).\n\n The Homeowners and Environmental Associations argue\nthat the Commission’s Certificate Order suffers from fatal\nsubstantive and procedural flaws. Specifically, they argue\nthat the Commission improperly conducted its environmental\nassessment under NEPA, failed to substantiate market need for\nthe Project as required by the Natural Gas Act, and denied them\ndue process by authorizing construction to commence before\nthe issuance of the Certificate Order could be judicially\nreviewed. None of those arguments succeed.\n\n A\n\n The administrative record forecloses the Homeowners’\nand Environmental Associations’ three NEPA arguments.\n\n First, the Homeowners and Environmental Associations\nargue that the Commission did not factor downstream\ngreenhouse-gas emissions into its evaluation of the Project’s\nenvironmental impacts.\n\n The Homeowners and Environmental Associations are\ncorrect that customers’ burning of the natural gas that the\nProject transports will produce greenhouse-gas emissions.\nSee Sierra Club, 867 F.3d at 1374. They are also correct that\nNEPA required the Commission to consider both the direct and\nindirect environmental effects of the Project, and that, despite\nwhat the Commission argues, the downstream greenhouse-gas\nemissions are just such an indirect effect. See id.; 40 C.F.R.\n§ 1502.16(b).\n\n But that is as far as the argument gets them, because the\nCommission already took the steps the Homeowners and\n\n 8\nEnvironmental Associations request. The Commission\naddressed downstream emissions in the Environmental Impact\nStatement, the Certificate Order, and the Certificate Rehearing\nOrder. In particular, the Commission estimated the amount of\nCO2 emissions resulting from the gas that the Project would\ntransport and predicted that those emissions would be partially\noffset by reductions in higher carbon-emitting fuel that the\nProject’s natural gas would replace. Neither the Homeowners\nnor the Environmental Associations have identified what more\nthe Commission should have said. That failure is fatal.\nUnsubstantiated objections are not enough to stop an agency’s\naction.\n\n Second, the Homeowners and Environmental Associations\nassert that the Commission impermissibly segmented its\nenvironmental review by failing to consider the synergistic\neffect of the Project on emissions associated with a different\npipeline—the Southeast Market Pipeline. Improper\nsegmentation occurs when FERC creates separate\nEnvironmental Impact Statements for interconnected pipelines\nthat should have been evaluated as a single project. See City\nof Boston Delegation v. FERC, 897 F.3d 241, 251–252 (D.C.\nCir. 2018).\n\n But NEPA requires more to make out an improper\nsegmentation claim than the Homeowners and Environmental\nAssociations have offered. Specifically, the Commission has\nno duty to consider the environmental effects of a separate\nproject if the project in question has “substantial independent\nutility.” City of Boston Delegation, 897 F.3d at 252 (internal\nquotation marks omitted); see also Coalition on Sensible\nTransp., Inc. v. Dole, 826 F.2d 60, 69 (D.C. Cir. 1987) (asking\n“whether one project will serve a significant purpose even if a\nsecond related project is not built”).\n\n 9\n The record supports the Commission’s determination that\nthe Southeast Market Pipeline and the Project are\nindependently justified and that each has its own distinct utility.\nCity of Boston Delegation, 897 F.3d at 252. That is because,\neven if the Project were never built, the Southeast Market\nPipeline still would be connected to enough natural gas supply\nto exceed its capacity. J.A. 843. On this record, that\ndisproves any claim of improper segmentation.\n\n Third, the Homeowners and Environmental Associations\ncontend that the Commission was deficient in its analysis of an\nalternative site for the Project called the Conestoga Route.\nUnder NEPA, agencies must “identify the reasonable\nalternatives to the contemplated action,” Minisink Residents for\nEnvtl. Pres. & Safety v. FERC, 762 F.3d 97, 102 (D.C. Cir.\n2014), and “discuss[] * * * the relevant issues and opposing\nviewpoints,” Myersville, 783 F.3d at 1324–1325 (quoting\nNevada v. Department of Energy, 457 F.3d 78, 93 (D.C. Cir.\n2006)). This requirement helps ensure that agency actions are\n“fully informed and well-considered.” Id. at 1325 (internal\nquotation marks omitted).\n\n The Commission’s consideration of the Conestoga Route\npasses NEPA muster. The Commission “look[ed] hard at the\nenvironmental effects of its decision” by comparing the\nproposed route with the Conestoga Route across highly\ndetailed criteria. Minisink, 762 F.3d at 102 (formatting\naltered). 2 In doing so, the Commission openly grappled with\nthe factors favoring the Conestoga Route and reasonably\nexplained why the proposed route was nonetheless superior.\n\n 2\n The “Waterbodies crossed” criterion, for example, quantifies\nwaterbodies for “Trout Stocking,” “High Quality Cold Water\nFishes,” “Cold Water Fishes,” “High Quality Warm Water Fishes,”\nand “Warm Water Fishes.” J.A. 318.\n\n 10\nIt acknowledged that the Conestoga Route was shorter and\nwould follow existing rights-of-way for more of its length than\nthe proposed route would, but it placed more weight on the\nproposed route’s virtues, including that it would cross fewer\nrecreational areas and would avoid potentially dangerous\nelevation changes. That suffices for NEPA. See Myersville,\n783 F.3d at 1308; Minisink, 762 F.3d at 112.\n\n B\n\n The Homeowners and Environmental Associations\nseparately claim that the Commission’s market-need\ndetermination violated the Natural Gas Act. A showing of\nmarket need is a precondition for the Commission to issue a\ncertificate of public convenience and necessity. See Sierra\nClub, 867 F.3d at 1379; Myersville, 783 F.3d at 1309; 90 FERC\nat ¶ 61,392; 88 FERC at ¶ 61,746. The certificate applicant\ncan make its required showing of market need for the pipeline\n“by presenting evidence of ‘preconstruction contracts’ for gas\ntransportation service.” Sierra Club, 867 F.3d at 1379.\n\n The Commission held Transco to this obligation. Its\nfinding of market need rested on the existence of contracts with\nshippers for 100% of the Project’s capacity. That alone is\nenough. See Sierra Club, 867 F.3d at 1379; Myersville, 783\nF.3d at 1311; Minisink, 762 F.3d at 111 n.10. And the\nCommission did not stop there. It also relied on comments by\ntwo shippers and one end-user, as well as a study submitted by\none of the Environmental Associations, all of which reinforced\nthe demand for the natural gas shipments.\n\n 11\n C\n\n Finally, the Environmental Associations and Homeowners\nargue that the Commission’s authorization for construction to\ngo forward while their rehearing petitions were still pending—\nand thus before the Commission’s decision was final and\njudicially reviewable—denied them due process. Circuit\nprecedent forecloses their claims.\n\n A due-process claim turns on two essential inquiries.\nFirst, is there a “liberty or property interest of which a person\nhas been deprived”? Swarthout v. Cooke, 562 U.S. 216, 219\n(2011). Second, were the “procedures followed” by the\ngovernment in encroaching on those interests “constitutionally\nsufficient”? Id.\n\n The Environmental Associations run into trouble on the\nfirst prong of that test. They assert that liberty and property\ninterests were bestowed upon them by the Pennsylvania\nConstitution’s Environmental Rights Amendment, Pa. Const.\nArt. 1, § 27, and the Natural Gas Act’s review procedures, 15\nU.S.C. §§ 717f(c)(1)(B), 717r(a)–(b). This court has already\nrejected those claims, and those decisions control here. See\nDelaware Riverkeeper Network v. FERC, 895 F.3d 102, 110\n(D.C. Cir. 2018) (Pennsylvania Constitution); Griffith v.\nFederal Labor Relations Auth., 842 F.2d 487, 495 (D.C. Cir.\n1988) (Natural Gas Act review procedures); see generally\nLaShawn A. v. Barry, 87 F.3d 1389, 1395 (D.C. Cir. 1996)\n(“One three-judge panel * * * does not have the authority to\noverrule another three-judge panel of the court.”).\n\n The Homeowners, for their part, lose on the second prong\nof the due-process analysis. They argue that the\nCommission’s delay in acting on their rehearing petitions while\nauthorizing construction to start “denied [them] the right to be\n\n 12\nheard on whether Transco’s taking of their property actually\nsatisfies the public use requirement of the Fifth Amendment.”\nPetitioners’ Opening Br. 47.\n\n But, again, circuit precedent says otherwise. We have\nheld that, as long as FERC’s public-convenience-and-necessity\ndetermination is not legally deficient, it necessarily satisfies the\nFifth Amendment’s public-use requirement. See Midcoast\nInterstate Transmission, Inc. v. FERC, 198 F.3d 960, 973 (D.C.\nCir. 2000). The Homeowners make no claim that they were\ndeprived of a meaningful opportunity to be heard as part of the\nCommission’s proceedings leading up to its issuance of the\nCertificate Order, and they make no effort to distinguish (or\neven acknowledge) our holding in Midcoast.\n\n * * *\n\n For the foregoing reasons, the petitions for review are\ndenied.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: ALLEGHENY DEFENSE PROJECT, Et Al., Petitioners v. FEDERAL ENERGY REGULATORY COMMISSION, Respondent Anadarko Energy Services Company, Et Al., Intervenors\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-02\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: Yes\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued December 7, 2018 Decided August 2, 2019\\n\\n No. 17-1098\\n\\n ALLEGHENY DEFENSE PROJECT, ET AL.,\\n PETITIONERS\\n\\n v.\\n\\n FEDERAL ENERGY REGULATORY COMMISSION,\\n RESPONDENT\\n\\n ANADARKO ENERGY SERVICES COMPANY, ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 17-1128, 17-1263, 18-1030\\n\\n\\n On Petitions for Review of Orders of the\\n Federal Energy Regulatory Commission\\n\\n\\n Elizabeth F. Benson argued the cause for petitioners\\nAllegheny Defense Project, et al. Siobhan K. Cole argued the\\ncause for petitioners Hilltop Hollow Limited Partnership, et al.\\nWith them on the joint briefs was Benjamin A. Luckett.\\nMichael N. Onufrak and Derek O. Teaney entered appearances.\\n\\n Beth G. Pacella, Deputy Solicitor, Federal Energy\\nRegulatory Commission, argued the cause for respondent.\\nWith her on the brief were James P. Danly, General Counsel,\\n\\n 2\\nand Robert H. Solomon, Solicitor. Anand Viswanathan,\\nAttorney, entered an appearance.\\n\\n John F. Stoviak argued the cause for intervenors. With\\nhim on the joint brief were Pamela S. Goodwin, Elizabeth U.\\nWitmer, Patrick F. Nugent, Kevin M. Sweeney, Jesse Stuart\\nUnkenholz, and Scott Borden Grover.\\n\\n Before: GARLAND, Chief Judge, and TATEL and MILLETT,\\nCircuit Judges.\\n\\n Opinion for the court filed PER CURIAM.\\n\\n Concurring opinion filed by Circuit Judge MILLETT.\\n\\n PER CURIAM: In February 2017, the Federal Energy\\nRegulatory Commission approved a natural gas pipeline\\nexpansion called the Atlantic Sunrise Project, which stretches\\nfrom northern Pennsylvania, across the Carolinas, and into\\nAlabama. Environmental Associations whose members live\\nand work in the areas affected by the Project and individual\\nHomeowners whose property was used for the Project seek\\nreview of the Commission’s orders permitting the\\nTranscontinental Gas Pipe Line Company to move forward\\nwith the pipeline expansion. Because the challenges to the\\nCommission’s decision cannot surmount the deferential\\nstandards of agency review and binding circuit precedent, we\\ndeny the petitions.\\n\\n I\\n\\n A\\n\\n Under the Natural Gas Act, 15 U.S.C. § 717 et seq., a\\ncompany wishing to construct a natural gas pipeline must first\\n\\n 3\\nobtain a certificate of “public convenience and necessity” from\\nthe Federal Energy Regulatory Commission. See id.\\n§ 717f(c); Myersville Citizens for a Rural Community, Inc. v.\\nFERC, 783 F.3d 1301, 1307 (D.C. Cir. 2015). The\\nCommission “shall * * * issue[]” the certificate if it finds that\\nthe proposed project “is or will be required by the present or\\nfuture public convenience and necessity.” 15 U.S.C.\\n§ 717f(e).\\n\\n When a pipeline company files a certificate application,\\nthe Commission reviews it under criteria set forth in its\\nCertificate Policy Statement, 88 FERC ¶ 61,227 (1999),\\nclarified, 90 FERC ¶ 61,128, further clarified, 92 FERC\\n¶ 61,094 (2000). The Certificate Policy Statement directs the\\nCommission to consider whether the project meets a market\\nneed and whether the public benefits of the project outweigh\\nthe harms. See Sierra Club v. FERC, 867 F.3d 1357, 1379\\n(D.C. Cir. 2017). If market need and public benefit are both\\nestablished, the Commission will issue a certificate authorizing\\nthe pipeline’s construction. Id. Once that certificate is\\ngranted, the Natural Gas Act empowers the private certificate\\nholder to exercise eminent domain authority if it “cannot\\nacquire by contract, or is unable to agree with the owner of\\nproperty to the compensation to be paid for, the necessary\\nright-of-way to construct, operate, and maintain a pipeline[.]”\\n15 U.S.C. § 717f(h).\\n\\n A party wishing to challenge the Commission’s issuance\\nof a certificate of public convenience and necessity must file a\\npetition for rehearing with the Commission. 15 U.S.C.\\n§ 717r(a). Until the Commission disposes of that rehearing\\npetition, the agency action is not final for purposes of judicial\\nreview. See id. § 717r(a)–(b); Clifton Power Corp. v. FERC,\\n294 F.3d 108, 110–111 (D.C. Cir. 2002). The filing and\\ndisposition of such a rehearing petition is thus a mandatory\\n\\n 4\\nprerequisite to obtaining judicial review of the Commission’s\\naction. See Delaware Riverkeeper Network v. FERC, 857\\nF.3d 388, 399 (D.C. Cir. 2017); Clifton Power Corp., 294 F.3d\\nat 110–111. Congress directed that petitions for rehearing\\nmay be “deemed to have been denied” if the Commission has\\nnot “act[ed] upon the application for rehearing within thirty\\ndays after it is filed[.]” 15 U.S.C. § 717r(a).\\n\\n The National Environmental Policy Act (“NEPA”), 42\\nU.S.C. § 4321 et seq., “require[s] the Commission to consider\\nand disclose the environmental effects of the actions it\\ncertifies.” Delaware Riverkeeper Network, 857 F.3d at 394\\n(internal quotation marks omitted). The Commission may\\nfulfill this requirement by compiling an Environmental Impact\\nStatement, which must consider, among other things, the\\nproposed project’s “indirect” environmental effects. See 42\\nU.S.C. § 4332(C); 40 C.F.R. §§ 1508.8, 1508.25(c).\\n\\n B\\n\\n In March 2015, the Transcontinental Gas Pipe Line\\nCompany (“Transco”) applied for a certificate of public\\nconvenience and necessity to build the Atlantic Sunrise Project.\\nAfter notice and public comment, the Commission issued a\\nfinal Environmental Impact Statement in December 2016. In\\nthat Statement, the Commission concluded that “neither\\nconstruction nor operation of the Project would significantly\\ncontribute to [greenhouse-gas] cumulative effects or climate\\nchange.” J.A. 323.\\n\\n On February 3, 2017, the Commission granted Transco its\\nrequested certificate of public convenience and necessity. 158\\nFERC ¶ 61,125 (2017) (“Certificate Order”). Environmental\\nAssociations and the private Homeowners whose land would\\nbecome subject to an easement for the Project both filed\\n\\n 5\\npetitions for rehearing with the Commission, along with\\nmotions for a stay of construction pending disposition of their\\npetitions. Before Congress’s 30-day deadline for action on the\\nrehearing petitions expired, the Commission “granted”\\nrehearing, but only “for the limited purpose of further\\nconsideration.” J.A. 600 (“Certificate Tolling Order”). The\\nCommission took no action on the stay motions for more than\\nfive months, and then denied them.\\n\\n In late August, a Pennsylvania federal district court\\npresiding over Transco’s eminent domain action entered an\\norder that declared Transco’s “right to immediate possession of\\nthe properties in question,” based on the presumed validity of\\nFERC’s Certificate Order. Transcontinental Gas Pipe Line\\nCo. v. Permanent Easements for 2.14 Acres & Temp.\\nEasements for 3.59 Acres in Conestoga Township, Lancaster\\nCounty, Pa., Tax Parcel No. 1201606900000, 2017 WL\\n3624250, at *1, *3 (E.D. Pa. Aug. 23, 2017) (rejecting the\\nHomeowners’ objections as “attacks on the FERC order itself,”\\nwhich “can only be challenged in front of FERC, and then in\\nthe United States Court of Appeals for the District of Columbia\\nCircuit”), aff’d, 907 F.3d 725 (3d Cir. 2018).\\n\\n On September 15, 2017, while the Environmental\\nAssociations’ and Homeowners’ petitions for rehearing were\\nstill pending, the Commission issued an order authorizing\\nTransco to begin construction of the Project. J.A. 616\\n(“Construction Order”). Transco broke ground that same day.\\n\\n The Environmental Associations promptly moved for\\nrehearing of the Construction Order and an order halting\\nconstruction. As its 30-day statutory deadline for action on\\nthe Construction Order rehearing petition approached, the\\nCommission again issued an order granting rehearing “for the\\n\\n 6\\nlimited purpose of further consideration.” J.A. 815\\n(“Construction Tolling Order”).\\n\\n The Commission eventually denied the Homeowners’ and\\nEnvironmental Associations’ petitions for rehearing in\\nDecember 2017, more than nine months after rehearing was\\nsought and three months after construction began. 161 FERC\\n¶ 61,250 (2017) (“Certificate Rehearing Order”). Three\\nmonths after that—and nearly six months after construction\\ncommenced—the Commission denied rehearing of the\\nConstruction Order. 162 FERC ¶ 61,192 (2018)\\n(“Construction Rehearing Order”).\\n\\n II\\n\\n This consolidated case arises out of four petitions for\\nreview collectively challenging the Certificate Order, the\\nCertificate Tolling Order, the Construction Order, the\\nConstruction Tolling Order, and the Certificate Rehearing\\nOrder. Because the Homeowners and Environmental\\nAssociations both properly sought rehearing of the Certificate\\nRehearing Order, which encompasses all of their claims for our\\nreview and is the final agency decision greenlighting the\\nProject, this court has jurisdiction. See 15 U.S.C. § 717r. 1\\n\\n The Commission’s factual findings are conclusive so long\\nas they are supported by substantial evidence. See 15 U.S.C.\\n§ 717(b). Where the Commission “has examined the relevant\\nconsiderations and articulated a satisfactory explanation for its\\naction, including a rational connection between the facts found\\nand the choice made,” we must uphold its decision. FERC v.\\n\\n 1\\n We therefore deny as moot the motions to dismiss the\\npetitions filed before the Certificate Rehearing Order issued, Nos.\\n17-1098 and 17-1128.\\n\\n 7\\nElectric Power Supply Ass’n, 136 S. Ct. 760, 782 (2016)\\n(formatting altered).\\n\\n The Homeowners and Environmental Associations argue\\nthat the Commission’s Certificate Order suffers from fatal\\nsubstantive and procedural flaws. Specifically, they argue\\nthat the Commission improperly conducted its environmental\\nassessment under NEPA, failed to substantiate market need for\\nthe Project as required by the Natural Gas Act, and denied them\\ndue process by authorizing construction to commence before\\nthe issuance of the Certificate Order could be judicially\\nreviewed. None of those arguments succeed.\\n\\n A\\n\\n The administrative record forecloses the Homeowners’\\nand Environmental Associations’ three NEPA arguments.\\n\\n First, the Homeowners and Environmental Associations\\nargue that the Commission did not factor downstream\\ngreenhouse-gas emissions into its evaluation of the Project’s\\nenvironmental impacts.\\n\\n The Homeowners and Environmental Associations are\\ncorrect that customers’ burning of the natural gas that the\\nProject transports will produce greenhouse-gas emissions.\\nSee Sierra Club, 867 F.3d at 1374. They are also correct that\\nNEPA required the Commission to consider both the direct and\\nindirect environmental effects of the Project, and that, despite\\nwhat the Commission argues, the downstream greenhouse-gas\\nemissions are just such an indirect effect. See id.; 40 C.F.R.\\n§ 1502.16(b).\\n\\n But that is as far as the argument gets them, because the\\nCommission already took the steps the Homeowners and\\n\\n 8\\nEnvironmental Associations request. The Commission\\naddressed downstream emissions in the Environmental Impact\\nStatement, the Certificate Order, and the Certificate Rehearing\\nOrder. In particular, the Commission estimated the amount of\\nCO2 emissions resulting from the gas that the Project would\\ntransport and predicted that those emissions would be partially\\noffset by reductions in higher carbon-emitting fuel that the\\nProject’s natural gas would replace. Neither the Homeowners\\nnor the Environmental Associations have identified what more\\nthe Commission should have said. That failure is fatal.\\nUnsubstantiated objections are not enough to stop an agency’s\\naction.\\n\\n Second, the Homeowners and Environmental Associations\\nassert that the Commission impermissibly segmented its\\nenvironmental review by failing to consider the synergistic\\neffect of the Project on emissions associated with a different\\npipeline—the Southeast Market Pipeline. Improper\\nsegmentation occurs when FERC creates separate\\nEnvironmental Impact Statements for interconnected pipelines\\nthat should have been evaluated as a single project. See City\\nof Boston Delegation v. FERC, 897 F.3d 241, 251–252 (D.C.\\nCir. 2018).\\n\\n But NEPA requires more to make out an improper\\nsegmentation claim than the Homeowners and Environmental\\nAssociations have offered. Specifically, the Commission has\\nno duty to consider the environmental effects of a separate\\nproject if the project in question has “substantial independent\\nutility.” City of Boston Delegation, 897 F.3d at 252 (internal\\nquotation marks omitted); see also Coalition on Sensible\\nTransp., Inc. v. Dole, 826 F.2d 60, 69 (D.C. Cir. 1987) (asking\\n“whether one project will serve a significant purpose even if a\\nsecond related project is not built”).\\n\\n 9\\n The record supports the Commission’s determination that\\nthe Southeast Market Pipeline and the Project are\\nindependently justified and that each has its own distinct utility.\\nCity of Boston Delegation, 897 F.3d at 252. That is because,\\neven if the Project were never built, the Southeast Market\\nPipeline still would be connected to enough natural gas supply\\nto exceed its capacity. J.A. 843. On this record, that\\ndisproves any claim of improper segmentation.\\n\\n Third, the Homeowners and Environmental Associations\\ncontend that the Commission was deficient in its analysis of an\\nalternative site for the Project called the Conestoga Route.\\nUnder NEPA, agencies must “identify the reasonable\\nalternatives to the contemplated action,” Minisink Residents for\\nEnvtl. Pres. & Safety v. FERC, 762 F.3d 97, 102 (D.C. Cir.\\n2014), and “discuss[] * * * the relevant issues and opposing\\nviewpoints,” Myersville, 783 F.3d at 1324–1325 (quoting\\nNevada v. Department of Energy, 457 F.3d 78, 93 (D.C. Cir.\\n2006)). This requirement helps ensure that agency actions are\\n“fully informed and well-considered.” Id. at 1325 (internal\\nquotation marks omitted).\\n\\n The Commission’s consideration of the Conestoga Route\\npasses NEPA muster. The Commission “look[ed] hard at the\\nenvironmental effects of its decision” by comparing the\\nproposed route with the Conestoga Route across highly\\ndetailed criteria. Minisink, 762 F.3d at 102 (formatting\\naltered). 2 In doing so, the Commission openly grappled with\\nthe factors favoring the Conestoga Route and reasonably\\nexplained why the proposed route was nonetheless superior.\\n\\n 2\\n The “Waterbodies crossed” criterion, for example, quantifies\\nwaterbodies for “Trout Stocking,” “High Quality Cold Water\\nFishes,” “Cold Water Fishes,” “High Quality Warm Water Fishes,”\\nand “Warm Water Fishes.” J.A. 318.\\n\\n 10\\nIt acknowledged that the Conestoga Route was shorter and\\nwould follow existing rights-of-way for more of its length than\\nthe proposed route would, but it placed more weight on the\\nproposed route’s virtues, including that it would cross fewer\\nrecreational areas and would avoid potentially dangerous\\nelevation changes. That suffices for NEPA. See Myersville,\\n783 F.3d at 1308; Minisink, 762 F.3d at 112.\\n\\n B\\n\\n The Homeowners and Environmental Associations\\nseparately claim that the Commission’s market-need\\ndetermination violated the Natural Gas Act. A showing of\\nmarket need is a precondition for the Commission to issue a\\ncertificate of public convenience and necessity. See Sierra\\nClub, 867 F.3d at 1379; Myersville, 783 F.3d at 1309; 90 FERC\\nat ¶ 61,392; 88 FERC at ¶ 61,746. The certificate applicant\\ncan make its required showing of market need for the pipeline\\n“by presenting evidence of ‘preconstruction contracts’ for gas\\ntransportation service.” Sierra Club, 867 F.3d at 1379.\\n\\n The Commission held Transco to this obligation. Its\\nfinding of market need rested on the existence of contracts with\\nshippers for 100% of the Project’s capacity. That alone is\\nenough. See Sierra Club, 867 F.3d at 1379; Myersville, 783\\nF.3d at 1311; Minisink, 762 F.3d at 111 n.10. And the\\nCommission did not stop there. It also relied on comments by\\ntwo shippers and one end-user, as well as a study submitted by\\none of the Environmental Associations, all of which reinforced\\nthe demand for the natural gas shipments.\\n\\n 11\\n C\\n\\n Finally, the Environmental Associations and Homeowners\\nargue that the Commission’s authorization for construction to\\ngo forward while their rehearing petitions were still pending—\\nand thus before the Commission’s decision was final and\\njudicially reviewable—denied them due process. Circuit\\nprecedent forecloses their claims.\\n\\n A due-process claim turns on two essential inquiries.\\nFirst, is there a “liberty or property interest of which a person\\nhas been deprived”? Swarthout v. Cooke, 562 U.S. 216, 219\\n(2011). Second, were the “procedures followed” by the\\ngovernment in encroaching on those interests “constitutionally\\nsufficient”? Id.\\n\\n The Environmental Associations run into trouble on the\\nfirst prong of that test. They assert that liberty and property\\ninterests were bestowed upon them by the Pennsylvania\\nConstitution’s Environmental Rights Amendment, Pa. Const.\\nArt. 1, § 27, and the Natural Gas Act’s review procedures, 15\\nU.S.C. §§ 717f(c)(1)(B), 717r(a)–(b). This court has already\\nrejected those claims, and those decisions control here. See\\nDelaware Riverkeeper Network v. FERC, 895 F.3d 102, 110\\n(D.C. Cir. 2018) (Pennsylvania Constitution); Griffith v.\\nFederal Labor Relations Auth., 842 F.2d 487, 495 (D.C. Cir.\\n1988) (Natural Gas Act review procedures); see generally\\nLaShawn A. v. Barry, 87 F.3d 1389, 1395 (D.C. Cir. 1996)\\n(“One three-judge panel * * * does not have the authority to\\noverrule another three-judge panel of the court.”).\\n\\n The Homeowners, for their part, lose on the second prong\\nof the due-process analysis. They argue that the\\nCommission’s delay in acting on their rehearing petitions while\\nauthorizing construction to start “denied [them] the right to be\\n\\n 12\\nheard on whether Transco’s taking of their property actually\\nsatisfies the public use requirement of the Fifth Amendment.”\\nPetitioners’ Opening Br. 47.\\n\\n But, again, circuit precedent says otherwise. We have\\nheld that, as long as FERC’s public-convenience-and-necessity\\ndetermination is not legally deficient, it necessarily satisfies the\\nFifth Amendment’s public-use requirement. See Midcoast\\nInterstate Transmission, Inc. v. FERC, 198 F.3d 960, 973 (D.C.\\nCir. 2000). The Homeowners make no claim that they were\\ndeprived of a meaningful opportunity to be heard as part of the\\nCommission’s proceedings leading up to its issuance of the\\nCertificate Order, and they make no effort to distinguish (or\\neven acknowledge) our holding in Midcoast.\\n\\n * * *\\n\\n For the foregoing reasons, the petitions for review are\\ndenied.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4170,"cells":{"id":{"kind":"number","value":4645796,"string":"4,645,796"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"Sealed Case (PUBLIC REDACTED OPINION)"},"decision_date":{"kind":"string","value":"2019-08-06"},"docket_number":{"kind":"string","value":"19-5068"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before: TATEL, MILLETT, and PILLARD, Circuit Judges.\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge TATEL\", \"type\": \"010combined\", \"text\": \"PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n njI $tat Iourt nf ppca1\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued July 12, 2019 Decided July 30, 2019\\n Reissued August 6, 2019\\n\\n No. 19-5068\\n\\n IN RE: SEALED CASE\\n\\n\\n\\n Consolidated with 19-5 100, 19-5 102, 19-5 103\\n\\n\\n Appeals from the United States District Court\\n for the District of Columbia\\n (No. 1:18-mc-00175)\\n (No. 1:18-mc-00176)\\n (No. 1:18-mc-00177)\\n\\n\\n Before: TATEL, MILLETT, and PILLARD, Circuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge TATEL.\\n\\n TATEL, Circuit Judge:* The three appellant banks,\\nheadquartered in China, hold records that the United States\\ngovernment thinks may clarify how North Korea finances its\\nnuclear weapons program. The government therefore procured\\nsubpoenas for those records, obtaining two from a grand jury\\n(for the two banks with U.S. branches) and one from the\\n *\\n NOTE: Portions of this opinion contain sealed\\ninformation, which has been redacted.\\n\\f 2\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nAttorney General under the Patriot Act (for the bank that has\\nno U.S. branch). The banks resisted on multiple fronts,\\nclaiming that the district court lacked personal jurisdiction, that\\nthe Patriot Act subpoena exceeded the government’s statutory\\nauthority, and that compelling production would run afoul of\\ncomity principles because turning over the records—even in\\nresponse to a U.S. court order—would trigger penalties under\\nChinese law. In a thorough and thoughtful pair of opinions, the\\ndistrict court overruled those objections and, when the first\\nopinion failed to elicit the records, held the banks in civil\\ncontempt. For the reasons set forth below, we affirm.\\n\\n I.\\n For decades, the Democratic People’s Republic of Korea\\n(better known as North Korea) has aspired to a battle-ready\\nnuclear weapons program. Seeking to defuse the threat, the\\nU.S. government has imposed stiff financial sanctions on the\\nNorth Korean regime and its instrumentalities, exercising\\nauthority located in the International Emergency Economic\\nPowers Act. See 50 U.S.C. §S 170 1—02 (granting the President\\nthe power to “prohibit” various “transactions” after declaring a\\n“national emergency” posed by “any unusual and extraordinary\\nthreat. to the national security, foreign policy, or economy\\n . .\\n\\n\\n\\n\\nof the United States”). Relevant to this case, those sanctions are\\nintended to cut off North Korea’s access to American markets\\nand currency. See Executive Order No. 13,382, § 1, 70 Fed.\\nReg. 38,567, 38,567 (June 28, 2005) (authorizing various\\ncabinet officers to “block[J” entities that further nuclear\\nproliferation by preventing any “property or interests in\\nproperty” from being “transferred, paid, exported, withdrawn,\\nor otherwise dealt in”). According to government investigators,\\nhowever, North Korea manages to evade the sanctions by using\\n\\f 3\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nChinese front companies that cloak the true ownership of the\\nfunds involved.\\n\\n The facts uncovered by this investigation so far are a case\\nin point. The U.S. government has collected substantial\\nevidence that a now-defunct Chinese company called\\n (“the Company”)\\nacted as a front for , a North Korea-\\nowned entity (“the NKE”), by facilitating transactions that\\nviolated the sanctions orders. In fact, to date the government’s\\n“investigation [has] not yield[ed] any evidence that” the\\nCompany “was ever used as a legitimate business.”\\nDeclaration of (“FBI\\nDeclaration”) ¶ 10, Joint Appendix (“J.A.”) 864.\\nUnsurprisingly, then, in addition to the NKE itself, the\\ngovernment has sanctioned the Company and two of its alleged\\nfounders. Actions Taken Pursuant to Executive Order 13382,\\n (designating the\\nNKE); Notice of [Office of Foreign Asset Control (OFAC)]\\nSanctions Actions,\\n (designating the Company); Notice of OFAC Sanctions\\nActions, •\\n(designating alleged co-founder\\n —\\n ); Notice of OFAC\\nSanctions Actions,\\n(designating alleged co-founder ).\\n According to the government, the scheme operated like\\nthis: Hobbled by the worthlessness of North Korea’s currency,\\nthe NKE would use the Company to make or receive payments\\nin U.S. dollars. These transactions helped North Korea access\\nresources that would otherwise have been beyond its reach. For\\nexample, the Company’s assistance allegedly enabled North\\nKorea to export hundreds of millions of dollars of coal and\\nother minerals, generating revenue in U.S. currency that North\\n\\f 4\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nKorea could then use to requisition other commodities vital to\\nits weapons program.\\n\\n In these transactions, the Company routinely took\\nadvantage of U.S. correspondent bank accounts—that is,\\naccounts held by banks outside the United States at banks\\nlocated inside the United States. See 31 U.S.C.\\n§ 53 18A(e)(1)(B) (defining “correspondent account” to mean\\n“an account established to receive deposits from, make\\npayments on behalf of a foreign financial institution, or handle\\nother financial transactions related to such institution”). Such\\naccounts allow foreign “banks to conduct business and provide\\nservices for their customers in” the United States even if “the\\nbanks have no physical presence” here. Minority Staff of the\\nSenate Permanent Subcommittee on Investigations,\\nCorrespondent Banking: A Gateway for Money Laundering,\\nS. Hrg. 107-84, 107th Cong., app. 287 (2001). This, in turn,\\nenables a foreign bank’s customers to “receive many or all of\\nthe services offered by [a] U.S. bank” without forming a direct-\\nclient relationship with an American institution. Id.\\nCorrespondent banking arrangements are fairly ubiquitous:\\n“[d]ue to U.S. prominence in international trade and the high\\ndemand for U.S. dollars due to their overall stability, most\\nforeign banks that wish to provide international services to\\ntheir customers have accounts in the United States capable of\\ntransacting business in U.S. dollars.” Id. In short, then,\\ncorrespondent bank accounts allow foreign banks to offer their\\nclients services—including obtaining and transacting in U.S.\\ncurrency—without opening a U.S. branch.\\n\\n These accounts are how the appellant banks (collectively,\\nthe “Banks”) enter the story. All three are China-based—\\nindeed, the Chinese government owns a substantial minority of\\neach—and hold correspondent accounts in the United States.\\n\\f 5\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nBanks One and Two also operate branches in the United States,\\nbut according to the government the Company relied on the\\nthree Banks’ U.S. correspondent accounts to execute its\\nscheme.\\n\\n To better grasp the full scope of the Company’s\\noperations, the government seeks certain records from the\\nBanks. The United States and China have a bilateral agreement\\nfor obtaining just these kinds of records. Known as the Mutual\\nLegal Assistance Agreement (MLAA), it obligates the two\\ncountries to assist each other “in proceedings related to\\ncriminal matters,” including by “providing . . records or\\n .\\n\\n\\n\\narticles of evidence” and “executing requests for inquiry,” and\\nspecifies the “[f]orm and [c]ontents” of any requests.\\nAgreement Between the Government of the United States of\\nAmerica and the Government of the People’s Republic of\\nChina on Mutual Legal Assistance in Criminal Matters, arts. 1,\\n4, June 19, 2000, T.I.A.S. No. 13,102. Absent an MLAA\\nrequest, Chinese law prohibits banks from disclosing records\\nto international investigators.\\n\\n In this case, however, the United States chose not to go\\nthrough the MLAA. Stymied by what it perceives as sluggish\\nand typically fruitless cooperation from Chinese authorities in\\nthe past, the government unilaterally issued subpoenas to all\\nthree Banks in December 2017. The two banks with branches\\nin the United States\\n (“Bank One” and “Bank Two,”\\nrespectively)—received grand jury subpoenas. See Federal\\nRule of Criminal Procedure 1 7(e)(1). The third\\n (“Bank Three”)—has no branch in\\nthe United States and received a subpoena from the Attorney\\nGeneral pursuant to the Patriot Act, which gives the Attorney\\nGeneral and the Treasury Secretary special investigatory tools\\n\\f 6\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nwhen it comes to U.S. correspondent bank accounts. See 31\\nU.S.C. § 53 18(k)(3)(A)(i). The subpoenas seek all records\\ndated 2012 to 2017 concerning all correspondent banking\\ntransactions associated with certain accounts linked to the\\nCompany, defining records to include signature cards, ledger\\ncards, account statements, and documentation of deposits,\\nwithdrawals, and transfers. Through this investigation, the\\ngovernment hopes to learn whether the Company or any other\\nentities have committed various federal crimes, including\\nengaging in unlicensed financial transactions (violating 50\\nU.S.C. § 1705), money laundering (violating 18 U.S.C.\\n § 1956), or failing to exercise due diligence in correspondent\\nbanking as required by the Bank Secrecy Act (violating 31\\nU.S.C. § 531$ and 5322). The government does not currently\\nsuspect the subpoenaed Banks of any wrongdoing.\\n\\n In correspondence with the U.S. government, the Banks\\nand the Chinese government protested that complying with the\\nsubpoenas would violate multiple Chinese laws, including the\\nlaw designating the MLAA process as the exclusive\\nmechanism for disclosures. They claimed that failure to use the\\nMLAA channel would expose the Banks to administrative and\\ncriminal penalties. In response, the U.S. government took the\\nposition that such a request would be futile, given China’s slow\\nand spotty history of providing records requested through that\\nprocess.\\n\\n There matters stood for about a year. Despite traveling\\ntwice to China for what proved to be unproductive negotiations\\nwith the Chinese government, there were no major\\nbreakthroughs: the United States refused to submit an MLAA\\nrequest, and the Banks continued to withhold the records. At\\nlast, in November 2018, things kicked back into gear when the\\ngovernment moved the district court to compel enforcement.\\n\\f 7\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nStill resisting, the Banks disputed the district court’s personal\\njurisdiction and contended that enforcing the subpoenas would\\nviolate deeply ingrained principles of international comity.\\nBank Three also protested that the scope of its subpoena\\nexceeded the Attorney General’s Patriot Act authority.\\n\\n The district court granted the government’s motion to\\ncompel. In re Grand Jury Investigation ofPossible Violations\\nof 18 USC. §‘ 1956 and 50 US.C. 1705, Nos. 1:18-mc-175,\\n-176, -177, slip op. at 3 (D.D.C. Mar. 18, 2019) (“March op.”).\\nThe court found that Banks One and Two had consented to\\npersonal jurisdiction when they signed an agreement with the\\nfederal Reserve permitting them to open branches in the\\nUnited States. Id. at 11—13. As for Bank Three, which has no\\nbranch in the United States, the court determined that (along\\nwith Banks One and Two) it had sufficient contact with the\\nUnited States as a whole and the subpoenas sufficiently related\\nto that contact so as to support the court’s personal jurisdiction.\\nId. at 13—27; see also International Shoe Co. v. State of\\nWashington, Office of Unemployment Compensation and\\nPlacement, 326 U.S. 310, 316 (1945) (personal jurisdiction\\nrequires “minimum contacts” with the forum such that\\njurisdiction is consistent with “traditional notions of fair play\\nand substantial justice” (internal quotation marks omitted)).\\nThe district court also perceived no problems with the scope of\\nthe Patriot Act subpoena. March Op. 27—31. Finally, although\\nthe government conceded that compliance with the subpoenas\\nwould entail violating at least some Chinese laws, the district\\ncourt concluded that comity principles did not require the\\nsubpoenas to be quashed. Id. at 32—5 8. The court found that,\\nbased on past practice, China’s preferred MLAA procedure\\nwould likely be futile and determined that the U.S.\\ngovernment’s strong national security interests outweighed\\nChina’s stake in enforcing the identified laws. Id. at 42—51. The\\n\\f 8\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ncourt was unmoved by a series of letters from the Chinese\\nMinistry of Justice promising to “promptly review and\\nprocess” any MLAA request in this case. Id. at 45 (quoting\\nLetter from Du Yaling, Director General of International\\nCooperation Department of the Ministry of Justice, to Chief\\nJudge Beryl A. Howell (Feb. 26, 2019), J.A. 966).\\n\\n About a month later, after the Banks made clear they\\nwould not comply with the compulsion order, the district court\\nheld them in contempt. In re Grand Jury Investigation of\\nPossible Violations of 18 U.S.C. § 1956 and 50 US.C. § 1705,\\nNos. 1:18-mc-175, -176, -177, slip op. at 1—3 (D.D.C. Apr. 10,\\n2018) (“April Op.”). It imposed a $50,000 per day fine against\\neach bank but stayed that penalty contingent on the parties\\nseeking an expedited appeal. Id. at 14—15. They obliged, and\\nwe agreed to accelerate our consideration of the case. Per\\nCuriam Order, In re: Sealed Case, No. 19-5068 (D.C. Cir. Apr.\\n25, 2019). We now turn to the Banks’ arguments, beginning\\nwith personal jurisdiction before examining the merits.\\n\\n II.\\n The Constitution forbids federal courts from exercising\\npersonal jurisdiction absent sufficient “minimum contacts” to\\nsatisfy “traditional notions of fair play and substantial\\njustice.” International Shoe, 326 U.S. at 316 (quoting Milliken\\nv. Meyer, 311 U.S. 457, 463 (1940)). That protection extends\\nto private foreign entities—such as the Banks—just as to\\nAmerican citizens. See Livnat v. Palestinian Authority, 851\\nF.3d 45,48 (D.C. Cir. 2017) (delineating the “general rule” that\\nrestrictions on personal jurisdiction “protect{J all litigants in\\nour courts”). Unlike subject-matter jurisdiction, however,\\npersonal jurisdiction “can. be waived,” meaning a party\\n . .\\n\\n\\n\\nmay “consent” to a court’s personal jurisdiction. Insurance\\n\\f 9\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nCorp. of Ireland v. Compagnie des Bauxites de Guinee, 456\\nU.s. 694, 703 (1982). Because neither the government nor the\\nBanks suggests that these principles operate differently in\\ngrand jury investigations than in ordinary civil litigation, we\\nshall treat these two contexts as identical for purposes of this\\ncase. But see Boydv. Meachum, 77 f.3d 60,66 (2d Cir. 1996)\\n(“While the federal constitutional requirements of personal\\njurisdiction in a civil setting are reasonably well-defined,\\nthis is not so in a criminal case.”).\\n\\n All three Banks contest personal jurisdiction. “Reviewing\\nthe district court’s assertion of personal jurisdiction de novo,”\\nUnited States ex rel. Miller v. Bill Harbert International\\nConstruction, Inc., 608 F.3d 871, 886 (D.C. Cir. 2010), we\\nconclude that the Banks’ challenges are meritless. Banks One\\nand Two consented to jurisdiction when they opened branches\\nin the United States. And although Bank Three has no U.S.\\nbranch and executed no such agreement, its choice to maintain\\ncorrespondent accounts in the United States establishes an\\nadequate connection to the forum and the enforcement action\\nto sustain jurisdiction.\\n\\n A.\\n Congress requires any “foreign bank” to acquire “prior\\napproval of the” Federal Reserve before it “may establish a\\nbranch or an agency” in the United States. 12 U.S.C.\\n§ 3 105(d)(1). The Federal Reserve, in turn, has authority to\\n“impose such conditions on its approval” of a foreign bank’s\\napplication “as it deems necessary,” Id. § 3 105(d)(5), to,\\namong other things, assure itself that the bank will be “in\\ncompliance with applicable United States law” and mitigate\\nany “risk to the stability of the United States financial system,”\\nid. § 350 1(d)(3)(D)—(E). See also 12 C.F.R. § 21 1.24(c)(2)(vii)\\n\\f 10\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n(Federal Reserve scrutinizes “[wjhether the foreign bank...\\nhas established adequate controls and procedures. to ensure . .\\n\\n\\n\\ncontinuing compliance with U.S. law, including controls\\ndirected to detection of money laundering”).\\n\\n When Banks One and Two applied to open their U.S.\\nbranches, both executed agreements with the Federal Reserve\\nto assure compliance with relevant provisions of American\\nlaw. As part of that deal, they “consent[edj to the jurisdiction\\nof the federal courts of the United States for purposes of\\n . . .\\n\\n\\n\\nany and all. proceedings initiated by. the United\\n . . . .\\n\\n\\n\\nStates. in any matter arising under U.S. Banking Law.”\\n . .\\n\\n\\n\\nLetter from Vice Chairman & President,\\n ,\\n\\n\\n to Board of Governors of the\\nFederal Reserve System 1 l.A. 1659 (“Bank\\n ,\\n\\n\\nOne Consent”); Assurances Commitment and Consent to\\nJurisdiction Commitment\\n 1, J.A. 1663 (“Bank Two Consent”). The agreements\\ndefine “U.S. Banking Law” expansively to include “all federal\\ncriminal laws of which violation(s) arise(s): under .the . . . . .\\n\\n\\n\\nBank Secrecy Act.” Bank One Consent 3 n.2, l.A. 1661; Bank\\nTwo Consent 1 n.2, l.A. 1663.\\n\\n The plain language of those agreements covers this case.\\nThis investigation is unquestionably a “proceeding[] initiated\\nby. the United States” and, as the government explains, “the\\n . .\\n\\n\\n\\ngrand jury is investigating whether those financial institutions\\nthat provided services in the United States”—for example, the\\nAmerican institutions hosting the foreign Banks’\\ncorrespondent accounts—”exercised due diligence.”\\nAppellee’s Br. 18 (citing 31 U.S.C. § 5318, 5318A, 5322).\\nThe Bank Secrecy Act makes willful dereliction of that\\nobligation a federal criminal violation. 31 U.S.C. § 5322. In\\nshort, then, the government has assembled the necessary\\n\\f 11\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ningredients for jurisdiction: a criminal investigation initiated by\\nthe United States (the grand jury investigation) arising under\\nU.S. Banking Law (the Bank Secrecy Act).\\n\\n The banks’ contrary arguments amount to nothing more\\nthan an unsuccessful effort to move the goalposts. They insist\\nthat there must be a reasonable showing that someone actually\\nviolated the Bank Secrecy Act. Bank Two Br. 8. Not so. All\\nour precedent requires is a “reasonable probability that. the. .\\n\\n\\n\\nfacts necessary for the exercise of jurisdiction” exist. In re\\nSealed Case, $32 F.2d 1268, 1274 (D.C. Cir. 1987) (quoting In\\nre Marc Rich & Co., 707 F.2d 663, 670 (2d Cir. 1983)). Here,\\nnothing in the consent agreements requires the government to\\nestablish liability; they require only that the proceedings\\n“aris[e] under U.S. Banking Law.” Bank One Consent 1,\\nJ.A. 1659 (emphasis added); Bank Two Consent 1, J.A. 1663\\n(emphasis added). Whether a proceeding “arises under” a\\nparticular law is a question about its subject matter at the outset,\\nnot a question about its outcome. To look at the same phrase in\\na more typical context, when we ask whether a lawsuit “aris[es]\\nunder” federal law for purposes of subject-matter jurisdiction,\\nU.S. Const. art. III, § 2, we look to whether the plaintiffs\\n“assert a claim” based on “federal law,” not whether they will\\nultimately prevail, National farmers Union Insurance Cos. v.\\nCrow Tribe of Indians, 471 U.S. 845, 850 (1985) (emphasis\\nadded); see also American Well Works Co. v. Layne & Bowler\\nCo., 241 U.S. 257, 260 (1916) (“A suit arises under the law that\\ncreates the cause of action.”). Given this “legal lineage,” Hall\\nv. Hall, 13$ S. Ct. 1118, 1125 (201$),agrand jury investigation\\nplainly “arises under” a particular statute when it probes\\nwhether that statute has been violated.\\n\\n Read properly, the consent agreements demand only that\\n“U.S. Banking Law” form the basis of the investigation. The\\n\\f 12\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ngovernment has cleared that threshold here. Even if, as the\\nbanks suggest, this investigation’s potential to yield successful\\nprosecutions is uncertain, the banks have offered no basis for\\ndoubting that “U.S. Banking Law” constitutes this\\ninvestigation’s subject matter.\\n\\n The banks also suggest that there is a problem with\\npersonal jurisdiction because the subpoenaed records are\\nlocated outside the United States. But that fact has no bearing\\non personal jurisdiction given that the banks’ consent\\nagreements supply the basis for jurisdiction. The argument is\\nbetter conceived as a challenge to the scope of the subpoenas—\\nan understanding not evident from the banks’ briefs. Even were\\nwe to entertain the argument, however, it would go nowhere.\\nFederal courts have long required the production of documents\\nheld abroad. See, e.g., In re Marc Rich, 707 F.2d at 667 (A\\nwitness may not “resist the production of documents on the\\nground that the documents are located abroad.”); United States\\nv. first National City Bank, 396 F.2d 897, 900—01 (2d Cir.\\n196$) (“[A] federal court has the power to require the\\nproduction of documents located in foreign countries if the\\ncourt has in personam jurisdiction of the person in possession\\nor control of the material.”). That practice makes sense: grand\\njuries often investigate violations of criminal laws that\\nobviously cover extraterritorial conduct. “It would be strange,\\nindeed, if the United States could punish a foreign corporation\\nfor violating its criminal laws . . but a federal grand jury\\n . ,\\n\\n\\ncould not investigate to ascertain the probability that a crime\\nhad taken place.” In re Marc Rich, 707 F.2d at 666.\\n\\n B.\\n Bank Three, by contrast, has neither opened a U.S. branch\\nnor consented to personal jurisdiction. We therefore must\\n\\f 13\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ndetermine whether it “has purposefully directed” its relevant\\n“activities” at “the forum.” Burger King Corp v. Rudzewicz,\\n471 U.S. 462, 472 (1985) (internal quotation marks omitted).\\n\\n That, of course, requires us to correctly identify “the\\nforum,” a task ordinarily linked to where Congress has\\nauthorized service of process. See BNSF Railway Co. v.\\nTyrrell, 137 S. Ct. 1549, 1555—56 (2017) (“Congress’ typical\\nmode of providing for the exercise of personal jurisdiction has\\nbeen to authorize service of process” because “a basis for\\nservice of a summons on the defendant is prerequisite to the\\nexercise of personal jurisdiction”). In many situations,\\nCongress limits the territorial reach of federal district courts to\\nthat of the state in which they sit. See Wright & Miller,\\n4 Federal Practice & Procedure § 1068.1 (4th ed.); see, e.g.,\\nFederal Rule of Civil Procedure 4(k)(1)(A). For example,\\nunless a federal statute otherwise provides, by virtue of Rule\\n4(k)(1)(A) the jurisdictional reach of a federal court in a civil\\naction is keyed to that of a court of general jurisdiction in the\\nstate in which it sits. So in reviewing the constitutionality of\\ntheir assertions of jurisdiction, federal courts oflen perform\\nstate-specific contacts analyses. Although the constitutional\\nrestraint in those situations necessarily flows from the Fifth\\nAmendment, the scope of the minimum-contacts inquiry is the\\nsame as it would be under the Fourteenth Amendment for a\\nclaim in state court: confined to contacts with the forum state.\\n\\n According to Bank Three, a state-specific contacts\\nanalysis is appropriate for the Patriot Act subpoenas at issue\\nhere. And, the Bank argues, since its agent was served in New\\nYork and the Bank lacks any meaningful contact with the\\nDistrict of Columbia, the district court here lacked jurisdiction.\\nFor its part, the government maintains that the Patriot Act\\nauthorizes nationwide service of process. Under those\\n\\f 14\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ncircumstances, the Fifth Amendment requires only “minimum\\ncontacts with the United States” as a whole—rather than with\\nthe forum state. SEC v. Bilzerian, 378 F.3d 1100, 1106 n.8\\n(D.C. Cir. 2004); accord Livnat, 851 F.3d at 55 (“Under the\\nFifth Amendment, which defines the reach of federal courts,\\ncontacts with the United States as a whole are relevant.”).\\nTherefore, the government argues, the district court properly\\nconsidered Bank Three’s contacts with the United States as a\\nwhole, including its maintenance and use of correspondent\\naccounts in New York.\\n\\n We begin by observing that the usual process of looking to\\nwhere federal law authorizes service of process as a proxy for\\nwhether Congress sought to trigger a federal court’s nationwide\\nterritorial reach is, perhaps, not a perfect fit for this provision\\nof the Patriot Act. Because the statute confers authority to issue\\na subpoena directly on the Attorney General and the Treasury\\nSecretary, rather than on a federal district court, Congress may\\nhave given little consideration to the territorial reach of a\\ndistrict court considering a motion to compel enforcement of\\nsuch a subpoena. But given that Bank Three assumes that\\nauthorizing nationwide service in this context would trigger a\\nnationwide contacts analysis, we shall too. See Sickle v. Torres\\nAdvanced Enterprise Solutions, LLC, 884 F.3d 338, 344 (D.C.\\nCir. 2018) (accepting a party’s decision “not to brief or argue\\nthe question of personal jurisdiction” because that “defense .\\n\\n\\n\\n\\ncan be waived or forfeited”).\\n\\n Here, that service-of-process provision, codified in 31\\nU.S.C. § 5318(k), authorizes the kind of subpoena served on\\nBank Three to “be served on the foreign bank in the United\\nStates if the foreign bank has a representative in the United\\nStates.” 31 U.S.C. § 5318(k)(3)(A)(ii). We know of no other\\nservice-of-process provision in the U.S. Code phrased exactly\\n\\f 15\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nthis way. Still, we are not entirely without guideposts as we\\nevaluate this question.\\n\\n We know, for instance, that Congress has many different\\nways of authorizing nationwide service of process. Sometimes\\nit makes its intention abundantly clear by authorizing service\\n“anywhere in the United States.” See, e.g., 7 U.S.C. § 499g(b)\\n(service for enforcing in federal district courts certain orders of\\nthe Secretary of Agriculture); 2$ U.S.C. § 232 1(c) (same for\\nenforcing orders of the Surface Transportation Board); 33\\nU.S.C. § 921(c) (service for review proceedings in the courts\\nof appeals of decisions of the Benefits Review Board for\\nlongshore and harbor workers). On other occasions, Congress\\nemploys some riff on the phrase “wherever the defendant may\\nbe found,” language generally understood to refer to\\nnationwide service. See, e.g., 7 U.S.C. § 25(c) (service in\\nprivate lawsuits under the Commodity Exchange Act); see also\\nOmni Capital International, Ltd. v. Rudolf Wolff & Co., 484\\nU.S. 97, 105—06 (1987) (recognizing that such language\\n“explicitly authorize[sJ nationwide service of process”);\\nAppellee’s Br. 25 n.20 (collecting additional examples). We\\nalso know that Congress can very clearly limit service to a\\nparticular judicial district when it wants to. See, e.g., 45 U.S.C.\\n § 3 62(b) (authorizing process issued by most courts in Railroad\\nRetirement Board proceedings to be served “in the judicial\\ndistrict of the district court issuing such. process”).\\n . .\\n\\n\\n\\n\\n At first glance, the language in section 5318(k) looks more\\nlike the provisions that authorize nationwide service. So long\\nas a foreign bank has designated a representative, the statute\\nallows for service “in the United States,” without specifying\\nany state or judicial district. True, the provision lacks the\\nunmistakable clarity of the statutes that expressly authorize\\nservice “anywhere” in the country; nonetheless, the text\\n\\f 16\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ncontains no limitation other than the territorial boundaries of\\nthe entire United States. Bank Three, however, seizes on the\\nrequirement to designate a representative to argue that, in fact,\\nCongress only meant to authorize service in the judicial district\\nwhere that representative is located.\\n\\n Bank Three’s position makes some intuitive sense. After\\nall, as a matter of physics, the designated representative will\\nhave a single location. On further reflection, however, this\\nlogic fails to resolve things in the Bank’s favor. After all, both\\nBank Three and the government agree that language\\nauthorizing service “wherever a defendant may be found”\\nauthorizes nationwide service, see Bank Three Br. 51,\\nAppellee’s Br. 25 n.20, even though in many instances that\\ndefendant, too, will only be located in one place at a time.\\n\\n Even more instructive, in other service-of-process regimes\\ninvolving designated representatives, when Congress wants to\\nspecify a particular location or district for service, it says so\\ndirectly. For example, in certain trademark actions, Congress\\nhas instructed foreign mark holders to designate a U.S. agent\\nand authorized service of process “by leaving with that person\\nor mailing to that person a copy [of the process] at the address\\nspecfied in the last designation so filed.” 15 U.S.C. § 1141k(f)\\n(emphasis added). Congress has even provided a backup\\noption, explaining that “[i]f the person so designated cannot be\\nfound at the last designated address” then “process may be\\nserved on the Director” of the United States Patent and\\nTrademark Office. Id.; see also 7 U.S.C. § 2569 (containing a\\nsimilar backup provision for when a designee “cannot be found\\nat the address given”). The provision at issue in this case is\\nnoticeably devoid of any similar language, leading to the\\ninference that Congress sought to impose no similar\\ngeographical limitation here.\\n\\f 17\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n\\n Reading section 5318(k) to authorize nationwide service\\nalso aligns with the Patriot Act’s structure and purpose. Cf\\nUnited States International Trade Commission v. ASAT, Inc.,\\n411 f.3d 245, 251 (D.C. Cir. 2005) (examining, when\\nconsidering whether a statute authorizes nationwide service,\\nwhether “Congress would have desired nationwide service of\\nprocess to effectuate the underlying statute’s purpose”). Rather\\nthan requiring the Attorney General or Treasury Secretary to\\ngo to court to obtain a subpoena, Congress authorized those\\nofficers to issue process themselves. Unlike federal district\\ncourts, which customarily serve a particular geographical area,\\nCabinet officers serve the entire country. Structurally speaking,\\nthen, Congress’s choice to give two such officers the subpoena\\npower evokes an intent to rely on their nationwide reach. As\\nfor purpose, having found that correspondent accounts were a\\nsource of special concern for money laundering, Congress\\nsought “to provide a clear national mandate for subjecting to\\nspecial scrutiny those. . financial institutions operating\\n .\\n\\n\\n\\noutside of the United States. that pose particular, identifiable\\n . .\\n\\n\\n\\nopportunities for criminal abuse.” USA PATRIOT Act of 2001,\\nPub. L. No. 107-56, § 302(a)(6), (b)(4), 115 Stat. 272, 296—97.\\nIt would be odd if, in establishing this “clear national mandate,”\\nCongress had geographically subdivided the government’s\\ninvestigative powers.\\n\\n Apparently recognizing that its preferred interpretation of\\nsection 5318(k) may not represent “the most natural reading of\\nthe provision,” Bank Three Br. 4$, Bank Three falls back on\\nthe presumption against extraterritoriality, which teaches that\\nwithout a “clearly expressed congressional intent to the\\ncontrary, federal laws will be construed to have only domestic\\napplication.” RJR Nabisco, Inc. v. European Community, 136\\nS. Ct. 2090, 2100 (2016). According to the Bank, reading the\\n\\f 18\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nPatriot Act to authorize nationwide service “would vastly\\nexpand the extraterritorial scope of federal regulatory power.”\\nBank Three Br. 49. But that does not follow. This service-of-\\nprocess provision has no impact on the substantive\\nextraterritorial reach of federal law; it merely helps sort out\\nwhich federal court has authority to enforce the subpoena. After\\nall, even under Bank Three’s preferred interpretation, at least\\none federal court could assert jurisdiction over Bank Three (the\\none in the district where the Bank’s agent resides). The\\npresumption against extraterritoriality therefore offers no basis\\nto upset the most straightforward reading of the text: that once\\na foreign bank has designated a representative in the United\\nStates, service may be accomplished throughout “the United\\nStates.”\\n\\n Having defined the forum as the entire United States, we\\nwould usually go on to assess whether the Bank’s contacts with\\nthat forum are sufficient. In this case, however, Bank Three\\nplaced all its eggs in the forum-identification basket; it claims\\nonly that it lacks the necessary contacts with the District of\\nColumbia, offering no argument that its contacts with the\\nUnited States as a whole are insufficient. See Bank Three\\nBr. 46—47 (“The only way the government can prevail is by\\nshowing that the relevant forum is the United States as a\\nwhole.”). We therefore have no occasion to question the district\\ncourt’s conclusion that Bank Three’s maintenance of\\ncorrespondent accounts in the United States supplies the\\nnecessary nexus. March op. 2 1—24. With no further\\njurisdictional hurdles in our path, we proceed to the merits.\\n\\n III.\\n Bank Three, the only bank to receive a Patriot Act\\nsubpoena, argues that its subpoena exceeds the Attorney\\n\\f 19\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nGeneral’s statutory authority. Although “[w]e generally review\\ndistrict court decisions enforcing document subpoenas only for\\narbitrariness or abuse of discretion,” we afford no deference to\\na decision that applies “the wrong legal standard.” In re Sealed\\nCase, 146 F.3d 881, $83 (D.C. Cir. 199$). We therefore review\\n“de novo” the district court’s “legal conclusions,” including its\\nstatutory interpretation, even as we review “its factual findings\\nfor clear error.” United States v. Ahn, 231 F.3d 26, 38 (D.C.\\nCir. 2000).\\n\\n Congress enacted section 531 8(k)(3) “to increase the\\nstrength of United States measures to prevent, detect, and\\nprosecute international money laundering and the financing of\\nterrorism.” Pub. L. No. 107-56, § 302(b)(1), 115 Stat. at 297.\\nIn particular, Congress found that correspondent accounts held\\nby foreign banks were “one of the banking mechanisms\\nsusceptible in some circumstances to manipulation by foreign\\nbanks to permit the laundering of funds.” § 302(a)(6), 115 Stat.\\nat 296. Thus, it authorized both the Attorney General and the\\nTreasury Secretary to issue a “subpoena to any foreign bank\\nthat maintains a correspondent account in the United States and\\nrequest records related to such correspondent account,\\nincluding records maintained outside of the United States\\nrelating to the deposit of funds into the foreign bank.” 31\\nU.S.C. § 5318(k)(3)(A)(i).\\n\\n Bank Three claims the subpoena it received strays outside\\nthe statutory lines because it is “not expressly confine[d]” to\\ndocuments related to the Bank’s correspondent accounts in the\\nUnited States. Bank Three Br. 23. To be sure, the subpoena is\\ncomprehensive: it seeks “[a]1l documents relating to [the\\nCompany’s] correspondent banking transactions”—including,\\nit would seem, documents related to transactions using\\ncorrespondent accounts in countries other than the United\\n\\f 20\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nStates. Bank Three Subpoena, J.A. 771; see also id. (requesting\\n“signature cards,” “account ledger cards,” “periodic account\\nstatements,” and “records. of all items deposited,\\n . .\\n\\n\\n\\n\\nwithdrawn, or transferred”). And we know Bank Three does,\\nin fact, have correspondent accounts outside the United States.\\nIf we knew nothing else about the Company, Bank Three’s\\nargument might therefore have some force: where, as here, a\\nbank maintains correspondent accounts in multiple countries,\\nit is by no means a given that all of an account holder’s\\ncorrespondent banking transactions “relate to” the bank’s US.\\ncorrespondent accounts, as the statute requires.\\n\\n With respect to this particular case, however, the\\ngovernment has a compelling response. It argues, as it did\\nbefore the district court, that evidence already collected during\\nthe investigation shows that the Company’s “exclusive raison\\nd’etre” was to serve as “a progenitor of U.S. dollars for” the\\nNKE. Appellee’s Br. 62; see also Id. at 64 (describing the\\nCompany as “a front company created for the sole purpose of\\nexecuting U.S.-dollar transactions”); March 5, 2019 Hearing\\nTr. 95:17—25, l.A. 1650—5 1 (“[The Company] existed only for\\none reason. [T]he [NKE] has to set up entities like this,\\n . .\\n\\n\\n\\n\\nbecause [it] want[s] access to the U.S. financial system.”). In\\nthe government’s view, such a showing means that all of the\\nCompany’s records are “related to” the Bank’s U.S.\\ncorrespondent accounts because those accounts were part of the\\nCompany’s means of obtaining U.S. dollars.\\n\\n The parties’ dispute turns on the meaning of the term\\n“related to” as used in section 531 8(k)(3). Arguing for a narrow\\ninterpretation, Bank Three reads the statute to authorize\\ncollecting only those records of transactions that themselves\\npassed through Bank Three’s U.S. correspondent accounts. For\\nits part, the government advocates a more expansive reading—\\n\\f 21\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\none that would cover all records that have a “connection with”\\nthe Company’s use of Bank Three’s U.S. correspondent\\naccounts, even if some individual transactions made no use of\\na U.S. correspondent account. Appellee’s Br. 60 (internal\\nquotation marks omitted).\\n\\n To assess how broadly the phrase “related to” sweeps, we\\nlook to the “traditional tools of statutory interpretation—text,\\nstructure, purpose, and legislative history.” Tax Analysts v. IRS,\\n350 F.3d 100, 103 (D.C. Cir. 2003) (internal quotation marks\\nomitted). The “first step in interpreting a statute is to determine\\nwhether the language at issue has a plain and unambiguous\\nmeaning”; “[o]ur inquiry must cease if the statutory language\\nis unambiguous and the statutory scheme is coherent and\\nconsistent.” Robinson v. Shell Oil Co., 519 U.S. 337, 340\\n(1997) (internal quotation marks omitted).\\n\\n “[W]hen asked to interpret statutory language including the\\nphrase ‘relating to,’ . .th[e] [Supreme] Court has typically\\n .\\n\\n\\n\\nread the relevant text expansively.” Lamar, Archer & Cofrmn,\\nLLP v. Appling, 13$ S. Ct. 1752, 1760 (201$). We see no\\nreason to deviate from that approach here. The ordinary\\nmeaning of that phrase encompasses the broad relationships\\ndescribed by the government. See, e.g., Merriam-Webster’s\\nCollegiate Dictionary 987 (10th ed. 1997) (defining “relate” as\\n“to have [a] relationship or connection”); Black’s Law\\nDictionary 1288 (6th ed. 1990) (defining “relate” as “[tb stand\\nin some relation; to have bearing or concern; to pertain; refer;\\nto bring into association with or connection with”); see also BF\\nAmerica Production Co. v. Burton, 549 U.S. 84, 91(2006)\\n(“Unless otherwise defined, statutory terms are generally\\ninterpreted in accordance with their ordinary meaning.”).\\n\\f 22\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n The government’s interpretation finds further support in the\\none example that Congress offered of a record “related to” a\\nU.S. correspondent account: “records maintained outside of the\\nUnited States relating to the deposit of funds into the foreign\\nbank.” 31 U.S.C. § 5318(k)(3)(A)(i). The final phrase is\\nimportant. Merely depositing funds “into the foreign bank”\\nwould not necessarily entail passing those funds through the\\nforeign bank’s U.S. correspondent account; rather, it may be a\\npreparatory step for using the correspondent account. For\\ninstance, an aspiring money launderer might deposit illicitly-\\nobtained foreign currency into the foreign bank before using\\nthe U.S. correspondent account to facilitate a payment in U.S.\\ndollars, helping the guilty party conceal the origin of the funds.\\nBy specifying that its view of records “related to” a U.S.\\ncorrespondent account includes such a predicate transaction,\\nCongress embraced an expansive view of the term in line with\\nthe government’s. And, of course, that view accords with\\nCongress’s stated purpose in enacting the provision: “to\\nincrease the strength of United States measures to prevent,\\ndetect, and prosecute international money laundering and the\\nfinancing of terrorism.” Pub. L. No. 107-56, § 302(b)(1), 115\\nStat. at 297.\\n\\n Resisting this expansive text, Bank Three offers two\\nreasons to prefer a more circumscribed understanding of the\\nstatute. First, it again looks to the presumption against\\nextraterritoriality. But here, the statute contains no ambiguity\\nrequiring us to resort to that presumption. See Validus\\nReinsurance, Ltd. v. United States, 786 F.3d 1039, 1046 (D.C.\\nCir. 2015) (explaining that “ambiguity is resolved by\\n[applyingj the presumption against extraterritoriality”\\n(emphasis added)). To be sure, the statutory text is capacious,\\nbut as the Supreme Court has observed, “[b]road. language\\n . .\\n\\n\\n\\nis not necessarily ambiguous when congressional objectives\\n\\f 23\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nrequire broad terms.” Diamond v. Chakrabarty, 447 U.S. 303,\\n315 (1980).\\n\\n Second, Bank Three points to the Justice Department’s\\nattempt to persuade Congress to amend the statute to allow the\\ngovernment to obtain not just “records relating to” a foreign\\nbank’s U.S. correspondent account, but also “any records”\\npertaining to “any related account at the foreign bank,\\nincluding records maintained outside of the United States, that\\nare the subject of any investigation of a criminal violation of\\nUnited States law.” Department of Justice, Title III, Anti-\\nCorruption Legislative Proposals, 114th Cong., at 4 (May 5,\\n2016), https://www.justice.gov/opa/file/849986/download\\n(emphases added). The Bank would have us infer from\\nCongress’s failure to adopt the amendment that section\\n53 18(k)(3) as currently written can never reach “[a]l1\\ndocuments” relating to “all” of a company’s “correspondent\\nbanking transactions,” as this subpoena does. Bank Three\\nSubpoena, J.A. 771. But the Supreme Court has long cautioned\\nagainst reading too much into congressional inaction given that\\n“[l]ogically, several equally tenable inferences could be drawn\\nfrom the failure of the Congress to adopt an amendment.\\nincluding the inference that the existing legislation already\\nincorporated the offered change.” United States v. Wise, 370\\nU.S. 405, 411 (1962). In any event, the unenacted amendment\\nBank Three cites does not expand the definition of what records\\ncount as “related to” a U.S. correspondent account; rather, it\\nwould allow the government to reach other accounts,\\nregardless of their relationship to a U.S. correspondent account.\\nSuch a proposal tells us little about what Congress currently\\nconsiders to be “related to” a correspondent account.\\n\\n We therefore conclude that records “related to” a U.S.\\ncorrespondent account include records of transactions that do\\n\\f 24\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nnot themselves pass through a correspondent account when\\nthose transactions are in service of an enterprise entirely\\ndedicated to obtaining access to U.S. currency and markets\\nusing a U.S. correspondent account. All that remains is to\\ndetermine whether the government has adequately shown that\\nthe Company was such an enterprise.\\n\\n The district court found that the government had made the\\nnecessary showing. Specifically, it found that “[n]o part” of the\\nCompany could “be separated from the ploy to put money\\nthrough the United States financial market.” March Op. 29. On\\nthe record before us, that finding was not clearly erroneous. See\\nBoca Investerings Partnership v. United States, 314 F.3d 625,\\n629—30 (D.C. Cir. 2003), as amended (Mar. 26, 2003) (a\\nconclusion is clearly erroneous when “the reviewing court on\\nthe entire evidence is left with the definite and firm conviction\\nthat a mistake has been committed”) (internal quotation marks\\nomitted). The investigation has turned up considerable\\nevidence supporting the court’s conclusion. The Company\\nengaged in an enormous volume of U.S. dollar transactions:\\n“[B]etween October 2012 and November 2015 alone, [the\\nCompany] made over $100 million U.S.-dollar payments.” FBI\\nDeclaration ¶ 13, J.A. 866. And the Company “was a\\ncounterparty to over 680 U.S.-dollar international transfers\\n(which transited through the United States financial system via\\ncorrespondent U.S. banks).” Id., J.A. 867. The government has\\nuncovered no “evidence that” any of these transactions were\\npart of “a legitimate business”; rather, all were part of the\\nNKE’s scheme to launder money. Id. ¶ 10, J.A. 864—65.\\n\\n Nor does any record evidence suggest that the NKE used\\nthe Company to launder money through currencies other than\\nU.S. dollars. Indeed, the NKE had a special need for dollars\\nbecause “commodity companies, such as those operating in\\n\\f 25\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nChina, Thailand, and Singapore, frequently prefer to take\\npayments in U.S. dollars because of the dollar’s global demand\\nand stability.” Id. ¶ 26, l.A. 871; see also Correspondent\\nBanking: A Gateway for Money Laundering, S. Hrg. 107-84\\napp. 287 n.12 (“U.S. dollars are one of a handful of major\\ncurrencies accepted throughout the world. They are also\\nviewed as a stable currency, less likely to lose value over time\\nand, thus, a preferred vehicle for savings, trade and\\ninvestment.”). Further, the government explains, and Bank\\nThree nowhere contests, that before 2017—and consequently\\nduring the vast majority of the period covered by this\\nsubpoena—the NKE “could legally conduct transactions in\\nother currencies” besides the U.S. dollar (such as the euro),\\nAppellee’ s Br. 61, signaling that the NKE would have had no\\nneed of a front company to conduct transactions in those\\ncurrencies.\\n\\n Taken together, these facts support the district court’s\\nfinding that “[nb part” of the Company—and thus, none of its\\nbanking records—”can be separated from the ploy” to launder\\nmoney through U.S. financial markets, March op. 29, because\\nthe Company’s sole purpose was to give the NKE access to\\nU.S. dollars through U.S. correspondent accounts. We\\ntherefore conclude that on the facts of this case, the records\\nrequested by the Patriot Act subpoena all “relate to” Bank\\nThree’s U.S correspondent account. To be clear, nothing in this\\nopinion authorizes, as the Bank fears, the government “for any\\nreason, [to] obtain essentially any document or record from any\\nforeign bank with a U.S. correspondent account,” Bank Three\\nBr. 22, regardless of whether those records are “related to” that\\nU.S. correspondent account. The government’s power to\\nsubpoena the records in this case derives from the evidence that\\nthis particular company operated exclusively as a U.S. dollar\\nclearinghouse for the NKE. Under such circumstances, all\\n\\f 26\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nrecords pertaining to the Company’s Bank Three account and\\nits correspondent banking transactions, no matter where they\\noccurred, are “related to” the Bank’s U.S. correspondent\\naccounts.\\n\\n Iv.\\n “Comity refers to the spirit of cooperation in which a\\ndomestic tribunal approaches the resolution of cases touching\\nthe laws and interests of other sovereign states.” Société\\nNationale Industrielle Aérospatiale v. U.S. District Court for\\nthe Southern District of Iowa (“Société Nationale”), 482 U.S.\\n522, 543 n.27 (1987). Invoking this “spirit of cooperation,” Id.,\\nall three Banks argue that “ordering a foreign-sovereign-owned\\nbank to violate foreign law on foreign soil violates basic\\n‘comity’ principles” and, as such, demands reversal, Bank One\\nBr. 11; see also Bank Two Br. 25 (making this argument);\\nBank Three Br. 32—33 (same). We begin by providing some\\nadditional background before explaining why the Banks’\\nargument fails.\\n\\n A.\\n A comity analysis typically starts by determining whether\\na “true conflict” exists between the implicated legal systems.\\nHartford Fire Insurance Co. v. Calfornia, 509 U.S. 764, 798\\n(1993) (internal quotation marks omitted). Absent a clash of\\nforeign and domestic law, American courts may press forward\\nfree from worry that their rulings will threaten the\\n“international legal ties that advance the rule of law within and\\namong nations.” Laker Airways Ltd. v. Sabena, Belgian World\\nAirlines, 731 F.2d 909, 937 (D.C. Cir. 1984), as amended\\n(Mar. 6 and 9, 1984). As the district court recognized, however,\\nin this case that determination “is the easy part,” given that\\n“[tJhe [U.S.] government concedes that complying with the\\n\\f 27\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nrespective subpoenas exposes each bank to legal penalties in\\nChina.” March op. 32.\\n\\n After this initial inquiry, the question is whether “the\\nfactual circumstances” of the case at hand counsel against\\napplying American law. Laker Airways, 731 f.2d at 937. A\\ncollection of factors helps us to delineate the “inherently\\nuncertain” borders set by the principles of comity in the context\\nof subpoena enforcement. Id. Factors include “the importance\\nto the litigation of the documents . requested,” “the degree\\n . .\\n\\n\\n\\nof specificity of the request,” “whether the information\\noriginated in the United States,” “the extent to which\\nnoncompliance with the request would undermine important\\ninterests of the United States, or compliance with the request\\nwould undermine important interests of the state where the\\ninformation is located,” and whether “alternative means of\\nsecuring the information” exist. Société Nationale, 482 U.S. at\\n544 n.28 (alteration omitted) (quoting Restatement of foreign\\nRelations Law of the United States (Revised) § 437(1)(c)).\\nCourts also look to whether compliance would impose hardship\\non the party targeted by the subpoena and whether that party\\nhas acted in good faith. See In re Sealed Case (“Sealed\\nCaseT’), 825 F.2d 494, 498—99 (D.C. Cir. 1987)\\n(distinguishing cases based on “good faith” considerations);\\nLinde v. Arab Bank, PLC, 706 f.3d 92, 110 (2d Cir. 2013)\\n (“[A] district court should also examine the hardship of the\\nparty facing conflicting legal obligations... .“).\\n\\n The district court diligently applied this “constellation of\\nfactors” to the facts at issue here. March Op. 33. In so doing, it\\nobserved that two factors indisputably weigh against enforcing\\nthe subpoenas: the documents originated outside the United\\nStates, see id. at 42, and the Banks have acted in good faith\\nthroughout this litigation, see id. at 57 (“[T]he government\\n\\f 2$\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nagrees that none of the banks has acted in bad faith.”).\\nAccording to the district court, another pair of factors just as\\nclearly weighs in favor of enforcement: the documents contain\\nunique information vital to the investigation—indeed, “the\\ngovernment avows the subpoenaed documents are the\\nfoundation of the United States’ investigation,” Id. at 39\\n(alteration and internal quotation marks omitted), and the\\nsubpoenas target an appropriately specific collection of\\nrecords—that is, “the subpoenas are tailored to specific\\nfinancial records, for a defined date range, at banks [the\\nCompany] is known to have used to launder funds,” id. at 41.\\n\\n Parsing the more complex remaining factors, the district\\ncourt emphasized the government’s goal of learning how North\\nKorea has thwarted American sanctions in pursuing weapons\\nof mass destruction. “[NJon-enforcement” of the subpoenas,\\nthe district court stressed, “would undermine a critical national\\ninterest.” Id. at 4$. Although acknowledging that China has a\\n“legitimate” “interest in the development of a sound banking\\nsystem,” the district court concluded that the United States’\\nnational security interest easily outweighs China’s interest in\\nsafeguarding its banking secrecy laws. Id. at 50; see also Id.\\n(observing that because “even Chinese authorities recognize\\nthat bank secrecy can co-exist with limited disclosure to\\ngovernment agencies,” “China’s interest would not be\\nundermined by enforcement of the subpoenas”).\\n\\n The district court recognized that the U.S. government had\\ndeclined to take advantage of an alternative channel through\\nwhich the United States might obtain the documents at issue:\\nthe Mutual Legal Assistance Agreement, which creates a\\nprocess to exchange “documents” and “records” relevant to\\ncriminal investigations. MLAA art. l(2)(c); see March Op. 42—\\n47 (discussing the MLAA). But based on the U.S.\\n\\f 29\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ngovernment’s declarations that China has failed to\\nsatisfactorily engage in the MLAA process in many other cases\\nover the past decade, the district court accepted the Executive\\nBranch’s representation that requiring the government to first\\nexhaust the MLAA process would just “[aJllow[J China to gum\\nup [the] United States’ investigations” and, therefore, that\\n“[tJhe Department of Justice no longer needs to take overtures\\nof China’s willingness to assist the United $tates[] with\\ninvestigations into North Korea seriously.” March op. 47.\\n\\n finally, with respect to the hardship facing the Banks, the\\ndistrict court acknowledged that compliance with the\\nsubpoenas would violate Chinese law. Though the government\\ndoubts the Banks’ claim that any criminal laws would be\\nbroken, it concedes that compliance would at least violate those\\nlaws prohibiting Chinese companies from handing over records\\nthat have not been approved for release through the MLAA\\nprocess. See Law of the People’s Republic of China on\\nInternational Criminal Judicial Assistance (promulgated by the\\nStanding Committee of the Thirteenth National People’s\\nCongress, Oct. 26, 201$, effective Oct. 26, 201$), ch. I, art. 4,\\nJ.A. 355 (forbidding any “institution [or] organization\\nwithin the territory of [China]” from “provid[ing] evidentiary\\nmaterials and assistance prescribed by this Law to foreign\\ncountries”); Id. ch. II, § 2, art. 13, J.A. 35$ (requiring a “foreign\\ncountry” to request “criminal judicial assistance” “in\\naccordance with the provisions of the criminal judicial\\nassistance treaty”). The district court also recognized that the\\nMinistry of Justice has advised that these violations would have\\nconsequences in the form of administrative penalties against\\nthe Banks and civil or criminal penalties against individual\\nofficers. See March Op. 51—53. The district court, however,\\nfound the hardships accompanying compliance both\\nspeculative and minimal. See id. at 53—57. It based this\\n\\f 30\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nconclusion in part on historical Chinese enforcement trends:\\npenalties are rare and relatively modest when they do occur.\\nThe court found, for example, no instances of “a state-owned\\nChinese enterprise . fac{ing] severe repercussions for\\n . .\\n\\n\\n\\n\\nresponding to the order of a foreign court,” Id. at 53, and that\\nbank officers had apparently faced penalties only for stealing\\nand selling “client information for profit,” id. at 55. The district\\ncourt also relied on the logical supposition that the Chinese\\ngovernment is unlikely to severely sanction banks that it\\nlargely owns, id., and concluded that the Ministry’s warnings\\nfailed to “suggest[] the administrative penalties will amount to\\nmore than nominal fines” or provide any “concrete indication\\ncriminal penalties are likely,” Id. at 56—57.\\n\\n Affording the most weight to the United States’\\nunassailable interest in successfully investigating and, with any\\nluck, frustrating North Korea’s arms programs, see id. at 58\\n(calling “the interests of the relevant countries” the “most\\nimportant factor”), the district court concluded that “[ojn\\nbalance, international comity is not a reason to refrain from\\ncompelling compliance with the subpoenas,” Id. On appeal, the\\nBanks argue that the district court applied the incorrect legal\\nstandard, that enforcing the subpoenas runs afoul of our case\\nlaw, that “[c]omity dictates that the Chinese government at\\nleast be given an opportunity to comply with an MLAA request\\nbefore forcing a Chinese bank to violate Chinese law on\\nChinese soil,” Bank Two Br. 19, and, finally, that the district\\ncourt incorrectly balanced the relevant factors.\\n\\n B.\\n Before considering these arguments, we must first\\nascertain the appropriate standard of review. Although we\\n“ordinarily review{] a district court’s grant of a motion to\\n\\f 31\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ncompel.. for abuse of discretion,” Recording Industry Ass ‘n\\n .\\n\\n\\n\\n\\nof America, Inc. v. Verizon Internet Services, Inc., 351 F.3d\\n 1229, 1233 (D.C. Cir. 2003), we have yet to expressly\\ndetermine the standard of review applied to a comity analysis\\nin this context, cf Sealed Case I, $25 f.2d at 497—99 (not\\nidentifying the standard of review); de Csepel v. Republic of\\nHungary, 714 F.3d 591, 606 (D.C. Cir. 2013) (“This court has\\nnever expressly addressed the standard by which we review a\\ndistrict court’s decision to grant comity to a foreign\\njudgment.”). The Supreme Court, however, has observed that\\n“[t]he exact line between reasonableness and\\nunreasonableness” when weighing comity principles “must be\\ndrawn by the trial court, based on its knowledge of the case,”\\nSociétë Nationale, 482 U.S. at 546, and we generally review\\nfor abuse of discretion when the district court finds itself\\nresponsible for making such a fact-bound reasonableness call,\\nsee, e.g., Ideal Electronic Security Co. v. International Fidelity\\nInsurance Co., 129 F.3d 143, 150 (D.C. Cir. 1997) (“[TJhe\\nreasonableness of an attorney’s fees award is within the sound\\n discretion of the trial court and is reviewed only for abuse of\\n discretion.”); United States v. Law, $06 F.3d 1103, 1105 (D.C.\\n Cir. 2015) (“We. consider the substantive reasonableness of\\n . .\\n\\n\\n\\n\\n the sentence imposed under an abuse-of-discretion standard.”\\n (internal quotation marks omitted)). This approach comports\\n with our sister circuits, which review similar questions for\\n abuse of discretion. See, e.g., Allstate Life Insurance Co. v.\\n Linter Group Ltd., 994 F.2d 996, 999 (2d Cir. 1993) (“[S]ince\\n the extension or denial of comity is within the court’s\\n discretion, we will reverse the court’s decision only when we\\n find an abuse of discretion.”); Reinsurance Co. ofAmerica v.\\nAdministratia Asigurarilor de Stat, 902 F.2d 1275, 1283 (7th\\n Cir. 1990) (“The denial of [the company’s] motion to compel\\n responses was . . not an abuse of discretion.”). The Banks\\n .\\n\\n\\n\\n\\n offer no argument for departing from that approach here. See\\n\\f 32\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nBank Two Br. 5 (suggesting “an abuse of discretion standard”);\\nOral Arg. Rec. 13:33—13:40 (Bank One counsel agreeing that\\nif “the district court did apply the correct legal standard, then\\nit’s an abuse of discretion standard”). Accordingly, we shall\\nreview the district court’s comity analysis for abuse of\\ndiscretion.\\n\\n Employing that standard, we ask whether the district court\\n“based its ruling on an error of law, a clearly erroneous\\nassessment of the evidence, or an improper weighing of the\\nfactors limiting its discretion.” fraenkel v. Islamic Republic of\\nIran, $92 F.3d 348, 356 (D.C. Cir. 201$) (citation and internal\\nquotation marks omitted). We of course review any embedded\\n“legal conclusions de novo.” Guedes v. Bureau of Alcohol,\\nTobacco, firearms & Explosives, 920 F.3d 1, 10 (D.C. Cir.\\n2019).\\n\\n Bank One first argues that “the District Court used the\\nwrong legal standard to review the Government’s request to\\nenforce the Subpoenas”—that is, the court should not have\\nadopted “the multifactorial approach of . Restatement\\n . .\\n\\n\\n\\n(Third) of Foreign Relations.” Bank One Reply Br. 1. This\\nargument gets the Banks nowhere. In Sociëtë Nationale, the\\nSupreme Court explained that the Restatement’s “factors are\\nrelevant to any comity analysis.” 482 U.S. 522, 544 n.2$; see\\nOverby v. National Ass ‘n of Letter Carriers, 595 F.3d 1290,\\n1295 (D.C. Cir. 2010) (“[C]arefully considered language of the\\nSupreme Court, even if technically dictum, generally must be\\ntreated as authoritative.” (internal quotation marks omitted)).\\nThe Bank points out that our decision in Sealed Case I, issued\\na few months after Société Nationale, never expressly\\nreferences the Restatement’s factors. In that case we reversed\\nan order holding a bank owned by (unidentified) Country X in\\ncontempt for failure to comply with a subpoena that would\\n\\f 33\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nhave required the bank to produce documents in contravention\\nof the banking privacy laws of Country Y, where the branch at\\nissue was located. Sealed Case I, 825 F.2d at 497—98. But\\nSealed Case I neither refutes Société Nationale’ s ratification of\\nthe Restatement factors nor offers a different test to apply—\\nnor, of course, could it. See We the People foundation, Inc. v.\\nUnited States, 485 F.3d 140, 144 (D.C. Cir. 2007) (circuit\\ncourts must follow Supreme Court precedent). As the district\\ncourt recognized, Sealed Case I “comports with the Supreme\\nCourt’s treatment of international comity as a principle that\\ndepends on ‘prior scrutiny in each case of the particular facts,\\nsovereign interests, and likelihood that resort to [alternative]\\nprocedures will prove effective.” March Op. 34 (alteration in\\noriginal) (quoting Société Nationale, 482 U.S. at 544).\\n\\n This brings us to the Banks’ related assertion that “[t]he\\npresent case is controlled by” Sealed Case I. Bank One Br. 14.\\nBut that case by its own terms constricts its holding to its facts:\\n“If any of the facts we rest on here were different,” we\\ncautioned, “our holding could well be different.” Sealed Case I,\\n825 F.2d at 499. We therefore agree with the district court that\\nthis case merits its own factual inquiry; Sealed Case I does not,\\nas a legal matter, control the outcome here.\\n\\n The Banks’ more powerful argument is that, due to the\\nfactual parallels between Sealed Case I and this case, “it was\\nan abuse of discretion to enforce the Subpoena[sJ.” Bank Two\\nBr. 25. The Sealed Case I court identified several facts central\\nto its decision not to enforce the contempt order at issue, and\\nthey might sound familiar, first, and “[m]ost important,” the\\n“sanctions represent[ed] an attempt by an American court to\\ncompel a foreign person to violate the laws of a different\\nforeign sovereign on that sovereign’s own territory.” 825 F.2d\\nat 498; see also Id. (observing that the bank had no way to\\n\\f 34\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n“comply with the contempt order without violating the laws of\\nCountry Y on Country Y’s soil”). We also recognized that the\\nbank “had acted in good faith,” was “a third party . not...\\n . .\\n\\n\\n\\naccused of any wrongdoing,” and constituted “not merely a\\nprivate foreign entity,” but also “an entity owned by the\\ngovernment of Country X.” Id. finally, because the bank’s\\nmanager would “be available and able to testify as to many of\\nthe facts that the grand jury may [have] wish[edj to ascertain,”\\nthe government was not “left empty-handed” by the decision\\nnot to enforce the subpoena. Id. at 499. But see Ed.\\n(acknowledging “that the grand jury’s investigation may\\nnonetheless be hampered, perhaps significantly” by the\\ndecision not to enforce).\\n\\n As the Banks point out, each of those facts maps neatly\\nonto this case, in which the government asks us to compel\\nChinese banks owned in significant part by the Chinese\\ngovernment to violate Chinese law on Chinese territory. The\\nBanks have acted in good faith, they are currently third parties\\nto the criminal investigation, and the MLAA, at least on its\\nface, provides an alternative path that would allow the\\ngovernment to obtain the records at issue.\\n\\n What’s more, to the extent there are differences, there is\\nsome reason to believe that this case is an even better candidate\\nfor comity than Sealed Case I. For one thing, where the\\nsubpoena recipient in Sealed Case I was being asked to violate\\nthe laws of a “different foreign sovereign,” Id. at 498, here the\\ngovernment demands that the Banks violate their home\\ncountry’s laws. Although the Country X-based bank in Sealed\\nCase I might have closed its branch in Country Y to avoid\\nsanctions, the Banks here have no such workaround. As the\\ndistrict court recognized, “[a] Country X bank can shutter the\\nCountry Y portion of its business; a Chinese bank cannot leave\\n\\f 35\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nChina.” March Op. 36 n.13; see also Id. (“[C]ompelling a bank\\nto violate its own nation’s law is more prejudicial to\\ninternational comity.” (internal quotation marks omitted)).\\nMoreover, the Chinese government in this case “has warned\\nthe Bank[s] that [they] will be sanctioned if [they] ignore[J” its\\n“express instructions not to produce these documents outside\\nthe MLAA process,” and “[t]here is no indication that the\\nforeign government in Sealed Case I did anything similar.”\\nBank Three Br. 39.\\n\\n The district court ignored none of these considerations.\\nQuite to the contrary, it grappled with each and instead found\\npersuasive two other “critical” differences between the cases.\\nMarch Op. 34.\\n\\n The district court first pointed out that Sealed Case I\\nconcerned a “domestic law enforcement matter,” which\\n“hardly compares to the national security interests and\\nassociated potential harm caused by non-compliance with a\\nsubpoena related to an investigation into funding a state-\\nsponsor of terrorism’s nuclear weapons program.” Id. at 35.\\nThis is an important distinction. The government justifiably\\nbelieves that the subpoenas will unearth information\\nconcerning North Korea’s use of third-party front companies\\nto evade economic sanctions, and nothing in Sealed Case I\\nsuggests any “sovereign interest[],” Société Nationale, 482\\nU.S. at 544, remotely comparable to impeding North Korea’s\\nstockpiling of nuclear weapons.\\n\\n We acknowledge, as the Banks point out, that the\\ngovernment here has failed to proceed as though North Korea\\nwere about to push the nuclear launch button. Indeed, the\\ngovernment took nearly a year to enforce the subpoenas in\\ncourt and never sought to expedite this case until the district\\n\\f 36\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ncourt instructed it to do so. It did not, however, sit on its hands\\nduring that time, as the Banks repeatedly assert, but “engaged\\nin intra- and inter-agency deliberations” and “twice sent a\\ndelegation to China” during that period. March op. 40. More\\nto the point, we are reluctant to second-guess the Executive\\nBranch’s assessments of national security interests and how to\\nbest serve them. See frugone v. Central Intelligence Agency,\\n 169 f.3d 772, 775 (D.C. Cir. 1999) (cautioning “that courts\\nhave little expertise in either international diplomacy or\\ncounterintelligence operations”). We of course agree with\\nBank One that “the judiciary has an independent obligation to\\nevaluate” the diplomatic ripple effects of its decisions. Bank\\nOne Br. 25; see also Sealed Case I, 225 f.2d at 499 (“[WJe see\\ngood reason for courts not to act on their own, even at the\\nurging ofthe executive branch, when their actions ‘may hinder\\nrather than further this country’s pursuit of goals both for itself\\nand for the community of nations as a whole in the international\\n sphere.” (emphasis added) (quoting Banco Nacional de Cuba\\nv. Sabbatino, 376 U.S. 398, 423 (1964))). But we are confident\\nthat the district court respected that obligation and that it acted\\nwithin its discretion by recognizing that the heightened national\\n interest here, as articulated by the Executive Branch,\\n distinguishes this case from Sealed Case I.\\n\\n The Banks next observe that “if the Government receives\\nthe documents via the MLAA, its national security interests\\nwill be fully vindicated.” Bank Two Br. 26. The district court,\\nhowever, concluded that the futility of the MLAA process sets\\nthis case apart from Sealed Case I, in which testimony from the\\nbank’s manager appeared to offer a sure-fire way of securing\\nat least some of the desired information. “[un this\\ninvestigation,” by contrast, the district court observed, “the\\ngovernment insists that without compelling compliance, the\\nrequested information is out of reach.” March Op. 35.\\n\\f 37\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nforcefully disagreeing with this conclusion, the Banks insist\\nthat the MLAA process provides a channel specifically\\ndesigned to allow the U.S. government to obtain exactly the\\ntypes of records it seeks. Allowing it to eschew that process\\nwithout so much as making an initial MLAA request, they\\nargue, neglects comity’ s “spirit of cooperation.” $ociété\\nNationale, 482 U.S. at 543 n.27. For its part, the government\\nargues in support of the district court that “the historical record\\ndemonstrates . an MLAA request of China for such bank\\n . .\\n\\n\\n\\nrecords is not a viable alternative.” Appellee’s Br. 37 (internal\\nquotation marks omitted).\\n\\n Recall that the MLAA establishes a procedure for the U.S.\\nand Chinese governments to obtain documents from each other\\nin criminal investigations. Specifically, the process provides\\nthat “insofar as national law permits,” if, say, the United States\\nmakes a request, China’s “Central Authority” must search for\\nand then send the requested records. MLAA\\narts. 2(1), 14(1), (3). Nothing in the MLAA, however,\\ndesignates it as the exclusive means of obtaining evidence in a\\ncriminal investigation.\\n\\n In this case, China’s Central Authority, the Ministry of\\nJustice, Id. art. 2(2), expressed a “willingness to cooperate with\\na DOJ MLAA request” as early as March 2018, Letter from\\nHank B. Walther & Samidh Guha to Zia Faruqui & An\\nRedboard 3 (Mar. 23, 2018), J.A. 830. Bank Three\\ncommunicated China’s overture to the U.S. Department of\\nJustice, observing that the MLAA process “would allow DOJ\\nto obtain the documents it seeks, while avoiding the time, cost,\\nand risk of litigating the Patriot Act subpoena.” Id. As we now\\nknow, the United States declined that invitation—even as it\\nfiled at least ten other MLAA requests in 2017 and 2012—and\\ninstead sought to enforce the subpoenas in court.\\n\\f 38\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n\\n The Ministry then directly communicated with the district\\ncourt multiple times, assuring that court that the Ministry\\n“would timely review and handle the requests for assistance\\nsought by the DOJ in accordance with the [MLAA] and\\napplicable domestic laws.” Letter from Du Yaling, Director\\nGeneral of International Cooperation Department of the\\nMinistry of Justice to Chief Judge Beryl A. Howell 4 (Jan. 6,\\n2019), J.A. 76. If the government submits a “request in line\\nwith the [MLAAJ,” the Ministry pledged, “China will provide\\nthe assistance to the United States accordingly.” Id. Confident\\nthat the MLAA “mechanism works effectively,” the Ministry\\nreminded the district court that China has “provided effective\\nassistance to the US in many cases” and that “[tJhe avenue of\\nthe {MLAA] is the only legal means to obtain criminal\\nevidence in China by foreign judicial agencies under Chinese\\nlaws.” Letter from Du Yaling, Director General of International\\nCooperation Department of the Ministry of Justice to Chief\\nJudge Beryl A. Howell 2—4 (Feb. 26, 2019), J.A. 965—67. The\\ndistrict judge observed that this marked “the first time” she had\\nseen “correspondence directly from a foreign government\\nrelated to an ongoing matter.” March 5, 2019 Hearing Tr. 18:2—\\n4, J.A. 1573. The government points to no other case where the\\nMinistry has made such a representation and reneged on it and\\nadmits that in some cases, “China has produced bank records\\nto the general satisfaction of the U.S. prosecutors and agents.”\\nDeclaration of Associate Director of the Office of International\\nAffairs Jeffrey M. Olson (“OIA Declaration”) ¶ 14(e),\\nJ.A. 951.\\n\\n The district court considered all these facts but ultimately\\nconcluded that pursuing the MLAA process would result in an\\ninadequate production of documents. According to the\\nAssociate Director of the Office of International Affairs—\\n\\f 39\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nwhich serves as the United States’ Central Authority under the\\nMLAA—the United States has submitted fifty MLAA requests\\nfor bank records over the past decade and has received no\\nresponsive records to thirty-five; the government found most\\nof China’s responses in the remaining fifteen cases inadequate\\nor untimely. And of the ten requests submitted within the last\\ntwo years, the United States has received at most one response.\\nCf OIA Declaration ¶ 13(b), J.A. 950 (describing that, for\\nexample, China failed to produce documents in one of those\\ncases even though the United States informed the Ministry that\\nthe request was time-sensitive due to an upcoming trial).\\n\\n Moreover, the district court recognized that the\\ngovernment is especially pessimistic about the chances of\\nsuccess in this case given that the investigation involves a\\nperfect storm of disagreements between the United States and\\nChina: “Chinese authorities have a demonstrated failure in\\nsupporting foreign and U.S. sanctions and U.S. law\\nenforcement actions involving North Korea,” and “the Chinese\\ngovernment has specifically objected to the imposition of\\nsanctions on [the Company] and other Chinese front\\ncompanies.” FBI Declaration ¶ 80, J.A. $88. Drawing the\\ndistrict court’s attention to a case “nearly identical” to this one,\\nId. at 37, J.A. 884, an FBI special agent submitted a declaration\\npointing out that although in that case, China had “nominally\\nagreed to work jointly with the U.S. authorities” on MLAA\\nrequests for bank records involving North Korea, id. ¶ 72,\\nJ.A. $85, nothing ever came of that initial agreement; offering\\nvarious excuses, China never handed over any records. “Based\\non these facts,” the agent “concluded that the MLAA process\\nis not an effective way to obtain bank records from Chinese\\nauthorities with respect to investigations involving North\\nKorea.” Id. ¶ 75, J.A. 886; see Id. (“[OJne reason that Chinese\\nauthorities do not want to assist with North Korean\\n\\f 40\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ninvestigations is that producing such records could reflect\\nbadly on the Chinese government and the Chinese financial\\nindustry.”).\\n\\n Given this record, we see no abuse of discretion in the\\ndistrict court’s decision to enforce the subpoenas despite the\\nfact that the United States chose not to pursue the MLAA\\nprocess. The Supreme Court has made clear that evidence-\\nsharing arrangements set up by international agreements,\\nalthough often valuable, can also prove “unduly time\\nconsuming and expensive, as well as less certain” than court-\\nenforced subpoenas “to produce needed evidence.” $ociété\\nNationale, 482 U.S. at 542. As such, whether comity “requires\\nresort” to an agreement’s procedures calls for an inquiry into\\nthe “likelihood that resort to those procedures will prove\\neffective.” Id. at 544. The district court addressed just that\\nquestion: it reviewed declarations from U.S. and Chinese\\nofficials, identified facts that undermined the Chinese\\ngovernment’s assurances about the MLAA system, and\\nconcluded that the U.S. government had no obligation to\\nsubmit a voluntary MLAA request because it would likely\\nprove ineffective.\\n\\n We need not tarry over the Banks’ remaining arguments\\nthat the district court abused its considerable discretion in\\nbalancing the comity factors. Cf Morley v. Central Intelligence\\nAgency, 894 F.3d 389, 391 (D.C. Cir. 2018) (recognizing that\\nwhen “factors point in different directions,” “it will be the rare\\ncase when we can reverse a district court’s balancing ofthe.\\nfactors”). Specifically, Bank Three reiterates its claim that “the\\nPatriot Act subpoena is facially overbroad,” Bank Three Reply\\nBr. 20, and, as such, the “degree of specificity of the request,”\\nSociétë Nationale, 482 U.S. at 544 n.28 (internal quotation\\nmarks omitted), also weighs in favor of the Bank. But we agree\\n\\f 41\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nwith the district court that given that the Company “solely\\nexisted as a front company,” “all [the Company’s] records\\nrelate to the conduct under investigation,” March Op. 41\\n(internal quotation marks omitted). With respect to the\\nhardship factor, Bank One takes issue with the district court’s\\nconclusion that the Banks would likely face minimal penalties\\nfor complying with the subpoenas. In a civil law country such\\nas China, the Bank contends, “[t]he District Court cannot know\\nhow the Chinese government will exercise its prosecutorial\\ndiscretion.” Bank One Br. 2$. But the district court never\\npresumed to declare itself a soothsayer; it assessed a cluster of\\ndatapoints presented and substantiated by the Executive\\nBranch and ultimately determined that the risk of harsh\\npunishment was slim. In any event, it recognized that “this\\nfactor tips ever so slightly toward the banks.” March Op. 57.\\nThe district court exercised its ‘judgment” in drawing these\\nconclusions and was “guided by sound legal principles.”\\n United States v. Taylor, 487 U.S. 326, 336 (198$) (internal\\nquotation marks omitted). Our analysis ends there.\\n\\n We proceed with extreme caution when enforcing\\nsubpoenas that would require recipients to violate a foreign\\nsovereign’s domestic laws. As we said in Sealed Case I, “[w]e\\nhave little doubt. that our government and our people would\\n . .\\n\\n\\n\\nbe affronted if a foreign court tried to compel someone to\\nviolate our laws within our borders,” 825 F.2d at 498—99, all\\nthe more so if our government assured that court that it was\\nready and willing to engage in a mutually agreed-upon bilateral\\nprocess to provide the precise information requested. But we\\ntake heart that at every stage of this process decisionmakers\\nhave recognized that whether to enforce a subpoena “in cases\\nlike this one raises grave difficulties for courts.” Id. at 498. The\\ngovernment decided to pursue the records at issue through the\\ncourts rather than the MLAA process only after twice sending\\n\\f 42\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ndelegates to China and engaging in extensive “internal\\ndiscussions with subject matter experts within the FBI, as\\n . . .\\n\\n\\n\\nwell as consultations involving senior FBI management and\\nDepartment of Justice and Department of Treasury officials.”\\nFBI Declaration ¶ 80, J.A. $88. That decision was then put to\\n“the wise discretion of our judicial colleague[] on the District\\nCourt,” fraenkel, 892 F.3d at 362, and that court consented to\\nthe government’s course of action only after undertaking a\\nthorough, considered analysis that frankly acknowledged the\\nBanks’ comity concerns. Whether to enforce the subpoenas\\nwas a hard call, the district court made that call, and we have\\nno reason to reverse its fact-bound conclusion.\\n\\n V.\\n Finally, Banks One and Three challenge the district court’s\\nApril 2019 contempt order, arguing that it was unjustified\\ngiven the government’s concession that the Banks acted in\\ngood faith to obtain the Chinese government’s permission to\\nproduce the records covered by the subpoenas. “[W]e\\nreview. the contempt finding. for abuse of discretion.” In\\n . . . .\\n\\n\\n\\nre Fannie Mae Securities Litigation, 552 F.3d 814, 818 (D.C.\\nCir. 2009).\\n\\n A civil contempt citation is appropriate “only if the putative\\ncontemnor has violated an order that is clear and unambiguous,\\nand the violation must be proved by clear and convincing\\nevidence.” Broderickv. Donaldson, 437 F.3d 1226, 1234 (D.C.\\nCir. 2006) (internal quotation marks omitted). Banks One and\\nThree nowhere dispute that by refusing to produce the\\ndocuments covered by the subpoenas, they have violated a\\nclear and unambiguous order. Instead, they rest their challenge\\nto the contempt order on the Restatement (Third) of Foreign\\nRelations Law, which suggests that a court “should not\\n\\f 43\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nordinarily impose sanctions of contempt, dismissal, or default\\non a party that has failed to comply with [an] order for\\nproduction” that would violate another state’s laws or\\nregulations, “except in cases of deliberate concealment or\\nremoval of information or of failure to make a good faith\\neffort” “to secure permission from the foreign authorities to\\nmake the information available.” Restatement (Third) of\\nForeign Relations Law § 442(2)(a)—(b) (1927) (emphasis\\nadded).\\n\\n The banks insist that the district court should not have held\\nthem in contempt because they acted in good faith (albeit\\nunsuccessfully) to secure the Chinese government’s\\npermission to produce the records at issue. This is a rather thin\\nreed to begin with, rendered more slender still by the fact that\\nthe more recent Restatement (fourth) of foreign Relations Law\\nsignificantly diminishes the importance of a putative\\ncontemnor’s “good faith”: the Restatement (Fourth) simply\\nstates that courts “have discretion to excuse violations of law,\\nor moderate the sanctions imposed for such violations, on the\\nground that the violations are compelled by another state’s\\nlaw” if “the person in question appears likely to suffer severe\\nsanctions for failing to comply with foreign law” and “has\\nacted in good faith to avoid the conflict.” Restatement (Fourth)\\nof foreign Relations Law § 442 (2018) (emphasis added)). In\\nany event, this court has long held that “[t]he intent of the\\nrecalcitrant party is irrelevant in a civil contempt proceeding\\nbecause, unlike a criminal contempt proceeding, a civil\\ncontempt action is a remedial sanction used to obtain\\ncompliance with a court order or to compensate for damage\\nsustained as a result of noncompliance.” Food Lion, Inc. v.\\n United Food & Commercial Workers International Union,\\nAFL-CIO-CLC, 103 F.3d 1007, 1016 (D.C. Cir. 1997) (internal\\nquotation marks omitted).\\n\\f 44\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n\\n Confronted with clear non-compliance in the face of a\\nperfectly pellucid order, the district court reasonably found that\\n“the banks have not demonstrated good faith in a way relevant\\nto contempt” and that, because “the requested records are\\nessential to an investigation into a matter of national\\nsecurity[,].. continued contumacy threatens tremendous\\n .\\n\\n\\n\\n\\nharm.” April op. 9 (internal quotation marks omitted). Under\\nthese circumstances, the banks have given us no basis for\\nconcluding that the district court abused its discretion.\\n\\n VI.\\n For the foregoing reasons, we affirm the district court’s\\ncontempt orders against all three Banks.\\n\\n So ordered\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4423049/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge TATEL\"}]}"},"judge1":{"kind":"string","value":"TATEL"},"judge2":{"kind":"string","value":"MILLETT"},"judge3":{"kind":"string","value":"PILLARD"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"TATEL\": \", Circuit\", \"MILLETT\": \", Circuit\", \"PILLARD\": \", Circuit\"}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4645796/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Sealed Case (PUBLIC REDACTED OPINION)\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-06\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nPUBLIC COPY - SEALED INFORMATION DELETED\n\n\n njI $tat Iourt nf ppca1\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued July 12, 2019 Decided July 30, 2019\n Reissued August 6, 2019\n\n No. 19-5068\n\n IN RE: SEALED CASE\n\n\n\n Consolidated with 19-5 100, 19-5 102, 19-5 103\n\n\n Appeals from the United States District Court\n for the District of Columbia\n (No. 1:18-mc-00175)\n (No. 1:18-mc-00176)\n (No. 1:18-mc-00177)\n\n\n Before: TATEL, MILLETT, and PILLARD, Circuit Judges.\n\n Opinion for the Court filed by Circuit Judge TATEL.\n\n TATEL, Circuit Judge:* The three appellant banks,\nheadquartered in China, hold records that the United States\ngovernment thinks may clarify how North Korea finances its\nnuclear weapons program. The government therefore procured\nsubpoenas for those records, obtaining two from a grand jury\n(for the two banks with U.S. branches) and one from the\n *\n NOTE: Portions of this opinion contain sealed\ninformation, which has been redacted.\n\f 2\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nAttorney General under the Patriot Act (for the bank that has\nno U.S. branch). The banks resisted on multiple fronts,\nclaiming that the district court lacked personal jurisdiction, that\nthe Patriot Act subpoena exceeded the government’s statutory\nauthority, and that compelling production would run afoul of\ncomity principles because turning over the records—even in\nresponse to a U.S. court order—would trigger penalties under\nChinese law. In a thorough and thoughtful pair of opinions, the\ndistrict court overruled those objections and, when the first\nopinion failed to elicit the records, held the banks in civil\ncontempt. For the reasons set forth below, we affirm.\n\n I.\n For decades, the Democratic People’s Republic of Korea\n(better known as North Korea) has aspired to a battle-ready\nnuclear weapons program. Seeking to defuse the threat, the\nU.S. government has imposed stiff financial sanctions on the\nNorth Korean regime and its instrumentalities, exercising\nauthority located in the International Emergency Economic\nPowers Act. See 50 U.S.C. §S 170 1—02 (granting the President\nthe power to “prohibit” various “transactions” after declaring a\n“national emergency” posed by “any unusual and extraordinary\nthreat. to the national security, foreign policy, or economy\n . .\n\n\n\n\nof the United States”). Relevant to this case, those sanctions are\nintended to cut off North Korea’s access to American markets\nand currency. See Executive Order No. 13,382, § 1, 70 Fed.\nReg. 38,567, 38,567 (June 28, 2005) (authorizing various\ncabinet officers to “block[J” entities that further nuclear\nproliferation by preventing any “property or interests in\nproperty” from being “transferred, paid, exported, withdrawn,\nor otherwise dealt in”). According to government investigators,\nhowever, North Korea manages to evade the sanctions by using\n\f 3\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nChinese front companies that cloak the true ownership of the\nfunds involved.\n\n The facts uncovered by this investigation so far are a case\nin point. The U.S. government has collected substantial\nevidence that a now-defunct Chinese company called\n (“the Company”)\nacted as a front for , a North Korea-\nowned entity (“the NKE”), by facilitating transactions that\nviolated the sanctions orders. In fact, to date the government’s\n“investigation [has] not yield[ed] any evidence that” the\nCompany “was ever used as a legitimate business.”\nDeclaration of (“FBI\nDeclaration”) ¶ 10, Joint Appendix (“J.A.”) 864.\nUnsurprisingly, then, in addition to the NKE itself, the\ngovernment has sanctioned the Company and two of its alleged\nfounders. Actions Taken Pursuant to Executive Order 13382,\n (designating the\nNKE); Notice of [Office of Foreign Asset Control (OFAC)]\nSanctions Actions,\n (designating the Company); Notice of OFAC Sanctions\nActions, •\n(designating alleged co-founder\n —\n ); Notice of OFAC\nSanctions Actions,\n(designating alleged co-founder ).\n According to the government, the scheme operated like\nthis: Hobbled by the worthlessness of North Korea’s currency,\nthe NKE would use the Company to make or receive payments\nin U.S. dollars. These transactions helped North Korea access\nresources that would otherwise have been beyond its reach. For\nexample, the Company’s assistance allegedly enabled North\nKorea to export hundreds of millions of dollars of coal and\nother minerals, generating revenue in U.S. currency that North\n\f 4\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nKorea could then use to requisition other commodities vital to\nits weapons program.\n\n In these transactions, the Company routinely took\nadvantage of U.S. correspondent bank accounts—that is,\naccounts held by banks outside the United States at banks\nlocated inside the United States. See 31 U.S.C.\n§ 53 18A(e)(1)(B) (defining “correspondent account” to mean\n“an account established to receive deposits from, make\npayments on behalf of a foreign financial institution, or handle\nother financial transactions related to such institution”). Such\naccounts allow foreign “banks to conduct business and provide\nservices for their customers in” the United States even if “the\nbanks have no physical presence” here. Minority Staff of the\nSenate Permanent Subcommittee on Investigations,\nCorrespondent Banking: A Gateway for Money Laundering,\nS. Hrg. 107-84, 107th Cong., app. 287 (2001). This, in turn,\nenables a foreign bank’s customers to “receive many or all of\nthe services offered by [a] U.S. bank” without forming a direct-\nclient relationship with an American institution. Id.\nCorrespondent banking arrangements are fairly ubiquitous:\n“[d]ue to U.S. prominence in international trade and the high\ndemand for U.S. dollars due to their overall stability, most\nforeign banks that wish to provide international services to\ntheir customers have accounts in the United States capable of\ntransacting business in U.S. dollars.” Id. In short, then,\ncorrespondent bank accounts allow foreign banks to offer their\nclients services—including obtaining and transacting in U.S.\ncurrency—without opening a U.S. branch.\n\n These accounts are how the appellant banks (collectively,\nthe “Banks”) enter the story. All three are China-based—\nindeed, the Chinese government owns a substantial minority of\neach—and hold correspondent accounts in the United States.\n\f 5\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nBanks One and Two also operate branches in the United States,\nbut according to the government the Company relied on the\nthree Banks’ U.S. correspondent accounts to execute its\nscheme.\n\n To better grasp the full scope of the Company’s\noperations, the government seeks certain records from the\nBanks. The United States and China have a bilateral agreement\nfor obtaining just these kinds of records. Known as the Mutual\nLegal Assistance Agreement (MLAA), it obligates the two\ncountries to assist each other “in proceedings related to\ncriminal matters,” including by “providing . . records or\n .\n\n\n\narticles of evidence” and “executing requests for inquiry,” and\nspecifies the “[f]orm and [c]ontents” of any requests.\nAgreement Between the Government of the United States of\nAmerica and the Government of the People’s Republic of\nChina on Mutual Legal Assistance in Criminal Matters, arts. 1,\n4, June 19, 2000, T.I.A.S. No. 13,102. Absent an MLAA\nrequest, Chinese law prohibits banks from disclosing records\nto international investigators.\n\n In this case, however, the United States chose not to go\nthrough the MLAA. Stymied by what it perceives as sluggish\nand typically fruitless cooperation from Chinese authorities in\nthe past, the government unilaterally issued subpoenas to all\nthree Banks in December 2017. The two banks with branches\nin the United States\n (“Bank One” and “Bank Two,”\nrespectively)—received grand jury subpoenas. See Federal\nRule of Criminal Procedure 1 7(e)(1). The third\n (“Bank Three”)—has no branch in\nthe United States and received a subpoena from the Attorney\nGeneral pursuant to the Patriot Act, which gives the Attorney\nGeneral and the Treasury Secretary special investigatory tools\n\f 6\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nwhen it comes to U.S. correspondent bank accounts. See 31\nU.S.C. § 53 18(k)(3)(A)(i). The subpoenas seek all records\ndated 2012 to 2017 concerning all correspondent banking\ntransactions associated with certain accounts linked to the\nCompany, defining records to include signature cards, ledger\ncards, account statements, and documentation of deposits,\nwithdrawals, and transfers. Through this investigation, the\ngovernment hopes to learn whether the Company or any other\nentities have committed various federal crimes, including\nengaging in unlicensed financial transactions (violating 50\nU.S.C. § 1705), money laundering (violating 18 U.S.C.\n § 1956), or failing to exercise due diligence in correspondent\nbanking as required by the Bank Secrecy Act (violating 31\nU.S.C. § 531$ and 5322). The government does not currently\nsuspect the subpoenaed Banks of any wrongdoing.\n\n In correspondence with the U.S. government, the Banks\nand the Chinese government protested that complying with the\nsubpoenas would violate multiple Chinese laws, including the\nlaw designating the MLAA process as the exclusive\nmechanism for disclosures. They claimed that failure to use the\nMLAA channel would expose the Banks to administrative and\ncriminal penalties. In response, the U.S. government took the\nposition that such a request would be futile, given China’s slow\nand spotty history of providing records requested through that\nprocess.\n\n There matters stood for about a year. Despite traveling\ntwice to China for what proved to be unproductive negotiations\nwith the Chinese government, there were no major\nbreakthroughs: the United States refused to submit an MLAA\nrequest, and the Banks continued to withhold the records. At\nlast, in November 2018, things kicked back into gear when the\ngovernment moved the district court to compel enforcement.\n\f 7\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nStill resisting, the Banks disputed the district court’s personal\njurisdiction and contended that enforcing the subpoenas would\nviolate deeply ingrained principles of international comity.\nBank Three also protested that the scope of its subpoena\nexceeded the Attorney General’s Patriot Act authority.\n\n The district court granted the government’s motion to\ncompel. In re Grand Jury Investigation ofPossible Violations\nof 18 USC. §‘ 1956 and 50 US.C. 1705, Nos. 1:18-mc-175,\n-176, -177, slip op. at 3 (D.D.C. Mar. 18, 2019) (“March op.”).\nThe court found that Banks One and Two had consented to\npersonal jurisdiction when they signed an agreement with the\nfederal Reserve permitting them to open branches in the\nUnited States. Id. at 11—13. As for Bank Three, which has no\nbranch in the United States, the court determined that (along\nwith Banks One and Two) it had sufficient contact with the\nUnited States as a whole and the subpoenas sufficiently related\nto that contact so as to support the court’s personal jurisdiction.\nId. at 13—27; see also International Shoe Co. v. State of\nWashington, Office of Unemployment Compensation and\nPlacement, 326 U.S. 310, 316 (1945) (personal jurisdiction\nrequires “minimum contacts” with the forum such that\njurisdiction is consistent with “traditional notions of fair play\nand substantial justice” (internal quotation marks omitted)).\nThe district court also perceived no problems with the scope of\nthe Patriot Act subpoena. March Op. 27—31. Finally, although\nthe government conceded that compliance with the subpoenas\nwould entail violating at least some Chinese laws, the district\ncourt concluded that comity principles did not require the\nsubpoenas to be quashed. Id. at 32—5 8. The court found that,\nbased on past practice, China’s preferred MLAA procedure\nwould likely be futile and determined that the U.S.\ngovernment’s strong national security interests outweighed\nChina’s stake in enforcing the identified laws. Id. at 42—51. The\n\f 8\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ncourt was unmoved by a series of letters from the Chinese\nMinistry of Justice promising to “promptly review and\nprocess” any MLAA request in this case. Id. at 45 (quoting\nLetter from Du Yaling, Director General of International\nCooperation Department of the Ministry of Justice, to Chief\nJudge Beryl A. Howell (Feb. 26, 2019), J.A. 966).\n\n About a month later, after the Banks made clear they\nwould not comply with the compulsion order, the district court\nheld them in contempt. In re Grand Jury Investigation of\nPossible Violations of 18 U.S.C. § 1956 and 50 US.C. § 1705,\nNos. 1:18-mc-175, -176, -177, slip op. at 1—3 (D.D.C. Apr. 10,\n2018) (“April Op.”). It imposed a $50,000 per day fine against\neach bank but stayed that penalty contingent on the parties\nseeking an expedited appeal. Id. at 14—15. They obliged, and\nwe agreed to accelerate our consideration of the case. Per\nCuriam Order, In re: Sealed Case, No. 19-5068 (D.C. Cir. Apr.\n25, 2019). We now turn to the Banks’ arguments, beginning\nwith personal jurisdiction before examining the merits.\n\n II.\n The Constitution forbids federal courts from exercising\npersonal jurisdiction absent sufficient “minimum contacts” to\nsatisfy “traditional notions of fair play and substantial\njustice.” International Shoe, 326 U.S. at 316 (quoting Milliken\nv. Meyer, 311 U.S. 457, 463 (1940)). That protection extends\nto private foreign entities—such as the Banks—just as to\nAmerican citizens. See Livnat v. Palestinian Authority, 851\nF.3d 45,48 (D.C. Cir. 2017) (delineating the “general rule” that\nrestrictions on personal jurisdiction “protect{J all litigants in\nour courts”). Unlike subject-matter jurisdiction, however,\npersonal jurisdiction “can. be waived,” meaning a party\n . .\n\n\n\nmay “consent” to a court’s personal jurisdiction. Insurance\n\f 9\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nCorp. of Ireland v. Compagnie des Bauxites de Guinee, 456\nU.s. 694, 703 (1982). Because neither the government nor the\nBanks suggests that these principles operate differently in\ngrand jury investigations than in ordinary civil litigation, we\nshall treat these two contexts as identical for purposes of this\ncase. But see Boydv. Meachum, 77 f.3d 60,66 (2d Cir. 1996)\n(“While the federal constitutional requirements of personal\njurisdiction in a civil setting are reasonably well-defined,\nthis is not so in a criminal case.”).\n\n All three Banks contest personal jurisdiction. “Reviewing\nthe district court’s assertion of personal jurisdiction de novo,”\nUnited States ex rel. Miller v. Bill Harbert International\nConstruction, Inc., 608 F.3d 871, 886 (D.C. Cir. 2010), we\nconclude that the Banks’ challenges are meritless. Banks One\nand Two consented to jurisdiction when they opened branches\nin the United States. And although Bank Three has no U.S.\nbranch and executed no such agreement, its choice to maintain\ncorrespondent accounts in the United States establishes an\nadequate connection to the forum and the enforcement action\nto sustain jurisdiction.\n\n A.\n Congress requires any “foreign bank” to acquire “prior\napproval of the” Federal Reserve before it “may establish a\nbranch or an agency” in the United States. 12 U.S.C.\n§ 3 105(d)(1). The Federal Reserve, in turn, has authority to\n“impose such conditions on its approval” of a foreign bank’s\napplication “as it deems necessary,” Id. § 3 105(d)(5), to,\namong other things, assure itself that the bank will be “in\ncompliance with applicable United States law” and mitigate\nany “risk to the stability of the United States financial system,”\nid. § 350 1(d)(3)(D)—(E). See also 12 C.F.R. § 21 1.24(c)(2)(vii)\n\f 10\n PUBLIC COPY - SEALED INFORMATION DELETED\n\n(Federal Reserve scrutinizes “[wjhether the foreign bank...\nhas established adequate controls and procedures. to ensure . .\n\n\n\ncontinuing compliance with U.S. law, including controls\ndirected to detection of money laundering”).\n\n When Banks One and Two applied to open their U.S.\nbranches, both executed agreements with the Federal Reserve\nto assure compliance with relevant provisions of American\nlaw. As part of that deal, they “consent[edj to the jurisdiction\nof the federal courts of the United States for purposes of\n . . .\n\n\n\nany and all. proceedings initiated by. the United\n . . . .\n\n\n\nStates. in any matter arising under U.S. Banking Law.”\n . .\n\n\n\nLetter from Vice Chairman & President,\n ,\n\n\n to Board of Governors of the\nFederal Reserve System 1 l.A. 1659 (“Bank\n ,\n\n\nOne Consent”); Assurances Commitment and Consent to\nJurisdiction Commitment\n 1, J.A. 1663 (“Bank Two Consent”). The agreements\ndefine “U.S. Banking Law” expansively to include “all federal\ncriminal laws of which violation(s) arise(s): under .the . . . . .\n\n\n\nBank Secrecy Act.” Bank One Consent 3 n.2, l.A. 1661; Bank\nTwo Consent 1 n.2, l.A. 1663.\n\n The plain language of those agreements covers this case.\nThis investigation is unquestionably a “proceeding[] initiated\nby. the United States” and, as the government explains, “the\n . .\n\n\n\ngrand jury is investigating whether those financial institutions\nthat provided services in the United States”—for example, the\nAmerican institutions hosting the foreign Banks’\ncorrespondent accounts—”exercised due diligence.”\nAppellee’s Br. 18 (citing 31 U.S.C. § 5318, 5318A, 5322).\nThe Bank Secrecy Act makes willful dereliction of that\nobligation a federal criminal violation. 31 U.S.C. § 5322. In\nshort, then, the government has assembled the necessary\n\f 11\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ningredients for jurisdiction: a criminal investigation initiated by\nthe United States (the grand jury investigation) arising under\nU.S. Banking Law (the Bank Secrecy Act).\n\n The banks’ contrary arguments amount to nothing more\nthan an unsuccessful effort to move the goalposts. They insist\nthat there must be a reasonable showing that someone actually\nviolated the Bank Secrecy Act. Bank Two Br. 8. Not so. All\nour precedent requires is a “reasonable probability that. the. .\n\n\n\nfacts necessary for the exercise of jurisdiction” exist. In re\nSealed Case, $32 F.2d 1268, 1274 (D.C. Cir. 1987) (quoting In\nre Marc Rich & Co., 707 F.2d 663, 670 (2d Cir. 1983)). Here,\nnothing in the consent agreements requires the government to\nestablish liability; they require only that the proceedings\n“aris[e] under U.S. Banking Law.” Bank One Consent 1,\nJ.A. 1659 (emphasis added); Bank Two Consent 1, J.A. 1663\n(emphasis added). Whether a proceeding “arises under” a\nparticular law is a question about its subject matter at the outset,\nnot a question about its outcome. To look at the same phrase in\na more typical context, when we ask whether a lawsuit “aris[es]\nunder” federal law for purposes of subject-matter jurisdiction,\nU.S. Const. art. III, § 2, we look to whether the plaintiffs\n“assert a claim” based on “federal law,” not whether they will\nultimately prevail, National farmers Union Insurance Cos. v.\nCrow Tribe of Indians, 471 U.S. 845, 850 (1985) (emphasis\nadded); see also American Well Works Co. v. Layne & Bowler\nCo., 241 U.S. 257, 260 (1916) (“A suit arises under the law that\ncreates the cause of action.”). Given this “legal lineage,” Hall\nv. Hall, 13$ S. Ct. 1118, 1125 (201$),agrand jury investigation\nplainly “arises under” a particular statute when it probes\nwhether that statute has been violated.\n\n Read properly, the consent agreements demand only that\n“U.S. Banking Law” form the basis of the investigation. The\n\f 12\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ngovernment has cleared that threshold here. Even if, as the\nbanks suggest, this investigation’s potential to yield successful\nprosecutions is uncertain, the banks have offered no basis for\ndoubting that “U.S. Banking Law” constitutes this\ninvestigation’s subject matter.\n\n The banks also suggest that there is a problem with\npersonal jurisdiction because the subpoenaed records are\nlocated outside the United States. But that fact has no bearing\non personal jurisdiction given that the banks’ consent\nagreements supply the basis for jurisdiction. The argument is\nbetter conceived as a challenge to the scope of the subpoenas—\nan understanding not evident from the banks’ briefs. Even were\nwe to entertain the argument, however, it would go nowhere.\nFederal courts have long required the production of documents\nheld abroad. See, e.g., In re Marc Rich, 707 F.2d at 667 (A\nwitness may not “resist the production of documents on the\nground that the documents are located abroad.”); United States\nv. first National City Bank, 396 F.2d 897, 900—01 (2d Cir.\n196$) (“[A] federal court has the power to require the\nproduction of documents located in foreign countries if the\ncourt has in personam jurisdiction of the person in possession\nor control of the material.”). That practice makes sense: grand\njuries often investigate violations of criminal laws that\nobviously cover extraterritorial conduct. “It would be strange,\nindeed, if the United States could punish a foreign corporation\nfor violating its criminal laws . . but a federal grand jury\n . ,\n\n\ncould not investigate to ascertain the probability that a crime\nhad taken place.” In re Marc Rich, 707 F.2d at 666.\n\n B.\n Bank Three, by contrast, has neither opened a U.S. branch\nnor consented to personal jurisdiction. We therefore must\n\f 13\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ndetermine whether it “has purposefully directed” its relevant\n“activities” at “the forum.” Burger King Corp v. Rudzewicz,\n471 U.S. 462, 472 (1985) (internal quotation marks omitted).\n\n That, of course, requires us to correctly identify “the\nforum,” a task ordinarily linked to where Congress has\nauthorized service of process. See BNSF Railway Co. v.\nTyrrell, 137 S. Ct. 1549, 1555—56 (2017) (“Congress’ typical\nmode of providing for the exercise of personal jurisdiction has\nbeen to authorize service of process” because “a basis for\nservice of a summons on the defendant is prerequisite to the\nexercise of personal jurisdiction”). In many situations,\nCongress limits the territorial reach of federal district courts to\nthat of the state in which they sit. See Wright & Miller,\n4 Federal Practice & Procedure § 1068.1 (4th ed.); see, e.g.,\nFederal Rule of Civil Procedure 4(k)(1)(A). For example,\nunless a federal statute otherwise provides, by virtue of Rule\n4(k)(1)(A) the jurisdictional reach of a federal court in a civil\naction is keyed to that of a court of general jurisdiction in the\nstate in which it sits. So in reviewing the constitutionality of\ntheir assertions of jurisdiction, federal courts oflen perform\nstate-specific contacts analyses. Although the constitutional\nrestraint in those situations necessarily flows from the Fifth\nAmendment, the scope of the minimum-contacts inquiry is the\nsame as it would be under the Fourteenth Amendment for a\nclaim in state court: confined to contacts with the forum state.\n\n According to Bank Three, a state-specific contacts\nanalysis is appropriate for the Patriot Act subpoenas at issue\nhere. And, the Bank argues, since its agent was served in New\nYork and the Bank lacks any meaningful contact with the\nDistrict of Columbia, the district court here lacked jurisdiction.\nFor its part, the government maintains that the Patriot Act\nauthorizes nationwide service of process. Under those\n\f 14\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ncircumstances, the Fifth Amendment requires only “minimum\ncontacts with the United States” as a whole—rather than with\nthe forum state. SEC v. Bilzerian, 378 F.3d 1100, 1106 n.8\n(D.C. Cir. 2004); accord Livnat, 851 F.3d at 55 (“Under the\nFifth Amendment, which defines the reach of federal courts,\ncontacts with the United States as a whole are relevant.”).\nTherefore, the government argues, the district court properly\nconsidered Bank Three’s contacts with the United States as a\nwhole, including its maintenance and use of correspondent\naccounts in New York.\n\n We begin by observing that the usual process of looking to\nwhere federal law authorizes service of process as a proxy for\nwhether Congress sought to trigger a federal court’s nationwide\nterritorial reach is, perhaps, not a perfect fit for this provision\nof the Patriot Act. Because the statute confers authority to issue\na subpoena directly on the Attorney General and the Treasury\nSecretary, rather than on a federal district court, Congress may\nhave given little consideration to the territorial reach of a\ndistrict court considering a motion to compel enforcement of\nsuch a subpoena. But given that Bank Three assumes that\nauthorizing nationwide service in this context would trigger a\nnationwide contacts analysis, we shall too. See Sickle v. Torres\nAdvanced Enterprise Solutions, LLC, 884 F.3d 338, 344 (D.C.\nCir. 2018) (accepting a party’s decision “not to brief or argue\nthe question of personal jurisdiction” because that “defense .\n\n\n\n\ncan be waived or forfeited”).\n\n Here, that service-of-process provision, codified in 31\nU.S.C. § 5318(k), authorizes the kind of subpoena served on\nBank Three to “be served on the foreign bank in the United\nStates if the foreign bank has a representative in the United\nStates.” 31 U.S.C. § 5318(k)(3)(A)(ii). We know of no other\nservice-of-process provision in the U.S. Code phrased exactly\n\f 15\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nthis way. Still, we are not entirely without guideposts as we\nevaluate this question.\n\n We know, for instance, that Congress has many different\nways of authorizing nationwide service of process. Sometimes\nit makes its intention abundantly clear by authorizing service\n“anywhere in the United States.” See, e.g., 7 U.S.C. § 499g(b)\n(service for enforcing in federal district courts certain orders of\nthe Secretary of Agriculture); 2$ U.S.C. § 232 1(c) (same for\nenforcing orders of the Surface Transportation Board); 33\nU.S.C. § 921(c) (service for review proceedings in the courts\nof appeals of decisions of the Benefits Review Board for\nlongshore and harbor workers). On other occasions, Congress\nemploys some riff on the phrase “wherever the defendant may\nbe found,” language generally understood to refer to\nnationwide service. See, e.g., 7 U.S.C. § 25(c) (service in\nprivate lawsuits under the Commodity Exchange Act); see also\nOmni Capital International, Ltd. v. Rudolf Wolff & Co., 484\nU.S. 97, 105—06 (1987) (recognizing that such language\n“explicitly authorize[sJ nationwide service of process”);\nAppellee’s Br. 25 n.20 (collecting additional examples). We\nalso know that Congress can very clearly limit service to a\nparticular judicial district when it wants to. See, e.g., 45 U.S.C.\n § 3 62(b) (authorizing process issued by most courts in Railroad\nRetirement Board proceedings to be served “in the judicial\ndistrict of the district court issuing such. process”).\n . .\n\n\n\n\n At first glance, the language in section 5318(k) looks more\nlike the provisions that authorize nationwide service. So long\nas a foreign bank has designated a representative, the statute\nallows for service “in the United States,” without specifying\nany state or judicial district. True, the provision lacks the\nunmistakable clarity of the statutes that expressly authorize\nservice “anywhere” in the country; nonetheless, the text\n\f 16\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ncontains no limitation other than the territorial boundaries of\nthe entire United States. Bank Three, however, seizes on the\nrequirement to designate a representative to argue that, in fact,\nCongress only meant to authorize service in the judicial district\nwhere that representative is located.\n\n Bank Three’s position makes some intuitive sense. After\nall, as a matter of physics, the designated representative will\nhave a single location. On further reflection, however, this\nlogic fails to resolve things in the Bank’s favor. After all, both\nBank Three and the government agree that language\nauthorizing service “wherever a defendant may be found”\nauthorizes nationwide service, see Bank Three Br. 51,\nAppellee’s Br. 25 n.20, even though in many instances that\ndefendant, too, will only be located in one place at a time.\n\n Even more instructive, in other service-of-process regimes\ninvolving designated representatives, when Congress wants to\nspecify a particular location or district for service, it says so\ndirectly. For example, in certain trademark actions, Congress\nhas instructed foreign mark holders to designate a U.S. agent\nand authorized service of process “by leaving with that person\nor mailing to that person a copy [of the process] at the address\nspecfied in the last designation so filed.” 15 U.S.C. § 1141k(f)\n(emphasis added). Congress has even provided a backup\noption, explaining that “[i]f the person so designated cannot be\nfound at the last designated address” then “process may be\nserved on the Director” of the United States Patent and\nTrademark Office. Id.; see also 7 U.S.C. § 2569 (containing a\nsimilar backup provision for when a designee “cannot be found\nat the address given”). The provision at issue in this case is\nnoticeably devoid of any similar language, leading to the\ninference that Congress sought to impose no similar\ngeographical limitation here.\n\f 17\n PUBLIC COPY - SEALED INFORMATION DELETED\n\n\n\n Reading section 5318(k) to authorize nationwide service\nalso aligns with the Patriot Act’s structure and purpose. Cf\nUnited States International Trade Commission v. ASAT, Inc.,\n411 f.3d 245, 251 (D.C. Cir. 2005) (examining, when\nconsidering whether a statute authorizes nationwide service,\nwhether “Congress would have desired nationwide service of\nprocess to effectuate the underlying statute’s purpose”). Rather\nthan requiring the Attorney General or Treasury Secretary to\ngo to court to obtain a subpoena, Congress authorized those\nofficers to issue process themselves. Unlike federal district\ncourts, which customarily serve a particular geographical area,\nCabinet officers serve the entire country. Structurally speaking,\nthen, Congress’s choice to give two such officers the subpoena\npower evokes an intent to rely on their nationwide reach. As\nfor purpose, having found that correspondent accounts were a\nsource of special concern for money laundering, Congress\nsought “to provide a clear national mandate for subjecting to\nspecial scrutiny those. . financial institutions operating\n .\n\n\n\noutside of the United States. that pose particular, identifiable\n . .\n\n\n\nopportunities for criminal abuse.” USA PATRIOT Act of 2001,\nPub. L. No. 107-56, § 302(a)(6), (b)(4), 115 Stat. 272, 296—97.\nIt would be odd if, in establishing this “clear national mandate,”\nCongress had geographically subdivided the government’s\ninvestigative powers.\n\n Apparently recognizing that its preferred interpretation of\nsection 5318(k) may not represent “the most natural reading of\nthe provision,” Bank Three Br. 4$, Bank Three falls back on\nthe presumption against extraterritoriality, which teaches that\nwithout a “clearly expressed congressional intent to the\ncontrary, federal laws will be construed to have only domestic\napplication.” RJR Nabisco, Inc. v. European Community, 136\nS. Ct. 2090, 2100 (2016). According to the Bank, reading the\n\f 18\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nPatriot Act to authorize nationwide service “would vastly\nexpand the extraterritorial scope of federal regulatory power.”\nBank Three Br. 49. But that does not follow. This service-of-\nprocess provision has no impact on the substantive\nextraterritorial reach of federal law; it merely helps sort out\nwhich federal court has authority to enforce the subpoena. After\nall, even under Bank Three’s preferred interpretation, at least\none federal court could assert jurisdiction over Bank Three (the\none in the district where the Bank’s agent resides). The\npresumption against extraterritoriality therefore offers no basis\nto upset the most straightforward reading of the text: that once\na foreign bank has designated a representative in the United\nStates, service may be accomplished throughout “the United\nStates.”\n\n Having defined the forum as the entire United States, we\nwould usually go on to assess whether the Bank’s contacts with\nthat forum are sufficient. In this case, however, Bank Three\nplaced all its eggs in the forum-identification basket; it claims\nonly that it lacks the necessary contacts with the District of\nColumbia, offering no argument that its contacts with the\nUnited States as a whole are insufficient. See Bank Three\nBr. 46—47 (“The only way the government can prevail is by\nshowing that the relevant forum is the United States as a\nwhole.”). We therefore have no occasion to question the district\ncourt’s conclusion that Bank Three’s maintenance of\ncorrespondent accounts in the United States supplies the\nnecessary nexus. March op. 2 1—24. With no further\njurisdictional hurdles in our path, we proceed to the merits.\n\n III.\n Bank Three, the only bank to receive a Patriot Act\nsubpoena, argues that its subpoena exceeds the Attorney\n\f 19\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nGeneral’s statutory authority. Although “[w]e generally review\ndistrict court decisions enforcing document subpoenas only for\narbitrariness or abuse of discretion,” we afford no deference to\na decision that applies “the wrong legal standard.” In re Sealed\nCase, 146 F.3d 881, $83 (D.C. Cir. 199$). We therefore review\n“de novo” the district court’s “legal conclusions,” including its\nstatutory interpretation, even as we review “its factual findings\nfor clear error.” United States v. Ahn, 231 F.3d 26, 38 (D.C.\nCir. 2000).\n\n Congress enacted section 531 8(k)(3) “to increase the\nstrength of United States measures to prevent, detect, and\nprosecute international money laundering and the financing of\nterrorism.” Pub. L. No. 107-56, § 302(b)(1), 115 Stat. at 297.\nIn particular, Congress found that correspondent accounts held\nby foreign banks were “one of the banking mechanisms\nsusceptible in some circumstances to manipulation by foreign\nbanks to permit the laundering of funds.” § 302(a)(6), 115 Stat.\nat 296. Thus, it authorized both the Attorney General and the\nTreasury Secretary to issue a “subpoena to any foreign bank\nthat maintains a correspondent account in the United States and\nrequest records related to such correspondent account,\nincluding records maintained outside of the United States\nrelating to the deposit of funds into the foreign bank.” 31\nU.S.C. § 5318(k)(3)(A)(i).\n\n Bank Three claims the subpoena it received strays outside\nthe statutory lines because it is “not expressly confine[d]” to\ndocuments related to the Bank’s correspondent accounts in the\nUnited States. Bank Three Br. 23. To be sure, the subpoena is\ncomprehensive: it seeks “[a]1l documents relating to [the\nCompany’s] correspondent banking transactions”—including,\nit would seem, documents related to transactions using\ncorrespondent accounts in countries other than the United\n\f 20\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nStates. Bank Three Subpoena, J.A. 771; see also id. (requesting\n“signature cards,” “account ledger cards,” “periodic account\nstatements,” and “records. of all items deposited,\n . .\n\n\n\n\nwithdrawn, or transferred”). And we know Bank Three does,\nin fact, have correspondent accounts outside the United States.\nIf we knew nothing else about the Company, Bank Three’s\nargument might therefore have some force: where, as here, a\nbank maintains correspondent accounts in multiple countries,\nit is by no means a given that all of an account holder’s\ncorrespondent banking transactions “relate to” the bank’s US.\ncorrespondent accounts, as the statute requires.\n\n With respect to this particular case, however, the\ngovernment has a compelling response. It argues, as it did\nbefore the district court, that evidence already collected during\nthe investigation shows that the Company’s “exclusive raison\nd’etre” was to serve as “a progenitor of U.S. dollars for” the\nNKE. Appellee’s Br. 62; see also Id. at 64 (describing the\nCompany as “a front company created for the sole purpose of\nexecuting U.S.-dollar transactions”); March 5, 2019 Hearing\nTr. 95:17—25, l.A. 1650—5 1 (“[The Company] existed only for\none reason. [T]he [NKE] has to set up entities like this,\n . .\n\n\n\n\nbecause [it] want[s] access to the U.S. financial system.”). In\nthe government’s view, such a showing means that all of the\nCompany’s records are “related to” the Bank’s U.S.\ncorrespondent accounts because those accounts were part of the\nCompany’s means of obtaining U.S. dollars.\n\n The parties’ dispute turns on the meaning of the term\n“related to” as used in section 531 8(k)(3). Arguing for a narrow\ninterpretation, Bank Three reads the statute to authorize\ncollecting only those records of transactions that themselves\npassed through Bank Three’s U.S. correspondent accounts. For\nits part, the government advocates a more expansive reading—\n\f 21\n PUBLIC COPY - SEALED INFORMATION DELETED\n\none that would cover all records that have a “connection with”\nthe Company’s use of Bank Three’s U.S. correspondent\naccounts, even if some individual transactions made no use of\na U.S. correspondent account. Appellee’s Br. 60 (internal\nquotation marks omitted).\n\n To assess how broadly the phrase “related to” sweeps, we\nlook to the “traditional tools of statutory interpretation—text,\nstructure, purpose, and legislative history.” Tax Analysts v. IRS,\n350 F.3d 100, 103 (D.C. Cir. 2003) (internal quotation marks\nomitted). The “first step in interpreting a statute is to determine\nwhether the language at issue has a plain and unambiguous\nmeaning”; “[o]ur inquiry must cease if the statutory language\nis unambiguous and the statutory scheme is coherent and\nconsistent.” Robinson v. Shell Oil Co., 519 U.S. 337, 340\n(1997) (internal quotation marks omitted).\n\n “[W]hen asked to interpret statutory language including the\nphrase ‘relating to,’ . .th[e] [Supreme] Court has typically\n .\n\n\n\nread the relevant text expansively.” Lamar, Archer & Cofrmn,\nLLP v. Appling, 13$ S. Ct. 1752, 1760 (201$). We see no\nreason to deviate from that approach here. The ordinary\nmeaning of that phrase encompasses the broad relationships\ndescribed by the government. See, e.g., Merriam-Webster’s\nCollegiate Dictionary 987 (10th ed. 1997) (defining “relate” as\n“to have [a] relationship or connection”); Black’s Law\nDictionary 1288 (6th ed. 1990) (defining “relate” as “[tb stand\nin some relation; to have bearing or concern; to pertain; refer;\nto bring into association with or connection with”); see also BF\nAmerica Production Co. v. Burton, 549 U.S. 84, 91(2006)\n(“Unless otherwise defined, statutory terms are generally\ninterpreted in accordance with their ordinary meaning.”).\n\f 22\n PUBLIC COPY - SEALED INFORMATION DELETED\n\n The government’s interpretation finds further support in the\none example that Congress offered of a record “related to” a\nU.S. correspondent account: “records maintained outside of the\nUnited States relating to the deposit of funds into the foreign\nbank.” 31 U.S.C. § 5318(k)(3)(A)(i). The final phrase is\nimportant. Merely depositing funds “into the foreign bank”\nwould not necessarily entail passing those funds through the\nforeign bank’s U.S. correspondent account; rather, it may be a\npreparatory step for using the correspondent account. For\ninstance, an aspiring money launderer might deposit illicitly-\nobtained foreign currency into the foreign bank before using\nthe U.S. correspondent account to facilitate a payment in U.S.\ndollars, helping the guilty party conceal the origin of the funds.\nBy specifying that its view of records “related to” a U.S.\ncorrespondent account includes such a predicate transaction,\nCongress embraced an expansive view of the term in line with\nthe government’s. And, of course, that view accords with\nCongress’s stated purpose in enacting the provision: “to\nincrease the strength of United States measures to prevent,\ndetect, and prosecute international money laundering and the\nfinancing of terrorism.” Pub. L. No. 107-56, § 302(b)(1), 115\nStat. at 297.\n\n Resisting this expansive text, Bank Three offers two\nreasons to prefer a more circumscribed understanding of the\nstatute. First, it again looks to the presumption against\nextraterritoriality. But here, the statute contains no ambiguity\nrequiring us to resort to that presumption. See Validus\nReinsurance, Ltd. v. United States, 786 F.3d 1039, 1046 (D.C.\nCir. 2015) (explaining that “ambiguity is resolved by\n[applyingj the presumption against extraterritoriality”\n(emphasis added)). To be sure, the statutory text is capacious,\nbut as the Supreme Court has observed, “[b]road. language\n . .\n\n\n\nis not necessarily ambiguous when congressional objectives\n\f 23\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nrequire broad terms.” Diamond v. Chakrabarty, 447 U.S. 303,\n315 (1980).\n\n Second, Bank Three points to the Justice Department’s\nattempt to persuade Congress to amend the statute to allow the\ngovernment to obtain not just “records relating to” a foreign\nbank’s U.S. correspondent account, but also “any records”\npertaining to “any related account at the foreign bank,\nincluding records maintained outside of the United States, that\nare the subject of any investigation of a criminal violation of\nUnited States law.” Department of Justice, Title III, Anti-\nCorruption Legislative Proposals, 114th Cong., at 4 (May 5,\n2016), https://www.justice.gov/opa/file/849986/download\n(emphases added). The Bank would have us infer from\nCongress’s failure to adopt the amendment that section\n53 18(k)(3) as currently written can never reach “[a]l1\ndocuments” relating to “all” of a company’s “correspondent\nbanking transactions,” as this subpoena does. Bank Three\nSubpoena, J.A. 771. But the Supreme Court has long cautioned\nagainst reading too much into congressional inaction given that\n“[l]ogically, several equally tenable inferences could be drawn\nfrom the failure of the Congress to adopt an amendment.\nincluding the inference that the existing legislation already\nincorporated the offered change.” United States v. Wise, 370\nU.S. 405, 411 (1962). In any event, the unenacted amendment\nBank Three cites does not expand the definition of what records\ncount as “related to” a U.S. correspondent account; rather, it\nwould allow the government to reach other accounts,\nregardless of their relationship to a U.S. correspondent account.\nSuch a proposal tells us little about what Congress currently\nconsiders to be “related to” a correspondent account.\n\n We therefore conclude that records “related to” a U.S.\ncorrespondent account include records of transactions that do\n\f 24\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nnot themselves pass through a correspondent account when\nthose transactions are in service of an enterprise entirely\ndedicated to obtaining access to U.S. currency and markets\nusing a U.S. correspondent account. All that remains is to\ndetermine whether the government has adequately shown that\nthe Company was such an enterprise.\n\n The district court found that the government had made the\nnecessary showing. Specifically, it found that “[n]o part” of the\nCompany could “be separated from the ploy to put money\nthrough the United States financial market.” March Op. 29. On\nthe record before us, that finding was not clearly erroneous. See\nBoca Investerings Partnership v. United States, 314 F.3d 625,\n629—30 (D.C. Cir. 2003), as amended (Mar. 26, 2003) (a\nconclusion is clearly erroneous when “the reviewing court on\nthe entire evidence is left with the definite and firm conviction\nthat a mistake has been committed”) (internal quotation marks\nomitted). The investigation has turned up considerable\nevidence supporting the court’s conclusion. The Company\nengaged in an enormous volume of U.S. dollar transactions:\n“[B]etween October 2012 and November 2015 alone, [the\nCompany] made over $100 million U.S.-dollar payments.” FBI\nDeclaration ¶ 13, J.A. 866. And the Company “was a\ncounterparty to over 680 U.S.-dollar international transfers\n(which transited through the United States financial system via\ncorrespondent U.S. banks).” Id., J.A. 867. The government has\nuncovered no “evidence that” any of these transactions were\npart of “a legitimate business”; rather, all were part of the\nNKE’s scheme to launder money. Id. ¶ 10, J.A. 864—65.\n\n Nor does any record evidence suggest that the NKE used\nthe Company to launder money through currencies other than\nU.S. dollars. Indeed, the NKE had a special need for dollars\nbecause “commodity companies, such as those operating in\n\f 25\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nChina, Thailand, and Singapore, frequently prefer to take\npayments in U.S. dollars because of the dollar’s global demand\nand stability.” Id. ¶ 26, l.A. 871; see also Correspondent\nBanking: A Gateway for Money Laundering, S. Hrg. 107-84\napp. 287 n.12 (“U.S. dollars are one of a handful of major\ncurrencies accepted throughout the world. They are also\nviewed as a stable currency, less likely to lose value over time\nand, thus, a preferred vehicle for savings, trade and\ninvestment.”). Further, the government explains, and Bank\nThree nowhere contests, that before 2017—and consequently\nduring the vast majority of the period covered by this\nsubpoena—the NKE “could legally conduct transactions in\nother currencies” besides the U.S. dollar (such as the euro),\nAppellee’ s Br. 61, signaling that the NKE would have had no\nneed of a front company to conduct transactions in those\ncurrencies.\n\n Taken together, these facts support the district court’s\nfinding that “[nb part” of the Company—and thus, none of its\nbanking records—”can be separated from the ploy” to launder\nmoney through U.S. financial markets, March op. 29, because\nthe Company’s sole purpose was to give the NKE access to\nU.S. dollars through U.S. correspondent accounts. We\ntherefore conclude that on the facts of this case, the records\nrequested by the Patriot Act subpoena all “relate to” Bank\nThree’s U.S correspondent account. To be clear, nothing in this\nopinion authorizes, as the Bank fears, the government “for any\nreason, [to] obtain essentially any document or record from any\nforeign bank with a U.S. correspondent account,” Bank Three\nBr. 22, regardless of whether those records are “related to” that\nU.S. correspondent account. The government’s power to\nsubpoena the records in this case derives from the evidence that\nthis particular company operated exclusively as a U.S. dollar\nclearinghouse for the NKE. Under such circumstances, all\n\f 26\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nrecords pertaining to the Company’s Bank Three account and\nits correspondent banking transactions, no matter where they\noccurred, are “related to” the Bank’s U.S. correspondent\naccounts.\n\n Iv.\n “Comity refers to the spirit of cooperation in which a\ndomestic tribunal approaches the resolution of cases touching\nthe laws and interests of other sovereign states.” Société\nNationale Industrielle Aérospatiale v. U.S. District Court for\nthe Southern District of Iowa (“Société Nationale”), 482 U.S.\n522, 543 n.27 (1987). Invoking this “spirit of cooperation,” Id.,\nall three Banks argue that “ordering a foreign-sovereign-owned\nbank to violate foreign law on foreign soil violates basic\n‘comity’ principles” and, as such, demands reversal, Bank One\nBr. 11; see also Bank Two Br. 25 (making this argument);\nBank Three Br. 32—33 (same). We begin by providing some\nadditional background before explaining why the Banks’\nargument fails.\n\n A.\n A comity analysis typically starts by determining whether\na “true conflict” exists between the implicated legal systems.\nHartford Fire Insurance Co. v. Calfornia, 509 U.S. 764, 798\n(1993) (internal quotation marks omitted). Absent a clash of\nforeign and domestic law, American courts may press forward\nfree from worry that their rulings will threaten the\n“international legal ties that advance the rule of law within and\namong nations.” Laker Airways Ltd. v. Sabena, Belgian World\nAirlines, 731 F.2d 909, 937 (D.C. Cir. 1984), as amended\n(Mar. 6 and 9, 1984). As the district court recognized, however,\nin this case that determination “is the easy part,” given that\n“[tJhe [U.S.] government concedes that complying with the\n\f 27\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nrespective subpoenas exposes each bank to legal penalties in\nChina.” March op. 32.\n\n After this initial inquiry, the question is whether “the\nfactual circumstances” of the case at hand counsel against\napplying American law. Laker Airways, 731 f.2d at 937. A\ncollection of factors helps us to delineate the “inherently\nuncertain” borders set by the principles of comity in the context\nof subpoena enforcement. Id. Factors include “the importance\nto the litigation of the documents . requested,” “the degree\n . .\n\n\n\nof specificity of the request,” “whether the information\noriginated in the United States,” “the extent to which\nnoncompliance with the request would undermine important\ninterests of the United States, or compliance with the request\nwould undermine important interests of the state where the\ninformation is located,” and whether “alternative means of\nsecuring the information” exist. Société Nationale, 482 U.S. at\n544 n.28 (alteration omitted) (quoting Restatement of foreign\nRelations Law of the United States (Revised) § 437(1)(c)).\nCourts also look to whether compliance would impose hardship\non the party targeted by the subpoena and whether that party\nhas acted in good faith. See In re Sealed Case (“Sealed\nCaseT’), 825 F.2d 494, 498—99 (D.C. Cir. 1987)\n(distinguishing cases based on “good faith” considerations);\nLinde v. Arab Bank, PLC, 706 f.3d 92, 110 (2d Cir. 2013)\n (“[A] district court should also examine the hardship of the\nparty facing conflicting legal obligations... .“).\n\n The district court diligently applied this “constellation of\nfactors” to the facts at issue here. March Op. 33. In so doing, it\nobserved that two factors indisputably weigh against enforcing\nthe subpoenas: the documents originated outside the United\nStates, see id. at 42, and the Banks have acted in good faith\nthroughout this litigation, see id. at 57 (“[T]he government\n\f 2$\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nagrees that none of the banks has acted in bad faith.”).\nAccording to the district court, another pair of factors just as\nclearly weighs in favor of enforcement: the documents contain\nunique information vital to the investigation—indeed, “the\ngovernment avows the subpoenaed documents are the\nfoundation of the United States’ investigation,” Id. at 39\n(alteration and internal quotation marks omitted), and the\nsubpoenas target an appropriately specific collection of\nrecords—that is, “the subpoenas are tailored to specific\nfinancial records, for a defined date range, at banks [the\nCompany] is known to have used to launder funds,” id. at 41.\n\n Parsing the more complex remaining factors, the district\ncourt emphasized the government’s goal of learning how North\nKorea has thwarted American sanctions in pursuing weapons\nof mass destruction. “[NJon-enforcement” of the subpoenas,\nthe district court stressed, “would undermine a critical national\ninterest.” Id. at 4$. Although acknowledging that China has a\n“legitimate” “interest in the development of a sound banking\nsystem,” the district court concluded that the United States’\nnational security interest easily outweighs China’s interest in\nsafeguarding its banking secrecy laws. Id. at 50; see also Id.\n(observing that because “even Chinese authorities recognize\nthat bank secrecy can co-exist with limited disclosure to\ngovernment agencies,” “China’s interest would not be\nundermined by enforcement of the subpoenas”).\n\n The district court recognized that the U.S. government had\ndeclined to take advantage of an alternative channel through\nwhich the United States might obtain the documents at issue:\nthe Mutual Legal Assistance Agreement, which creates a\nprocess to exchange “documents” and “records” relevant to\ncriminal investigations. MLAA art. l(2)(c); see March Op. 42—\n47 (discussing the MLAA). But based on the U.S.\n\f 29\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ngovernment’s declarations that China has failed to\nsatisfactorily engage in the MLAA process in many other cases\nover the past decade, the district court accepted the Executive\nBranch’s representation that requiring the government to first\nexhaust the MLAA process would just “[aJllow[J China to gum\nup [the] United States’ investigations” and, therefore, that\n“[tJhe Department of Justice no longer needs to take overtures\nof China’s willingness to assist the United $tates[] with\ninvestigations into North Korea seriously.” March op. 47.\n\n finally, with respect to the hardship facing the Banks, the\ndistrict court acknowledged that compliance with the\nsubpoenas would violate Chinese law. Though the government\ndoubts the Banks’ claim that any criminal laws would be\nbroken, it concedes that compliance would at least violate those\nlaws prohibiting Chinese companies from handing over records\nthat have not been approved for release through the MLAA\nprocess. See Law of the People’s Republic of China on\nInternational Criminal Judicial Assistance (promulgated by the\nStanding Committee of the Thirteenth National People’s\nCongress, Oct. 26, 201$, effective Oct. 26, 201$), ch. I, art. 4,\nJ.A. 355 (forbidding any “institution [or] organization\nwithin the territory of [China]” from “provid[ing] evidentiary\nmaterials and assistance prescribed by this Law to foreign\ncountries”); Id. ch. II, § 2, art. 13, J.A. 35$ (requiring a “foreign\ncountry” to request “criminal judicial assistance” “in\naccordance with the provisions of the criminal judicial\nassistance treaty”). The district court also recognized that the\nMinistry of Justice has advised that these violations would have\nconsequences in the form of administrative penalties against\nthe Banks and civil or criminal penalties against individual\nofficers. See March Op. 51—53. The district court, however,\nfound the hardships accompanying compliance both\nspeculative and minimal. See id. at 53—57. It based this\n\f 30\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nconclusion in part on historical Chinese enforcement trends:\npenalties are rare and relatively modest when they do occur.\nThe court found, for example, no instances of “a state-owned\nChinese enterprise . fac{ing] severe repercussions for\n . .\n\n\n\n\nresponding to the order of a foreign court,” Id. at 53, and that\nbank officers had apparently faced penalties only for stealing\nand selling “client information for profit,” id. at 55. The district\ncourt also relied on the logical supposition that the Chinese\ngovernment is unlikely to severely sanction banks that it\nlargely owns, id., and concluded that the Ministry’s warnings\nfailed to “suggest[] the administrative penalties will amount to\nmore than nominal fines” or provide any “concrete indication\ncriminal penalties are likely,” Id. at 56—57.\n\n Affording the most weight to the United States’\nunassailable interest in successfully investigating and, with any\nluck, frustrating North Korea’s arms programs, see id. at 58\n(calling “the interests of the relevant countries” the “most\nimportant factor”), the district court concluded that “[ojn\nbalance, international comity is not a reason to refrain from\ncompelling compliance with the subpoenas,” Id. On appeal, the\nBanks argue that the district court applied the incorrect legal\nstandard, that enforcing the subpoenas runs afoul of our case\nlaw, that “[c]omity dictates that the Chinese government at\nleast be given an opportunity to comply with an MLAA request\nbefore forcing a Chinese bank to violate Chinese law on\nChinese soil,” Bank Two Br. 19, and, finally, that the district\ncourt incorrectly balanced the relevant factors.\n\n B.\n Before considering these arguments, we must first\nascertain the appropriate standard of review. Although we\n“ordinarily review{] a district court’s grant of a motion to\n\f 31\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ncompel.. for abuse of discretion,” Recording Industry Ass ‘n\n .\n\n\n\n\nof America, Inc. v. Verizon Internet Services, Inc., 351 F.3d\n 1229, 1233 (D.C. Cir. 2003), we have yet to expressly\ndetermine the standard of review applied to a comity analysis\nin this context, cf Sealed Case I, $25 f.2d at 497—99 (not\nidentifying the standard of review); de Csepel v. Republic of\nHungary, 714 F.3d 591, 606 (D.C. Cir. 2013) (“This court has\nnever expressly addressed the standard by which we review a\ndistrict court’s decision to grant comity to a foreign\njudgment.”). The Supreme Court, however, has observed that\n“[t]he exact line between reasonableness and\nunreasonableness” when weighing comity principles “must be\ndrawn by the trial court, based on its knowledge of the case,”\nSociétë Nationale, 482 U.S. at 546, and we generally review\nfor abuse of discretion when the district court finds itself\nresponsible for making such a fact-bound reasonableness call,\nsee, e.g., Ideal Electronic Security Co. v. International Fidelity\nInsurance Co., 129 F.3d 143, 150 (D.C. Cir. 1997) (“[TJhe\nreasonableness of an attorney’s fees award is within the sound\n discretion of the trial court and is reviewed only for abuse of\n discretion.”); United States v. Law, $06 F.3d 1103, 1105 (D.C.\n Cir. 2015) (“We. consider the substantive reasonableness of\n . .\n\n\n\n\n the sentence imposed under an abuse-of-discretion standard.”\n (internal quotation marks omitted)). This approach comports\n with our sister circuits, which review similar questions for\n abuse of discretion. See, e.g., Allstate Life Insurance Co. v.\n Linter Group Ltd., 994 F.2d 996, 999 (2d Cir. 1993) (“[S]ince\n the extension or denial of comity is within the court’s\n discretion, we will reverse the court’s decision only when we\n find an abuse of discretion.”); Reinsurance Co. ofAmerica v.\nAdministratia Asigurarilor de Stat, 902 F.2d 1275, 1283 (7th\n Cir. 1990) (“The denial of [the company’s] motion to compel\n responses was . . not an abuse of discretion.”). The Banks\n .\n\n\n\n\n offer no argument for departing from that approach here. See\n\f 32\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nBank Two Br. 5 (suggesting “an abuse of discretion standard”);\nOral Arg. Rec. 13:33—13:40 (Bank One counsel agreeing that\nif “the district court did apply the correct legal standard, then\nit’s an abuse of discretion standard”). Accordingly, we shall\nreview the district court’s comity analysis for abuse of\ndiscretion.\n\n Employing that standard, we ask whether the district court\n“based its ruling on an error of law, a clearly erroneous\nassessment of the evidence, or an improper weighing of the\nfactors limiting its discretion.” fraenkel v. Islamic Republic of\nIran, $92 F.3d 348, 356 (D.C. Cir. 201$) (citation and internal\nquotation marks omitted). We of course review any embedded\n“legal conclusions de novo.” Guedes v. Bureau of Alcohol,\nTobacco, firearms & Explosives, 920 F.3d 1, 10 (D.C. Cir.\n2019).\n\n Bank One first argues that “the District Court used the\nwrong legal standard to review the Government’s request to\nenforce the Subpoenas”—that is, the court should not have\nadopted “the multifactorial approach of . Restatement\n . .\n\n\n\n(Third) of Foreign Relations.” Bank One Reply Br. 1. This\nargument gets the Banks nowhere. In Sociëtë Nationale, the\nSupreme Court explained that the Restatement’s “factors are\nrelevant to any comity analysis.” 482 U.S. 522, 544 n.2$; see\nOverby v. National Ass ‘n of Letter Carriers, 595 F.3d 1290,\n1295 (D.C. Cir. 2010) (“[C]arefully considered language of the\nSupreme Court, even if technically dictum, generally must be\ntreated as authoritative.” (internal quotation marks omitted)).\nThe Bank points out that our decision in Sealed Case I, issued\na few months after Société Nationale, never expressly\nreferences the Restatement’s factors. In that case we reversed\nan order holding a bank owned by (unidentified) Country X in\ncontempt for failure to comply with a subpoena that would\n\f 33\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nhave required the bank to produce documents in contravention\nof the banking privacy laws of Country Y, where the branch at\nissue was located. Sealed Case I, 825 F.2d at 497—98. But\nSealed Case I neither refutes Société Nationale’ s ratification of\nthe Restatement factors nor offers a different test to apply—\nnor, of course, could it. See We the People foundation, Inc. v.\nUnited States, 485 F.3d 140, 144 (D.C. Cir. 2007) (circuit\ncourts must follow Supreme Court precedent). As the district\ncourt recognized, Sealed Case I “comports with the Supreme\nCourt’s treatment of international comity as a principle that\ndepends on ‘prior scrutiny in each case of the particular facts,\nsovereign interests, and likelihood that resort to [alternative]\nprocedures will prove effective.” March Op. 34 (alteration in\noriginal) (quoting Société Nationale, 482 U.S. at 544).\n\n This brings us to the Banks’ related assertion that “[t]he\npresent case is controlled by” Sealed Case I. Bank One Br. 14.\nBut that case by its own terms constricts its holding to its facts:\n“If any of the facts we rest on here were different,” we\ncautioned, “our holding could well be different.” Sealed Case I,\n825 F.2d at 499. We therefore agree with the district court that\nthis case merits its own factual inquiry; Sealed Case I does not,\nas a legal matter, control the outcome here.\n\n The Banks’ more powerful argument is that, due to the\nfactual parallels between Sealed Case I and this case, “it was\nan abuse of discretion to enforce the Subpoena[sJ.” Bank Two\nBr. 25. The Sealed Case I court identified several facts central\nto its decision not to enforce the contempt order at issue, and\nthey might sound familiar, first, and “[m]ost important,” the\n“sanctions represent[ed] an attempt by an American court to\ncompel a foreign person to violate the laws of a different\nforeign sovereign on that sovereign’s own territory.” 825 F.2d\nat 498; see also Id. (observing that the bank had no way to\n\f 34\n PUBLIC COPY - SEALED INFORMATION DELETED\n\n“comply with the contempt order without violating the laws of\nCountry Y on Country Y’s soil”). We also recognized that the\nbank “had acted in good faith,” was “a third party . not...\n . .\n\n\n\naccused of any wrongdoing,” and constituted “not merely a\nprivate foreign entity,” but also “an entity owned by the\ngovernment of Country X.” Id. finally, because the bank’s\nmanager would “be available and able to testify as to many of\nthe facts that the grand jury may [have] wish[edj to ascertain,”\nthe government was not “left empty-handed” by the decision\nnot to enforce the subpoena. Id. at 499. But see Ed.\n(acknowledging “that the grand jury’s investigation may\nnonetheless be hampered, perhaps significantly” by the\ndecision not to enforce).\n\n As the Banks point out, each of those facts maps neatly\nonto this case, in which the government asks us to compel\nChinese banks owned in significant part by the Chinese\ngovernment to violate Chinese law on Chinese territory. The\nBanks have acted in good faith, they are currently third parties\nto the criminal investigation, and the MLAA, at least on its\nface, provides an alternative path that would allow the\ngovernment to obtain the records at issue.\n\n What’s more, to the extent there are differences, there is\nsome reason to believe that this case is an even better candidate\nfor comity than Sealed Case I. For one thing, where the\nsubpoena recipient in Sealed Case I was being asked to violate\nthe laws of a “different foreign sovereign,” Id. at 498, here the\ngovernment demands that the Banks violate their home\ncountry’s laws. Although the Country X-based bank in Sealed\nCase I might have closed its branch in Country Y to avoid\nsanctions, the Banks here have no such workaround. As the\ndistrict court recognized, “[a] Country X bank can shutter the\nCountry Y portion of its business; a Chinese bank cannot leave\n\f 35\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nChina.” March Op. 36 n.13; see also Id. (“[C]ompelling a bank\nto violate its own nation’s law is more prejudicial to\ninternational comity.” (internal quotation marks omitted)).\nMoreover, the Chinese government in this case “has warned\nthe Bank[s] that [they] will be sanctioned if [they] ignore[J” its\n“express instructions not to produce these documents outside\nthe MLAA process,” and “[t]here is no indication that the\nforeign government in Sealed Case I did anything similar.”\nBank Three Br. 39.\n\n The district court ignored none of these considerations.\nQuite to the contrary, it grappled with each and instead found\npersuasive two other “critical” differences between the cases.\nMarch Op. 34.\n\n The district court first pointed out that Sealed Case I\nconcerned a “domestic law enforcement matter,” which\n“hardly compares to the national security interests and\nassociated potential harm caused by non-compliance with a\nsubpoena related to an investigation into funding a state-\nsponsor of terrorism’s nuclear weapons program.” Id. at 35.\nThis is an important distinction. The government justifiably\nbelieves that the subpoenas will unearth information\nconcerning North Korea’s use of third-party front companies\nto evade economic sanctions, and nothing in Sealed Case I\nsuggests any “sovereign interest[],” Société Nationale, 482\nU.S. at 544, remotely comparable to impeding North Korea’s\nstockpiling of nuclear weapons.\n\n We acknowledge, as the Banks point out, that the\ngovernment here has failed to proceed as though North Korea\nwere about to push the nuclear launch button. Indeed, the\ngovernment took nearly a year to enforce the subpoenas in\ncourt and never sought to expedite this case until the district\n\f 36\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ncourt instructed it to do so. It did not, however, sit on its hands\nduring that time, as the Banks repeatedly assert, but “engaged\nin intra- and inter-agency deliberations” and “twice sent a\ndelegation to China” during that period. March op. 40. More\nto the point, we are reluctant to second-guess the Executive\nBranch’s assessments of national security interests and how to\nbest serve them. See frugone v. Central Intelligence Agency,\n 169 f.3d 772, 775 (D.C. Cir. 1999) (cautioning “that courts\nhave little expertise in either international diplomacy or\ncounterintelligence operations”). We of course agree with\nBank One that “the judiciary has an independent obligation to\nevaluate” the diplomatic ripple effects of its decisions. Bank\nOne Br. 25; see also Sealed Case I, 225 f.2d at 499 (“[WJe see\ngood reason for courts not to act on their own, even at the\nurging ofthe executive branch, when their actions ‘may hinder\nrather than further this country’s pursuit of goals both for itself\nand for the community of nations as a whole in the international\n sphere.” (emphasis added) (quoting Banco Nacional de Cuba\nv. Sabbatino, 376 U.S. 398, 423 (1964))). But we are confident\nthat the district court respected that obligation and that it acted\nwithin its discretion by recognizing that the heightened national\n interest here, as articulated by the Executive Branch,\n distinguishes this case from Sealed Case I.\n\n The Banks next observe that “if the Government receives\nthe documents via the MLAA, its national security interests\nwill be fully vindicated.” Bank Two Br. 26. The district court,\nhowever, concluded that the futility of the MLAA process sets\nthis case apart from Sealed Case I, in which testimony from the\nbank’s manager appeared to offer a sure-fire way of securing\nat least some of the desired information. “[un this\ninvestigation,” by contrast, the district court observed, “the\ngovernment insists that without compelling compliance, the\nrequested information is out of reach.” March Op. 35.\n\f 37\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nforcefully disagreeing with this conclusion, the Banks insist\nthat the MLAA process provides a channel specifically\ndesigned to allow the U.S. government to obtain exactly the\ntypes of records it seeks. Allowing it to eschew that process\nwithout so much as making an initial MLAA request, they\nargue, neglects comity’ s “spirit of cooperation.” $ociété\nNationale, 482 U.S. at 543 n.27. For its part, the government\nargues in support of the district court that “the historical record\ndemonstrates . an MLAA request of China for such bank\n . .\n\n\n\nrecords is not a viable alternative.” Appellee’s Br. 37 (internal\nquotation marks omitted).\n\n Recall that the MLAA establishes a procedure for the U.S.\nand Chinese governments to obtain documents from each other\nin criminal investigations. Specifically, the process provides\nthat “insofar as national law permits,” if, say, the United States\nmakes a request, China’s “Central Authority” must search for\nand then send the requested records. MLAA\narts. 2(1), 14(1), (3). Nothing in the MLAA, however,\ndesignates it as the exclusive means of obtaining evidence in a\ncriminal investigation.\n\n In this case, China’s Central Authority, the Ministry of\nJustice, Id. art. 2(2), expressed a “willingness to cooperate with\na DOJ MLAA request” as early as March 2018, Letter from\nHank B. Walther & Samidh Guha to Zia Faruqui & An\nRedboard 3 (Mar. 23, 2018), J.A. 830. Bank Three\ncommunicated China’s overture to the U.S. Department of\nJustice, observing that the MLAA process “would allow DOJ\nto obtain the documents it seeks, while avoiding the time, cost,\nand risk of litigating the Patriot Act subpoena.” Id. As we now\nknow, the United States declined that invitation—even as it\nfiled at least ten other MLAA requests in 2017 and 2012—and\ninstead sought to enforce the subpoenas in court.\n\f 38\n PUBLIC COPY - SEALED INFORMATION DELETED\n\n\n\n The Ministry then directly communicated with the district\ncourt multiple times, assuring that court that the Ministry\n“would timely review and handle the requests for assistance\nsought by the DOJ in accordance with the [MLAA] and\napplicable domestic laws.” Letter from Du Yaling, Director\nGeneral of International Cooperation Department of the\nMinistry of Justice to Chief Judge Beryl A. Howell 4 (Jan. 6,\n2019), J.A. 76. If the government submits a “request in line\nwith the [MLAAJ,” the Ministry pledged, “China will provide\nthe assistance to the United States accordingly.” Id. Confident\nthat the MLAA “mechanism works effectively,” the Ministry\nreminded the district court that China has “provided effective\nassistance to the US in many cases” and that “[tJhe avenue of\nthe {MLAA] is the only legal means to obtain criminal\nevidence in China by foreign judicial agencies under Chinese\nlaws.” Letter from Du Yaling, Director General of International\nCooperation Department of the Ministry of Justice to Chief\nJudge Beryl A. Howell 2—4 (Feb. 26, 2019), J.A. 965—67. The\ndistrict judge observed that this marked “the first time” she had\nseen “correspondence directly from a foreign government\nrelated to an ongoing matter.” March 5, 2019 Hearing Tr. 18:2—\n4, J.A. 1573. The government points to no other case where the\nMinistry has made such a representation and reneged on it and\nadmits that in some cases, “China has produced bank records\nto the general satisfaction of the U.S. prosecutors and agents.”\nDeclaration of Associate Director of the Office of International\nAffairs Jeffrey M. Olson (“OIA Declaration”) ¶ 14(e),\nJ.A. 951.\n\n The district court considered all these facts but ultimately\nconcluded that pursuing the MLAA process would result in an\ninadequate production of documents. According to the\nAssociate Director of the Office of International Affairs—\n\f 39\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nwhich serves as the United States’ Central Authority under the\nMLAA—the United States has submitted fifty MLAA requests\nfor bank records over the past decade and has received no\nresponsive records to thirty-five; the government found most\nof China’s responses in the remaining fifteen cases inadequate\nor untimely. And of the ten requests submitted within the last\ntwo years, the United States has received at most one response.\nCf OIA Declaration ¶ 13(b), J.A. 950 (describing that, for\nexample, China failed to produce documents in one of those\ncases even though the United States informed the Ministry that\nthe request was time-sensitive due to an upcoming trial).\n\n Moreover, the district court recognized that the\ngovernment is especially pessimistic about the chances of\nsuccess in this case given that the investigation involves a\nperfect storm of disagreements between the United States and\nChina: “Chinese authorities have a demonstrated failure in\nsupporting foreign and U.S. sanctions and U.S. law\nenforcement actions involving North Korea,” and “the Chinese\ngovernment has specifically objected to the imposition of\nsanctions on [the Company] and other Chinese front\ncompanies.” FBI Declaration ¶ 80, J.A. $88. Drawing the\ndistrict court’s attention to a case “nearly identical” to this one,\nId. at 37, J.A. 884, an FBI special agent submitted a declaration\npointing out that although in that case, China had “nominally\nagreed to work jointly with the U.S. authorities” on MLAA\nrequests for bank records involving North Korea, id. ¶ 72,\nJ.A. $85, nothing ever came of that initial agreement; offering\nvarious excuses, China never handed over any records. “Based\non these facts,” the agent “concluded that the MLAA process\nis not an effective way to obtain bank records from Chinese\nauthorities with respect to investigations involving North\nKorea.” Id. ¶ 75, J.A. 886; see Id. (“[OJne reason that Chinese\nauthorities do not want to assist with North Korean\n\f 40\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ninvestigations is that producing such records could reflect\nbadly on the Chinese government and the Chinese financial\nindustry.”).\n\n Given this record, we see no abuse of discretion in the\ndistrict court’s decision to enforce the subpoenas despite the\nfact that the United States chose not to pursue the MLAA\nprocess. The Supreme Court has made clear that evidence-\nsharing arrangements set up by international agreements,\nalthough often valuable, can also prove “unduly time\nconsuming and expensive, as well as less certain” than court-\nenforced subpoenas “to produce needed evidence.” $ociété\nNationale, 482 U.S. at 542. As such, whether comity “requires\nresort” to an agreement’s procedures calls for an inquiry into\nthe “likelihood that resort to those procedures will prove\neffective.” Id. at 544. The district court addressed just that\nquestion: it reviewed declarations from U.S. and Chinese\nofficials, identified facts that undermined the Chinese\ngovernment’s assurances about the MLAA system, and\nconcluded that the U.S. government had no obligation to\nsubmit a voluntary MLAA request because it would likely\nprove ineffective.\n\n We need not tarry over the Banks’ remaining arguments\nthat the district court abused its considerable discretion in\nbalancing the comity factors. Cf Morley v. Central Intelligence\nAgency, 894 F.3d 389, 391 (D.C. Cir. 2018) (recognizing that\nwhen “factors point in different directions,” “it will be the rare\ncase when we can reverse a district court’s balancing ofthe.\nfactors”). Specifically, Bank Three reiterates its claim that “the\nPatriot Act subpoena is facially overbroad,” Bank Three Reply\nBr. 20, and, as such, the “degree of specificity of the request,”\nSociétë Nationale, 482 U.S. at 544 n.28 (internal quotation\nmarks omitted), also weighs in favor of the Bank. But we agree\n\f 41\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nwith the district court that given that the Company “solely\nexisted as a front company,” “all [the Company’s] records\nrelate to the conduct under investigation,” March Op. 41\n(internal quotation marks omitted). With respect to the\nhardship factor, Bank One takes issue with the district court’s\nconclusion that the Banks would likely face minimal penalties\nfor complying with the subpoenas. In a civil law country such\nas China, the Bank contends, “[t]he District Court cannot know\nhow the Chinese government will exercise its prosecutorial\ndiscretion.” Bank One Br. 2$. But the district court never\npresumed to declare itself a soothsayer; it assessed a cluster of\ndatapoints presented and substantiated by the Executive\nBranch and ultimately determined that the risk of harsh\npunishment was slim. In any event, it recognized that “this\nfactor tips ever so slightly toward the banks.” March Op. 57.\nThe district court exercised its ‘judgment” in drawing these\nconclusions and was “guided by sound legal principles.”\n United States v. Taylor, 487 U.S. 326, 336 (198$) (internal\nquotation marks omitted). Our analysis ends there.\n\n We proceed with extreme caution when enforcing\nsubpoenas that would require recipients to violate a foreign\nsovereign’s domestic laws. As we said in Sealed Case I, “[w]e\nhave little doubt. that our government and our people would\n . .\n\n\n\nbe affronted if a foreign court tried to compel someone to\nviolate our laws within our borders,” 825 F.2d at 498—99, all\nthe more so if our government assured that court that it was\nready and willing to engage in a mutually agreed-upon bilateral\nprocess to provide the precise information requested. But we\ntake heart that at every stage of this process decisionmakers\nhave recognized that whether to enforce a subpoena “in cases\nlike this one raises grave difficulties for courts.” Id. at 498. The\ngovernment decided to pursue the records at issue through the\ncourts rather than the MLAA process only after twice sending\n\f 42\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ndelegates to China and engaging in extensive “internal\ndiscussions with subject matter experts within the FBI, as\n . . .\n\n\n\nwell as consultations involving senior FBI management and\nDepartment of Justice and Department of Treasury officials.”\nFBI Declaration ¶ 80, J.A. $88. That decision was then put to\n“the wise discretion of our judicial colleague[] on the District\nCourt,” fraenkel, 892 F.3d at 362, and that court consented to\nthe government’s course of action only after undertaking a\nthorough, considered analysis that frankly acknowledged the\nBanks’ comity concerns. Whether to enforce the subpoenas\nwas a hard call, the district court made that call, and we have\nno reason to reverse its fact-bound conclusion.\n\n V.\n Finally, Banks One and Three challenge the district court’s\nApril 2019 contempt order, arguing that it was unjustified\ngiven the government’s concession that the Banks acted in\ngood faith to obtain the Chinese government’s permission to\nproduce the records covered by the subpoenas. “[W]e\nreview. the contempt finding. for abuse of discretion.” In\n . . . .\n\n\n\nre Fannie Mae Securities Litigation, 552 F.3d 814, 818 (D.C.\nCir. 2009).\n\n A civil contempt citation is appropriate “only if the putative\ncontemnor has violated an order that is clear and unambiguous,\nand the violation must be proved by clear and convincing\nevidence.” Broderickv. Donaldson, 437 F.3d 1226, 1234 (D.C.\nCir. 2006) (internal quotation marks omitted). Banks One and\nThree nowhere dispute that by refusing to produce the\ndocuments covered by the subpoenas, they have violated a\nclear and unambiguous order. Instead, they rest their challenge\nto the contempt order on the Restatement (Third) of Foreign\nRelations Law, which suggests that a court “should not\n\f 43\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nordinarily impose sanctions of contempt, dismissal, or default\non a party that has failed to comply with [an] order for\nproduction” that would violate another state’s laws or\nregulations, “except in cases of deliberate concealment or\nremoval of information or of failure to make a good faith\neffort” “to secure permission from the foreign authorities to\nmake the information available.” Restatement (Third) of\nForeign Relations Law § 442(2)(a)—(b) (1927) (emphasis\nadded).\n\n The banks insist that the district court should not have held\nthem in contempt because they acted in good faith (albeit\nunsuccessfully) to secure the Chinese government’s\npermission to produce the records at issue. This is a rather thin\nreed to begin with, rendered more slender still by the fact that\nthe more recent Restatement (fourth) of foreign Relations Law\nsignificantly diminishes the importance of a putative\ncontemnor’s “good faith”: the Restatement (Fourth) simply\nstates that courts “have discretion to excuse violations of law,\nor moderate the sanctions imposed for such violations, on the\nground that the violations are compelled by another state’s\nlaw” if “the person in question appears likely to suffer severe\nsanctions for failing to comply with foreign law” and “has\nacted in good faith to avoid the conflict.” Restatement (Fourth)\nof foreign Relations Law § 442 (2018) (emphasis added)). In\nany event, this court has long held that “[t]he intent of the\nrecalcitrant party is irrelevant in a civil contempt proceeding\nbecause, unlike a criminal contempt proceeding, a civil\ncontempt action is a remedial sanction used to obtain\ncompliance with a court order or to compensate for damage\nsustained as a result of noncompliance.” Food Lion, Inc. v.\n United Food & Commercial Workers International Union,\nAFL-CIO-CLC, 103 F.3d 1007, 1016 (D.C. Cir. 1997) (internal\nquotation marks omitted).\n\f 44\n PUBLIC COPY - SEALED INFORMATION DELETED\n\n\n\n Confronted with clear non-compliance in the face of a\nperfectly pellucid order, the district court reasonably found that\n“the banks have not demonstrated good faith in a way relevant\nto contempt” and that, because “the requested records are\nessential to an investigation into a matter of national\nsecurity[,].. continued contumacy threatens tremendous\n .\n\n\n\n\nharm.” April op. 9 (internal quotation marks omitted). Under\nthese circumstances, the banks have given us no basis for\nconcluding that the district court abused its discretion.\n\n VI.\n For the foregoing reasons, we affirm the district court’s\ncontempt orders against all three Banks.\n\n So ordered\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Sealed Case (PUBLIC REDACTED OPINION)\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-06\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nPUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n njI $tat Iourt nf ppca1\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued July 12, 2019 Decided July 30, 2019\\n Reissued August 6, 2019\\n\\n No. 19-5068\\n\\n IN RE: SEALED CASE\\n\\n\\n\\n Consolidated with 19-5 100, 19-5 102, 19-5 103\\n\\n\\n Appeals from the United States District Court\\n for the District of Columbia\\n (No. 1:18-mc-00175)\\n (No. 1:18-mc-00176)\\n (No. 1:18-mc-00177)\\n\\n\\n Before: TATEL, MILLETT, and PILLARD, Circuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge TATEL.\\n\\n TATEL, Circuit Judge:* The three appellant banks,\\nheadquartered in China, hold records that the United States\\ngovernment thinks may clarify how North Korea finances its\\nnuclear weapons program. The government therefore procured\\nsubpoenas for those records, obtaining two from a grand jury\\n(for the two banks with U.S. branches) and one from the\\n *\\n NOTE: Portions of this opinion contain sealed\\ninformation, which has been redacted.\\n\\f 2\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nAttorney General under the Patriot Act (for the bank that has\\nno U.S. branch). The banks resisted on multiple fronts,\\nclaiming that the district court lacked personal jurisdiction, that\\nthe Patriot Act subpoena exceeded the government’s statutory\\nauthority, and that compelling production would run afoul of\\ncomity principles because turning over the records—even in\\nresponse to a U.S. court order—would trigger penalties under\\nChinese law. In a thorough and thoughtful pair of opinions, the\\ndistrict court overruled those objections and, when the first\\nopinion failed to elicit the records, held the banks in civil\\ncontempt. For the reasons set forth below, we affirm.\\n\\n I.\\n For decades, the Democratic People’s Republic of Korea\\n(better known as North Korea) has aspired to a battle-ready\\nnuclear weapons program. Seeking to defuse the threat, the\\nU.S. government has imposed stiff financial sanctions on the\\nNorth Korean regime and its instrumentalities, exercising\\nauthority located in the International Emergency Economic\\nPowers Act. See 50 U.S.C. §S 170 1—02 (granting the President\\nthe power to “prohibit” various “transactions” after declaring a\\n“national emergency” posed by “any unusual and extraordinary\\nthreat. to the national security, foreign policy, or economy\\n . .\\n\\n\\n\\n\\nof the United States”). Relevant to this case, those sanctions are\\nintended to cut off North Korea’s access to American markets\\nand currency. See Executive Order No. 13,382, § 1, 70 Fed.\\nReg. 38,567, 38,567 (June 28, 2005) (authorizing various\\ncabinet officers to “block[J” entities that further nuclear\\nproliferation by preventing any “property or interests in\\nproperty” from being “transferred, paid, exported, withdrawn,\\nor otherwise dealt in”). According to government investigators,\\nhowever, North Korea manages to evade the sanctions by using\\n\\f 3\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nChinese front companies that cloak the true ownership of the\\nfunds involved.\\n\\n The facts uncovered by this investigation so far are a case\\nin point. The U.S. government has collected substantial\\nevidence that a now-defunct Chinese company called\\n (“the Company”)\\nacted as a front for , a North Korea-\\nowned entity (“the NKE”), by facilitating transactions that\\nviolated the sanctions orders. In fact, to date the government’s\\n“investigation [has] not yield[ed] any evidence that” the\\nCompany “was ever used as a legitimate business.”\\nDeclaration of (“FBI\\nDeclaration”) ¶ 10, Joint Appendix (“J.A.”) 864.\\nUnsurprisingly, then, in addition to the NKE itself, the\\ngovernment has sanctioned the Company and two of its alleged\\nfounders. Actions Taken Pursuant to Executive Order 13382,\\n (designating the\\nNKE); Notice of [Office of Foreign Asset Control (OFAC)]\\nSanctions Actions,\\n (designating the Company); Notice of OFAC Sanctions\\nActions, •\\n(designating alleged co-founder\\n —\\n ); Notice of OFAC\\nSanctions Actions,\\n(designating alleged co-founder ).\\n According to the government, the scheme operated like\\nthis: Hobbled by the worthlessness of North Korea’s currency,\\nthe NKE would use the Company to make or receive payments\\nin U.S. dollars. These transactions helped North Korea access\\nresources that would otherwise have been beyond its reach. For\\nexample, the Company’s assistance allegedly enabled North\\nKorea to export hundreds of millions of dollars of coal and\\nother minerals, generating revenue in U.S. currency that North\\n\\f 4\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nKorea could then use to requisition other commodities vital to\\nits weapons program.\\n\\n In these transactions, the Company routinely took\\nadvantage of U.S. correspondent bank accounts—that is,\\naccounts held by banks outside the United States at banks\\nlocated inside the United States. See 31 U.S.C.\\n§ 53 18A(e)(1)(B) (defining “correspondent account” to mean\\n“an account established to receive deposits from, make\\npayments on behalf of a foreign financial institution, or handle\\nother financial transactions related to such institution”). Such\\naccounts allow foreign “banks to conduct business and provide\\nservices for their customers in” the United States even if “the\\nbanks have no physical presence” here. Minority Staff of the\\nSenate Permanent Subcommittee on Investigations,\\nCorrespondent Banking: A Gateway for Money Laundering,\\nS. Hrg. 107-84, 107th Cong., app. 287 (2001). This, in turn,\\nenables a foreign bank’s customers to “receive many or all of\\nthe services offered by [a] U.S. bank” without forming a direct-\\nclient relationship with an American institution. Id.\\nCorrespondent banking arrangements are fairly ubiquitous:\\n“[d]ue to U.S. prominence in international trade and the high\\ndemand for U.S. dollars due to their overall stability, most\\nforeign banks that wish to provide international services to\\ntheir customers have accounts in the United States capable of\\ntransacting business in U.S. dollars.” Id. In short, then,\\ncorrespondent bank accounts allow foreign banks to offer their\\nclients services—including obtaining and transacting in U.S.\\ncurrency—without opening a U.S. branch.\\n\\n These accounts are how the appellant banks (collectively,\\nthe “Banks”) enter the story. All three are China-based—\\nindeed, the Chinese government owns a substantial minority of\\neach—and hold correspondent accounts in the United States.\\n\\f 5\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nBanks One and Two also operate branches in the United States,\\nbut according to the government the Company relied on the\\nthree Banks’ U.S. correspondent accounts to execute its\\nscheme.\\n\\n To better grasp the full scope of the Company’s\\noperations, the government seeks certain records from the\\nBanks. The United States and China have a bilateral agreement\\nfor obtaining just these kinds of records. Known as the Mutual\\nLegal Assistance Agreement (MLAA), it obligates the two\\ncountries to assist each other “in proceedings related to\\ncriminal matters,” including by “providing . . records or\\n .\\n\\n\\n\\narticles of evidence” and “executing requests for inquiry,” and\\nspecifies the “[f]orm and [c]ontents” of any requests.\\nAgreement Between the Government of the United States of\\nAmerica and the Government of the People’s Republic of\\nChina on Mutual Legal Assistance in Criminal Matters, arts. 1,\\n4, June 19, 2000, T.I.A.S. No. 13,102. Absent an MLAA\\nrequest, Chinese law prohibits banks from disclosing records\\nto international investigators.\\n\\n In this case, however, the United States chose not to go\\nthrough the MLAA. Stymied by what it perceives as sluggish\\nand typically fruitless cooperation from Chinese authorities in\\nthe past, the government unilaterally issued subpoenas to all\\nthree Banks in December 2017. The two banks with branches\\nin the United States\\n (“Bank One” and “Bank Two,”\\nrespectively)—received grand jury subpoenas. See Federal\\nRule of Criminal Procedure 1 7(e)(1). The third\\n (“Bank Three”)—has no branch in\\nthe United States and received a subpoena from the Attorney\\nGeneral pursuant to the Patriot Act, which gives the Attorney\\nGeneral and the Treasury Secretary special investigatory tools\\n\\f 6\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nwhen it comes to U.S. correspondent bank accounts. See 31\\nU.S.C. § 53 18(k)(3)(A)(i). The subpoenas seek all records\\ndated 2012 to 2017 concerning all correspondent banking\\ntransactions associated with certain accounts linked to the\\nCompany, defining records to include signature cards, ledger\\ncards, account statements, and documentation of deposits,\\nwithdrawals, and transfers. Through this investigation, the\\ngovernment hopes to learn whether the Company or any other\\nentities have committed various federal crimes, including\\nengaging in unlicensed financial transactions (violating 50\\nU.S.C. § 1705), money laundering (violating 18 U.S.C.\\n § 1956), or failing to exercise due diligence in correspondent\\nbanking as required by the Bank Secrecy Act (violating 31\\nU.S.C. § 531$ and 5322). The government does not currently\\nsuspect the subpoenaed Banks of any wrongdoing.\\n\\n In correspondence with the U.S. government, the Banks\\nand the Chinese government protested that complying with the\\nsubpoenas would violate multiple Chinese laws, including the\\nlaw designating the MLAA process as the exclusive\\nmechanism for disclosures. They claimed that failure to use the\\nMLAA channel would expose the Banks to administrative and\\ncriminal penalties. In response, the U.S. government took the\\nposition that such a request would be futile, given China’s slow\\nand spotty history of providing records requested through that\\nprocess.\\n\\n There matters stood for about a year. Despite traveling\\ntwice to China for what proved to be unproductive negotiations\\nwith the Chinese government, there were no major\\nbreakthroughs: the United States refused to submit an MLAA\\nrequest, and the Banks continued to withhold the records. At\\nlast, in November 2018, things kicked back into gear when the\\ngovernment moved the district court to compel enforcement.\\n\\f 7\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nStill resisting, the Banks disputed the district court’s personal\\njurisdiction and contended that enforcing the subpoenas would\\nviolate deeply ingrained principles of international comity.\\nBank Three also protested that the scope of its subpoena\\nexceeded the Attorney General’s Patriot Act authority.\\n\\n The district court granted the government’s motion to\\ncompel. In re Grand Jury Investigation ofPossible Violations\\nof 18 USC. §‘ 1956 and 50 US.C. 1705, Nos. 1:18-mc-175,\\n-176, -177, slip op. at 3 (D.D.C. Mar. 18, 2019) (“March op.”).\\nThe court found that Banks One and Two had consented to\\npersonal jurisdiction when they signed an agreement with the\\nfederal Reserve permitting them to open branches in the\\nUnited States. Id. at 11—13. As for Bank Three, which has no\\nbranch in the United States, the court determined that (along\\nwith Banks One and Two) it had sufficient contact with the\\nUnited States as a whole and the subpoenas sufficiently related\\nto that contact so as to support the court’s personal jurisdiction.\\nId. at 13—27; see also International Shoe Co. v. State of\\nWashington, Office of Unemployment Compensation and\\nPlacement, 326 U.S. 310, 316 (1945) (personal jurisdiction\\nrequires “minimum contacts” with the forum such that\\njurisdiction is consistent with “traditional notions of fair play\\nand substantial justice” (internal quotation marks omitted)).\\nThe district court also perceived no problems with the scope of\\nthe Patriot Act subpoena. March Op. 27—31. Finally, although\\nthe government conceded that compliance with the subpoenas\\nwould entail violating at least some Chinese laws, the district\\ncourt concluded that comity principles did not require the\\nsubpoenas to be quashed. Id. at 32—5 8. The court found that,\\nbased on past practice, China’s preferred MLAA procedure\\nwould likely be futile and determined that the U.S.\\ngovernment’s strong national security interests outweighed\\nChina’s stake in enforcing the identified laws. Id. at 42—51. The\\n\\f 8\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ncourt was unmoved by a series of letters from the Chinese\\nMinistry of Justice promising to “promptly review and\\nprocess” any MLAA request in this case. Id. at 45 (quoting\\nLetter from Du Yaling, Director General of International\\nCooperation Department of the Ministry of Justice, to Chief\\nJudge Beryl A. Howell (Feb. 26, 2019), J.A. 966).\\n\\n About a month later, after the Banks made clear they\\nwould not comply with the compulsion order, the district court\\nheld them in contempt. In re Grand Jury Investigation of\\nPossible Violations of 18 U.S.C. § 1956 and 50 US.C. § 1705,\\nNos. 1:18-mc-175, -176, -177, slip op. at 1—3 (D.D.C. Apr. 10,\\n2018) (“April Op.”). It imposed a $50,000 per day fine against\\neach bank but stayed that penalty contingent on the parties\\nseeking an expedited appeal. Id. at 14—15. They obliged, and\\nwe agreed to accelerate our consideration of the case. Per\\nCuriam Order, In re: Sealed Case, No. 19-5068 (D.C. Cir. Apr.\\n25, 2019). We now turn to the Banks’ arguments, beginning\\nwith personal jurisdiction before examining the merits.\\n\\n II.\\n The Constitution forbids federal courts from exercising\\npersonal jurisdiction absent sufficient “minimum contacts” to\\nsatisfy “traditional notions of fair play and substantial\\njustice.” International Shoe, 326 U.S. at 316 (quoting Milliken\\nv. Meyer, 311 U.S. 457, 463 (1940)). That protection extends\\nto private foreign entities—such as the Banks—just as to\\nAmerican citizens. See Livnat v. Palestinian Authority, 851\\nF.3d 45,48 (D.C. Cir. 2017) (delineating the “general rule” that\\nrestrictions on personal jurisdiction “protect{J all litigants in\\nour courts”). Unlike subject-matter jurisdiction, however,\\npersonal jurisdiction “can. be waived,” meaning a party\\n . .\\n\\n\\n\\nmay “consent” to a court’s personal jurisdiction. Insurance\\n\\f 9\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nCorp. of Ireland v. Compagnie des Bauxites de Guinee, 456\\nU.s. 694, 703 (1982). Because neither the government nor the\\nBanks suggests that these principles operate differently in\\ngrand jury investigations than in ordinary civil litigation, we\\nshall treat these two contexts as identical for purposes of this\\ncase. But see Boydv. Meachum, 77 f.3d 60,66 (2d Cir. 1996)\\n(“While the federal constitutional requirements of personal\\njurisdiction in a civil setting are reasonably well-defined,\\nthis is not so in a criminal case.”).\\n\\n All three Banks contest personal jurisdiction. “Reviewing\\nthe district court’s assertion of personal jurisdiction de novo,”\\nUnited States ex rel. Miller v. Bill Harbert International\\nConstruction, Inc., 608 F.3d 871, 886 (D.C. Cir. 2010), we\\nconclude that the Banks’ challenges are meritless. Banks One\\nand Two consented to jurisdiction when they opened branches\\nin the United States. And although Bank Three has no U.S.\\nbranch and executed no such agreement, its choice to maintain\\ncorrespondent accounts in the United States establishes an\\nadequate connection to the forum and the enforcement action\\nto sustain jurisdiction.\\n\\n A.\\n Congress requires any “foreign bank” to acquire “prior\\napproval of the” Federal Reserve before it “may establish a\\nbranch or an agency” in the United States. 12 U.S.C.\\n§ 3 105(d)(1). The Federal Reserve, in turn, has authority to\\n“impose such conditions on its approval” of a foreign bank’s\\napplication “as it deems necessary,” Id. § 3 105(d)(5), to,\\namong other things, assure itself that the bank will be “in\\ncompliance with applicable United States law” and mitigate\\nany “risk to the stability of the United States financial system,”\\nid. § 350 1(d)(3)(D)—(E). See also 12 C.F.R. § 21 1.24(c)(2)(vii)\\n\\f 10\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n(Federal Reserve scrutinizes “[wjhether the foreign bank...\\nhas established adequate controls and procedures. to ensure . .\\n\\n\\n\\ncontinuing compliance with U.S. law, including controls\\ndirected to detection of money laundering”).\\n\\n When Banks One and Two applied to open their U.S.\\nbranches, both executed agreements with the Federal Reserve\\nto assure compliance with relevant provisions of American\\nlaw. As part of that deal, they “consent[edj to the jurisdiction\\nof the federal courts of the United States for purposes of\\n . . .\\n\\n\\n\\nany and all. proceedings initiated by. the United\\n . . . .\\n\\n\\n\\nStates. in any matter arising under U.S. Banking Law.”\\n . .\\n\\n\\n\\nLetter from Vice Chairman & President,\\n ,\\n\\n\\n to Board of Governors of the\\nFederal Reserve System 1 l.A. 1659 (“Bank\\n ,\\n\\n\\nOne Consent”); Assurances Commitment and Consent to\\nJurisdiction Commitment\\n 1, J.A. 1663 (“Bank Two Consent”). The agreements\\ndefine “U.S. Banking Law” expansively to include “all federal\\ncriminal laws of which violation(s) arise(s): under .the . . . . .\\n\\n\\n\\nBank Secrecy Act.” Bank One Consent 3 n.2, l.A. 1661; Bank\\nTwo Consent 1 n.2, l.A. 1663.\\n\\n The plain language of those agreements covers this case.\\nThis investigation is unquestionably a “proceeding[] initiated\\nby. the United States” and, as the government explains, “the\\n . .\\n\\n\\n\\ngrand jury is investigating whether those financial institutions\\nthat provided services in the United States”—for example, the\\nAmerican institutions hosting the foreign Banks’\\ncorrespondent accounts—”exercised due diligence.”\\nAppellee’s Br. 18 (citing 31 U.S.C. § 5318, 5318A, 5322).\\nThe Bank Secrecy Act makes willful dereliction of that\\nobligation a federal criminal violation. 31 U.S.C. § 5322. In\\nshort, then, the government has assembled the necessary\\n\\f 11\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ningredients for jurisdiction: a criminal investigation initiated by\\nthe United States (the grand jury investigation) arising under\\nU.S. Banking Law (the Bank Secrecy Act).\\n\\n The banks’ contrary arguments amount to nothing more\\nthan an unsuccessful effort to move the goalposts. They insist\\nthat there must be a reasonable showing that someone actually\\nviolated the Bank Secrecy Act. Bank Two Br. 8. Not so. All\\nour precedent requires is a “reasonable probability that. the. .\\n\\n\\n\\nfacts necessary for the exercise of jurisdiction” exist. In re\\nSealed Case, $32 F.2d 1268, 1274 (D.C. Cir. 1987) (quoting In\\nre Marc Rich & Co., 707 F.2d 663, 670 (2d Cir. 1983)). Here,\\nnothing in the consent agreements requires the government to\\nestablish liability; they require only that the proceedings\\n“aris[e] under U.S. Banking Law.” Bank One Consent 1,\\nJ.A. 1659 (emphasis added); Bank Two Consent 1, J.A. 1663\\n(emphasis added). Whether a proceeding “arises under” a\\nparticular law is a question about its subject matter at the outset,\\nnot a question about its outcome. To look at the same phrase in\\na more typical context, when we ask whether a lawsuit “aris[es]\\nunder” federal law for purposes of subject-matter jurisdiction,\\nU.S. Const. art. III, § 2, we look to whether the plaintiffs\\n“assert a claim” based on “federal law,” not whether they will\\nultimately prevail, National farmers Union Insurance Cos. v.\\nCrow Tribe of Indians, 471 U.S. 845, 850 (1985) (emphasis\\nadded); see also American Well Works Co. v. Layne & Bowler\\nCo., 241 U.S. 257, 260 (1916) (“A suit arises under the law that\\ncreates the cause of action.”). Given this “legal lineage,” Hall\\nv. Hall, 13$ S. Ct. 1118, 1125 (201$),agrand jury investigation\\nplainly “arises under” a particular statute when it probes\\nwhether that statute has been violated.\\n\\n Read properly, the consent agreements demand only that\\n“U.S. Banking Law” form the basis of the investigation. The\\n\\f 12\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ngovernment has cleared that threshold here. Even if, as the\\nbanks suggest, this investigation’s potential to yield successful\\nprosecutions is uncertain, the banks have offered no basis for\\ndoubting that “U.S. Banking Law” constitutes this\\ninvestigation’s subject matter.\\n\\n The banks also suggest that there is a problem with\\npersonal jurisdiction because the subpoenaed records are\\nlocated outside the United States. But that fact has no bearing\\non personal jurisdiction given that the banks’ consent\\nagreements supply the basis for jurisdiction. The argument is\\nbetter conceived as a challenge to the scope of the subpoenas—\\nan understanding not evident from the banks’ briefs. Even were\\nwe to entertain the argument, however, it would go nowhere.\\nFederal courts have long required the production of documents\\nheld abroad. See, e.g., In re Marc Rich, 707 F.2d at 667 (A\\nwitness may not “resist the production of documents on the\\nground that the documents are located abroad.”); United States\\nv. first National City Bank, 396 F.2d 897, 900—01 (2d Cir.\\n196$) (“[A] federal court has the power to require the\\nproduction of documents located in foreign countries if the\\ncourt has in personam jurisdiction of the person in possession\\nor control of the material.”). That practice makes sense: grand\\njuries often investigate violations of criminal laws that\\nobviously cover extraterritorial conduct. “It would be strange,\\nindeed, if the United States could punish a foreign corporation\\nfor violating its criminal laws . . but a federal grand jury\\n . ,\\n\\n\\ncould not investigate to ascertain the probability that a crime\\nhad taken place.” In re Marc Rich, 707 F.2d at 666.\\n\\n B.\\n Bank Three, by contrast, has neither opened a U.S. branch\\nnor consented to personal jurisdiction. We therefore must\\n\\f 13\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ndetermine whether it “has purposefully directed” its relevant\\n“activities” at “the forum.” Burger King Corp v. Rudzewicz,\\n471 U.S. 462, 472 (1985) (internal quotation marks omitted).\\n\\n That, of course, requires us to correctly identify “the\\nforum,” a task ordinarily linked to where Congress has\\nauthorized service of process. See BNSF Railway Co. v.\\nTyrrell, 137 S. Ct. 1549, 1555—56 (2017) (“Congress’ typical\\nmode of providing for the exercise of personal jurisdiction has\\nbeen to authorize service of process” because “a basis for\\nservice of a summons on the defendant is prerequisite to the\\nexercise of personal jurisdiction”). In many situations,\\nCongress limits the territorial reach of federal district courts to\\nthat of the state in which they sit. See Wright & Miller,\\n4 Federal Practice & Procedure § 1068.1 (4th ed.); see, e.g.,\\nFederal Rule of Civil Procedure 4(k)(1)(A). For example,\\nunless a federal statute otherwise provides, by virtue of Rule\\n4(k)(1)(A) the jurisdictional reach of a federal court in a civil\\naction is keyed to that of a court of general jurisdiction in the\\nstate in which it sits. So in reviewing the constitutionality of\\ntheir assertions of jurisdiction, federal courts oflen perform\\nstate-specific contacts analyses. Although the constitutional\\nrestraint in those situations necessarily flows from the Fifth\\nAmendment, the scope of the minimum-contacts inquiry is the\\nsame as it would be under the Fourteenth Amendment for a\\nclaim in state court: confined to contacts with the forum state.\\n\\n According to Bank Three, a state-specific contacts\\nanalysis is appropriate for the Patriot Act subpoenas at issue\\nhere. And, the Bank argues, since its agent was served in New\\nYork and the Bank lacks any meaningful contact with the\\nDistrict of Columbia, the district court here lacked jurisdiction.\\nFor its part, the government maintains that the Patriot Act\\nauthorizes nationwide service of process. Under those\\n\\f 14\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ncircumstances, the Fifth Amendment requires only “minimum\\ncontacts with the United States” as a whole—rather than with\\nthe forum state. SEC v. Bilzerian, 378 F.3d 1100, 1106 n.8\\n(D.C. Cir. 2004); accord Livnat, 851 F.3d at 55 (“Under the\\nFifth Amendment, which defines the reach of federal courts,\\ncontacts with the United States as a whole are relevant.”).\\nTherefore, the government argues, the district court properly\\nconsidered Bank Three’s contacts with the United States as a\\nwhole, including its maintenance and use of correspondent\\naccounts in New York.\\n\\n We begin by observing that the usual process of looking to\\nwhere federal law authorizes service of process as a proxy for\\nwhether Congress sought to trigger a federal court’s nationwide\\nterritorial reach is, perhaps, not a perfect fit for this provision\\nof the Patriot Act. Because the statute confers authority to issue\\na subpoena directly on the Attorney General and the Treasury\\nSecretary, rather than on a federal district court, Congress may\\nhave given little consideration to the territorial reach of a\\ndistrict court considering a motion to compel enforcement of\\nsuch a subpoena. But given that Bank Three assumes that\\nauthorizing nationwide service in this context would trigger a\\nnationwide contacts analysis, we shall too. See Sickle v. Torres\\nAdvanced Enterprise Solutions, LLC, 884 F.3d 338, 344 (D.C.\\nCir. 2018) (accepting a party’s decision “not to brief or argue\\nthe question of personal jurisdiction” because that “defense .\\n\\n\\n\\n\\ncan be waived or forfeited”).\\n\\n Here, that service-of-process provision, codified in 31\\nU.S.C. § 5318(k), authorizes the kind of subpoena served on\\nBank Three to “be served on the foreign bank in the United\\nStates if the foreign bank has a representative in the United\\nStates.” 31 U.S.C. § 5318(k)(3)(A)(ii). We know of no other\\nservice-of-process provision in the U.S. Code phrased exactly\\n\\f 15\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nthis way. Still, we are not entirely without guideposts as we\\nevaluate this question.\\n\\n We know, for instance, that Congress has many different\\nways of authorizing nationwide service of process. Sometimes\\nit makes its intention abundantly clear by authorizing service\\n“anywhere in the United States.” See, e.g., 7 U.S.C. § 499g(b)\\n(service for enforcing in federal district courts certain orders of\\nthe Secretary of Agriculture); 2$ U.S.C. § 232 1(c) (same for\\nenforcing orders of the Surface Transportation Board); 33\\nU.S.C. § 921(c) (service for review proceedings in the courts\\nof appeals of decisions of the Benefits Review Board for\\nlongshore and harbor workers). On other occasions, Congress\\nemploys some riff on the phrase “wherever the defendant may\\nbe found,” language generally understood to refer to\\nnationwide service. See, e.g., 7 U.S.C. § 25(c) (service in\\nprivate lawsuits under the Commodity Exchange Act); see also\\nOmni Capital International, Ltd. v. Rudolf Wolff & Co., 484\\nU.S. 97, 105—06 (1987) (recognizing that such language\\n“explicitly authorize[sJ nationwide service of process”);\\nAppellee’s Br. 25 n.20 (collecting additional examples). We\\nalso know that Congress can very clearly limit service to a\\nparticular judicial district when it wants to. See, e.g., 45 U.S.C.\\n § 3 62(b) (authorizing process issued by most courts in Railroad\\nRetirement Board proceedings to be served “in the judicial\\ndistrict of the district court issuing such. process”).\\n . .\\n\\n\\n\\n\\n At first glance, the language in section 5318(k) looks more\\nlike the provisions that authorize nationwide service. So long\\nas a foreign bank has designated a representative, the statute\\nallows for service “in the United States,” without specifying\\nany state or judicial district. True, the provision lacks the\\nunmistakable clarity of the statutes that expressly authorize\\nservice “anywhere” in the country; nonetheless, the text\\n\\f 16\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ncontains no limitation other than the territorial boundaries of\\nthe entire United States. Bank Three, however, seizes on the\\nrequirement to designate a representative to argue that, in fact,\\nCongress only meant to authorize service in the judicial district\\nwhere that representative is located.\\n\\n Bank Three’s position makes some intuitive sense. After\\nall, as a matter of physics, the designated representative will\\nhave a single location. On further reflection, however, this\\nlogic fails to resolve things in the Bank’s favor. After all, both\\nBank Three and the government agree that language\\nauthorizing service “wherever a defendant may be found”\\nauthorizes nationwide service, see Bank Three Br. 51,\\nAppellee’s Br. 25 n.20, even though in many instances that\\ndefendant, too, will only be located in one place at a time.\\n\\n Even more instructive, in other service-of-process regimes\\ninvolving designated representatives, when Congress wants to\\nspecify a particular location or district for service, it says so\\ndirectly. For example, in certain trademark actions, Congress\\nhas instructed foreign mark holders to designate a U.S. agent\\nand authorized service of process “by leaving with that person\\nor mailing to that person a copy [of the process] at the address\\nspecfied in the last designation so filed.” 15 U.S.C. § 1141k(f)\\n(emphasis added). Congress has even provided a backup\\noption, explaining that “[i]f the person so designated cannot be\\nfound at the last designated address” then “process may be\\nserved on the Director” of the United States Patent and\\nTrademark Office. Id.; see also 7 U.S.C. § 2569 (containing a\\nsimilar backup provision for when a designee “cannot be found\\nat the address given”). The provision at issue in this case is\\nnoticeably devoid of any similar language, leading to the\\ninference that Congress sought to impose no similar\\ngeographical limitation here.\\n\\f 17\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n\\n Reading section 5318(k) to authorize nationwide service\\nalso aligns with the Patriot Act’s structure and purpose. Cf\\nUnited States International Trade Commission v. ASAT, Inc.,\\n411 f.3d 245, 251 (D.C. Cir. 2005) (examining, when\\nconsidering whether a statute authorizes nationwide service,\\nwhether “Congress would have desired nationwide service of\\nprocess to effectuate the underlying statute’s purpose”). Rather\\nthan requiring the Attorney General or Treasury Secretary to\\ngo to court to obtain a subpoena, Congress authorized those\\nofficers to issue process themselves. Unlike federal district\\ncourts, which customarily serve a particular geographical area,\\nCabinet officers serve the entire country. Structurally speaking,\\nthen, Congress’s choice to give two such officers the subpoena\\npower evokes an intent to rely on their nationwide reach. As\\nfor purpose, having found that correspondent accounts were a\\nsource of special concern for money laundering, Congress\\nsought “to provide a clear national mandate for subjecting to\\nspecial scrutiny those. . financial institutions operating\\n .\\n\\n\\n\\noutside of the United States. that pose particular, identifiable\\n . .\\n\\n\\n\\nopportunities for criminal abuse.” USA PATRIOT Act of 2001,\\nPub. L. No. 107-56, § 302(a)(6), (b)(4), 115 Stat. 272, 296—97.\\nIt would be odd if, in establishing this “clear national mandate,”\\nCongress had geographically subdivided the government’s\\ninvestigative powers.\\n\\n Apparently recognizing that its preferred interpretation of\\nsection 5318(k) may not represent “the most natural reading of\\nthe provision,” Bank Three Br. 4$, Bank Three falls back on\\nthe presumption against extraterritoriality, which teaches that\\nwithout a “clearly expressed congressional intent to the\\ncontrary, federal laws will be construed to have only domestic\\napplication.” RJR Nabisco, Inc. v. European Community, 136\\nS. Ct. 2090, 2100 (2016). According to the Bank, reading the\\n\\f 18\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nPatriot Act to authorize nationwide service “would vastly\\nexpand the extraterritorial scope of federal regulatory power.”\\nBank Three Br. 49. But that does not follow. This service-of-\\nprocess provision has no impact on the substantive\\nextraterritorial reach of federal law; it merely helps sort out\\nwhich federal court has authority to enforce the subpoena. After\\nall, even under Bank Three’s preferred interpretation, at least\\none federal court could assert jurisdiction over Bank Three (the\\none in the district where the Bank’s agent resides). The\\npresumption against extraterritoriality therefore offers no basis\\nto upset the most straightforward reading of the text: that once\\na foreign bank has designated a representative in the United\\nStates, service may be accomplished throughout “the United\\nStates.”\\n\\n Having defined the forum as the entire United States, we\\nwould usually go on to assess whether the Bank’s contacts with\\nthat forum are sufficient. In this case, however, Bank Three\\nplaced all its eggs in the forum-identification basket; it claims\\nonly that it lacks the necessary contacts with the District of\\nColumbia, offering no argument that its contacts with the\\nUnited States as a whole are insufficient. See Bank Three\\nBr. 46—47 (“The only way the government can prevail is by\\nshowing that the relevant forum is the United States as a\\nwhole.”). We therefore have no occasion to question the district\\ncourt’s conclusion that Bank Three’s maintenance of\\ncorrespondent accounts in the United States supplies the\\nnecessary nexus. March op. 2 1—24. With no further\\njurisdictional hurdles in our path, we proceed to the merits.\\n\\n III.\\n Bank Three, the only bank to receive a Patriot Act\\nsubpoena, argues that its subpoena exceeds the Attorney\\n\\f 19\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nGeneral’s statutory authority. Although “[w]e generally review\\ndistrict court decisions enforcing document subpoenas only for\\narbitrariness or abuse of discretion,” we afford no deference to\\na decision that applies “the wrong legal standard.” In re Sealed\\nCase, 146 F.3d 881, $83 (D.C. Cir. 199$). We therefore review\\n“de novo” the district court’s “legal conclusions,” including its\\nstatutory interpretation, even as we review “its factual findings\\nfor clear error.” United States v. Ahn, 231 F.3d 26, 38 (D.C.\\nCir. 2000).\\n\\n Congress enacted section 531 8(k)(3) “to increase the\\nstrength of United States measures to prevent, detect, and\\nprosecute international money laundering and the financing of\\nterrorism.” Pub. L. No. 107-56, § 302(b)(1), 115 Stat. at 297.\\nIn particular, Congress found that correspondent accounts held\\nby foreign banks were “one of the banking mechanisms\\nsusceptible in some circumstances to manipulation by foreign\\nbanks to permit the laundering of funds.” § 302(a)(6), 115 Stat.\\nat 296. Thus, it authorized both the Attorney General and the\\nTreasury Secretary to issue a “subpoena to any foreign bank\\nthat maintains a correspondent account in the United States and\\nrequest records related to such correspondent account,\\nincluding records maintained outside of the United States\\nrelating to the deposit of funds into the foreign bank.” 31\\nU.S.C. § 5318(k)(3)(A)(i).\\n\\n Bank Three claims the subpoena it received strays outside\\nthe statutory lines because it is “not expressly confine[d]” to\\ndocuments related to the Bank’s correspondent accounts in the\\nUnited States. Bank Three Br. 23. To be sure, the subpoena is\\ncomprehensive: it seeks “[a]1l documents relating to [the\\nCompany’s] correspondent banking transactions”—including,\\nit would seem, documents related to transactions using\\ncorrespondent accounts in countries other than the United\\n\\f 20\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nStates. Bank Three Subpoena, J.A. 771; see also id. (requesting\\n“signature cards,” “account ledger cards,” “periodic account\\nstatements,” and “records. of all items deposited,\\n . .\\n\\n\\n\\n\\nwithdrawn, or transferred”). And we know Bank Three does,\\nin fact, have correspondent accounts outside the United States.\\nIf we knew nothing else about the Company, Bank Three’s\\nargument might therefore have some force: where, as here, a\\nbank maintains correspondent accounts in multiple countries,\\nit is by no means a given that all of an account holder’s\\ncorrespondent banking transactions “relate to” the bank’s US.\\ncorrespondent accounts, as the statute requires.\\n\\n With respect to this particular case, however, the\\ngovernment has a compelling response. It argues, as it did\\nbefore the district court, that evidence already collected during\\nthe investigation shows that the Company’s “exclusive raison\\nd’etre” was to serve as “a progenitor of U.S. dollars for” the\\nNKE. Appellee’s Br. 62; see also Id. at 64 (describing the\\nCompany as “a front company created for the sole purpose of\\nexecuting U.S.-dollar transactions”); March 5, 2019 Hearing\\nTr. 95:17—25, l.A. 1650—5 1 (“[The Company] existed only for\\none reason. [T]he [NKE] has to set up entities like this,\\n . .\\n\\n\\n\\n\\nbecause [it] want[s] access to the U.S. financial system.”). In\\nthe government’s view, such a showing means that all of the\\nCompany’s records are “related to” the Bank’s U.S.\\ncorrespondent accounts because those accounts were part of the\\nCompany’s means of obtaining U.S. dollars.\\n\\n The parties’ dispute turns on the meaning of the term\\n“related to” as used in section 531 8(k)(3). Arguing for a narrow\\ninterpretation, Bank Three reads the statute to authorize\\ncollecting only those records of transactions that themselves\\npassed through Bank Three’s U.S. correspondent accounts. For\\nits part, the government advocates a more expansive reading—\\n\\f 21\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\none that would cover all records that have a “connection with”\\nthe Company’s use of Bank Three’s U.S. correspondent\\naccounts, even if some individual transactions made no use of\\na U.S. correspondent account. Appellee’s Br. 60 (internal\\nquotation marks omitted).\\n\\n To assess how broadly the phrase “related to” sweeps, we\\nlook to the “traditional tools of statutory interpretation—text,\\nstructure, purpose, and legislative history.” Tax Analysts v. IRS,\\n350 F.3d 100, 103 (D.C. Cir. 2003) (internal quotation marks\\nomitted). The “first step in interpreting a statute is to determine\\nwhether the language at issue has a plain and unambiguous\\nmeaning”; “[o]ur inquiry must cease if the statutory language\\nis unambiguous and the statutory scheme is coherent and\\nconsistent.” Robinson v. Shell Oil Co., 519 U.S. 337, 340\\n(1997) (internal quotation marks omitted).\\n\\n “[W]hen asked to interpret statutory language including the\\nphrase ‘relating to,’ . .th[e] [Supreme] Court has typically\\n .\\n\\n\\n\\nread the relevant text expansively.” Lamar, Archer & Cofrmn,\\nLLP v. Appling, 13$ S. Ct. 1752, 1760 (201$). We see no\\nreason to deviate from that approach here. The ordinary\\nmeaning of that phrase encompasses the broad relationships\\ndescribed by the government. See, e.g., Merriam-Webster’s\\nCollegiate Dictionary 987 (10th ed. 1997) (defining “relate” as\\n“to have [a] relationship or connection”); Black’s Law\\nDictionary 1288 (6th ed. 1990) (defining “relate” as “[tb stand\\nin some relation; to have bearing or concern; to pertain; refer;\\nto bring into association with or connection with”); see also BF\\nAmerica Production Co. v. Burton, 549 U.S. 84, 91(2006)\\n(“Unless otherwise defined, statutory terms are generally\\ninterpreted in accordance with their ordinary meaning.”).\\n\\f 22\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n The government’s interpretation finds further support in the\\none example that Congress offered of a record “related to” a\\nU.S. correspondent account: “records maintained outside of the\\nUnited States relating to the deposit of funds into the foreign\\nbank.” 31 U.S.C. § 5318(k)(3)(A)(i). The final phrase is\\nimportant. Merely depositing funds “into the foreign bank”\\nwould not necessarily entail passing those funds through the\\nforeign bank’s U.S. correspondent account; rather, it may be a\\npreparatory step for using the correspondent account. For\\ninstance, an aspiring money launderer might deposit illicitly-\\nobtained foreign currency into the foreign bank before using\\nthe U.S. correspondent account to facilitate a payment in U.S.\\ndollars, helping the guilty party conceal the origin of the funds.\\nBy specifying that its view of records “related to” a U.S.\\ncorrespondent account includes such a predicate transaction,\\nCongress embraced an expansive view of the term in line with\\nthe government’s. And, of course, that view accords with\\nCongress’s stated purpose in enacting the provision: “to\\nincrease the strength of United States measures to prevent,\\ndetect, and prosecute international money laundering and the\\nfinancing of terrorism.” Pub. L. No. 107-56, § 302(b)(1), 115\\nStat. at 297.\\n\\n Resisting this expansive text, Bank Three offers two\\nreasons to prefer a more circumscribed understanding of the\\nstatute. First, it again looks to the presumption against\\nextraterritoriality. But here, the statute contains no ambiguity\\nrequiring us to resort to that presumption. See Validus\\nReinsurance, Ltd. v. United States, 786 F.3d 1039, 1046 (D.C.\\nCir. 2015) (explaining that “ambiguity is resolved by\\n[applyingj the presumption against extraterritoriality”\\n(emphasis added)). To be sure, the statutory text is capacious,\\nbut as the Supreme Court has observed, “[b]road. language\\n . .\\n\\n\\n\\nis not necessarily ambiguous when congressional objectives\\n\\f 23\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nrequire broad terms.” Diamond v. Chakrabarty, 447 U.S. 303,\\n315 (1980).\\n\\n Second, Bank Three points to the Justice Department’s\\nattempt to persuade Congress to amend the statute to allow the\\ngovernment to obtain not just “records relating to” a foreign\\nbank’s U.S. correspondent account, but also “any records”\\npertaining to “any related account at the foreign bank,\\nincluding records maintained outside of the United States, that\\nare the subject of any investigation of a criminal violation of\\nUnited States law.” Department of Justice, Title III, Anti-\\nCorruption Legislative Proposals, 114th Cong., at 4 (May 5,\\n2016), https://www.justice.gov/opa/file/849986/download\\n(emphases added). The Bank would have us infer from\\nCongress’s failure to adopt the amendment that section\\n53 18(k)(3) as currently written can never reach “[a]l1\\ndocuments” relating to “all” of a company’s “correspondent\\nbanking transactions,” as this subpoena does. Bank Three\\nSubpoena, J.A. 771. But the Supreme Court has long cautioned\\nagainst reading too much into congressional inaction given that\\n“[l]ogically, several equally tenable inferences could be drawn\\nfrom the failure of the Congress to adopt an amendment.\\nincluding the inference that the existing legislation already\\nincorporated the offered change.” United States v. Wise, 370\\nU.S. 405, 411 (1962). In any event, the unenacted amendment\\nBank Three cites does not expand the definition of what records\\ncount as “related to” a U.S. correspondent account; rather, it\\nwould allow the government to reach other accounts,\\nregardless of their relationship to a U.S. correspondent account.\\nSuch a proposal tells us little about what Congress currently\\nconsiders to be “related to” a correspondent account.\\n\\n We therefore conclude that records “related to” a U.S.\\ncorrespondent account include records of transactions that do\\n\\f 24\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nnot themselves pass through a correspondent account when\\nthose transactions are in service of an enterprise entirely\\ndedicated to obtaining access to U.S. currency and markets\\nusing a U.S. correspondent account. All that remains is to\\ndetermine whether the government has adequately shown that\\nthe Company was such an enterprise.\\n\\n The district court found that the government had made the\\nnecessary showing. Specifically, it found that “[n]o part” of the\\nCompany could “be separated from the ploy to put money\\nthrough the United States financial market.” March Op. 29. On\\nthe record before us, that finding was not clearly erroneous. See\\nBoca Investerings Partnership v. United States, 314 F.3d 625,\\n629—30 (D.C. Cir. 2003), as amended (Mar. 26, 2003) (a\\nconclusion is clearly erroneous when “the reviewing court on\\nthe entire evidence is left with the definite and firm conviction\\nthat a mistake has been committed”) (internal quotation marks\\nomitted). The investigation has turned up considerable\\nevidence supporting the court’s conclusion. The Company\\nengaged in an enormous volume of U.S. dollar transactions:\\n“[B]etween October 2012 and November 2015 alone, [the\\nCompany] made over $100 million U.S.-dollar payments.” FBI\\nDeclaration ¶ 13, J.A. 866. And the Company “was a\\ncounterparty to over 680 U.S.-dollar international transfers\\n(which transited through the United States financial system via\\ncorrespondent U.S. banks).” Id., J.A. 867. The government has\\nuncovered no “evidence that” any of these transactions were\\npart of “a legitimate business”; rather, all were part of the\\nNKE’s scheme to launder money. Id. ¶ 10, J.A. 864—65.\\n\\n Nor does any record evidence suggest that the NKE used\\nthe Company to launder money through currencies other than\\nU.S. dollars. Indeed, the NKE had a special need for dollars\\nbecause “commodity companies, such as those operating in\\n\\f 25\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nChina, Thailand, and Singapore, frequently prefer to take\\npayments in U.S. dollars because of the dollar’s global demand\\nand stability.” Id. ¶ 26, l.A. 871; see also Correspondent\\nBanking: A Gateway for Money Laundering, S. Hrg. 107-84\\napp. 287 n.12 (“U.S. dollars are one of a handful of major\\ncurrencies accepted throughout the world. They are also\\nviewed as a stable currency, less likely to lose value over time\\nand, thus, a preferred vehicle for savings, trade and\\ninvestment.”). Further, the government explains, and Bank\\nThree nowhere contests, that before 2017—and consequently\\nduring the vast majority of the period covered by this\\nsubpoena—the NKE “could legally conduct transactions in\\nother currencies” besides the U.S. dollar (such as the euro),\\nAppellee’ s Br. 61, signaling that the NKE would have had no\\nneed of a front company to conduct transactions in those\\ncurrencies.\\n\\n Taken together, these facts support the district court’s\\nfinding that “[nb part” of the Company—and thus, none of its\\nbanking records—”can be separated from the ploy” to launder\\nmoney through U.S. financial markets, March op. 29, because\\nthe Company’s sole purpose was to give the NKE access to\\nU.S. dollars through U.S. correspondent accounts. We\\ntherefore conclude that on the facts of this case, the records\\nrequested by the Patriot Act subpoena all “relate to” Bank\\nThree’s U.S correspondent account. To be clear, nothing in this\\nopinion authorizes, as the Bank fears, the government “for any\\nreason, [to] obtain essentially any document or record from any\\nforeign bank with a U.S. correspondent account,” Bank Three\\nBr. 22, regardless of whether those records are “related to” that\\nU.S. correspondent account. The government’s power to\\nsubpoena the records in this case derives from the evidence that\\nthis particular company operated exclusively as a U.S. dollar\\nclearinghouse for the NKE. Under such circumstances, all\\n\\f 26\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nrecords pertaining to the Company’s Bank Three account and\\nits correspondent banking transactions, no matter where they\\noccurred, are “related to” the Bank’s U.S. correspondent\\naccounts.\\n\\n Iv.\\n “Comity refers to the spirit of cooperation in which a\\ndomestic tribunal approaches the resolution of cases touching\\nthe laws and interests of other sovereign states.” Société\\nNationale Industrielle Aérospatiale v. U.S. District Court for\\nthe Southern District of Iowa (“Société Nationale”), 482 U.S.\\n522, 543 n.27 (1987). Invoking this “spirit of cooperation,” Id.,\\nall three Banks argue that “ordering a foreign-sovereign-owned\\nbank to violate foreign law on foreign soil violates basic\\n‘comity’ principles” and, as such, demands reversal, Bank One\\nBr. 11; see also Bank Two Br. 25 (making this argument);\\nBank Three Br. 32—33 (same). We begin by providing some\\nadditional background before explaining why the Banks’\\nargument fails.\\n\\n A.\\n A comity analysis typically starts by determining whether\\na “true conflict” exists between the implicated legal systems.\\nHartford Fire Insurance Co. v. Calfornia, 509 U.S. 764, 798\\n(1993) (internal quotation marks omitted). Absent a clash of\\nforeign and domestic law, American courts may press forward\\nfree from worry that their rulings will threaten the\\n“international legal ties that advance the rule of law within and\\namong nations.” Laker Airways Ltd. v. Sabena, Belgian World\\nAirlines, 731 F.2d 909, 937 (D.C. Cir. 1984), as amended\\n(Mar. 6 and 9, 1984). As the district court recognized, however,\\nin this case that determination “is the easy part,” given that\\n“[tJhe [U.S.] government concedes that complying with the\\n\\f 27\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nrespective subpoenas exposes each bank to legal penalties in\\nChina.” March op. 32.\\n\\n After this initial inquiry, the question is whether “the\\nfactual circumstances” of the case at hand counsel against\\napplying American law. Laker Airways, 731 f.2d at 937. A\\ncollection of factors helps us to delineate the “inherently\\nuncertain” borders set by the principles of comity in the context\\nof subpoena enforcement. Id. Factors include “the importance\\nto the litigation of the documents . requested,” “the degree\\n . .\\n\\n\\n\\nof specificity of the request,” “whether the information\\noriginated in the United States,” “the extent to which\\nnoncompliance with the request would undermine important\\ninterests of the United States, or compliance with the request\\nwould undermine important interests of the state where the\\ninformation is located,” and whether “alternative means of\\nsecuring the information” exist. Société Nationale, 482 U.S. at\\n544 n.28 (alteration omitted) (quoting Restatement of foreign\\nRelations Law of the United States (Revised) § 437(1)(c)).\\nCourts also look to whether compliance would impose hardship\\non the party targeted by the subpoena and whether that party\\nhas acted in good faith. See In re Sealed Case (“Sealed\\nCaseT’), 825 F.2d 494, 498—99 (D.C. Cir. 1987)\\n(distinguishing cases based on “good faith” considerations);\\nLinde v. Arab Bank, PLC, 706 f.3d 92, 110 (2d Cir. 2013)\\n (“[A] district court should also examine the hardship of the\\nparty facing conflicting legal obligations... .“).\\n\\n The district court diligently applied this “constellation of\\nfactors” to the facts at issue here. March Op. 33. In so doing, it\\nobserved that two factors indisputably weigh against enforcing\\nthe subpoenas: the documents originated outside the United\\nStates, see id. at 42, and the Banks have acted in good faith\\nthroughout this litigation, see id. at 57 (“[T]he government\\n\\f 2$\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nagrees that none of the banks has acted in bad faith.”).\\nAccording to the district court, another pair of factors just as\\nclearly weighs in favor of enforcement: the documents contain\\nunique information vital to the investigation—indeed, “the\\ngovernment avows the subpoenaed documents are the\\nfoundation of the United States’ investigation,” Id. at 39\\n(alteration and internal quotation marks omitted), and the\\nsubpoenas target an appropriately specific collection of\\nrecords—that is, “the subpoenas are tailored to specific\\nfinancial records, for a defined date range, at banks [the\\nCompany] is known to have used to launder funds,” id. at 41.\\n\\n Parsing the more complex remaining factors, the district\\ncourt emphasized the government’s goal of learning how North\\nKorea has thwarted American sanctions in pursuing weapons\\nof mass destruction. “[NJon-enforcement” of the subpoenas,\\nthe district court stressed, “would undermine a critical national\\ninterest.” Id. at 4$. Although acknowledging that China has a\\n“legitimate” “interest in the development of a sound banking\\nsystem,” the district court concluded that the United States’\\nnational security interest easily outweighs China’s interest in\\nsafeguarding its banking secrecy laws. Id. at 50; see also Id.\\n(observing that because “even Chinese authorities recognize\\nthat bank secrecy can co-exist with limited disclosure to\\ngovernment agencies,” “China’s interest would not be\\nundermined by enforcement of the subpoenas”).\\n\\n The district court recognized that the U.S. government had\\ndeclined to take advantage of an alternative channel through\\nwhich the United States might obtain the documents at issue:\\nthe Mutual Legal Assistance Agreement, which creates a\\nprocess to exchange “documents” and “records” relevant to\\ncriminal investigations. MLAA art. l(2)(c); see March Op. 42—\\n47 (discussing the MLAA). But based on the U.S.\\n\\f 29\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ngovernment’s declarations that China has failed to\\nsatisfactorily engage in the MLAA process in many other cases\\nover the past decade, the district court accepted the Executive\\nBranch’s representation that requiring the government to first\\nexhaust the MLAA process would just “[aJllow[J China to gum\\nup [the] United States’ investigations” and, therefore, that\\n“[tJhe Department of Justice no longer needs to take overtures\\nof China’s willingness to assist the United $tates[] with\\ninvestigations into North Korea seriously.” March op. 47.\\n\\n finally, with respect to the hardship facing the Banks, the\\ndistrict court acknowledged that compliance with the\\nsubpoenas would violate Chinese law. Though the government\\ndoubts the Banks’ claim that any criminal laws would be\\nbroken, it concedes that compliance would at least violate those\\nlaws prohibiting Chinese companies from handing over records\\nthat have not been approved for release through the MLAA\\nprocess. See Law of the People’s Republic of China on\\nInternational Criminal Judicial Assistance (promulgated by the\\nStanding Committee of the Thirteenth National People’s\\nCongress, Oct. 26, 201$, effective Oct. 26, 201$), ch. I, art. 4,\\nJ.A. 355 (forbidding any “institution [or] organization\\nwithin the territory of [China]” from “provid[ing] evidentiary\\nmaterials and assistance prescribed by this Law to foreign\\ncountries”); Id. ch. II, § 2, art. 13, J.A. 35$ (requiring a “foreign\\ncountry” to request “criminal judicial assistance” “in\\naccordance with the provisions of the criminal judicial\\nassistance treaty”). The district court also recognized that the\\nMinistry of Justice has advised that these violations would have\\nconsequences in the form of administrative penalties against\\nthe Banks and civil or criminal penalties against individual\\nofficers. See March Op. 51—53. The district court, however,\\nfound the hardships accompanying compliance both\\nspeculative and minimal. See id. at 53—57. It based this\\n\\f 30\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nconclusion in part on historical Chinese enforcement trends:\\npenalties are rare and relatively modest when they do occur.\\nThe court found, for example, no instances of “a state-owned\\nChinese enterprise . fac{ing] severe repercussions for\\n . .\\n\\n\\n\\n\\nresponding to the order of a foreign court,” Id. at 53, and that\\nbank officers had apparently faced penalties only for stealing\\nand selling “client information for profit,” id. at 55. The district\\ncourt also relied on the logical supposition that the Chinese\\ngovernment is unlikely to severely sanction banks that it\\nlargely owns, id., and concluded that the Ministry’s warnings\\nfailed to “suggest[] the administrative penalties will amount to\\nmore than nominal fines” or provide any “concrete indication\\ncriminal penalties are likely,” Id. at 56—57.\\n\\n Affording the most weight to the United States’\\nunassailable interest in successfully investigating and, with any\\nluck, frustrating North Korea’s arms programs, see id. at 58\\n(calling “the interests of the relevant countries” the “most\\nimportant factor”), the district court concluded that “[ojn\\nbalance, international comity is not a reason to refrain from\\ncompelling compliance with the subpoenas,” Id. On appeal, the\\nBanks argue that the district court applied the incorrect legal\\nstandard, that enforcing the subpoenas runs afoul of our case\\nlaw, that “[c]omity dictates that the Chinese government at\\nleast be given an opportunity to comply with an MLAA request\\nbefore forcing a Chinese bank to violate Chinese law on\\nChinese soil,” Bank Two Br. 19, and, finally, that the district\\ncourt incorrectly balanced the relevant factors.\\n\\n B.\\n Before considering these arguments, we must first\\nascertain the appropriate standard of review. Although we\\n“ordinarily review{] a district court’s grant of a motion to\\n\\f 31\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ncompel.. for abuse of discretion,” Recording Industry Ass ‘n\\n .\\n\\n\\n\\n\\nof America, Inc. v. Verizon Internet Services, Inc., 351 F.3d\\n 1229, 1233 (D.C. Cir. 2003), we have yet to expressly\\ndetermine the standard of review applied to a comity analysis\\nin this context, cf Sealed Case I, $25 f.2d at 497—99 (not\\nidentifying the standard of review); de Csepel v. Republic of\\nHungary, 714 F.3d 591, 606 (D.C. Cir. 2013) (“This court has\\nnever expressly addressed the standard by which we review a\\ndistrict court’s decision to grant comity to a foreign\\njudgment.”). The Supreme Court, however, has observed that\\n“[t]he exact line between reasonableness and\\nunreasonableness” when weighing comity principles “must be\\ndrawn by the trial court, based on its knowledge of the case,”\\nSociétë Nationale, 482 U.S. at 546, and we generally review\\nfor abuse of discretion when the district court finds itself\\nresponsible for making such a fact-bound reasonableness call,\\nsee, e.g., Ideal Electronic Security Co. v. International Fidelity\\nInsurance Co., 129 F.3d 143, 150 (D.C. Cir. 1997) (“[TJhe\\nreasonableness of an attorney’s fees award is within the sound\\n discretion of the trial court and is reviewed only for abuse of\\n discretion.”); United States v. Law, $06 F.3d 1103, 1105 (D.C.\\n Cir. 2015) (“We. consider the substantive reasonableness of\\n . .\\n\\n\\n\\n\\n the sentence imposed under an abuse-of-discretion standard.”\\n (internal quotation marks omitted)). This approach comports\\n with our sister circuits, which review similar questions for\\n abuse of discretion. See, e.g., Allstate Life Insurance Co. v.\\n Linter Group Ltd., 994 F.2d 996, 999 (2d Cir. 1993) (“[S]ince\\n the extension or denial of comity is within the court’s\\n discretion, we will reverse the court’s decision only when we\\n find an abuse of discretion.”); Reinsurance Co. ofAmerica v.\\nAdministratia Asigurarilor de Stat, 902 F.2d 1275, 1283 (7th\\n Cir. 1990) (“The denial of [the company’s] motion to compel\\n responses was . . not an abuse of discretion.”). The Banks\\n .\\n\\n\\n\\n\\n offer no argument for departing from that approach here. See\\n\\f 32\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nBank Two Br. 5 (suggesting “an abuse of discretion standard”);\\nOral Arg. Rec. 13:33—13:40 (Bank One counsel agreeing that\\nif “the district court did apply the correct legal standard, then\\nit’s an abuse of discretion standard”). Accordingly, we shall\\nreview the district court’s comity analysis for abuse of\\ndiscretion.\\n\\n Employing that standard, we ask whether the district court\\n“based its ruling on an error of law, a clearly erroneous\\nassessment of the evidence, or an improper weighing of the\\nfactors limiting its discretion.” fraenkel v. Islamic Republic of\\nIran, $92 F.3d 348, 356 (D.C. Cir. 201$) (citation and internal\\nquotation marks omitted). We of course review any embedded\\n“legal conclusions de novo.” Guedes v. Bureau of Alcohol,\\nTobacco, firearms & Explosives, 920 F.3d 1, 10 (D.C. Cir.\\n2019).\\n\\n Bank One first argues that “the District Court used the\\nwrong legal standard to review the Government’s request to\\nenforce the Subpoenas”—that is, the court should not have\\nadopted “the multifactorial approach of . Restatement\\n . .\\n\\n\\n\\n(Third) of Foreign Relations.” Bank One Reply Br. 1. This\\nargument gets the Banks nowhere. In Sociëtë Nationale, the\\nSupreme Court explained that the Restatement’s “factors are\\nrelevant to any comity analysis.” 482 U.S. 522, 544 n.2$; see\\nOverby v. National Ass ‘n of Letter Carriers, 595 F.3d 1290,\\n1295 (D.C. Cir. 2010) (“[C]arefully considered language of the\\nSupreme Court, even if technically dictum, generally must be\\ntreated as authoritative.” (internal quotation marks omitted)).\\nThe Bank points out that our decision in Sealed Case I, issued\\na few months after Société Nationale, never expressly\\nreferences the Restatement’s factors. In that case we reversed\\nan order holding a bank owned by (unidentified) Country X in\\ncontempt for failure to comply with a subpoena that would\\n\\f 33\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nhave required the bank to produce documents in contravention\\nof the banking privacy laws of Country Y, where the branch at\\nissue was located. Sealed Case I, 825 F.2d at 497—98. But\\nSealed Case I neither refutes Société Nationale’ s ratification of\\nthe Restatement factors nor offers a different test to apply—\\nnor, of course, could it. See We the People foundation, Inc. v.\\nUnited States, 485 F.3d 140, 144 (D.C. Cir. 2007) (circuit\\ncourts must follow Supreme Court precedent). As the district\\ncourt recognized, Sealed Case I “comports with the Supreme\\nCourt’s treatment of international comity as a principle that\\ndepends on ‘prior scrutiny in each case of the particular facts,\\nsovereign interests, and likelihood that resort to [alternative]\\nprocedures will prove effective.” March Op. 34 (alteration in\\noriginal) (quoting Société Nationale, 482 U.S. at 544).\\n\\n This brings us to the Banks’ related assertion that “[t]he\\npresent case is controlled by” Sealed Case I. Bank One Br. 14.\\nBut that case by its own terms constricts its holding to its facts:\\n“If any of the facts we rest on here were different,” we\\ncautioned, “our holding could well be different.” Sealed Case I,\\n825 F.2d at 499. We therefore agree with the district court that\\nthis case merits its own factual inquiry; Sealed Case I does not,\\nas a legal matter, control the outcome here.\\n\\n The Banks’ more powerful argument is that, due to the\\nfactual parallels between Sealed Case I and this case, “it was\\nan abuse of discretion to enforce the Subpoena[sJ.” Bank Two\\nBr. 25. The Sealed Case I court identified several facts central\\nto its decision not to enforce the contempt order at issue, and\\nthey might sound familiar, first, and “[m]ost important,” the\\n“sanctions represent[ed] an attempt by an American court to\\ncompel a foreign person to violate the laws of a different\\nforeign sovereign on that sovereign’s own territory.” 825 F.2d\\nat 498; see also Id. (observing that the bank had no way to\\n\\f 34\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n“comply with the contempt order without violating the laws of\\nCountry Y on Country Y’s soil”). We also recognized that the\\nbank “had acted in good faith,” was “a third party . not...\\n . .\\n\\n\\n\\naccused of any wrongdoing,” and constituted “not merely a\\nprivate foreign entity,” but also “an entity owned by the\\ngovernment of Country X.” Id. finally, because the bank’s\\nmanager would “be available and able to testify as to many of\\nthe facts that the grand jury may [have] wish[edj to ascertain,”\\nthe government was not “left empty-handed” by the decision\\nnot to enforce the subpoena. Id. at 499. But see Ed.\\n(acknowledging “that the grand jury’s investigation may\\nnonetheless be hampered, perhaps significantly” by the\\ndecision not to enforce).\\n\\n As the Banks point out, each of those facts maps neatly\\nonto this case, in which the government asks us to compel\\nChinese banks owned in significant part by the Chinese\\ngovernment to violate Chinese law on Chinese territory. The\\nBanks have acted in good faith, they are currently third parties\\nto the criminal investigation, and the MLAA, at least on its\\nface, provides an alternative path that would allow the\\ngovernment to obtain the records at issue.\\n\\n What’s more, to the extent there are differences, there is\\nsome reason to believe that this case is an even better candidate\\nfor comity than Sealed Case I. For one thing, where the\\nsubpoena recipient in Sealed Case I was being asked to violate\\nthe laws of a “different foreign sovereign,” Id. at 498, here the\\ngovernment demands that the Banks violate their home\\ncountry’s laws. Although the Country X-based bank in Sealed\\nCase I might have closed its branch in Country Y to avoid\\nsanctions, the Banks here have no such workaround. As the\\ndistrict court recognized, “[a] Country X bank can shutter the\\nCountry Y portion of its business; a Chinese bank cannot leave\\n\\f 35\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nChina.” March Op. 36 n.13; see also Id. (“[C]ompelling a bank\\nto violate its own nation’s law is more prejudicial to\\ninternational comity.” (internal quotation marks omitted)).\\nMoreover, the Chinese government in this case “has warned\\nthe Bank[s] that [they] will be sanctioned if [they] ignore[J” its\\n“express instructions not to produce these documents outside\\nthe MLAA process,” and “[t]here is no indication that the\\nforeign government in Sealed Case I did anything similar.”\\nBank Three Br. 39.\\n\\n The district court ignored none of these considerations.\\nQuite to the contrary, it grappled with each and instead found\\npersuasive two other “critical” differences between the cases.\\nMarch Op. 34.\\n\\n The district court first pointed out that Sealed Case I\\nconcerned a “domestic law enforcement matter,” which\\n“hardly compares to the national security interests and\\nassociated potential harm caused by non-compliance with a\\nsubpoena related to an investigation into funding a state-\\nsponsor of terrorism’s nuclear weapons program.” Id. at 35.\\nThis is an important distinction. The government justifiably\\nbelieves that the subpoenas will unearth information\\nconcerning North Korea’s use of third-party front companies\\nto evade economic sanctions, and nothing in Sealed Case I\\nsuggests any “sovereign interest[],” Société Nationale, 482\\nU.S. at 544, remotely comparable to impeding North Korea’s\\nstockpiling of nuclear weapons.\\n\\n We acknowledge, as the Banks point out, that the\\ngovernment here has failed to proceed as though North Korea\\nwere about to push the nuclear launch button. Indeed, the\\ngovernment took nearly a year to enforce the subpoenas in\\ncourt and never sought to expedite this case until the district\\n\\f 36\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ncourt instructed it to do so. It did not, however, sit on its hands\\nduring that time, as the Banks repeatedly assert, but “engaged\\nin intra- and inter-agency deliberations” and “twice sent a\\ndelegation to China” during that period. March op. 40. More\\nto the point, we are reluctant to second-guess the Executive\\nBranch’s assessments of national security interests and how to\\nbest serve them. See frugone v. Central Intelligence Agency,\\n 169 f.3d 772, 775 (D.C. Cir. 1999) (cautioning “that courts\\nhave little expertise in either international diplomacy or\\ncounterintelligence operations”). We of course agree with\\nBank One that “the judiciary has an independent obligation to\\nevaluate” the diplomatic ripple effects of its decisions. Bank\\nOne Br. 25; see also Sealed Case I, 225 f.2d at 499 (“[WJe see\\ngood reason for courts not to act on their own, even at the\\nurging ofthe executive branch, when their actions ‘may hinder\\nrather than further this country’s pursuit of goals both for itself\\nand for the community of nations as a whole in the international\\n sphere.” (emphasis added) (quoting Banco Nacional de Cuba\\nv. Sabbatino, 376 U.S. 398, 423 (1964))). But we are confident\\nthat the district court respected that obligation and that it acted\\nwithin its discretion by recognizing that the heightened national\\n interest here, as articulated by the Executive Branch,\\n distinguishes this case from Sealed Case I.\\n\\n The Banks next observe that “if the Government receives\\nthe documents via the MLAA, its national security interests\\nwill be fully vindicated.” Bank Two Br. 26. The district court,\\nhowever, concluded that the futility of the MLAA process sets\\nthis case apart from Sealed Case I, in which testimony from the\\nbank’s manager appeared to offer a sure-fire way of securing\\nat least some of the desired information. “[un this\\ninvestigation,” by contrast, the district court observed, “the\\ngovernment insists that without compelling compliance, the\\nrequested information is out of reach.” March Op. 35.\\n\\f 37\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nforcefully disagreeing with this conclusion, the Banks insist\\nthat the MLAA process provides a channel specifically\\ndesigned to allow the U.S. government to obtain exactly the\\ntypes of records it seeks. Allowing it to eschew that process\\nwithout so much as making an initial MLAA request, they\\nargue, neglects comity’ s “spirit of cooperation.” $ociété\\nNationale, 482 U.S. at 543 n.27. For its part, the government\\nargues in support of the district court that “the historical record\\ndemonstrates . an MLAA request of China for such bank\\n . .\\n\\n\\n\\nrecords is not a viable alternative.” Appellee’s Br. 37 (internal\\nquotation marks omitted).\\n\\n Recall that the MLAA establishes a procedure for the U.S.\\nand Chinese governments to obtain documents from each other\\nin criminal investigations. Specifically, the process provides\\nthat “insofar as national law permits,” if, say, the United States\\nmakes a request, China’s “Central Authority” must search for\\nand then send the requested records. MLAA\\narts. 2(1), 14(1), (3). Nothing in the MLAA, however,\\ndesignates it as the exclusive means of obtaining evidence in a\\ncriminal investigation.\\n\\n In this case, China’s Central Authority, the Ministry of\\nJustice, Id. art. 2(2), expressed a “willingness to cooperate with\\na DOJ MLAA request” as early as March 2018, Letter from\\nHank B. Walther & Samidh Guha to Zia Faruqui & An\\nRedboard 3 (Mar. 23, 2018), J.A. 830. Bank Three\\ncommunicated China’s overture to the U.S. Department of\\nJustice, observing that the MLAA process “would allow DOJ\\nto obtain the documents it seeks, while avoiding the time, cost,\\nand risk of litigating the Patriot Act subpoena.” Id. As we now\\nknow, the United States declined that invitation—even as it\\nfiled at least ten other MLAA requests in 2017 and 2012—and\\ninstead sought to enforce the subpoenas in court.\\n\\f 38\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n\\n The Ministry then directly communicated with the district\\ncourt multiple times, assuring that court that the Ministry\\n“would timely review and handle the requests for assistance\\nsought by the DOJ in accordance with the [MLAA] and\\napplicable domestic laws.” Letter from Du Yaling, Director\\nGeneral of International Cooperation Department of the\\nMinistry of Justice to Chief Judge Beryl A. Howell 4 (Jan. 6,\\n2019), J.A. 76. If the government submits a “request in line\\nwith the [MLAAJ,” the Ministry pledged, “China will provide\\nthe assistance to the United States accordingly.” Id. Confident\\nthat the MLAA “mechanism works effectively,” the Ministry\\nreminded the district court that China has “provided effective\\nassistance to the US in many cases” and that “[tJhe avenue of\\nthe {MLAA] is the only legal means to obtain criminal\\nevidence in China by foreign judicial agencies under Chinese\\nlaws.” Letter from Du Yaling, Director General of International\\nCooperation Department of the Ministry of Justice to Chief\\nJudge Beryl A. Howell 2—4 (Feb. 26, 2019), J.A. 965—67. The\\ndistrict judge observed that this marked “the first time” she had\\nseen “correspondence directly from a foreign government\\nrelated to an ongoing matter.” March 5, 2019 Hearing Tr. 18:2—\\n4, J.A. 1573. The government points to no other case where the\\nMinistry has made such a representation and reneged on it and\\nadmits that in some cases, “China has produced bank records\\nto the general satisfaction of the U.S. prosecutors and agents.”\\nDeclaration of Associate Director of the Office of International\\nAffairs Jeffrey M. Olson (“OIA Declaration”) ¶ 14(e),\\nJ.A. 951.\\n\\n The district court considered all these facts but ultimately\\nconcluded that pursuing the MLAA process would result in an\\ninadequate production of documents. According to the\\nAssociate Director of the Office of International Affairs—\\n\\f 39\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nwhich serves as the United States’ Central Authority under the\\nMLAA—the United States has submitted fifty MLAA requests\\nfor bank records over the past decade and has received no\\nresponsive records to thirty-five; the government found most\\nof China’s responses in the remaining fifteen cases inadequate\\nor untimely. And of the ten requests submitted within the last\\ntwo years, the United States has received at most one response.\\nCf OIA Declaration ¶ 13(b), J.A. 950 (describing that, for\\nexample, China failed to produce documents in one of those\\ncases even though the United States informed the Ministry that\\nthe request was time-sensitive due to an upcoming trial).\\n\\n Moreover, the district court recognized that the\\ngovernment is especially pessimistic about the chances of\\nsuccess in this case given that the investigation involves a\\nperfect storm of disagreements between the United States and\\nChina: “Chinese authorities have a demonstrated failure in\\nsupporting foreign and U.S. sanctions and U.S. law\\nenforcement actions involving North Korea,” and “the Chinese\\ngovernment has specifically objected to the imposition of\\nsanctions on [the Company] and other Chinese front\\ncompanies.” FBI Declaration ¶ 80, J.A. $88. Drawing the\\ndistrict court’s attention to a case “nearly identical” to this one,\\nId. at 37, J.A. 884, an FBI special agent submitted a declaration\\npointing out that although in that case, China had “nominally\\nagreed to work jointly with the U.S. authorities” on MLAA\\nrequests for bank records involving North Korea, id. ¶ 72,\\nJ.A. $85, nothing ever came of that initial agreement; offering\\nvarious excuses, China never handed over any records. “Based\\non these facts,” the agent “concluded that the MLAA process\\nis not an effective way to obtain bank records from Chinese\\nauthorities with respect to investigations involving North\\nKorea.” Id. ¶ 75, J.A. 886; see Id. (“[OJne reason that Chinese\\nauthorities do not want to assist with North Korean\\n\\f 40\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ninvestigations is that producing such records could reflect\\nbadly on the Chinese government and the Chinese financial\\nindustry.”).\\n\\n Given this record, we see no abuse of discretion in the\\ndistrict court’s decision to enforce the subpoenas despite the\\nfact that the United States chose not to pursue the MLAA\\nprocess. The Supreme Court has made clear that evidence-\\nsharing arrangements set up by international agreements,\\nalthough often valuable, can also prove “unduly time\\nconsuming and expensive, as well as less certain” than court-\\nenforced subpoenas “to produce needed evidence.” $ociété\\nNationale, 482 U.S. at 542. As such, whether comity “requires\\nresort” to an agreement’s procedures calls for an inquiry into\\nthe “likelihood that resort to those procedures will prove\\neffective.” Id. at 544. The district court addressed just that\\nquestion: it reviewed declarations from U.S. and Chinese\\nofficials, identified facts that undermined the Chinese\\ngovernment’s assurances about the MLAA system, and\\nconcluded that the U.S. government had no obligation to\\nsubmit a voluntary MLAA request because it would likely\\nprove ineffective.\\n\\n We need not tarry over the Banks’ remaining arguments\\nthat the district court abused its considerable discretion in\\nbalancing the comity factors. Cf Morley v. Central Intelligence\\nAgency, 894 F.3d 389, 391 (D.C. Cir. 2018) (recognizing that\\nwhen “factors point in different directions,” “it will be the rare\\ncase when we can reverse a district court’s balancing ofthe.\\nfactors”). Specifically, Bank Three reiterates its claim that “the\\nPatriot Act subpoena is facially overbroad,” Bank Three Reply\\nBr. 20, and, as such, the “degree of specificity of the request,”\\nSociétë Nationale, 482 U.S. at 544 n.28 (internal quotation\\nmarks omitted), also weighs in favor of the Bank. But we agree\\n\\f 41\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nwith the district court that given that the Company “solely\\nexisted as a front company,” “all [the Company’s] records\\nrelate to the conduct under investigation,” March Op. 41\\n(internal quotation marks omitted). With respect to the\\nhardship factor, Bank One takes issue with the district court’s\\nconclusion that the Banks would likely face minimal penalties\\nfor complying with the subpoenas. In a civil law country such\\nas China, the Bank contends, “[t]he District Court cannot know\\nhow the Chinese government will exercise its prosecutorial\\ndiscretion.” Bank One Br. 2$. But the district court never\\npresumed to declare itself a soothsayer; it assessed a cluster of\\ndatapoints presented and substantiated by the Executive\\nBranch and ultimately determined that the risk of harsh\\npunishment was slim. In any event, it recognized that “this\\nfactor tips ever so slightly toward the banks.” March Op. 57.\\nThe district court exercised its ‘judgment” in drawing these\\nconclusions and was “guided by sound legal principles.”\\n United States v. Taylor, 487 U.S. 326, 336 (198$) (internal\\nquotation marks omitted). Our analysis ends there.\\n\\n We proceed with extreme caution when enforcing\\nsubpoenas that would require recipients to violate a foreign\\nsovereign’s domestic laws. As we said in Sealed Case I, “[w]e\\nhave little doubt. that our government and our people would\\n . .\\n\\n\\n\\nbe affronted if a foreign court tried to compel someone to\\nviolate our laws within our borders,” 825 F.2d at 498—99, all\\nthe more so if our government assured that court that it was\\nready and willing to engage in a mutually agreed-upon bilateral\\nprocess to provide the precise information requested. But we\\ntake heart that at every stage of this process decisionmakers\\nhave recognized that whether to enforce a subpoena “in cases\\nlike this one raises grave difficulties for courts.” Id. at 498. The\\ngovernment decided to pursue the records at issue through the\\ncourts rather than the MLAA process only after twice sending\\n\\f 42\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ndelegates to China and engaging in extensive “internal\\ndiscussions with subject matter experts within the FBI, as\\n . . .\\n\\n\\n\\nwell as consultations involving senior FBI management and\\nDepartment of Justice and Department of Treasury officials.”\\nFBI Declaration ¶ 80, J.A. $88. That decision was then put to\\n“the wise discretion of our judicial colleague[] on the District\\nCourt,” fraenkel, 892 F.3d at 362, and that court consented to\\nthe government’s course of action only after undertaking a\\nthorough, considered analysis that frankly acknowledged the\\nBanks’ comity concerns. Whether to enforce the subpoenas\\nwas a hard call, the district court made that call, and we have\\nno reason to reverse its fact-bound conclusion.\\n\\n V.\\n Finally, Banks One and Three challenge the district court’s\\nApril 2019 contempt order, arguing that it was unjustified\\ngiven the government’s concession that the Banks acted in\\ngood faith to obtain the Chinese government’s permission to\\nproduce the records covered by the subpoenas. “[W]e\\nreview. the contempt finding. for abuse of discretion.” In\\n . . . .\\n\\n\\n\\nre Fannie Mae Securities Litigation, 552 F.3d 814, 818 (D.C.\\nCir. 2009).\\n\\n A civil contempt citation is appropriate “only if the putative\\ncontemnor has violated an order that is clear and unambiguous,\\nand the violation must be proved by clear and convincing\\nevidence.” Broderickv. Donaldson, 437 F.3d 1226, 1234 (D.C.\\nCir. 2006) (internal quotation marks omitted). Banks One and\\nThree nowhere dispute that by refusing to produce the\\ndocuments covered by the subpoenas, they have violated a\\nclear and unambiguous order. Instead, they rest their challenge\\nto the contempt order on the Restatement (Third) of Foreign\\nRelations Law, which suggests that a court “should not\\n\\f 43\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nordinarily impose sanctions of contempt, dismissal, or default\\non a party that has failed to comply with [an] order for\\nproduction” that would violate another state’s laws or\\nregulations, “except in cases of deliberate concealment or\\nremoval of information or of failure to make a good faith\\neffort” “to secure permission from the foreign authorities to\\nmake the information available.” Restatement (Third) of\\nForeign Relations Law § 442(2)(a)—(b) (1927) (emphasis\\nadded).\\n\\n The banks insist that the district court should not have held\\nthem in contempt because they acted in good faith (albeit\\nunsuccessfully) to secure the Chinese government’s\\npermission to produce the records at issue. This is a rather thin\\nreed to begin with, rendered more slender still by the fact that\\nthe more recent Restatement (fourth) of foreign Relations Law\\nsignificantly diminishes the importance of a putative\\ncontemnor’s “good faith”: the Restatement (Fourth) simply\\nstates that courts “have discretion to excuse violations of law,\\nor moderate the sanctions imposed for such violations, on the\\nground that the violations are compelled by another state’s\\nlaw” if “the person in question appears likely to suffer severe\\nsanctions for failing to comply with foreign law” and “has\\nacted in good faith to avoid the conflict.” Restatement (Fourth)\\nof foreign Relations Law § 442 (2018) (emphasis added)). In\\nany event, this court has long held that “[t]he intent of the\\nrecalcitrant party is irrelevant in a civil contempt proceeding\\nbecause, unlike a criminal contempt proceeding, a civil\\ncontempt action is a remedial sanction used to obtain\\ncompliance with a court order or to compensate for damage\\nsustained as a result of noncompliance.” Food Lion, Inc. v.\\n United Food & Commercial Workers International Union,\\nAFL-CIO-CLC, 103 F.3d 1007, 1016 (D.C. Cir. 1997) (internal\\nquotation marks omitted).\\n\\f 44\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n\\n Confronted with clear non-compliance in the face of a\\nperfectly pellucid order, the district court reasonably found that\\n“the banks have not demonstrated good faith in a way relevant\\nto contempt” and that, because “the requested records are\\nessential to an investigation into a matter of national\\nsecurity[,].. continued contumacy threatens tremendous\\n .\\n\\n\\n\\n\\nharm.” April op. 9 (internal quotation marks omitted). Under\\nthese circumstances, the banks have given us no basis for\\nconcluding that the district court abused its discretion.\\n\\n VI.\\n For the foregoing reasons, we affirm the district court’s\\ncontempt orders against all three Banks.\\n\\n So ordered\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4171,"cells":{"id":{"kind":"number","value":4647052,"string":"4,647,052"},"name":{"kind":"string","value":"UNITED KEETOOWAH BAND OF CHEROKEE INDIANS IN OKLAHOMA, Individually and on Behalf of All Other Native American Indian Tribes and Tribal Organizations, Et Al., Petitioners v. FEDERAL COMMUNICATIONS COMMISSION and United States of America, Respondents National Association of Tribal Historic Preservation Officers, Et Al., Intervenors"},"name_abbreviation":{"kind":"string","value":"United Keetoowah Band of Cherokee Indians in Okla v. Fed. Commc'ns Comm'n"},"decision_date":{"kind":"string","value":"2019-08-09"},"docket_number":{"kind":"string","value":"18-1129; C/w 18-1135, 18-1148, 18-1159, 18-1184"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Tatel, Pillard, Edwards\", \"parties\": \"\", \"opinions\": [{\"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued March 15, 2019 Decided August 9, 2019\\n\\n No. 18-1129\\n\\n UNITED KEETOOWAH BAND OF CHEROKEE INDIANS IN\\n OKLAHOMA, INDIVIDUALLY AND ON BEHALF OF ALL OTHER\\n NATIVE AMERICAN INDIAN TRIBES AND TRIBAL\\n ORGANIZATIONS, ET AL.,\\n PETITIONERS\\n\\n v.\\n\\n FEDERAL COMMUNICATIONS COMMISSION AND UNITED\\n STATES OF AMERICA,\\n RESPONDENTS\\n\\nNATIONAL ASSOCIATION OF TRIBAL HISTORIC PRESERVATION\\n OFFICERS, ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 18-1135, 18-1148, 18-1159, 18-1184\\n\\n\\n On Petitions for Review of an Order of\\n the Federal Communications Commission\\n\\n\\n Stephen Díaz Gavin argued the cause for petitioners\\nUnited Keetoowah Band of Cherokee Indians in Oklahoma, et\\nal., and supporting intervenors. With him on the briefs were J.\\nScott Sypolt, Joel D. Bertocchi, Joseph H. Webster, F. Michael\\n\\f 2\\nWillis, Andrew Jay Schwartzman, James T. Graves, and\\nElizabeth S. Merritt. Angela J. Campbell entered an\\nappearance.\\n\\n Sharon Buccino argued the cause for petitioner Natural\\nResources Defense Council and intervenor Edward B. Myers.\\nWith her on the briefs was Edward B. Myers.\\n\\n Natalie A. Landreth argued the cause for petitioners\\nBlackfeet Tribe, et al. With her on the briefs were Wesley J.\\nFurlong, Joel West Williams, Troy A. Eid, Jennifer H. Weddle,\\nand Heather D. Thompson.\\n\\n Jacob M. Lewis, Associate General Counsel, Federal\\nCommunications Commission, argued the cause for\\nrespondents. With him on the brief were Jeffrey Bossert Clark,\\nAssistant Attorney General, U.S. Department of Justice, Eric\\nA. Grant, Deputy Assistant Attorney General, Andrew C.\\nMergen and Allen M. Brabender, Attorneys, Thomas M.\\nJohnson Jr., General Counsel, Federal Communications\\nCommission, David M. Gossett, Deputy General Counsel, and\\nC. Grey Pash Jr., Counsel. Jonathan H. Laskin and Robert B.\\nNicholson, Attorneys, U.S. Department of Justice, and Richard\\nK. Welch, Deputy Associate General Counsel, Federal\\nCommunications Commission, entered appearances.\\n\\n Joshua Turner argued the cause for intervenors in support\\nof respondents. With him on the brief were Christopher J.\\nWright and E. Austin Bonner.\\n\\n Before: TATEL and PILLARD, Circuit Judges, and\\nEDWARDS, Senior Circuit Judge.\\n\\n Opinion for the Court filed by Circuit Judge PILLARD.\\n\\f 3\\n PILLARD, Circuit Judge: Cellular wireless services,\\nincluding telephone and other forms of wireless data\\ntransmission, depend on facilities that transmit their radio\\nsignals on bands of electromagnetic spectrum. The Federal\\nCommunications Commission (FCC or Commission) has\\nexclusive control over the spectrum, and wireless providers\\nmust obtain licenses from the FCC to transmit. Wireless\\nservice in the United States has mostly depended on large,\\n“macrocell” radio towers to transmit cell signal, but companies\\noffering the next generation of wireless service—known as\\n5G—are in the process of shifting to transmission via hundreds\\nof thousands of densely spaced small wireless facilities, or\\n“small cells.” As part of an effort to expedite the rollout of 5G\\nservice, the Commission has removed some regulatory\\nrequirements for the construction of wireless facilities. These\\npetitions challenge one of the FCC’s orders paring back such\\nregulations, In re Accelerating Wireless Broadband\\nDeployment by Removing Barriers to Infrastructure\\nInvestment (Second Report & Order) (Order), FCC 18-30,\\n2018 WL 1559856 (F.C.C.) (Mar. 30, 2018).\\n\\n The Order exempted most small cell construction from\\ntwo kinds of previously required review: historic-preservation\\nreview under the National Historic Preservation Act (NHPA)\\nand environmental review under the National Environmental\\nPolicy Act (NEPA). Together, these reviews assess the effects\\nof new construction on, among other things, sites of religious\\nand cultural importance to federally recognized Indian Tribes.\\nThe Order also effectively reduced Tribes’ role in reviewing\\nproposed construction of macrocell towers and other wireless\\nfacilities that remain subject to cultural and environmental\\nreview.\\n\\n Three groups of petitioners challenge the Order as\\nviolating the NHPA, NEPA, and the Administrative Procedure\\n\\f 4\\nAct on several grounds: that its elimination of historic-\\npreservation and environmental review of small cell\\nconstruction was arbitrary and capricious, an unjustified policy\\nreversal, and contrary to the NHPA and NEPA; that the\\nchanges to Tribes’ role in reviewing new construction was\\narbitrary and capricious; that the Commission arbitrarily and\\ncapriciously failed to engage in meaningful consultations with\\nTribes in promulgating the Order; and that the Order itself\\nrequired NEPA review.\\n\\n We grant in part the petitions for review because the Order\\ndoes not justify the Commission’s determination that it was not\\nin the public interest to require review of small cell\\ndeployments. In particular, the Commission failed to justify its\\nconfidence that small cell deployments pose little to no\\ncognizable religious, cultural, or environmental risk,\\nparticularly given the vast number of proposed deployments\\nand the reality that the Order will principally affect small cells\\nthat require new construction. The Commission accordingly\\ndid not, pursuant to its public interest authority, 47 U.S.C.\\n§ 319(d), adequately address possible harms of deregulation\\nand benefits of environmental and historic-preservation\\nreview. The Order’s deregulation of small cells is thus\\narbitrary and capricious. We do not reach the alternative\\nobjections to the elimination of review on small cell\\nconstruction. We deny the petitions for review on the\\nremaining grounds.\\n\\n BACKGROUND\\n\\n I. Statutory and Regulatory Background\\n\\n A. National Historic Preservation Act (NHPA)\\n\\n Congress enacted the NHPA to “foster conditions under\\nwhich our modern society and our historic property can exist\\n\\f 5\\nin productive harmony” and “contribute to the preservation of\\nnonfederally owned historic property and give maximum\\nencouragement to organizations and individuals undertaking\\npreservation by private means.” 54 U.S.C. § 300101(1), (4).\\nAs part of that mission, NHPA’s Section 106 requires federal\\nagencies to “take into account the effect of” their\\n“undertaking[s] on any historic property.” Id. § 306108.\\n\\n Both “historic property” and “undertaking” have specific\\nmeanings under the statute. Historic properties include myriad\\nmonuments, buildings, and sites of historic importance,\\nincluding “[p]roperty of traditional religious and cultural\\nimportance to an Indian tribe.” Id. §§ 302706, 300308. Insofar\\nas Tribal heritage is concerned, the Section 106 process\\nrequires federal agencies to “consult with any Indian tribe . . .\\nthat attaches religious and cultural significance to” a historic\\nproperty potentially affected by a federal undertaking. Id.\\n§§ 302706, 306102. To count as “historic,” such properties\\nneed not be on Tribal land; in fact, they “are commonly located\\noutside Tribal lands and may include Tribal burial grounds,\\nland vistas, and other sites that Tribal Nations . . . regard as\\nsacred or otherwise culturally significant.” Order ¶ 97. Only\\na federal “undertaking,” not a state or purely private one,\\ntriggers the Section 106 Tribal consultation process. A federal\\n“undertaking,” as relevant here, is “a project, activity, or\\nprogram funded in whole or in part under the direct or indirect\\njurisdiction of a Federal agency, including . . . those requiring\\na Federal permit, license, or approval.” 54 U.S.C. § 300320.\\nWe have construed the statute to mean that, for an action to be\\na federal undertaking, “only a ‘Federal permit, license or\\napproval’ is required,” not necessarily federal funding. CTIA-\\nWireless Ass’n v. FCC, 466 F.3d 105, 112 (D.C. Cir. 2006).\\n\\n The Section 106 process requires that an agency “consider\\nthe impacts of its undertaking” and consult various parties, not\\n\\f 6\\nthat it necessarily “engage in any particular preservation\\nactivities.” Id. at 107 (quoting Davis v. Latschar, 202 F.3d 359,\\n370 (D.C. Cir. 2000)). The NHPA established an independent\\nagency, the Advisory Council on Historic Preservation\\n(Advisory Council), 54 U.S.C. § 304101, which is responsible\\nfor promulgating regulations “to govern the implementation\\nof” Section 106, id. § 304108(a). Agencies must consult with\\nthe Advisory Council, State Historic Preservation Officers, and\\nTribal Historic Preservation Officers, the last of which adopt\\nthe responsibilities of State Historic Preservation Officers on\\nTribal lands. 54 U.S.C. §§ 302303, 302702; 36 C.F.R.\\n§§ 800.3(c), 800.16(v)-(w) (defining State and Tribal Historic\\nPreservation Officers).\\n\\n The Advisory Council’s regulations authorize the use of\\nalternatives to the ordinary Section 106 procedures, called\\n“programmatic agreements.” 36 C.F.R. § 800.14(b). The\\nCommission develops programmatic agreements in\\nconsultation with the Advisory Council, Tribes, and other\\ninterested parties, “to govern the implementation of a particular\\nprogram or the resolution of adverse effects from certain\\ncomplex project situations or multiple undertakings” in certain\\ncircumstances, such as when “effects on historic properties are\\nsimilar and repetitive” or “effects on historic properties cannot\\nbe fully determined prior to approval of an undertaking.” Id.\\n§ 800.14(1)(i)-(ii). Tribes’ views must be taken into account\\nwhere the agreement “has the potential to affect historic\\nproperties on tribal lands or historic properties of religious and\\ncultural significance to an Indian tribe.” Id. § 800.14(b)(1)(i),\\n(f). For instance, the Commission has consulted with Tribes to\\nuse programmatic agreements to exclude from individualized\\nreview entire categories of undertakings that are unlikely to\\naffect historic properties. See In re Nationwide Programmatic\\nAgreement Regarding the Section 106 [NHPA] Review Process\\n(Section 106 Agreement), 20 FCC Rcd. 1073, 1075 ¶ 2 (2004).\\n\\f 7\\n B. National Environmental Policy Act (NEPA)\\n\\n Congress enacted NEPA to “encourage productive and\\nenjoyable harmony between man and his environment” and\\n“promote efforts which will prevent or eliminate damage to the\\nenvironment and biosphere and stimulate the health and\\nwelfare of man,” among other purposes. 42 U.S.C. § 4321.\\nLike the NHPA, NEPA mandates a review process that “does\\nnot dictate particular decisional outcomes, but ‘merely\\nprohibits uninformed—rather than unwise—agency action.’”\\nSierra Club v. U.S. Army Corps of Eng’rs, 803 F.3d 31, 37\\n(D.C. Cir. 2015) (quoting Robertson v. Methow Valley Citizens\\nCouncil, 490 U.S. 332, 351 (1989)).\\n\\n All “major Federal actions significantly affecting the\\nquality of the human environment” trigger environmental\\nreview under NEPA, just as federal “undertakings” trigger\\nhistoric preservation review under the NHPA. 42 U.S.C.\\n§ 4332(C). Major federal actions “include[] actions . . . which\\nare potentially subject to Federal control and responsibility.”\\n40 C.F.R. § 1508.18. Under the Commission’s procedures\\nimplementing NEPA, if an action may significantly affect the\\nenvironment, applicants must conduct a preliminary\\nEnvironmental Assessment to help the Commission determine\\nwhether “the proposal will have a significant environmental\\nimpact upon the quality of the human environment,” and so\\nperhaps necessitate a more detailed Environmental Impact\\nStatement. 47 C.F.R. § 1.1308; see also 40 C.F.R. § 1508.9.\\nIf, after reviewing the Environmental Assessment, the\\nCommission determines that the action will not have a\\nsignificant environmental impact, it will make a “finding of no\\nsignificant impact” and process the application “without\\nfurther documentation of environmental effect.” 47 C.F.R.\\n§ 1.1308(d).\\n\\f 8\\n NEPA also has an analogue to the NHPA’s Advisory\\nCouncil. In enacting NEPA, Congress established the Council\\non Environmental Quality, in the Executive Office of the\\nPresident, to oversee implementation of NEPA across the\\nentire federal government. 42 U.S.C. §§ 4342, 4344. With the\\nendorsement of the Council on Environmental Quality and by\\nfollowing a series of mandated procedures, agencies can\\nestablish “categorical exclusions” for federal actions that\\nrequire neither an Environmental Assessment nor an\\nEnvironmental Impact Statement. 40 C.F.R. § 1508.4.\\nCategorical exclusions are appropriate for “a category of\\nactions which do not individually or cumulatively have a\\nsignificant effect on the human environment and which have\\nbeen found to have no such effect in procedures adopted by a\\nFederal agency.” Id. “Categorical exclusions are not\\nexemptions or waivers of NEPA review; they are simply one\\ntype of NEPA review.” Council on Environmental Quality,\\nMemorandum for Heads of Federal Dep’ts and Agencies:\\nEstablishing, Applying & Revising Categorical Exclusions\\nunder [NEPA] (Categorical Exclusion Memo) 2 (2010).\\n\\n C. Legal Framework for Wireless Infrastructure\\n\\n The Communications Act of 1934 established the FCC to\\nmake available a “rapid, efficient . . . wire and radio\\ncommunication service with adequate facilities at reasonable\\ncharges.” 47 U.S.C. § 151. In licensing use of the spectrum,\\nthe Commission is tasked with promoting “the development\\nand rapid deployment of new technologies, products and\\nservices for the benefit of the public . . . without administrative\\nor judicial delays,” id. § 309, and “maintain[ing] the control of\\nthe United States over all the channels of radio transmission,”\\nid. § 301.\\n\\f 9\\n The Commission generally does not require construction\\npermits before private parties can build wireless facilities.\\nCongress largely eliminated the FCC’s site-specific\\nconstruction permits in 1982, and the Commission has since\\nrequired construction permits only where it finds that the public\\ninterest would be served by such permitting. See Pub. L. 97-\\n259, 96 Stat. 1087, § 119 (1982) (codified at 47 U.S.C.\\n§ 319(d)). It has not made such a finding for the wireless\\nfacilities at issue here.\\n\\n The FCC does, however, require licensing of the spectrum\\nused by wireless small cells. It does so by issuing geographic\\narea licenses, which allow wireless providers to operate on\\ncertain frequency bands in a wide geographic area. See 47\\nU.S.C. § 309(j). Those licenses authorize using spectrum\\nrather than building wireless facilities, but they necessarily\\ncontemplate facility construction. They have coverage\\nrequirements—for instance, one type of geographic area\\nlicense required licensees to provide service to at least 40% of\\nthe population in their geographic service area by June 2013.\\nSee 47 C.F.R. § 27.14(h). If they fail to meet the coverage\\nrequirements, they can be stripped of authority to operate for\\nthe license’s full term or serve part of its geographic area, and\\nthey “may be subject to enforcement action, including\\nforfeitures.” Id. The Commission also exercises continuing\\nauthority to inspect radio installations to ascertain their\\ncompliance with any and all applicable laws, whether or not the\\nlicensee itself constructed those installations. See 47 U.S.C.\\n§ 303(n); 47 C.F.R. § 1.9020(c)(5).\\n\\n The Commission has not identified any period since the\\nenactment of the NHPA (in 1966) and NEPA (in 1970) when\\nit did not require historic-preservation and environmental\\nreview of wireless facilities. After Congress eliminated the\\nconstruction permit requirement, the Commission for a time\\n\\f 10\\nrequired NEPA and NHPA review of facilities before it granted\\ntheir service licenses. See, e.g., In re Amendment of Envtl.\\nRules in Response to New Regulations Issued by [CEQ], FCC\\n85-626, 1986 WL 292182, at *5 ¶ 18 (F.C.C.) (Mar. 26, 1986)\\n(requiring review “during the period prior to grant of a station\\nlicense”); id. at *8 App’x ¶ 7 (requiring NEPA review on\\n“[f]acilities that will affect districts, sites, buildings, structures\\nor objects . . . that are listed in the National Register of Historic\\nPlaces or are eligible for listing,” which includes property of\\nreligious or cultural significance to Indian Tribes, 54 U.S.C.\\n§ 302706(a)). In 1990, the Commission shifted review from\\nthe licensing stage to the construction stage by establishing a\\n“limited approval authority” over construction of wireless\\nfacilities. In re Amendment of Envtl. Rules (1990 Order), 5\\nFCC Rcd. 2942 (1990). Limited approval authority required\\nthat, “where construction of a Commission-regulated radio\\ncommunications facility is permitted without prior\\nCommission authorization (i.e., without a construction permit),\\nthe licensee must nonetheless comply with historic\\npreservation and environmental review procedures.” Order\\n¶ 51; see also 47 C.F.R. § 1.1312. The authority was “limited”\\nin that it allowed “the Commission [to] exercise[] control over\\ndeployment solely to conduct federal historic and\\nenvironmental review.” Resp’t Br. 12. The Commission\\nemphasized that shifting review to the pre-construction stage\\nserved a practical function: Before it had established its limited\\napproval authority, the FCC’s rules “provide[d] that any\\nrequired submission of [Environmental Assessments] and any\\nrequired Commission environmental review take place at the\\nlicensing stage rather than prior to construction,” with the result\\nthat “[a]pplicants who ha[d] already constructed their\\nfacilities” could “subsequently be denied licenses on\\nenvironmental grounds.” 1990 Order 2942 ¶ 3. The\\nCommission explained that it continued to require review “to\\nensure that the Commission fully complies with Federal\\n\\f 11\\nenvironmental laws in connection with facilities that do not\\nrequire pre-construction authorization.” Id. ¶ 4. It announced\\nthe changes as “necessary to ensure that the Commission\\naddresses environmental issues early enough in the licensing\\nprocess to ensure that it fully meets its obligations under\\nFederal environmental laws,” including NEPA and the NHPA.\\nId. at 2943 ¶ 9 & n.16.\\n\\n The Commission has never required individualized review\\nof each separate facility, however. A long series of regulations,\\nprogrammatic agreements, and categorical exclusions has\\naggregated facilities for joint consideration and focused NHPA\\nand NEPA review on those deployments most likely to have\\ncultural or environmental effects. For instance, most\\ncollocations—deployments on existing structures—are\\nexcluded from individualized review under NHPA\\nprogrammatic agreements and NEPA categorical exclusions.\\nSee In re Implementation of the National Environmental Policy\\nAct of 1969 (Implementation of NEPA), 49 F.C.C.2d 1313,\\n1319-20 (1974); Nationwide Programmatic Agreement for the\\nCollocation of Wireless Antennas (Collocation Agreement), 47\\nC.F.R. pt.1, app. B (2001); Section 106 Agreement, 20 FCC\\nRcd. at 1075 ¶ 2; Nationwide Programmatic Agreement for\\nReview Under the National Historic Preservation Act, 70 Fed.\\nReg. 556 (2005); In re Acceleration of Broadband Deployment\\nby Improving Wireless Facilities Siting Policies (Improving\\nWireless Facilities Siting Policies), 29 FCC Rcd. 12865, 12870\\n¶ 11 (2014); 47 C.F.R. § 1.1320(b)(4). Categorical exclusions\\ngo through notice and comment, 40 C.F.R. § 1507.3; include\\nimpact findings, Categorical Exclusion Memo 9; require the\\nCouncil on Environmental Quality to approve them as\\nconsistent with its regulations and NEPA, 40 C.F.R.\\n§ 1507.3(a); and reserve rights to interested parties to request\\nfurther review in the event that atypical adverse effects do\\noccur, 47 C.F.R. § 1.1307(c), (d). At the same time, they\\n\\f 12\\nachieve enormous efficiencies in the review processes for\\nclasses of actions or undertakings anticipated to have minimal\\nor no adverse cultural or environmental effects.\\n\\n Since 2004, the FCC has been conducting NHPA review\\nin accordance with a broad programmatic agreement, the\\nSection 106 Agreement, 20 FCC Rcd. 1073. Interested parties\\ndeveloped that agreement to “tailor the Section 106 review in\\nthe communications context in order to improve compliance\\nand streamline the review process for construction of towers\\nand other Commission undertakings, while at the same time\\nadvancing and preserving the goal of the NHPA to protect\\nhistoric properties, including historic properties to which\\nfederally recognized Indian tribes . . . attach religious and\\ncultural significance.” Id. at 1074-75 ¶ 1. In the Section 106\\nAgreement, the Commission adopted “procedures for\\nparticipation of federally recognized Indian tribes,” among\\nother changes. Id. at 1075 ¶ 2. It also formalized the use of the\\nelectronic Tower Construction Notification System, which\\nnotifies Tribes of proposed wireless construction in areas they\\nhave identified as containing properties of religious and\\ncultural significance, and allows them to give applicants\\ninformation on the potential effects of proposed construction.\\nId. at 1106-10 ¶¶ 89-100.\\n\\n II. Order Under Review\\n\\n The challenged Order eliminated NHPA and NEPA\\nreview on small cells that meet certain size and other\\nspecifications, based on the Commission’s conclusion that\\nsuch review was not statutorily required and would impede the\\nadvance of 5G networks, and that its costs outweighed any\\nbenefits. See Order ¶¶ 36-45. The Order also altered Tribal\\ninvolvement in those Section 106 reviews that are still\\nconducted on wireless facilities that were not encompassed in\\n\\f 13\\nthe small cell exemption. See id. ¶¶ 96-130. Two of the five\", \"author\": \"Opinion for the Court filed by Circuit Judge PILLARD\"}, {\"type\": \"dissent\", \"author\": \"Commissioners\", \"text\": \"Commissioners dissented. See Order, Dissenting Statement of\\nComm’r Mignon L. Clyburn; Dissenting Statement of Comm’r\\nJessica Rosenworcel.\\n\\n We consolidated five timely petitions for review of the\\nOrder into this action. They challenge the Commission’s\\nexclusion of small cell construction from NHPA and NEPA\\nreview, its changes to Tribal involvement in Section 106\\nreview, and its promulgation of the Order itself. Three groups\\nof petitioners and intervenors, each designated here by the\\nname of its lead petitioner, challenge the Order. United\\nKeetoowah Band of Cherokee Indians (Keetoowah) represents\\na group of Tribes and historic preservation organizations.\\nBlackfeet Tribe (Blackfeet) represents another group of Tribes\\nand the Native American Rights Fund. The Natural Resources\\nDefense Council (NRDC) represents itself and Maryland\\ncitizen Edward B. Myers. Two wireless industry groups\\n(jointly, CTIA) intervened to defend the order alongside the\\nFCC.\\n\\n ANALYSIS\\n\\n We set aside an agency order only if it is “arbitrary,\\ncapricious, an abuse of discretion, or otherwise not in\\naccordance with law.” 5 U.S.C. § 706(2)(A). Agencies’\\nobligation to engage in “reasoned decisionmaking” means that\\n“[n]ot only must an agency’s decreed result be within the scope\\nof its lawful authority, but the process by which it reaches that\\nresult must be logical and rational.” Michigan v. EPA, 135 S.\\nCt. 2699, 2706 (2015) (quoting Allentown Mack Sales & Serv.,\\nInc. v. NLRB, 522 U.S. 359, 374 (1998)). Although “a court is\\nnot to substitute its judgment for that of the agency,” the\\narbitrary and capricious standard demands that the agency\\n“examine the relevant data and articulate a satisfactory\\n\\f 14\\nexplanation for its action including a rational connection\\nbetween the facts found and the choice made.” Motor Vehicle\\nMfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463\\nU.S. 29, 43 (1983) (internal quotation marks omitted). An\\nagency action is arbitrary and capricious where the agency has\\n“entirely failed to consider an important aspect of the problem”\\nor “offered an explanation for its decision that runs counter to\\nthe evidence before the agency, or is so implausible that it\\ncould not be ascribed to a difference in view or the product of\\nagency expertise.” Id.\\n\\n The FCC is entitled to deference to its reasonable\\ninterpretations of ambiguous provisions of the\\nCommunications Act. See Chevron, USA, Inc. v. Nat. Res. Def.\\nCouncil, Inc., 467 U.S. 837, 844-45 (1984). We owe no\\ndeference to the FCC’s interpretations of the NHPA or NEPA,\\nwhich are primarily administered by the Advisory Council, see\\nMcMillan Park Comm. v. Nat’l Capital Planning Comm’n, 968\\nF.2d 1283, 1287-88 (D.C. Cir. 1992), and the Council on\\nEnvironmental Quality, see Grand Canyon Tr. v. FAA, 290\\nF.3d 339, 341 (D.C. Cir. 2002) (as amended Aug. 27, 2002),\\nrespectively.\\n\\n I. Eliminating NHPA and NEPA Review on Small Cells\\n\\n The Order did not follow the processes for a programmatic\\nagreement under the NHPA, a categorical exclusion from\\nNEPA, or any other wholesale or aggregated form of review,\\nbut simply eliminated NHPA and NEPA review on most small\\ncells by removing them from the FCC’s limited approval\\nauthority. Small cells had not previously been defined or\\nregulated separately from macrocell towers. The Commission\\ndefines the small cells that its Order deregulates as wireless\\nfacilities that are not on Tribal lands, do not require antenna\\nstructure registration because they could not constitute a\\n\\f 15\\nmenace to air navigation, do not result in human exposure to\\nradiofrequency radiation in excess of applicable safety\\nstandards, and that are “small” per the following conditions:\\n\\n (i) The facilities are mounted on structures 50\\n feet or less in height including their antennas . . .\\n or the facilities are mounted on structures no\\n more than 10 percent taller than other adjacent\\n structures, or the facilities do not extend\\n existing structures on which they are located to\\n a height of more than 50 feet or by more than 10\\n percent, whichever is greater;\\n\\n (ii) Each antenna associated with the\\n deployment, excluding the associated\\n equipment . . . is no more than three cubic feet\\n in volume;\\n\\n (iii) All other wireless equipment associated\\n with the structure, including the wireless\\n equipment associated with the antenna and any\\n pre-existing associated equipment on the\\n structure, is no more than 28 cubic feet in\\n volume.\\n\\n47 C.F.R. § 1.1312(e)(2). Small cells that meet those\\nrequirements are now outside the purview of the Commission’s\\nlimited approval authority, the mechanism by which it has\\nrequired NHPA and NEPA review since 1990.\\n\\n The Commission deregulated small cells as part of a\\nbroader effort to reduce regulations that the FCC says “are\\nunnecessarily impeding deployment of wireless broadband\\nnetworks” on which 5G service depends. Order ¶ 3. “Within\\nthe next few years,” the Commission explained, “5G networks\\n. . . will make possible once-unimaginable advances, such as\\n\\f 16\\nself-driving cars and growth of the Internet of Things,” i.e.\\nphysical objects controllable over the internet. Id. ¶ 1. 5G\\nnetworks “will increasingly need to rely on network\\ndensification,” which entails “the deployment of far more\\nnumerous, smaller, lower-powered base stations or nodes that\\nare much more densely spaced.” Id. According to the\\nCommission, rapid proliferation of hundreds of thousands of\\nsmall cells would be hindered by the significant time and cost\\nof NHPA and NEPA reviews, even as the benefits of such\\nreview—which it characterized as already minimal—would be\\nnegligible because small cells are “inherently unlikely to\\ntrigger environmental and historic preservation concerns.” Id.\\n¶ 92; see also id. ¶¶ 9, 11-16. It noted that the FCC’s baseline\\napproach to environmental and historic-preservation review,\\nwhich requires facility-specific review unless a programmatic\\nagreement or categorical exclusion applies, “was developed\\nwhen all or nearly all deployments involved large macrocell\\nfacilities and accordingly failed to consider both the relatively\\ndiminutive size of small wireless facilities and the proliferation\\nof these facilities necessary for deployment of advanced\\nwireless technologies.” Id. ¶ 9.\\n\\n In the Order, the Commission asserts that federal law does\\nnot independently require such review. The only basis for\\ntreating small cell construction as either a federal undertaking\\ntriggering NHPA review or a major federal action triggering\\nNEPA review was, the Commission says, the limited approval\\nauthority the Commission exercised over that construction—\\nwhich the Order eliminated. See Order ¶¶ 58-59. The\\nCommission reasons that removing small cell construction\\nfrom its limited approval authority removes the “sufficient\\ndegree of federal involvement” necessary to render an\\nundertaking or action “federal.” Id. ¶ 58. It now says its power\\nto exercise limited approval authority over construction derives\\nexclusively from its “public interest authority” under the\\n\\f 17\\nCommunications Act, see Order ¶¶ 39, 53, 61, rather than from\\n“its obligations under Federal environmental laws,” 1990\\nOrder at 2943 ¶ 9. In this context, the “public interest\\nauthority” refers to the FCC’s power to require pre-\\nconstruction permits for wireless facilities if it “determines that\\nthe public interest, convenience, and necessity would be served\\nby requiring such permits.” 47 U.S.C. § 319(d). While the\\nCommission has never made such a determination for the\\ncategory of facilities at issue here, it has previously interpreted\\nthe public interest authority “as allowing the Commission to\\nrequire covered entities [not requiring preconstruction permits]\\nto nonetheless comply with environmental and historic\\npreservation processing requirements.” Order ¶ 53. In the\\nOrder, the Commission made a new determination that it was\\nnot in the public interest to require NHPA and NEPA review\\non small cells, so simply removed them from its limited\\napproval authority.\\n\\n Petitioners all argue that the FCC unlawfully excluded\\nsmall cells from NHPA and NEPA review. They contend first\\nthat removing small cells from the FCC’s limited approval\\nauthority was arbitrary and capricious. See 5 U.S.C.\\n§ 706(2)(A). Keetoowah and the NRDC argue that the\\nCommission failed to adequately consider the harms of\\nmassive deployment and to justify its decision to completely\\nexempt small cells from review. Additionally, all petitioners\\nargue that the NHPA and NEPA mandate review of small cell\\nconstruction. They assert that the geographic licenses the\\nCommission grants, which allow wireless companies to\\noperate on spectrum, constitute sufficient federal control over\\nwireless facility construction to make the construction a federal\\nundertaking and a major federal action triggering review under\\nthose statutes. Keetoowah also contends that the exclusion\\nviolates the Administrative Procedure Act on various other\\ngrounds, including that it is an unjustified policy reversal. If\\n\\f 18\\npetitioners prevail on any one or more of those grounds, we\\nmust vacate the Order’s deregulation of small cells and remand\\nto the FCC.\\n\\n The Commission failed to justify its determination that it\\nis not in the public interest to require review of small cell\\ndeployments. We therefore grant the petitions in part because\\nthe Order’s deregulation of small cells is arbitrary and\\ncapricious. The Commission did not adequately address the\\nharms of deregulation or justify its portrayal of those harms as\\nnegligible. In light of its mischaracterization of small cells’\\nfootprint, the scale of the deployment it anticipates, the many\\nexpedients already in place for low-impact wireless\\nconstruction, and the Commission’s decades-long history of\\ncarefully tailored review, the FCC’s characterization of the\\nOrder as consistent with its longstanding policy was not\\n“logical and rational.” Michigan v. EPA, 135 S. Ct. at 2706.\\nFinally, the Commission did not satisfactorily consider the\\nbenefits of review.\\n\\n First, the Commission inadequately justified its portrayal\\nof deregulation’s harms as negligible. The FCC partly based\\nits public-interest conclusion on a picture of small cells that the\\nrecord does not support. It described small cells as “materially\\ndifferent from the deployment of macrocells in terms of . . . the\\nlower likelihood of impact on surrounding areas.” Order ¶ 41.\\nIn its brief, the Commission sums up its explanation of the\\ndifference: “small cells are primarily pizza-box sized, lower-\\npowered antennas that can be placed on existing structures.”\\nResp’t Br. 3; see also Order ¶¶ 66, 92. It likened small cells to\\nsmall household items that operate on radiofrequency such as\\n“consumer signal boosters [and] Wi-Fi routers,” which do not\\nundergo review. Order ¶ 66. Small cells are, to be sure, quite\\ndifferent from macrocells in many ways, but the Commission\\nfails to address that small cells are typically mounted on much\\n\\f 19\\nbigger structures, and the Order is not limited to deployments\\non structures that already exist or are independently subject to\\nreview. Small cells deregulated under the Order can be\\n“mounted on structures 50 feet or less in height including their\\nantennas” or “mounted on structures no more than 10 percent\\ntaller than other adjacent structures.” 47 C.F.R. § 1.1312(e)(i).\\nThat makes them crucially different from the consumer signal\\nboosters and Wi-Fi routers to which the FCC compares them.\\n\\n The scale of the deployment the FCC seeks to facilitate,\\nparticularly given its exemption of small cells that require new\\nconstruction, makes it impossible on this record to credit the\\nclaim that small cell deregulation will “leave little to no\\nenvironmental footprint.” Order ¶ 41. The Commission\\nanticipates that the needed “densification of small deployments\\nover large geographic areas,” id., could require 800,000\\ndeployments by 2026, FCC, Declaratory Ruling & Third\\nReport & Order, FCC 18-133 ¶ 126 (Sept. 26, 2018). Even if\\nonly twenty percent of small cells required new construction—\\nas one wireless company estimates and the FCC highlights in\\nits brief, see Resp’t Br. 54—that could entail as many as\\n160,000 densely spaced 50-foot towers (or 198-foot towers, as\\nlong as they are located near 180-foot adjacent structures). The\\nCommission does not grapple with that possibility. Instead, it\\nhighlights the small cells that can be collocated without\\naddressing the many thousands that cannot be.\\n\\n As Keetoowah points out, the FCC “offers no analysis of\\nthe footprint of” the new towers on which small cells can be\\nmounted, “what equipment will be used, what ongoing\\nmaintenance or security will be provided and how often towers\\nwill be updated or rebuilt.” Keetoowah Br. 15-16.\\nDeployment of new small cells requires not only new\\nconstruction but also wired infrastructure, such as electricity\\nhookups, communications cables, and wired “backhaul,”\\n\\f 20\\nwhich connects the new antenna to the core network. See, e.g.,\\nComment of Sprint, Joint Appendix (J.A.) 380 (describing\\nprocess of deploying small cells); Comment of the Cities of\\nBos., Mass., et al., J.A. 705-06 (describing the equipment\\nassociated with small cells), NRDC Br. Ex. A, Decl. of Warren\\nBetts ¶¶ 11-12 (describing concerns about disruption “by the\\nlaying of cables and wires, by the maintenance they require,\\n[and] by the sound of the maintenance vehicles” in otherwise\\ntranquil areas, and concerns “that trees may be cut down or\\ndamaged by the construction of small cells”). Construction,\\nconnection, and maintenance may entail excavation and\\nclearing of land. The Tribal Historic Preservation Officer for\\nthe Seminole Tribe of Florida expressed concern about effects\\nof anticipated “additional related infrastructure, such as\\nfencing, security, and access for periodic maintenance and\\ntroubleshooting.” Keetoowah Br. Add. 114, Decl. of Paul\\nBackhouse, ¶ 28. While the Commission asserted that\\n“deployment of small wireless facilities commonly (although\\nnot always) involves previously disturbed ground,” it\\neliminated review of small cells that will involve new ground\\ndisturbance without responding to concerns about such\\ndisturbance. Order ¶ 92; see also, e.g., Comment of the Nat’l\\nCong. of Am. Indians, et al. (NCAI), J.A. 430-31 (expressing\\nconcern about small cells that require ground disturbance);\\nComment of the Cities of Bos., Mass., et al., J.A. 707 (“No\\nexplanation is offered by the Commission for its exclusion of\\nany ground disturbance related conditions” in the draft Order).\\n\\n The Commission also failed to assess the harms that can\\nattend deployments that do not require new construction,\\nparticularly the cumulative harms from densification. While\\n“Tribal Nations are most concerned with federal undertakings\\nthat disturb the ground and turn up dirt,” even “[c]ollocations\\ncan affect cultural and historical properties th[r]ough\\ndisturbing view sheds” because “[t]he cultural and spiritual\\n\\f 21\\ntraditions of Tribal Nations across the United States frequently\\ninvolve the uninterrupted view of a particular landscape,\\nmountain range, or other view shed.” Comment of NCAI, J.A.\\n50. The FCC did not respond to historic-preservation\\ncommenters warning “that permanent, direct adverse effects\\nwill be more likely with small wireless facilities as in many\\ncases they are proposed for installation on or in historic\\nbuildings,” and “these multi-site deployments have a greater\\npotential to cause cumulative effects to historic properties,\\ncluttering historic districts with multiple towers, antennae, and\\nutility enclosures.” Comment of Tex. Historical Comm’n, J.A.\\n794; see also, e.g., Ex Parte Commc’n of Thlopthlocco Tribal\\nTown Tribal Historic Pres. Officer, J.A. 690 (noting that the\\nCommission did not discuss “the issue of multiple collocations\\non the same pole which cumulatively would exceed the volume\\nrestriction and would create an adverse impact”); Comment of\\nArk. State Historic Pres. Officer, J.A. 751 (“[A]lthough\\nindividual small cells are unlikely to adversely impact\\nindividual historic properties or districts, the FCC doesn’t\\naddress how the large scale, nationwide deployment of 5G and\\nsmall cells facilities will cumulatively impact cultural and\\nnatural resources.”). The Commission noted that all facilities\\nremain subject to its limits on radiofrequency exposure, Order\\n¶ 45, but failed to address concerns that it was speeding\\ndensification “without completing its investigation of . . . health\\neffects of low-intensity radiofrequency radiation,” which it is\\ncurrently reassessing. Comment of BioInitiative Working Grp.,\\nJ.A. 235.\\n\\n The FCC does not reconcile its assertion that planned\\nsmall cell densification does not warrant review because it will\\n“leave little to no environmental footprint” with the Order’s\\nprincipal deregulatory effect of eliminating review of precisely\\nthe new construction and other deployments that the\\nCommission previously considered likely to pose cultural and\\n\\f 22\\nenvironmental risks. The Commission already had in place\\nNEPA categorical exclusions and NHPA programmatic\\nagreements covering most collocations—as well as other kinds\\nof deployments unlikely to have cultural and environmental\\nimpacts. What the new Order accomplishes, then, is to sweep\\naway the review the Commission had concluded should not be\\nrelinquished.\\n\\n Since the 1970s, the Commission has explained that most\\ncollocations on existing towers or buildings are not “major”\\nfederal actions and therefore are not subject to NEPA review.\\nImplementation of NEPA, 49 F.C.C.2d at 1319-20; 47 C.F.R.\\n§§ 1.1301-1.1319. The FCC’s NEPA regulations limit\\nenvironmental review to a small subset of actions likely to have\\nsignificant environmental effects, see 47 C.F.R. § 1.1307, as\\nwell as those actions found through Section 106 review to have\\nadverse effects on historic properties, see id. § 1.1307(a)(4).\\nBefore it promulgated the challenged rule, the Commission had\\nfurther shrunk the category of actions that receive\\nindividualized NHPA or NEPA review by adopting\\nprogrammatic agreements and categorical exclusions. In\\nchronological order, it excluded most collocations from\\nindividualized review, see Collocation Agreement, 47 C.F.R.\\nPt.1, App. B; adopted “categories of undertakings that are\\nexcluded from the Section 106 process because they are\\nunlikely by their nature to have an impact upon historic\\nproperties,” Section 106 Agreement, 20 FCC Rcd. at 1075 ¶ 2;\\nexcluded from individualized review new categories of\\nwireless construction and modification unlikely to have\\nhistoric preservation effects, see Nationwide Programmatic\\nAgreement for Review Under the National Historic\\nPreservation Act, 70 Fed. Reg. at 558; and, most recently,\\nexpanded NHPA and NEPA exclusions for collocations, see\\nImproving Wireless Facilities Siting Policies, 29 FCC Rcd. at\\n12870 ¶ 11. In sum, the FCC had already streamlined and\\n\\f 23\\nminimized review of vast numbers of minor actions, focusing\\nattention only on subcategories of deployments likely to have\\ncultural or environmental effects.\\n\\n Second, in sweeping away wholesale the review it had\\npreserved for the small cell deployments most likely to be\\ndisruptive, the Order is not, as the FCC asserts, “consistent\\nwith the Commission’s treatment of small wireless facility\\ndeployments in other contexts,” but directly contrary to it.\\nOrder ¶ 42. We observe by way of example the Commission’s\\nassertion that “under the Collocation [Agreement], the\\nCommission already excludes” from NHPA review “many\\nfacilities that meet size limits similar to those” of small cells.\\nId. As the Commission sees it, the Order thus “builds upon the\\ninsight underlying these existing rules that small wireless\\nfacilities pose little or no risk of adverse environmental or\\nhistoric preservation effects.” Id. But the Collocation\\nAgreement exclusion was defined not just by size, but by other\\ncharacteristics that minimized the likelihood of cultural harm.\\nThe section of the Collocation Agreement the FCC cites in fact\\nonly excludes from individualized NHPA review “small\\nwireless antennas and associated equipment on building and\\nnon-tower structures that are outside of historic districts and are\\nnot historic properties,” which include property of religious\\nand cultural importance to Tribes. Collocation Agreement, 47\\nC.F.R. Pt.1, App. B § VI (formatting altered); see also 54\\nU.S.C. §§ 300308, 302706. A different section of the\\nCollocation Agreement, which did exempt certain collocations\\nof small antennas in historic districts or on historic properties,\\nlikewise included numerous conditions to minimize effects on\\nhistoric properties. An antenna could only be collocated on a\\nhistoric property if, for example, “a member of the public, an\\nIndian Tribe, a [State Historic Preservation Office] or the\\n[Advisory] Council” had not complained “that the collocation\\nha[d] an adverse effect on one or more historic properties,”\\n\\f 24\\nCollocation Agreement, 47 C.F.R. Pt.1, App. B § VII(A)(6),\\nand if the antenna was installed “using stealth techniques that\\nmatch or complement the structure on which or within which it\\nis deployed,” id. § VII(A)(2)(c), and “in a way that does not\\ndamage historic materials and permits removal of such\\nfacilities without damaging historic materials,” id. § VII(A)(4),\\namong other conditions. After the Order, none of those\\nlimiting conditions applies. The insight of the Collocation\\nAgreement was not that small cells by their nature “pose little\\nor no risk of adverse environmental or historic preservation\\neffects,” Order ¶ 42, but that small cells under certain carefully\\ndefined conditions pose little such risk.\\n\\n Similarly, the FCC explains its “conclusion that, as a class,\\nthe nature of small wireless facility deployments appears to\\nrender them inherently unlikely to trigger environmental and\\nhistoric preservation concerns” by reference to limiting criteria\\nthat it chose not to place on its small cell exemption. Id. ¶ 92.\\nIt notes, for example, that “deployment of small wireless\\nfacilities commonly (although not always) involves previously\\ndisturbed ground, where fewer concerns generally arise than on\\nundisturbed ground,” and reiterates that “use of existing\\nstructures, where feasible, can both promote efficiency and\\navoid adverse impacts on the human environment.” Id. But\\nthe Commission decided not to limit the Order’s exemption\\nonly to facilities sited on previously disturbed ground, or those\\nthat are collocated on existing structures. It therefore fails to\\njustify its conclusion that small cells “as a class” and by their\\n“nature” are “inherently unlikely” to trigger concerns.\\n\\n By ignoring the extent to which it had already streamlined\\nreview, the Commission also overstated the burdens of review.\\nIt said it could not “simply turn a blind eye to the reality that\\nthe mechanical application of [limited approval authority]\\nrequirements to each of [the] small deployments” necessary for\\n\\f 25\\n5G “would increase the burden of review both to regulated\\nentities and the Commission by multiples of tens or hundreds.”\\nId. ¶ 65. As the preceding discussion of the Collocation\\nAgreement illustrates, however, the FCC was not\\nindiscriminately or “mechanic[ally]” requiring full NHPA and\\nNEPA review for each individual small cell. The Commission\\nfails to explain why the categorical exclusions and\\nprogrammatic agreements in place did not already minimize\\nunnecessary costs while preserving review for deployments\\nwith greater potential cultural and environmental impacts.\\n\\n Third, given that only the most vulnerable cases were still\\nsubject to individualized NHPA or NEPA review, the\\nCommission did not adequately address either the possible\\nbenefits of retaining review, or the potential for further\\nstreamlining review without eliminating it altogether. It\\ndismissed the benefits of historic-preservation and\\nenvironmental review in a two-sentence paragraph, describing\\nmost of the comments that highlight those benefits as\\n“generalized” and the comments that point to specific benefits\\nas “few.” Id. ¶ 78. Characterizing a concern as “generalized”\\nwithout addressing that concern does not meet the standard of\\n“reasoned decisionmaking.” Michigan v. EPA, 135 S. Ct. at\\n2706.\\n\\n The Commission found that adverse effects are rare, but it\\nconsidered neither the importance of the sites review does save,\\nnor how that rarity depends on the very review it eliminates,\\nwhich forestalled adverse effects that otherwise would have\\noccurred. The FCC cited comments suggesting that only 0.3 or\\n0.4% of requests for Tribal review result in findings of adverse\\neffects or possible adverse effects. Order ¶ 79. Based on the\\nestimate of 800,000 small cell deployments, that could mean\\n3,200 adverse effects. The Order displayed no consideration\\nof the importance of the 3,200 Tribal sites that might be saved\\n\\f 26\\nthrough review except to describe that benefit as “de minimis\\nboth individually and in the aggregate.” Id. As counsel for\\npetitioner Blackfeet Tribe said at oral argument: “They may\\nthink that’s infinitesimal. To us, it means the world.” Oral\\nArgument at 1:16:16-20. The Commission also did not address\\ncomments that “no adverse effects in 99% of tower\\ndeployments shows that the current system is working”\\nbecause “[o]ften, after an applicant enters a location into” the\\nTower Construction Notification System, a Tribal\\nrepresentative “will notify the applicant of an issue and the\\napplicant will choose a new location or resolve that effect,”\\nwhich “gets counted as having no adverse effect.” Comment of\\nNat’l Ass’n of Tribal Historic Pres. Officers, J.A. 661. Other\\ncommenters agreed that “[t]he lack of significant impact should\\nbe a testament to the value of the review process in these\\ninstances, not negate its necessity.” Comment of Tex.\\nHistorical Comm’n, J.A. 794 (“In our experience, the vast\\nmajority of adverse effects for cell projects are resolved\\nthrough sensitive design modifications, including stealth\\nmeasures, modifying how equipment is attached if directly\\nmounted to a historic building or structure, or relocation to an\\nalternate site further removed from historic properties.”).\\n\\n Similarly, the Commission dismissed the point that its own\\noversight deters adverse effects by describing comments to that\\neffect as “generalized, and undercut by our conclusion that, as\\na class, the nature of small wireless facility deployments\\nappears to render them inherently unlikely to trigger\\nenvironmental and historic preservation concerns.” Order\\n¶ 92. For the reasons already explained, the FCC’s conclusion\\nthat small cells are inherently unlikely to trigger concerns is\\narbitrary and capricious, and describing comments as\\n“generalized” does not excuse the agency of its obligation to\\nconsider those comments as part of reasoned decisionmaking.\\n\\f 27\\n We hold that the Order’s deregulation of small cells is\\narbitrary and capricious because its public-interest analysis did\\nnot meet the standard of reasoned decisionmaking. We\\ntherefore decide neither the alternative grounds for holding that\\nthe Order is arbitrary and capricious or otherwise violated the\\nAdministrative Procedure Act, nor the claim that small cell\\nconstruction is a federal undertaking and a major federal action\\nrequiring NHPA and NEPA review.\\n\\n II. Tribal Involvement in Section 106 Review\\n\\n The Order also made three changes to Tribal involvement\\nin the Section 106 review not eliminated by the Order, such as\\nreview of macrocells and small wireless facilities on Tribal\\nland. The first two changes relate to two types of Tribal\\ninvolvement that the Commission and the Advisory Council\\ndistinguish from one another: (a) government-to-government\\nconsultation between the agency and the Tribes, in which\\nTribes function in their governmental capacity, and (b) the\\n“identification and evaluation phase of the Section 106 process\\nwhen the agency or applicant is carrying out its duty to identify\\nhistoric properties that may be significant to an Indian tribe.”\\nAdvisory Council, Consultation with Indian Tribes in the\\nSection 106 Review Process: A Handbook (Section 106\\nHandbook), J.A. 1015; see also FCC, Voluntary Best Practices\\nfor Expediting the Process of Communications Tower and\\nAntenna Siting Review Pursuant to Section 106 of the NHPA,\\nJ.A. 933; Order ¶¶ 118-19.\\n\\n Section 106 review comprises “four steps”: “initiation,\\nidentification, assessment [or evaluation], and resolution.”\\nSection 106 Handbook, J.A. 1018. Government-to-\\ngovernment consultation is a background requirement of\\nSection 106 review at every stage. See id. at J.A. 1014, 1018;\\nAdvisory Council, Fees in the Section 106 Review Process,\\n\\f 28\\nJ.A. 913; 36 C.F.R. § 800.2(c)(2)(ii)(A) (consultation requires\\ngiving the interested Tribe “a reasonable opportunity to\\nidentify its concerns about historic properties, advise on the\\nidentification and evaluation of historic properties, . . .\\narticulate its views on the undertaking’s effects on such\\nproperties, and participate in the resolution of adverse effects”).\\nIn the identification and evaluation period, however, applicants\\nhave often paid for expertise and assistance from Tribes acting\\n“in a role similar to that of a consultant or contractor” such as\\nby providing “specific information and documentation\\nregarding the location, nature, and condition of individual\\nsites” or even conducting surveys. Section 106 Handbook, J.A.\\n1015. The Order explains that identification and evaluation\\ninvolves “activities undertaken after the initial determination\\nthat historic properties are likely to be located in the site\\nvicinity,” and that it includes “monitoring and other activities\\ndirected toward completing the identification of historic\\nproperties as well as assessing and mitigating the project’s\\nimpacts on those properties.” Order ¶ 124.\\n\\n The “initial determination” falls into the government-to-\\ngovernment consultation category. See Section 106 Handbook,\\nJ.A. 1021 (explaining that initiating contact with Tribes is part\\nof the Commission’s “responsibilities to conduct government-\\nto-government Consultation”). In practice, however, Tribes\\nhave been allowing applicants to contact them directly, in lieu\\nof government-to-government consultation, to help make the\\ninitial determination. See Section 106 Agreement, 20 FCC Rcd.\\nat 1108 ¶¶ 95-96; Keetoowah Br. 37. The Section 106\\nAgreement “expresses the ambition that this initial contact will\\nlead to voluntary direct discussions through which applicants\\nand tribes . . . will resolve questions involving the presence of\\nrelevant historic properties and effects on such properties to the\\ntribe[’s] . . . satisfaction without Commission involvement.”\\n20 FCC Rcd. at 1108 ¶ 97. But “if an applicant and an Indian\\n\\f 29\\ntribe . . . disagree regarding whether an undertaking will have\\nan adverse effect on a historic property of religious and cultural\\nsignificance, or if the tribe . . . does not respond to the\\napplicant’s inquiries,” the Commission steps in to consult and\\nultimately “make a decision regarding the proposed\\nundertaking.” Id.\\n\\n The Advisory Council explains that “[t]hese two tribal\\nroles”—government-to-government consultation, and\\nassistance with identification and evaluation—“are not treated\\nthe same when it comes to compensation, although the line\\nbetween them may not be sharp.” Advisory Council, Fees in\\nthe Section 106 Review Process, J.A. 913. Advisory Council\\nguidance states that “agencies are strongly encouraged to use\\navailable resources to help overcome financial impediments to\\neffective tribal participation in the Section 106 process” and\\napplicants are likewise “encouraged to use available resources\\nto facilitate and support tribal participation.” Advisory\\nCouncil, Section 106 Handbook, J.A. 1015. At the same time,\\nit says that agencies and applicants should not expect to pay\\nfees for government-to-government consultation, which\\n“give[s] the Indian tribe an opportunity to get its interests and\\nconcerns before the agency,” Advisory Council, Fees in the\\nSection 106 Review Process, J.A. 913, but “should reasonably\\nexpect to pay” fees for the identification and evaluation, which\\nputs Tribes in a “consultant or contractor” role, Advisory\\nCouncil, Section 106 Handbook, J.A. 1015. It notes, however,\\nthat “this encouragement is not a legal mandate; nor does any\\nportion of the NHPA or the [Advisory Council’s] regulations\\nrequire an agency or an applicant to pay for any form of tribal\\ninvolvement.” Id.\\n\\n First, apparently because applicants had been consistently\\npaying upfront fees, see Keetoowah Br. 37, the Order made\\nclear that applicants’ payment of upfront fees to Tribes is\\n\\f 30\\nvoluntary. See Order ¶ 116. Upfront fees are payments made\\nto Tribes for the initial determination whether the Tribe\\nactually has religiously or culturally significant properties that\\nmight be affected by a proposed construction. See id. ¶ 116.\\nApplicants contact Tribes for that initial determination when\\nTribes have noted that properties in the general area of\\nproposed construction may have religious or cultural\\nsignificance for them. Id. When an applicant follows up “to\\nascertain whether there are in fact such properties that may be\\naffected,” some Tribes have requested upfront fees before they\\nwill respond. Id. As the Order describes the practice, the\\nupfront fees “do not compensate Tribal Nations for fulfilling\\nspecific requests for information and documentation, or for\\nfulfilling specific requests to conduct surveys,” but are “more\\nin the nature of a processing fee” to “obtain a response” to an\\napplicant’s initial Tower Construction Notification contact\\nwith a Tribal Nation. Id. ¶ 119.\\n\\n Second, while the Order approved of fees for identifying\\nand evaluating properties that may be significant to Tribes, as\\nopposed to upfront fees, see id. ¶ 123, it also authorized\\napplicants to consult with non-Tribal parties in the\\nidentification and evaluation phase, see id. ¶¶ 124-45. The\\nCommission found that, if an applicant asks a Tribe to perform\\nwork to aid it in documenting, surveying, or analyzing\\npotentially historic properties, “the applicant should expect to\\nnegotiate a fee for that work” and, if the parties are “unable to\\nagree on a fee, the applicant may seek other means to fulfill its\\nobligations.” Id. ¶ 125. “The agency or applicant is free to\\nrefuse just as it may refuse to pay for an archeological\\nconsultant, but the agency still retains the duties of obtaining\\nthe necessary information for the identification [and\\nevaluation] of historic properties . . . through reasonable\\nmeans.” Id. (quoting Advisory Council, Section 106\\nHandbook, J.A. 1015).\\n\\f 31\\n Third, the Order shortened from 60 to 45 days the timeline\\nfor Tribes to respond to notifications on the Tower\\nConstruction Notification System, eliminated the requirement\\nthat applicants make a second attempt to contact Tribes, and\\nshortened from 20 to 15 days the timeline for Tribal response\\nto Commission contact. Id. ¶¶ 110-11.\\n\\n Keetoowah and Blackfeet challenge those three changes as\\narbitrary and capricious and inconsistent with the NHPA.\\nKeetoowah complains that the Order “encourages applicants,\\nwhich have until this point voluntarily paid fees, to refuse\\npaying Tribes” upfront fees, Keetoowah Br. 37; that “FCC\\nimplementation goes far beyond the terms of the Order by\\nrefusing to even allow Tribes to request voluntary fees\\nthrough” the Tower Construction Notification System, id. at\\n37-38; that letting applications proceed where Tribes refuse to\\nparticipate without compensation or are not hired as\\nconsultants violates the Commission’s legal obligation to\\nconsult with Tribes, id. at 38; and that the shortened timelines\\nare unreasonable, id. at 40. Blackfeet asserts that the\\nCommission lacks “the authority to prohibit tribes from\\ncollecting fees” because only the Advisory Council may\\npromulgate regulations implementing Section 106. Blackfeet\\nBr. 16.\\n\\n None of those challenges is availing. The clarification that\\napplicants are not required to pay upfront fees is consistent with\\nthe Advisory Council’s preexisting guidance and does not\\nviolate the Commission’s duty to consult with Tribes. The\\nOrder permissibly authorizes applicants to contract with non-\\nTribal parties in the identification-and-evaluation phase\\nbecause it stipulates that contractors must be “properly\\nqualified,” which we understand does not authorize hiring\\nother contractors in any circumstance in which only Tribes are\\n\\f 32\\nqualified. Order ¶ 128. The shortened timeline for Tribal\\nresponse is reasonable and sufficiently explained.\\n\\n A. Upfront Fees\\n\\n The Order permissibly confirms that upfront fees for\\nTribes to comment on proposed deployments are voluntary.\\nUnchallenged Advisory Council regulations already make\\nclear that fees are voluntary, so the Order’s reiteration of the\\nsame point is not arbitrary and capricious. While applicants\\nhave apparently been uniformly paying upfront fees for Section\\n106 review, no party asserts that they have been required to do\\nso. See Keetoowah Reply Br. 20. The Advisory Council has\\nbeen explicit that no “portion of the NHPA or the [Advisory\\nCouncil’s] regulations require an agency or an applicant to pay\\nfor any form of tribal involvement.” Advisory Council, Section\\n106 Handbook, J.A. 1015; see also Advisory Council, Fees in\\nthe Section 106 Review Process, J.A. 913 (neither the NHPA\\nnor Advisory Council regulations “requires Federal agencies to\\npay for any aspect of tribal [or] other consulting party\\nparticipation in the Section 106 process”). Blackfeet’s\\ncomplaint that “[t]he FCC does not have the authority to\\nprohibit tribes from collecting fees” and that the Order is\\nimpermissibly “implementing and administering Section 106\\nthrough regulation” is misplaced. The challenged Order\\ncontains no such prohibition, but does no more than recognize\\nand reiterate the Advisory Council’s existing rule.\\n\\n The Commission has a non-delegable duty to consult with\\nTribes about the effect of federal undertakings on property\\nsignificant to the Tribes, which Tribes can invoke or waive as\\nthey choose. The NHPA mandates that, “[i]n carrying out its\\nresponsibilities under [Section 106], a Federal agency shall\\nconsult with any Indian tribe . . . that attaches religious and\\ncultural significance to property.” 54 U.S.C. § 302706(b). The\\n\\f 33\\nAdvisory Council has explained that “federal agencies cannot\\nunilaterally delegate their tribal consultation responsibilities to\\nan applicant,” but can only delegate if “expressly authorized by\\nthe Indian tribe to do so.” Advisory Council, Limitations on\\nthe Delegation of Authority by Federal Agencies to Initiate\\nTribal Consultation under Section 106 of the National Historic\\nPreservation Act (Limitations on Section 106 Delegation) 1\\n(2011), https://go.usa.gov/xyWGq. The Commission has also\\nrecognized that its “fiduciary responsibility and duty of\\nconsultation [to Tribes] rest with the Commission as an agency\\nof the federal government, not with licensees, applicants, or\\nother third parties.” Section 106 Agreement, 20 FCC Rcd. at\\n1106 ¶ 91.\\n\\n Keetoowah says its challenge is not to the “FCC’s\\nclarification that fees are voluntary,” but to “the Order’s\\ndetermination that FCC will process applications without tribal\\ninput if tribes insist on charging applicants for their reviews.”\\nKeetoowah Reply Br. 19-20. That determination, Keetoowah\\nasserts, violates the Commission’s “statutory obligation to\\nconsult with tribes.” Id. at 19. Under the Section 106\\nAgreement, Tribes can and do permit applicants to contact them\\nto request review of proposed construction—essentially\\nagreeing to accept that contact in satisfaction of the\\nCommission’s responsibility to consult with Tribes directly.\\n20 FCC Rcd. at 1108 ¶ 96; see also Keetoowah Br. 37;\\nComment of the Seminole Tribe of Florida, J.A. 743 (“[T]ribes\\nparticipate in review . . . on a voluntary basis” as a substitute\\nfor “direct Section 106 consultation with the FCC.”) But\\nTribes can request “the federal agency to reenter the\\nconsultation process at any time . . . since the federal agency\\nremains responsible for government-to-government\\nconsultation.” Limitations on Section 106 Delegation 2.\\nKeetoowah implies that Tribes have only agreed to accept\\ndirect contact from applicants under the condition that\\n\\f 34\\napplicants pay for Tribes’ responses—meaning that if Tribes\\nrefuse to respond without being paid upfront fees, they will not\\nhave waived the Commission’s responsibility to consult with\\nthem directly. Without having fulfilled its legal obligation to\\nconsult, Keetoowah contends, the Commission cannot permit\\napplicants to go ahead with construction.\\n\\n Keetoowah overlooks the fact that when a Tribe refuses to\\nreview an application without being paid, the Order requires\\nthe Commission to step in to ask the Tribe for a response before\\nallowing applicants to construct. Tribes’ refusal to respond\\ntriggers a process in which applicants can refer the matter to\\nthe Commission, the Commission must contact Tribes directly,\\nand Tribes have 15 days from Commission contact to respond.\\nSee Order ¶ 111. Only if the Tribe does not timely respond to\\nthe Commission are “the applicant’s pre-construction\\nobligations . . . discharged with respect to that Tribal Nation.”\\nId. The Tribe is guaranteed the opportunity to consult as a\\nsovereign—a capacity in which it need not be paid—and the\\nCommission cannot force an unwilling Tribe to respond.\\nTherefore, if a Tribe refuses to respond when the Commission\\nrequests its views on an application, the Commission has\\ndischarged its obligation of direct Commission-to-Tribe\\nconsultation. See id. ¶ 111. Apart from the shortened\\ntimeframe, discussed below, Keetoowah has not offered any\\nreason the Commission’s contacting Tribes directly with a\\nrequest to consult that the Tribe rejects does not satisfy the\\nCommission’s consultation obligation.\\n\\n Finally, the objection that the Commission is prohibiting\\nTribes from requesting voluntary fees on the Tower\\nConstruction Notification System, Keetoowah Br. 38-40, is not\\nproperly before us. That prohibition does not appear in the\\nOrder itself but seems to originate with a later decision of\\nCommission staff. See Resp’t Br. 64 n.19.\\n\\f 35\\n B. Non-Tribal Consultation\\n\\n The Order states that applicants need not contract with\\nTribes to identify which properties have historic or cultural\\nsignificance to Tribes and determine how to assess or mitigate\\nadverse effects of construction. Order ¶¶ 124-25, 128-29.\\nKeetoowah argues that allowing applicants to contract with\\nnon-Tribal parties is arbitrary and capricious because “only\\nTribes are qualified to perform” such services “based on their\\nunique, often sacred, knowledge.” Keetoowah Br. 23. Because\\nthe Order stipulates that contractors must be “properly\\nqualified,” we reject the arbitrary-and-capricious claim. Order\\n¶ 128.\\n\\n Advisory Council regulations require the agency to “make\\na reasonable and good faith effort to carry out appropriate\\nidentification efforts” under Section 106. 36 C.F.R.\\n§ 800.4(b)(1). The Order explains that “the applicant is not\\nbound to any particular method of gathering information,”\\nOrder ¶ 125, but it stipulates that contractors must be “properly\\nqualified,” id. ¶ 128. The “reasonable and good faith efforts”\\nstandard together with the Order’s mandate that parties be\\n“properly qualified” may sometimes require applicants to hire\\nTribes—for instance, where Tribes have “unique” and “sacred”\\nknowledge of historic properties. Advisory Council guidance\\nsupports that notion, explaining that “unless an archeologist\\nhas been specifically authorized by a tribe to speak on its behalf\\non the subject, it should not be assumed that the archaeologist\\npossesses the appropriate expertise to determine what\\nproperties are or are not of significance to an Indian tribe.”\\nSection 106 Handbook, J.A. 1022. The Order itself suggests\\nthat applicants should try to hire Tribes first: “[I]f an applicant\\nasks a Tribal Nation” to perform identification and evaluation\\nof historic properties, “the applicant should expect to negotiate\\na fee for that work,” but if the Tribe and applicant “are unable\\n\\f 36\\nto agree on a fee, the applicant may seek other means to fulfill\\nits obligations.” Order ¶ 125. We cannot say, ex ante, how\\noften as a practical matter applicants might find qualified non-\\nTribal contractors or whether, as applied, the law will\\nordinarily require hiring Tribes. If a Tribe believes an\\napplicant has hired an unqualified contractor, that issue can be\\nlitigated when it arises.\\n\\n C. Timeline Changes\\n\\n Keetoowah’s one-paragraph challenge to the Order’s\\nshortening the timeline for Tribal response to Tower\\nConstruction Notification System notifications provides no\\nbasis on which to hold the shortened timeline arbitrary and\\ncapricious. Keetoowah Br. 40. Its sole objection is that Tribes\\n“operate with limited staff and budget, making the shortening\\nof Tribal review time unreasonable.” Id. The Commission\\nacted within its discretion and “considered the relevant factors\\nand articulate[d] a rational connection between the facts found\\nand the choice made.” Vonage Holdings Corp. v. FCC, 489\\nF.3d 1232, 1241 (D.C. Cir. 2007) (quoting BellSouth\\nTelecomms., Inc. v. FCC, 469 F.3d 1052, 1056 (D.C. Cir. 2006)\\n(alteration in original)). It reasonably justified the decision as\\na compromise between industry requests for even shorter\\ntimelines to address delays, and Tribes’ need for adequate time\\nto review submissions. See Order ¶¶ 112 n.262, 113.\\n\\n III. Promulgation of the Order Itself\\n\\n All petitioners argue that the promulgation of the Order\\nitself violated the law. Keetoowah and Blackfeet argue that the\\nCommission violated its duty to consult with Tribes, as\\nestablished by the Tribes’ sovereign status and the government-\\nto-government relationship recognized in Article I, Section 8\\nof the Constitution, the NHPA, and the Commission’s\\nregulations. See Keetoowah Br. 40-42; Blackfeet Br. 20-21.\\n\\f 37\\nThe NRDC argues that the Order itself was a major federal\\naction that required NEPA review. See NRDC Br. 10-11.\\nBecause the Order documents extensive consultation with\\nTribes, we reject the first contention. We lack jurisdiction to\\nconsider the second because the NRDC forfeited it by failing\\nto raise it to the Commission.\\n\\n As for the Tribes’ contention that the Order is invalid\\nbecause the Commission did not meet its obligations to consult\\nwith Tribes, the Commission responds that it extensively\\nconsulted with Tribes, and that in any event its consultation\\nobligation is not judicially enforceable. Resp’t Br. 69-74. We\\nconclude that the Commission fulfilled its obligation to\\nconsult. The Commission presented abundant evidence that it\\n“consulted” Tribes in the ordinary sense of the word, and the\\nTribes have offered no other concrete standard by which to\\njudge the Commission’s efforts.\\n\\n On this record, we cannot say that the Commission failed\\nto consult with Tribes in its meetings and other\\ncommunications, which began in 2016 and continued through\\nearly 2018. See Order ¶¶ 19, 34. The Commission\\ndocumented extensive meetings it held with Tribes before it\\nissued the Order. See Order ¶¶ 19-35. Under Advisory\\nCouncil regulations, “[c]onsultation means the process of\\nseeking, discussing, and considering the views of other\\nparticipants, and, where feasible, seeking agreement with them\\nregarding matters arising in the section 106 process.” 36\\nC.F.R. § 800.16(f); see also 54 U.S.C. § 302706(b). The\\ndictionary definition of consulting is “seek[ing] advice or\\ninformation of.” Consult, American Heritage Dict. (5th ed.\\n2019). Keetoowah complains that the FCC’s efforts were\\n“listening sessions, briefings, conference calls, and delivery of\\nremarks by a Commissioner” rather than “consultations,” and\\npresents evidence that Tribes did not view these meetings as\\n\\f 38\\nconsultations. Keetoowah Br. 44. But it offers no standard by\\nwhich to judge which consultations were “listening sessions”\\nor whether a “listening session” or a conference call qualifies\\nas a consultation. The only case Keetoowah cites interpreting\\nan agency’s failure to consult is inapposite: there, an agency\\nofficial “acknowledged at trial” that the contested decision\\n“had already been made prior to” the first meeting between\\nTribal members and agency officials discussing the decision.\\nOglala Sioux Tribe of Indians v. Andrus, 603 F.2d 707, 710\\n(8th Cir. 1979). No evidence in this record suggests the\\nCommission had already determined the Order’s substance\\nbefore meeting with Tribes—and the series of communications\\nand meeting commenced even before the Commission issued\\nthe Notice of Proposed Rulemaking. See Order ¶ 19. The\\nCommission appeared to “seek[], discuss[], and consider[] the\\nviews of” the Tribes, even if it did not ultimately adopt those\\nviews.\\n\\n The NRDC argues that promulgating the Order was itself\\na major federal action that required NEPA review. See NRDC\\nBr. 10-11. But, as intervenor CTIA points out, the NRDC\\nforfeited that argument by failing to make it to the\\nCommission, see CTIA Br. 38, and we lack jurisdiction to\\nreview a claim that was not raised there. Free Access & Broad.\\nTelemedia, LLC v. FCC, 865 F.3d 615, 619 (D.C. Cir. 2017).\\nWhile the NRDC points to its own and others’ comments\\n“urg[ing] the Commission to conduct a NEPA analysis,”\\nNRDC Reply Br. 3, none of those comments said the\\nCommission was required to perform a NEPA analysis of the\\nOrder. The NRDC cites its own comment “that if the FCC\\nsought to exclude an entire category of wireless facilities from\\nNEPA, it was required to establish a categorical exclusion.” Id.\\n(citing J.A. 787-90). But the NRDC did not there contend, as\\nit now does, that the Order is a major federal action. Rather,\\nthe NRDC’s argument was that the federal character of the\\n\\f 39\\ngeographic area license meant that the Commission could not\\nentirely exempt wireless facility construction from NEPA\\nreview, J.A. 790—the same statutory argument it made here—\\nand that the proper approach to exempting federal “activities\\nthat by their nature do not have significant impacts on the\\nenvironment is with a categorical exclusion,” J.A. 789.\\nWhether the licenses or construction are federal, the basis of\\nthe NRDC’s argument, is irrelevant to the question whether the\\nOrder overall is a major federal action that requires NEPA\\nreview. One of the other two comments it cites asserted that\\nthe proposed rule failed to comply with NEPA, but again, not\\nbecause the Order required NEPA analysis—rather because\\nthe issuance of licenses constitutes a major federal action. See\\nComment of the Nat’l Trust for Historic Pres., J.A. 770. The\\nthird comment urged the Commission to consider the\\ncumulative effects of radiofrequency exposure, but did not\\neven mention NEPA. See Comment of BioInitiative Working\\nGrp., J.A. 235-38. The argument that the Order required\\nindependent NEPA review was never fairly before the\\nCommission.\\n\\n CONCLUSION\\n\\n We grant the petitions to vacate the Order’s removal of\\nsmall cells from its limited approval authority and remand to\\nthe FCC. We deny the petitions to vacate the Order’s changes\\nto Tribal involvement in Section 106 review and to vacate the\\nOrder in its entirety.\\n\\n So ordered.\"}]}"},"judge1":{"kind":"string","value":"TATEL"},"judge2":{"kind":"string","value":"PILLARD"},"judge3":{"kind":"string","value":"EDWARDS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":1,"string":"1"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4647052/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: UNITED KEETOOWAH BAND OF CHEROKEE INDIANS IN OKLAHOMA, Individually and on Behalf of All Other Native American Indian Tribes and Tribal Organizations, Et Al., Petitioners v. FEDERAL COMMUNICATIONS COMMISSION and United States of America, Respondents National Association of Tribal Historic Preservation Officers, Et Al., Intervenors\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-09\nOpinion type: majority\nHas dissenting opinion: Yes\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued March 15, 2019 Decided August 9, 2019\n\n No. 18-1129\n\n UNITED KEETOOWAH BAND OF CHEROKEE INDIANS IN\n OKLAHOMA, INDIVIDUALLY AND ON BEHALF OF ALL OTHER\n NATIVE AMERICAN INDIAN TRIBES AND TRIBAL\n ORGANIZATIONS, ET AL.,\n PETITIONERS\n\n v.\n\n FEDERAL COMMUNICATIONS COMMISSION AND UNITED\n STATES OF AMERICA,\n RESPONDENTS\n\nNATIONAL ASSOCIATION OF TRIBAL HISTORIC PRESERVATION\n OFFICERS, ET AL.,\n INTERVENORS\n\n\n Consolidated with 18-1135, 18-1148, 18-1159, 18-1184\n\n\n On Petitions for Review of an Order of\n the Federal Communications Commission\n\n\n Stephen Díaz Gavin argued the cause for petitioners\nUnited Keetoowah Band of Cherokee Indians in Oklahoma, et\nal., and supporting intervenors. With him on the briefs were J.\nScott Sypolt, Joel D. Bertocchi, Joseph H. Webster, F. Michael\n\f 2\nWillis, Andrew Jay Schwartzman, James T. Graves, and\nElizabeth S. Merritt. Angela J. Campbell entered an\nappearance.\n\n Sharon Buccino argued the cause for petitioner Natural\nResources Defense Council and intervenor Edward B. Myers.\nWith her on the briefs was Edward B. Myers.\n\n Natalie A. Landreth argued the cause for petitioners\nBlackfeet Tribe, et al. With her on the briefs were Wesley J.\nFurlong, Joel West Williams, Troy A. Eid, Jennifer H. Weddle,\nand Heather D. Thompson.\n\n Jacob M. Lewis, Associate General Counsel, Federal\nCommunications Commission, argued the cause for\nrespondents. With him on the brief were Jeffrey Bossert Clark,\nAssistant Attorney General, U.S. Department of Justice, Eric\nA. Grant, Deputy Assistant Attorney General, Andrew C.\nMergen and Allen M. Brabender, Attorneys, Thomas M.\nJohnson Jr., General Counsel, Federal Communications\nCommission, David M. Gossett, Deputy General Counsel, and\nC. Grey Pash Jr., Counsel. Jonathan H. Laskin and Robert B.\nNicholson, Attorneys, U.S. Department of Justice, and Richard\nK. Welch, Deputy Associate General Counsel, Federal\nCommunications Commission, entered appearances.\n\n Joshua Turner argued the cause for intervenors in support\nof respondents. With him on the brief were Christopher J.\nWright and E. Austin Bonner.\n\n Before: TATEL and PILLARD, Circuit Judges, and\nEDWARDS, Senior Circuit Judge.\n\n Opinion for the Court filed by Circuit Judge PILLARD.\n\f 3\n PILLARD, Circuit Judge: Cellular wireless services,\nincluding telephone and other forms of wireless data\ntransmission, depend on facilities that transmit their radio\nsignals on bands of electromagnetic spectrum. The Federal\nCommunications Commission (FCC or Commission) has\nexclusive control over the spectrum, and wireless providers\nmust obtain licenses from the FCC to transmit. Wireless\nservice in the United States has mostly depended on large,\n“macrocell” radio towers to transmit cell signal, but companies\noffering the next generation of wireless service—known as\n5G—are in the process of shifting to transmission via hundreds\nof thousands of densely spaced small wireless facilities, or\n“small cells.” As part of an effort to expedite the rollout of 5G\nservice, the Commission has removed some regulatory\nrequirements for the construction of wireless facilities. These\npetitions challenge one of the FCC’s orders paring back such\nregulations, In re Accelerating Wireless Broadband\nDeployment by Removing Barriers to Infrastructure\nInvestment (Second Report & Order) (Order), FCC 18-30,\n2018 WL 1559856 (F.C.C.) (Mar. 30, 2018).\n\n The Order exempted most small cell construction from\ntwo kinds of previously required review: historic-preservation\nreview under the National Historic Preservation Act (NHPA)\nand environmental review under the National Environmental\nPolicy Act (NEPA). Together, these reviews assess the effects\nof new construction on, among other things, sites of religious\nand cultural importance to federally recognized Indian Tribes.\nThe Order also effectively reduced Tribes’ role in reviewing\nproposed construction of macrocell towers and other wireless\nfacilities that remain subject to cultural and environmental\nreview.\n\n Three groups of petitioners challenge the Order as\nviolating the NHPA, NEPA, and the Administrative Procedure\n\f 4\nAct on several grounds: that its elimination of historic-\npreservation and environmental review of small cell\nconstruction was arbitrary and capricious, an unjustified policy\nreversal, and contrary to the NHPA and NEPA; that the\nchanges to Tribes’ role in reviewing new construction was\narbitrary and capricious; that the Commission arbitrarily and\ncapriciously failed to engage in meaningful consultations with\nTribes in promulgating the Order; and that the Order itself\nrequired NEPA review.\n\n We grant in part the petitions for review because the Order\ndoes not justify the Commission’s determination that it was not\nin the public interest to require review of small cell\ndeployments. In particular, the Commission failed to justify its\nconfidence that small cell deployments pose little to no\ncognizable religious, cultural, or environmental risk,\nparticularly given the vast number of proposed deployments\nand the reality that the Order will principally affect small cells\nthat require new construction. The Commission accordingly\ndid not, pursuant to its public interest authority, 47 U.S.C.\n§ 319(d), adequately address possible harms of deregulation\nand benefits of environmental and historic-preservation\nreview. The Order’s deregulation of small cells is thus\narbitrary and capricious. We do not reach the alternative\nobjections to the elimination of review on small cell\nconstruction. We deny the petitions for review on the\nremaining grounds.\n\n BACKGROUND\n\n I. Statutory and Regulatory Background\n\n A. National Historic Preservation Act (NHPA)\n\n Congress enacted the NHPA to “foster conditions under\nwhich our modern society and our historic property can exist\n\f 5\nin productive harmony” and “contribute to the preservation of\nnonfederally owned historic property and give maximum\nencouragement to organizations and individuals undertaking\npreservation by private means.” 54 U.S.C. § 300101(1), (4).\nAs part of that mission, NHPA’s Section 106 requires federal\nagencies to “take into account the effect of” their\n“undertaking[s] on any historic property.” Id. § 306108.\n\n Both “historic property” and “undertaking” have specific\nmeanings under the statute. Historic properties include myriad\nmonuments, buildings, and sites of historic importance,\nincluding “[p]roperty of traditional religious and cultural\nimportance to an Indian tribe.” Id. §§ 302706, 300308. Insofar\nas Tribal heritage is concerned, the Section 106 process\nrequires federal agencies to “consult with any Indian tribe . . .\nthat attaches religious and cultural significance to” a historic\nproperty potentially affected by a federal undertaking. Id.\n§§ 302706, 306102. To count as “historic,” such properties\nneed not be on Tribal land; in fact, they “are commonly located\noutside Tribal lands and may include Tribal burial grounds,\nland vistas, and other sites that Tribal Nations . . . regard as\nsacred or otherwise culturally significant.” Order ¶ 97. Only\na federal “undertaking,” not a state or purely private one,\ntriggers the Section 106 Tribal consultation process. A federal\n“undertaking,” as relevant here, is “a project, activity, or\nprogram funded in whole or in part under the direct or indirect\njurisdiction of a Federal agency, including . . . those requiring\na Federal permit, license, or approval.” 54 U.S.C. § 300320.\nWe have construed the statute to mean that, for an action to be\na federal undertaking, “only a ‘Federal permit, license or\napproval’ is required,” not necessarily federal funding. CTIA-\nWireless Ass’n v. FCC, 466 F.3d 105, 112 (D.C. Cir. 2006).\n\n The Section 106 process requires that an agency “consider\nthe impacts of its undertaking” and consult various parties, not\n\f 6\nthat it necessarily “engage in any particular preservation\nactivities.” Id. at 107 (quoting Davis v. Latschar, 202 F.3d 359,\n370 (D.C. Cir. 2000)). The NHPA established an independent\nagency, the Advisory Council on Historic Preservation\n(Advisory Council), 54 U.S.C. § 304101, which is responsible\nfor promulgating regulations “to govern the implementation\nof” Section 106, id. § 304108(a). Agencies must consult with\nthe Advisory Council, State Historic Preservation Officers, and\nTribal Historic Preservation Officers, the last of which adopt\nthe responsibilities of State Historic Preservation Officers on\nTribal lands. 54 U.S.C. §§ 302303, 302702; 36 C.F.R.\n§§ 800.3(c), 800.16(v)-(w) (defining State and Tribal Historic\nPreservation Officers).\n\n The Advisory Council’s regulations authorize the use of\nalternatives to the ordinary Section 106 procedures, called\n“programmatic agreements.” 36 C.F.R. § 800.14(b). The\nCommission develops programmatic agreements in\nconsultation with the Advisory Council, Tribes, and other\ninterested parties, “to govern the implementation of a particular\nprogram or the resolution of adverse effects from certain\ncomplex project situations or multiple undertakings” in certain\ncircumstances, such as when “effects on historic properties are\nsimilar and repetitive” or “effects on historic properties cannot\nbe fully determined prior to approval of an undertaking.” Id.\n§ 800.14(1)(i)-(ii). Tribes’ views must be taken into account\nwhere the agreement “has the potential to affect historic\nproperties on tribal lands or historic properties of religious and\ncultural significance to an Indian tribe.” Id. § 800.14(b)(1)(i),\n(f). For instance, the Commission has consulted with Tribes to\nuse programmatic agreements to exclude from individualized\nreview entire categories of undertakings that are unlikely to\naffect historic properties. See In re Nationwide Programmatic\nAgreement Regarding the Section 106 [NHPA] Review Process\n(Section 106 Agreement), 20 FCC Rcd. 1073, 1075 ¶ 2 (2004).\n\f 7\n B. National Environmental Policy Act (NEPA)\n\n Congress enacted NEPA to “encourage productive and\nenjoyable harmony between man and his environment” and\n“promote efforts which will prevent or eliminate damage to the\nenvironment and biosphere and stimulate the health and\nwelfare of man,” among other purposes. 42 U.S.C. § 4321.\nLike the NHPA, NEPA mandates a review process that “does\nnot dictate particular decisional outcomes, but ‘merely\nprohibits uninformed—rather than unwise—agency action.’”\nSierra Club v. U.S. Army Corps of Eng’rs, 803 F.3d 31, 37\n(D.C. Cir. 2015) (quoting Robertson v. Methow Valley Citizens\nCouncil, 490 U.S. 332, 351 (1989)).\n\n All “major Federal actions significantly affecting the\nquality of the human environment” trigger environmental\nreview under NEPA, just as federal “undertakings” trigger\nhistoric preservation review under the NHPA. 42 U.S.C.\n§ 4332(C). Major federal actions “include[] actions . . . which\nare potentially subject to Federal control and responsibility.”\n40 C.F.R. § 1508.18. Under the Commission’s procedures\nimplementing NEPA, if an action may significantly affect the\nenvironment, applicants must conduct a preliminary\nEnvironmental Assessment to help the Commission determine\nwhether “the proposal will have a significant environmental\nimpact upon the quality of the human environment,” and so\nperhaps necessitate a more detailed Environmental Impact\nStatement. 47 C.F.R. § 1.1308; see also 40 C.F.R. § 1508.9.\nIf, after reviewing the Environmental Assessment, the\nCommission determines that the action will not have a\nsignificant environmental impact, it will make a “finding of no\nsignificant impact” and process the application “without\nfurther documentation of environmental effect.” 47 C.F.R.\n§ 1.1308(d).\n\f 8\n NEPA also has an analogue to the NHPA’s Advisory\nCouncil. In enacting NEPA, Congress established the Council\non Environmental Quality, in the Executive Office of the\nPresident, to oversee implementation of NEPA across the\nentire federal government. 42 U.S.C. §§ 4342, 4344. With the\nendorsement of the Council on Environmental Quality and by\nfollowing a series of mandated procedures, agencies can\nestablish “categorical exclusions” for federal actions that\nrequire neither an Environmental Assessment nor an\nEnvironmental Impact Statement. 40 C.F.R. § 1508.4.\nCategorical exclusions are appropriate for “a category of\nactions which do not individually or cumulatively have a\nsignificant effect on the human environment and which have\nbeen found to have no such effect in procedures adopted by a\nFederal agency.” Id. “Categorical exclusions are not\nexemptions or waivers of NEPA review; they are simply one\ntype of NEPA review.” Council on Environmental Quality,\nMemorandum for Heads of Federal Dep’ts and Agencies:\nEstablishing, Applying & Revising Categorical Exclusions\nunder [NEPA] (Categorical Exclusion Memo) 2 (2010).\n\n C. Legal Framework for Wireless Infrastructure\n\n The Communications Act of 1934 established the FCC to\nmake available a “rapid, efficient . . . wire and radio\ncommunication service with adequate facilities at reasonable\ncharges.” 47 U.S.C. § 151. In licensing use of the spectrum,\nthe Commission is tasked with promoting “the development\nand rapid deployment of new technologies, products and\nservices for the benefit of the public . . . without administrative\nor judicial delays,” id. § 309, and “maintain[ing] the control of\nthe United States over all the channels of radio transmission,”\nid. § 301.\n\f 9\n The Commission generally does not require construction\npermits before private parties can build wireless facilities.\nCongress largely eliminated the FCC’s site-specific\nconstruction permits in 1982, and the Commission has since\nrequired construction permits only where it finds that the public\ninterest would be served by such permitting. See Pub. L. 97-\n259, 96 Stat. 1087, § 119 (1982) (codified at 47 U.S.C.\n§ 319(d)). It has not made such a finding for the wireless\nfacilities at issue here.\n\n The FCC does, however, require licensing of the spectrum\nused by wireless small cells. It does so by issuing geographic\narea licenses, which allow wireless providers to operate on\ncertain frequency bands in a wide geographic area. See 47\nU.S.C. § 309(j). Those licenses authorize using spectrum\nrather than building wireless facilities, but they necessarily\ncontemplate facility construction. They have coverage\nrequirements—for instance, one type of geographic area\nlicense required licensees to provide service to at least 40% of\nthe population in their geographic service area by June 2013.\nSee 47 C.F.R. § 27.14(h). If they fail to meet the coverage\nrequirements, they can be stripped of authority to operate for\nthe license’s full term or serve part of its geographic area, and\nthey “may be subject to enforcement action, including\nforfeitures.” Id. The Commission also exercises continuing\nauthority to inspect radio installations to ascertain their\ncompliance with any and all applicable laws, whether or not the\nlicensee itself constructed those installations. See 47 U.S.C.\n§ 303(n); 47 C.F.R. § 1.9020(c)(5).\n\n The Commission has not identified any period since the\nenactment of the NHPA (in 1966) and NEPA (in 1970) when\nit did not require historic-preservation and environmental\nreview of wireless facilities. After Congress eliminated the\nconstruction permit requirement, the Commission for a time\n\f 10\nrequired NEPA and NHPA review of facilities before it granted\ntheir service licenses. See, e.g., In re Amendment of Envtl.\nRules in Response to New Regulations Issued by [CEQ], FCC\n85-626, 1986 WL 292182, at *5 ¶ 18 (F.C.C.) (Mar. 26, 1986)\n(requiring review “during the period prior to grant of a station\nlicense”); id. at *8 App’x ¶ 7 (requiring NEPA review on\n“[f]acilities that will affect districts, sites, buildings, structures\nor objects . . . that are listed in the National Register of Historic\nPlaces or are eligible for listing,” which includes property of\nreligious or cultural significance to Indian Tribes, 54 U.S.C.\n§ 302706(a)). In 1990, the Commission shifted review from\nthe licensing stage to the construction stage by establishing a\n“limited approval authority” over construction of wireless\nfacilities. In re Amendment of Envtl. Rules (1990 Order), 5\nFCC Rcd. 2942 (1990). Limited approval authority required\nthat, “where construction of a Commission-regulated radio\ncommunications facility is permitted without prior\nCommission authorization (i.e., without a construction permit),\nthe licensee must nonetheless comply with historic\npreservation and environmental review procedures.” Order\n¶ 51; see also 47 C.F.R. § 1.1312. The authority was “limited”\nin that it allowed “the Commission [to] exercise[] control over\ndeployment solely to conduct federal historic and\nenvironmental review.” Resp’t Br. 12. The Commission\nemphasized that shifting review to the pre-construction stage\nserved a practical function: Before it had established its limited\napproval authority, the FCC’s rules “provide[d] that any\nrequired submission of [Environmental Assessments] and any\nrequired Commission environmental review take place at the\nlicensing stage rather than prior to construction,” with the result\nthat “[a]pplicants who ha[d] already constructed their\nfacilities” could “subsequently be denied licenses on\nenvironmental grounds.” 1990 Order 2942 ¶ 3. The\nCommission explained that it continued to require review “to\nensure that the Commission fully complies with Federal\n\f 11\nenvironmental laws in connection with facilities that do not\nrequire pre-construction authorization.” Id. ¶ 4. It announced\nthe changes as “necessary to ensure that the Commission\naddresses environmental issues early enough in the licensing\nprocess to ensure that it fully meets its obligations under\nFederal environmental laws,” including NEPA and the NHPA.\nId. at 2943 ¶ 9 & n.16.\n\n The Commission has never required individualized review\nof each separate facility, however. A long series of regulations,\nprogrammatic agreements, and categorical exclusions has\naggregated facilities for joint consideration and focused NHPA\nand NEPA review on those deployments most likely to have\ncultural or environmental effects. For instance, most\ncollocations—deployments on existing structures—are\nexcluded from individualized review under NHPA\nprogrammatic agreements and NEPA categorical exclusions.\nSee In re Implementation of the National Environmental Policy\nAct of 1969 (Implementation of NEPA), 49 F.C.C.2d 1313,\n1319-20 (1974); Nationwide Programmatic Agreement for the\nCollocation of Wireless Antennas (Collocation Agreement), 47\nC.F.R. pt.1, app. B (2001); Section 106 Agreement, 20 FCC\nRcd. at 1075 ¶ 2; Nationwide Programmatic Agreement for\nReview Under the National Historic Preservation Act, 70 Fed.\nReg. 556 (2005); In re Acceleration of Broadband Deployment\nby Improving Wireless Facilities Siting Policies (Improving\nWireless Facilities Siting Policies), 29 FCC Rcd. 12865, 12870\n¶ 11 (2014); 47 C.F.R. § 1.1320(b)(4). Categorical exclusions\ngo through notice and comment, 40 C.F.R. § 1507.3; include\nimpact findings, Categorical Exclusion Memo 9; require the\nCouncil on Environmental Quality to approve them as\nconsistent with its regulations and NEPA, 40 C.F.R.\n§ 1507.3(a); and reserve rights to interested parties to request\nfurther review in the event that atypical adverse effects do\noccur, 47 C.F.R. § 1.1307(c), (d). At the same time, they\n\f 12\nachieve enormous efficiencies in the review processes for\nclasses of actions or undertakings anticipated to have minimal\nor no adverse cultural or environmental effects.\n\n Since 2004, the FCC has been conducting NHPA review\nin accordance with a broad programmatic agreement, the\nSection 106 Agreement, 20 FCC Rcd. 1073. Interested parties\ndeveloped that agreement to “tailor the Section 106 review in\nthe communications context in order to improve compliance\nand streamline the review process for construction of towers\nand other Commission undertakings, while at the same time\nadvancing and preserving the goal of the NHPA to protect\nhistoric properties, including historic properties to which\nfederally recognized Indian tribes . . . attach religious and\ncultural significance.” Id. at 1074-75 ¶ 1. In the Section 106\nAgreement, the Commission adopted “procedures for\nparticipation of federally recognized Indian tribes,” among\nother changes. Id. at 1075 ¶ 2. It also formalized the use of the\nelectronic Tower Construction Notification System, which\nnotifies Tribes of proposed wireless construction in areas they\nhave identified as containing properties of religious and\ncultural significance, and allows them to give applicants\ninformation on the potential effects of proposed construction.\nId. at 1106-10 ¶¶ 89-100.\n\n II. Order Under Review\n\n The challenged Order eliminated NHPA and NEPA\nreview on small cells that meet certain size and other\nspecifications, based on the Commission’s conclusion that\nsuch review was not statutorily required and would impede the\nadvance of 5G networks, and that its costs outweighed any\nbenefits. See Order ¶¶ 36-45. The Order also altered Tribal\ninvolvement in those Section 106 reviews that are still\nconducted on wireless facilities that were not encompassed in\n\f 13\nthe small cell exemption. See id. ¶¶ 96-130. Two of the five\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: UNITED KEETOOWAH BAND OF CHEROKEE INDIANS IN OKLAHOMA, Individually and on Behalf of All Other Native American Indian Tribes and Tribal Organizations, Et Al., Petitioners v. FEDERAL COMMUNICATIONS COMMISSION and United States of America, Respondents National Association of Tribal Historic Preservation Officers, Et Al., Intervenors\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-09\\nOpinion type: majority\\nHas dissenting opinion: Yes\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued March 15, 2019 Decided August 9, 2019\\n\\n No. 18-1129\\n\\n UNITED KEETOOWAH BAND OF CHEROKEE INDIANS IN\\n OKLAHOMA, INDIVIDUALLY AND ON BEHALF OF ALL OTHER\\n NATIVE AMERICAN INDIAN TRIBES AND TRIBAL\\n ORGANIZATIONS, ET AL.,\\n PETITIONERS\\n\\n v.\\n\\n FEDERAL COMMUNICATIONS COMMISSION AND UNITED\\n STATES OF AMERICA,\\n RESPONDENTS\\n\\nNATIONAL ASSOCIATION OF TRIBAL HISTORIC PRESERVATION\\n OFFICERS, ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 18-1135, 18-1148, 18-1159, 18-1184\\n\\n\\n On Petitions for Review of an Order of\\n the Federal Communications Commission\\n\\n\\n Stephen Díaz Gavin argued the cause for petitioners\\nUnited Keetoowah Band of Cherokee Indians in Oklahoma, et\\nal., and supporting intervenors. With him on the briefs were J.\\nScott Sypolt, Joel D. Bertocchi, Joseph H. Webster, F. Michael\\n\\f 2\\nWillis, Andrew Jay Schwartzman, James T. Graves, and\\nElizabeth S. Merritt. Angela J. Campbell entered an\\nappearance.\\n\\n Sharon Buccino argued the cause for petitioner Natural\\nResources Defense Council and intervenor Edward B. Myers.\\nWith her on the briefs was Edward B. Myers.\\n\\n Natalie A. Landreth argued the cause for petitioners\\nBlackfeet Tribe, et al. With her on the briefs were Wesley J.\\nFurlong, Joel West Williams, Troy A. Eid, Jennifer H. Weddle,\\nand Heather D. Thompson.\\n\\n Jacob M. Lewis, Associate General Counsel, Federal\\nCommunications Commission, argued the cause for\\nrespondents. With him on the brief were Jeffrey Bossert Clark,\\nAssistant Attorney General, U.S. Department of Justice, Eric\\nA. Grant, Deputy Assistant Attorney General, Andrew C.\\nMergen and Allen M. Brabender, Attorneys, Thomas M.\\nJohnson Jr., General Counsel, Federal Communications\\nCommission, David M. Gossett, Deputy General Counsel, and\\nC. Grey Pash Jr., Counsel. Jonathan H. Laskin and Robert B.\\nNicholson, Attorneys, U.S. Department of Justice, and Richard\\nK. Welch, Deputy Associate General Counsel, Federal\\nCommunications Commission, entered appearances.\\n\\n Joshua Turner argued the cause for intervenors in support\\nof respondents. With him on the brief were Christopher J.\\nWright and E. Austin Bonner.\\n\\n Before: TATEL and PILLARD, Circuit Judges, and\\nEDWARDS, Senior Circuit Judge.\\n\\n Opinion for the Court filed by Circuit Judge PILLARD.\\n\\f 3\\n PILLARD, Circuit Judge: Cellular wireless services,\\nincluding telephone and other forms of wireless data\\ntransmission, depend on facilities that transmit their radio\\nsignals on bands of electromagnetic spectrum. The Federal\\nCommunications Commission (FCC or Commission) has\\nexclusive control over the spectrum, and wireless providers\\nmust obtain licenses from the FCC to transmit. Wireless\\nservice in the United States has mostly depended on large,\\n“macrocell” radio towers to transmit cell signal, but companies\\noffering the next generation of wireless service—known as\\n5G—are in the process of shifting to transmission via hundreds\\nof thousands of densely spaced small wireless facilities, or\\n“small cells.” As part of an effort to expedite the rollout of 5G\\nservice, the Commission has removed some regulatory\\nrequirements for the construction of wireless facilities. These\\npetitions challenge one of the FCC’s orders paring back such\\nregulations, In re Accelerating Wireless Broadband\\nDeployment by Removing Barriers to Infrastructure\\nInvestment (Second Report & Order) (Order), FCC 18-30,\\n2018 WL 1559856 (F.C.C.) (Mar. 30, 2018).\\n\\n The Order exempted most small cell construction from\\ntwo kinds of previously required review: historic-preservation\\nreview under the National Historic Preservation Act (NHPA)\\nand environmental review under the National Environmental\\nPolicy Act (NEPA). Together, these reviews assess the effects\\nof new construction on, among other things, sites of religious\\nand cultural importance to federally recognized Indian Tribes.\\nThe Order also effectively reduced Tribes’ role in reviewing\\nproposed construction of macrocell towers and other wireless\\nfacilities that remain subject to cultural and environmental\\nreview.\\n\\n Three groups of petitioners challenge the Order as\\nviolating the NHPA, NEPA, and the Administrative Procedure\\n\\f 4\\nAct on several grounds: that its elimination of historic-\\npreservation and environmental review of small cell\\nconstruction was arbitrary and capricious, an unjustified policy\\nreversal, and contrary to the NHPA and NEPA; that the\\nchanges to Tribes’ role in reviewing new construction was\\narbitrary and capricious; that the Commission arbitrarily and\\ncapriciously failed to engage in meaningful consultations with\\nTribes in promulgating the Order; and that the Order itself\\nrequired NEPA review.\\n\\n We grant in part the petitions for review because the Order\\ndoes not justify the Commission’s determination that it was not\\nin the public interest to require review of small cell\\ndeployments. In particular, the Commission failed to justify its\\nconfidence that small cell deployments pose little to no\\ncognizable religious, cultural, or environmental risk,\\nparticularly given the vast number of proposed deployments\\nand the reality that the Order will principally affect small cells\\nthat require new construction. The Commission accordingly\\ndid not, pursuant to its public interest authority, 47 U.S.C.\\n§ 319(d), adequately address possible harms of deregulation\\nand benefits of environmental and historic-preservation\\nreview. The Order’s deregulation of small cells is thus\\narbitrary and capricious. We do not reach the alternative\\nobjections to the elimination of review on small cell\\nconstruction. We deny the petitions for review on the\\nremaining grounds.\\n\\n BACKGROUND\\n\\n I. Statutory and Regulatory Background\\n\\n A. National Historic Preservation Act (NHPA)\\n\\n Congress enacted the NHPA to “foster conditions under\\nwhich our modern society and our historic property can exist\\n\\f 5\\nin productive harmony” and “contribute to the preservation of\\nnonfederally owned historic property and give maximum\\nencouragement to organizations and individuals undertaking\\npreservation by private means.” 54 U.S.C. § 300101(1), (4).\\nAs part of that mission, NHPA’s Section 106 requires federal\\nagencies to “take into account the effect of” their\\n“undertaking[s] on any historic property.” Id. § 306108.\\n\\n Both “historic property” and “undertaking” have specific\\nmeanings under the statute. Historic properties include myriad\\nmonuments, buildings, and sites of historic importance,\\nincluding “[p]roperty of traditional religious and cultural\\nimportance to an Indian tribe.” Id. §§ 302706, 300308. Insofar\\nas Tribal heritage is concerned, the Section 106 process\\nrequires federal agencies to “consult with any Indian tribe . . .\\nthat attaches religious and cultural significance to” a historic\\nproperty potentially affected by a federal undertaking. Id.\\n§§ 302706, 306102. To count as “historic,” such properties\\nneed not be on Tribal land; in fact, they “are commonly located\\noutside Tribal lands and may include Tribal burial grounds,\\nland vistas, and other sites that Tribal Nations . . . regard as\\nsacred or otherwise culturally significant.” Order ¶ 97. Only\\na federal “undertaking,” not a state or purely private one,\\ntriggers the Section 106 Tribal consultation process. A federal\\n“undertaking,” as relevant here, is “a project, activity, or\\nprogram funded in whole or in part under the direct or indirect\\njurisdiction of a Federal agency, including . . . those requiring\\na Federal permit, license, or approval.” 54 U.S.C. § 300320.\\nWe have construed the statute to mean that, for an action to be\\na federal undertaking, “only a ‘Federal permit, license or\\napproval’ is required,” not necessarily federal funding. CTIA-\\nWireless Ass’n v. FCC, 466 F.3d 105, 112 (D.C. Cir. 2006).\\n\\n The Section 106 process requires that an agency “consider\\nthe impacts of its undertaking” and consult various parties, not\\n\\f 6\\nthat it necessarily “engage in any particular preservation\\nactivities.” Id. at 107 (quoting Davis v. Latschar, 202 F.3d 359,\\n370 (D.C. Cir. 2000)). The NHPA established an independent\\nagency, the Advisory Council on Historic Preservation\\n(Advisory Council), 54 U.S.C. § 304101, which is responsible\\nfor promulgating regulations “to govern the implementation\\nof” Section 106, id. § 304108(a). Agencies must consult with\\nthe Advisory Council, State Historic Preservation Officers, and\\nTribal Historic Preservation Officers, the last of which adopt\\nthe responsibilities of State Historic Preservation Officers on\\nTribal lands. 54 U.S.C. §§ 302303, 302702; 36 C.F.R.\\n§§ 800.3(c), 800.16(v)-(w) (defining State and Tribal Historic\\nPreservation Officers).\\n\\n The Advisory Council’s regulations authorize the use of\\nalternatives to the ordinary Section 106 procedures, called\\n“programmatic agreements.” 36 C.F.R. § 800.14(b). The\\nCommission develops programmatic agreements in\\nconsultation with the Advisory Council, Tribes, and other\\ninterested parties, “to govern the implementation of a particular\\nprogram or the resolution of adverse effects from certain\\ncomplex project situations or multiple undertakings” in certain\\ncircumstances, such as when “effects on historic properties are\\nsimilar and repetitive” or “effects on historic properties cannot\\nbe fully determined prior to approval of an undertaking.” Id.\\n§ 800.14(1)(i)-(ii). Tribes’ views must be taken into account\\nwhere the agreement “has the potential to affect historic\\nproperties on tribal lands or historic properties of religious and\\ncultural significance to an Indian tribe.” Id. § 800.14(b)(1)(i),\\n(f). For instance, the Commission has consulted with Tribes to\\nuse programmatic agreements to exclude from individualized\\nreview entire categories of undertakings that are unlikely to\\naffect historic properties. See In re Nationwide Programmatic\\nAgreement Regarding the Section 106 [NHPA] Review Process\\n(Section 106 Agreement), 20 FCC Rcd. 1073, 1075 ¶ 2 (2004).\\n\\f 7\\n B. National Environmental Policy Act (NEPA)\\n\\n Congress enacted NEPA to “encourage productive and\\nenjoyable harmony between man and his environment” and\\n“promote efforts which will prevent or eliminate damage to the\\nenvironment and biosphere and stimulate the health and\\nwelfare of man,” among other purposes. 42 U.S.C. § 4321.\\nLike the NHPA, NEPA mandates a review process that “does\\nnot dictate particular decisional outcomes, but ‘merely\\nprohibits uninformed—rather than unwise—agency action.’”\\nSierra Club v. U.S. Army Corps of Eng’rs, 803 F.3d 31, 37\\n(D.C. Cir. 2015) (quoting Robertson v. Methow Valley Citizens\\nCouncil, 490 U.S. 332, 351 (1989)).\\n\\n All “major Federal actions significantly affecting the\\nquality of the human environment” trigger environmental\\nreview under NEPA, just as federal “undertakings” trigger\\nhistoric preservation review under the NHPA. 42 U.S.C.\\n§ 4332(C). Major federal actions “include[] actions . . . which\\nare potentially subject to Federal control and responsibility.”\\n40 C.F.R. § 1508.18. Under the Commission’s procedures\\nimplementing NEPA, if an action may significantly affect the\\nenvironment, applicants must conduct a preliminary\\nEnvironmental Assessment to help the Commission determine\\nwhether “the proposal will have a significant environmental\\nimpact upon the quality of the human environment,” and so\\nperhaps necessitate a more detailed Environmental Impact\\nStatement. 47 C.F.R. § 1.1308; see also 40 C.F.R. § 1508.9.\\nIf, after reviewing the Environmental Assessment, the\\nCommission determines that the action will not have a\\nsignificant environmental impact, it will make a “finding of no\\nsignificant impact” and process the application “without\\nfurther documentation of environmental effect.” 47 C.F.R.\\n§ 1.1308(d).\\n\\f 8\\n NEPA also has an analogue to the NHPA’s Advisory\\nCouncil. In enacting NEPA, Congress established the Council\\non Environmental Quality, in the Executive Office of the\\nPresident, to oversee implementation of NEPA across the\\nentire federal government. 42 U.S.C. §§ 4342, 4344. With the\\nendorsement of the Council on Environmental Quality and by\\nfollowing a series of mandated procedures, agencies can\\nestablish “categorical exclusions” for federal actions that\\nrequire neither an Environmental Assessment nor an\\nEnvironmental Impact Statement. 40 C.F.R. § 1508.4.\\nCategorical exclusions are appropriate for “a category of\\nactions which do not individually or cumulatively have a\\nsignificant effect on the human environment and which have\\nbeen found to have no such effect in procedures adopted by a\\nFederal agency.” Id. “Categorical exclusions are not\\nexemptions or waivers of NEPA review; they are simply one\\ntype of NEPA review.” Council on Environmental Quality,\\nMemorandum for Heads of Federal Dep’ts and Agencies:\\nEstablishing, Applying & Revising Categorical Exclusions\\nunder [NEPA] (Categorical Exclusion Memo) 2 (2010).\\n\\n C. Legal Framework for Wireless Infrastructure\\n\\n The Communications Act of 1934 established the FCC to\\nmake available a “rapid, efficient . . . wire and radio\\ncommunication service with adequate facilities at reasonable\\ncharges.” 47 U.S.C. § 151. In licensing use of the spectrum,\\nthe Commission is tasked with promoting “the development\\nand rapid deployment of new technologies, products and\\nservices for the benefit of the public . . . without administrative\\nor judicial delays,” id. § 309, and “maintain[ing] the control of\\nthe United States over all the channels of radio transmission,”\\nid. § 301.\\n\\f 9\\n The Commission generally does not require construction\\npermits before private parties can build wireless facilities.\\nCongress largely eliminated the FCC’s site-specific\\nconstruction permits in 1982, and the Commission has since\\nrequired construction permits only where it finds that the public\\ninterest would be served by such permitting. See Pub. L. 97-\\n259, 96 Stat. 1087, § 119 (1982) (codified at 47 U.S.C.\\n§ 319(d)). It has not made such a finding for the wireless\\nfacilities at issue here.\\n\\n The FCC does, however, require licensing of the spectrum\\nused by wireless small cells. It does so by issuing geographic\\narea licenses, which allow wireless providers to operate on\\ncertain frequency bands in a wide geographic area. See 47\\nU.S.C. § 309(j). Those licenses authorize using spectrum\\nrather than building wireless facilities, but they necessarily\\ncontemplate facility construction. They have coverage\\nrequirements—for instance, one type of geographic area\\nlicense required licensees to provide service to at least 40% of\\nthe population in their geographic service area by June 2013.\\nSee 47 C.F.R. § 27.14(h). If they fail to meet the coverage\\nrequirements, they can be stripped of authority to operate for\\nthe license’s full term or serve part of its geographic area, and\\nthey “may be subject to enforcement action, including\\nforfeitures.” Id. The Commission also exercises continuing\\nauthority to inspect radio installations to ascertain their\\ncompliance with any and all applicable laws, whether or not the\\nlicensee itself constructed those installations. See 47 U.S.C.\\n§ 303(n); 47 C.F.R. § 1.9020(c)(5).\\n\\n The Commission has not identified any period since the\\nenactment of the NHPA (in 1966) and NEPA (in 1970) when\\nit did not require historic-preservation and environmental\\nreview of wireless facilities. After Congress eliminated the\\nconstruction permit requirement, the Commission for a time\\n\\f 10\\nrequired NEPA and NHPA review of facilities before it granted\\ntheir service licenses. See, e.g., In re Amendment of Envtl.\\nRules in Response to New Regulations Issued by [CEQ], FCC\\n85-626, 1986 WL 292182, at *5 ¶ 18 (F.C.C.) (Mar. 26, 1986)\\n(requiring review “during the period prior to grant of a station\\nlicense”); id. at *8 App’x ¶ 7 (requiring NEPA review on\\n“[f]acilities that will affect districts, sites, buildings, structures\\nor objects . . . that are listed in the National Register of Historic\\nPlaces or are eligible for listing,” which includes property of\\nreligious or cultural significance to Indian Tribes, 54 U.S.C.\\n§ 302706(a)). In 1990, the Commission shifted review from\\nthe licensing stage to the construction stage by establishing a\\n“limited approval authority” over construction of wireless\\nfacilities. In re Amendment of Envtl. Rules (1990 Order), 5\\nFCC Rcd. 2942 (1990). Limited approval authority required\\nthat, “where construction of a Commission-regulated radio\\ncommunications facility is permitted without prior\\nCommission authorization (i.e., without a construction permit),\\nthe licensee must nonetheless comply with historic\\npreservation and environmental review procedures.” Order\\n¶ 51; see also 47 C.F.R. § 1.1312. The authority was “limited”\\nin that it allowed “the Commission [to] exercise[] control over\\ndeployment solely to conduct federal historic and\\nenvironmental review.” Resp’t Br. 12. The Commission\\nemphasized that shifting review to the pre-construction stage\\nserved a practical function: Before it had established its limited\\napproval authority, the FCC’s rules “provide[d] that any\\nrequired submission of [Environmental Assessments] and any\\nrequired Commission environmental review take place at the\\nlicensing stage rather than prior to construction,” with the result\\nthat “[a]pplicants who ha[d] already constructed their\\nfacilities” could “subsequently be denied licenses on\\nenvironmental grounds.” 1990 Order 2942 ¶ 3. The\\nCommission explained that it continued to require review “to\\nensure that the Commission fully complies with Federal\\n\\f 11\\nenvironmental laws in connection with facilities that do not\\nrequire pre-construction authorization.” Id. ¶ 4. It announced\\nthe changes as “necessary to ensure that the Commission\\naddresses environmental issues early enough in the licensing\\nprocess to ensure that it fully meets its obligations under\\nFederal environmental laws,” including NEPA and the NHPA.\\nId. at 2943 ¶ 9 & n.16.\\n\\n The Commission has never required individualized review\\nof each separate facility, however. A long series of regulations,\\nprogrammatic agreements, and categorical exclusions has\\naggregated facilities for joint consideration and focused NHPA\\nand NEPA review on those deployments most likely to have\\ncultural or environmental effects. For instance, most\\ncollocations—deployments on existing structures—are\\nexcluded from individualized review under NHPA\\nprogrammatic agreements and NEPA categorical exclusions.\\nSee In re Implementation of the National Environmental Policy\\nAct of 1969 (Implementation of NEPA), 49 F.C.C.2d 1313,\\n1319-20 (1974); Nationwide Programmatic Agreement for the\\nCollocation of Wireless Antennas (Collocation Agreement), 47\\nC.F.R. pt.1, app. B (2001); Section 106 Agreement, 20 FCC\\nRcd. at 1075 ¶ 2; Nationwide Programmatic Agreement for\\nReview Under the National Historic Preservation Act, 70 Fed.\\nReg. 556 (2005); In re Acceleration of Broadband Deployment\\nby Improving Wireless Facilities Siting Policies (Improving\\nWireless Facilities Siting Policies), 29 FCC Rcd. 12865, 12870\\n¶ 11 (2014); 47 C.F.R. § 1.1320(b)(4). Categorical exclusions\\ngo through notice and comment, 40 C.F.R. § 1507.3; include\\nimpact findings, Categorical Exclusion Memo 9; require the\\nCouncil on Environmental Quality to approve them as\\nconsistent with its regulations and NEPA, 40 C.F.R.\\n§ 1507.3(a); and reserve rights to interested parties to request\\nfurther review in the event that atypical adverse effects do\\noccur, 47 C.F.R. § 1.1307(c), (d). At the same time, they\\n\\f 12\\nachieve enormous efficiencies in the review processes for\\nclasses of actions or undertakings anticipated to have minimal\\nor no adverse cultural or environmental effects.\\n\\n Since 2004, the FCC has been conducting NHPA review\\nin accordance with a broad programmatic agreement, the\\nSection 106 Agreement, 20 FCC Rcd. 1073. Interested parties\\ndeveloped that agreement to “tailor the Section 106 review in\\nthe communications context in order to improve compliance\\nand streamline the review process for construction of towers\\nand other Commission undertakings, while at the same time\\nadvancing and preserving the goal of the NHPA to protect\\nhistoric properties, including historic properties to which\\nfederally recognized Indian tribes . . . attach religious and\\ncultural significance.” Id. at 1074-75 ¶ 1. In the Section 106\\nAgreement, the Commission adopted “procedures for\\nparticipation of federally recognized Indian tribes,” among\\nother changes. Id. at 1075 ¶ 2. It also formalized the use of the\\nelectronic Tower Construction Notification System, which\\nnotifies Tribes of proposed wireless construction in areas they\\nhave identified as containing properties of religious and\\ncultural significance, and allows them to give applicants\\ninformation on the potential effects of proposed construction.\\nId. at 1106-10 ¶¶ 89-100.\\n\\n II. Order Under Review\\n\\n The challenged Order eliminated NHPA and NEPA\\nreview on small cells that meet certain size and other\\nspecifications, based on the Commission’s conclusion that\\nsuch review was not statutorily required and would impede the\\nadvance of 5G networks, and that its costs outweighed any\\nbenefits. See Order ¶¶ 36-45. The Order also altered Tribal\\ninvolvement in those Section 106 reviews that are still\\nconducted on wireless facilities that were not encompassed in\\n\\f 13\\nthe small cell exemption. See id. ¶¶ 96-130. Two of the five\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4172,"cells":{"id":{"kind":"number","value":4647053,"string":"4,647,053"},"name":{"kind":"string","value":"DISTRICT NO. 1, PACIFIC COAST DISTRICT, MARINE ENGINEERS BENEFICIAL ASSOCIATION, AFL-CIO, Appellee v. LIBERTY MARITIME CORPORATION, Appellant"},"name_abbreviation":{"kind":"string","value":"District No. 1, Pacific Coast v. Liberty Maritime Corporation"},"decision_date":{"kind":"string","value":"2019-08-09"},"docket_number":{"kind":"string","value":"18-7148"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Millett, Katsas, Rao\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge RAO\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued May 10, 2019 Decided August 9, 2019\\n\\n No. 18-7148\\n\\n DISTRICT NO. 1, PACIFIC COAST DISTRICT, MARINE\\n ENGINEERS BENEFICIAL ASSOCIATION, AFL-CIO,\\n APPELLEE\\n\\n v.\\n\\n LIBERTY MARITIME CORPORATION,\\n APPELLANT\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:17-cv-02173)\\n\\n\\n Steffen N. Johnson argued the cause for appellant. With\\nhim on the briefs were William G. Miossi and Paul N. Harold.\\n\\n Mark J. Murphy argued the cause and filed the brief for\\nappellee.\\n\\n Before: MILLETT, KATSAS, and RAO, Circuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge RAO.\\n\\n RAO, Circuit Judge: Liberty Maritime Corporation\\n(Liberty) is a shipping company that has contracted over the\\npast thirty years with District No. 1, Pacific Coast District,\\n\\f 2\\nMarine Engineers Beneficial Association, AFL-CIO (MEBA),\\na labor union representing supervisory employees in the\\nmaritime industry. This case arises out of an underlying dispute\\nabout whether Liberty was contractually required to hire\\nMEBA employees on a new vessel managed by Liberty.\\nMEBA sued in the United States District Court for the District\\nof Columbia, claiming its contract with Liberty required the\\nparties to submit the dispute to arbitration. The district court\\nruled in favor of the union, granting judgment on the pleadings\\nunder Federal Rule of Civil Procedure 12(c) and compelling\\narbitration. Liberty timely appealed, arguing that the district\\ncourt lacked subject matter jurisdiction, or in the alternative,\\nerred in its application of the Rule 12(c) standards.\\n\\n For the reasons explained below, we agree that the district\\ncourt had jurisdiction over MEBA’s claim under Section 301\\nof the Labor Management Relations Act of 1947 (LMRA), 29\\nU.S.C. §§ 141 et seq., which provides federal jurisdiction over\\nsuits for “violation of contracts between an employer and a\\nlabor organization.” Id. § 185(a). MEBA raised contractual\\nissues regarding the arbitrability of the dispute and thus its\\nclaim clearly falls within the district court’s statutory\\njurisdiction. Although Liberty alleges that the dispute primarily\\nraised representational issues and thus should be within the\\nexclusive jurisdiction of the National Labor Relations Board\\n(NLRB) under the doctrine of “Garmon preemption,” federal\\ncourts retain jurisdiction over “hybrid” claims raising both\\ncontractual and representational issues. Dist. No. 1, Pac. Coast\\nDist., Marine Engineers’ Beneficial Ass’n, AFL-CIO v. Liberty\\nMar. Corp., 815 F.3d 834, 840 (D.C. Cir. 2016) (“Liberty\\nMaritime I”); see also William E. Arnold Co. v. Carpenters\\nDist. Council of Jacksonville & Vicinity, 417 U.S. 12, 18\\n(1974).\\n\\f 3\\n Although jurisdiction here was proper, we reverse and\\nremand because material facts remained in dispute regarding\\nthe existence of an applicable arbitration clause, and therefore\\nMEBA was not entitled to judgment on the pleadings under\\nRule 12(c).\\n\\n I.\\n\\n Appellant Liberty is a shipping company that transports\\ncommodities, vehicles, equipment, and other cargoes on the\\nseagoing vessels it manages. Liberty’s clients include the U.S.\\nGovernment, the United Nations, and commercial entities such\\nas automobile manufacturers. Liberty manages vessels\\ntransporting bulk cargo—including dry bulk, break bulk, and\\nbagged commodities—and “roll on/roll-off” vessels, like car\\nand truck carriers configured to transport vehicles that drive on\\nand off the vessel. Many of these vessels are enrolled in the\\nU.S. Maritime Security Program, a federal program that\\nsubsidizes shipping companies for national security\\npurposes—namely, to ensure a fleet of vessels is available in\\nthe event of a war or national emergency. See generally 46\\nU.S.C. §§ 53101 et seq. Appellee MEBA is a labor\\norganization that represents supervisory employees in the U.S.\\nmaritime industry at ports throughout the United States and on\\noceangoing vessels. On car and truck carrier vessels operated\\nby Liberty and enrolled in the U.S. Maritime Security Program,\\nMEBA represents licensed officers and engineers.\\n\\n The parties’ relationship began in 1988 when they signed\\ntwo agreements: the Tanker Vessels Master Agreement and the\\nDry Cargo Vessels Master Agreement. Although the\\nauthenticity of some of the documents attached to the pleadings\\nis disputed, the documents that purport to be current copies of\\nthese Master Agreements provide that “[a]ll disputes relating\\nto the interpretation or performance of this Agreement shall be\\n\\f 4\\ndetermined in accordance with the provisions of this Section.”\\n“[T]his Section” states that grievances will be presented to a\\nlicensed personnel board consisting of two persons appointed\\nby the union and two persons appointed by the company; if the\\nlicensed personnel board fails to resolve a grievance, an\\narbitrator will assume jurisdiction over the grievance.\\n\\n Over the past three decades, the parties have modified their\\ncontractual relationship on numerous occasions. At this stage\\nof the proceedings, the record includes only a few of these\\nagreements. Both parties agree, however, they were signatories\\nto a 2012 Memorandum of Understanding (MOU). This MOU\\nidentifies numerous prior agreements and states that prior\\nagreements will remain in effect except as expressly modified,\\nbut the MOU does not expressly modify any arbitration clause\\nin a manner relevant to this case.\\n\\n This suit arises out of a dispute between Liberty and\\nMEBA over a ship named the M/V Liberty Peace. On July 24,\\n2017, Liberty sent MEBA a letter stating its intention to\\ncommence managing this foreign flagged car and truck carrier\\nvessel and operate it as a U.S. flagged vessel. In the letter,\\nLiberty claimed the Liberty Peace would not fall under the\\nparties’ collective bargaining agreements and the various\\ncontractual modifications of those agreements because the\\nvessel would not be enrolled in the U.S. Maritime Security\\nProgram. MEBA disagreed, insisting the existing agreements\\ncovered the new vessel. Although the parties met to discuss the\\nmatter, they did not resolve their dispute. In the meantime,\\nLiberty began managing the Liberty Peace as the agent of a\\nthird party, and that third party entered into labor agreements\\nwith a different union.\\n\\n MEBA sent Liberty a grievance letter on August 31, 2017,\\nasserting Liberty was “in violation of the parties’ collective\\n\\f 5\\nbargaining agreement by failing to apply the terms and\\nconditions of the parties’ labor contract” to the Liberty Peace.\\nLiberty did not submit MEBA’s grievance to arbitration.\\n\\n MEBA subsequently filed a “Complaint to Compel\\nArbitration” in the United States District Court for the District\\nof Columbia. MEBA requested the district court compel\\nLiberty to participate in the arbitration process set forth in the\\nparties’ collective bargaining agreement and grant any other\\nappropriate relief, including attorneys’ fees and costs. MEBA\\nattached as exhibits several documents purporting to be the two\\noriginal Master Agreements, the MOU, MEBA’s August 31\\ngrievance letter, and some additional correspondence between\\nMEBA and Liberty.\\n\\n In its answer to MEBA’s complaint, Liberty admitted it\\nhad signed the Master Agreements and the MOU. Liberty\\nadmitted the authenticity of the MOU, but denied the\\nauthenticity of the exhibits MEBA claimed were copies of the\\nMaster Agreements. Liberty denied that the MOU incorporated\\nthe terms of the Master Agreements and that the arbitration\\nclauses covered the Liberty Peace. Liberty also denied that any\\nlabor contract or arbitration agreement with MEBA covered the\\nLiberty Peace. As an affirmative defense, Liberty alleged the\\ndistrict court lacked subject matter jurisdiction because the suit\\nconcerned representational rights and therefore was preempted\\nby the jurisdiction of the NLRB under the terms of the National\\nLabor Relations Act (NLRA).\\n\\n MEBA moved for judgment on the pleadings under\\nFederal Rule of Civil Procedure 12(c), and the district court\\ngranted the motion. The district court found that the Master\\nAgreements stated, “[a]ll disputes relating to the interpretation\\nor performance of this Agreement shall be determined in\\naccordance with the provisions of this Section.” Dist. Ct. Op.\\n\\f 6\\nat 3, 13. The district court concluded that this language created\\na presumption of arbitrability; Liberty failed to rebut the\\npresumption; and no agreement between the parties excluded\\nthis sort of dispute from arbitration. Id. at 13–14. The district\\ncourt also rejected Liberty’s preemption argument on the\\ngrounds that federal courts have jurisdiction over contractual\\nmatters and that MEBA’s suit “plainly requires deciding a\\ncontractual matter: whether the arbitration clause covers the\\ndispute at issue.” Id. at 10–11 n.7.\\n\\n Liberty timely appealed, challenging the district court’s\\norder on jurisdictional grounds and arguing the district court\\nviolated Rule 12(c) by making findings the pleadings did not\\nadequately support.\\n\\n II.\\n\\n “The ‘first and fundamental question’ that we are ‘bound\\nto ask and answer’ is whether the court has jurisdiction to\\ndecide the case.” Bancoult v. McNamara, 445 F.3d 427, 432\\n(D.C. Cir. 2006) (quoting Steel Co. v. Citizens for a Better\\nEnv’t, 523 U.S. 83, 94 (1998)). The district court held that\\nfederal courts have jurisdiction over contractual matters under\\nSection 301 of the LMRA, which provides:\\n\\n Suits for violation of contracts between an employer and a\\n labor organization representing employees in an industry\\n affecting commerce as defined in this chapter, . . . may be\\n brought in any district court of the United States having\\n jurisdiction of the parties . . . .\\n\\n29 U.S.C. § 185(a). Section 301 confers federal court\\njurisdiction over suits for breach of collective bargaining\\nagreements, which are contractual. “Congress deliberately\\nchose to leave the enforcement of collective agreements to the\\n\\f 7\\nusual processes of the law.” Charles Dowd Box Co. v.\\nCourtney, 368 U.S. 502, 513 (1962).\\n\\n Nevertheless, Liberty argues the district court lacked\\nsubject matter jurisdiction over this case under the judicially\\ncreated doctrine known as “Garmon preemption.” Washington\\nServ. Contractors Coal. v. Dist. Columbia, 54 F.3d 811, 815\\n(D.C. Cir. 1995) (citing San Diego Bldg. Trades Council v.\\nGarmon, 359 U.S. 236 (1959)). This doctrine holds that\\n“[w]hen an activity is arguably subject to § 7 or § 8 of the\\n[NLRA], . . . the federal courts must defer to the exclusive\\ncompetence of the [NLRB].” Garmon, 359 U.S. at 245. Suits\\nimplicating § 7 or § 8 of the NLRA are often described as\\n“representational.”\\n\\n Liberty attempts to rely on this court’s decision in a\\nprevious suit between Liberty and MEBA, in which Liberty\\nraised and lost a similar jurisdictional argument. Liberty cites\\nthis case for the proposition that three categories of legal claims\\nare preempted by the NLRA under Garmon: claims over which\\nthe NLRB “has already exercised jurisdiction,” claims that call\\nfor “an initial decision in the representation area,” and claims\\n“in which the center of the dispute is a representational\\nquestion.” Appellant Br. 35–36 (quoting Liberty Maritime I,\\n815 F.3d at 841 (citations and quotation marks omitted)).\\nLiberty urges this court to evaluate whether MEBA’s claim is\\n“primarily representational or primarily contractual,” as\\nseveral other circuits do. See, e.g., United Food & Commercial\\nWorkers Union, Local 400 v. Shoppers Food Warehouse Corp.,\\n35 F.3d 958, 961 (4th Cir. 1994); Local Union 204 of Int’l Bhd.\\nof Elec. Workers, AFL-CIO v. Iowa Elec. Light & Power Co.,\\n668 F.2d 413, 419 (8th Cir. 1982). Liberty claims such analysis\\nwould show this suit may fall into the purported third category\\nof claims preempted under Garmon: that the center of the\\ndispute may concern a representational matter, such as whether\\n\\f 8\\nMEBA or another union has representational rights over\\ncrewmembers of the Liberty Peace. In order to determine the\\ntrue center of this dispute, Liberty argues the district court\\nshould have considered MEBA’s grievance letter instead of\\nfocusing only on MEBA’s complaint. Liberty contends it was\\nlegal error for the district court to have concluded that, as a\\nmatter of law, it maintained jurisdiction over MEBA’s claims.\\n\\n MEBA responds that Liberty’s argument on appeal\\n“conflates the type of claim with the effect of a claim’s\\nenforcement.” Appellee Br. 26 (quoting Liberty Maritime I,\\n815 F.3d at 843). While the possible outcome of its suit may\\ntouch on representational issues, MEBA argues it has a contract\\nthat requires Liberty to arbitrate, and MEBA asked the district\\ncourt to compel compliance with that contract. MEBA\\nmaintains this type of contractual dispute is squarely covered\\nby Section 301 of the LMRA.\\n\\n The district court properly exercised jurisdiction over\\nMEBA’s claim under the plain meaning of Section 301 as well\\nas established Supreme Court and Circuit precedent. Section\\n301 covers “[s]uits for violations of contracts between an\\nemployer and a labor organization.” Liberty Maritime I, 815\\nF.3d at 840 (quoting 29 U.S.C. § 185(a)). As the Supreme\\nCourt has stated, Section 301 “permits suits for breach of a\\ncollective bargaining agreement regardless of whether the\\nparticular breach is also an unfair labor practice within the\\njurisdiction of the Board.” Vaca v. Sipes, 386 U.S. 171, 179–\\n80 (1967) (emphasis added); see also Carey v. Westinghouse\\nCorp., 375 U.S. 261, 267–68 (1964) (holding that Section 301\\ngives a federal court jurisdiction over a suit to enforce an\\narbitration clause in a collective bargaining agreement even if\\nthe case is “truly a representation case” that could also be heard\\nby the NLRB under Section 9 of the NLRA). Thus, the\\n“Garmon doctrine is ‘not relevant’ to actions within the\\n\\f 9\\npurview of § 301” of the LMRA. Arnold, 417 U.S. at 16 (citing\\nLocal 174, Teamsters v. Lucas Flour Co., 369 U.S. 95, 101 n.9\\n(1962)). We held in Liberty Maritime I that federal courts and\\nthe NLRB have concurrent jurisdiction over claims that are\\n“both contractual and representational.” 815 F.3d at 840\\n(emphasis original) (citing Arnold, 417 U.S. at 16, and Smith v.\\nEvening News Ass’n, 371 U.S. 195, 197 (1962)).\\n\\n Thus, if a case is both representational and contractual, it\\nis treated as a “hybrid” claim. See, e.g., United Parcel Serv.,\\nInc. v. Mitchell, 451 U.S. 56, 66 (1981) (Stewart, J., concurring\\nin the judgment) (describing “a hybrid ‘§ 301 and breach of\\nduty suit’”); DelCostello v. Int’l Bhd. of Teamsters, 462 U.S.\\n151, 165 (1983) (describing “hybrid § 301/fair representation\\nlitigation”); Cephas v. MVM, Inc., 520 F.3d 480, 485 (D.C. Cir.\\n2008) (same). Such “hybrid” claims create concurrent\\njurisdiction for the federal courts and the NLRB, but they do\\nnot divest courts of their statutory jurisdiction. Consistent with\\nSupreme Court precedent, that is precisely what this court has\\nheld: “Instead of forcing courts to shoehorn a hybrid claim into\\none category or the other, the Supreme Court has held that they\\nretain jurisdiction to hear a contractual claim even if the claim\\nis also representational.” Liberty Maritime I, 815 F.3d at 840\\n(citing Arnold, 417 U.S. at 16). “[F]ederal courts have\\nindependent jurisdiction to decide cases alleging a breach of\\ncollective bargaining agreements, even though that very breach\\nmay also be an unfair labor practice.” Mullins v. Kaiser Steel\\nCorp., 642 F.2d 1302, 1316 (D.C. Cir. 1980), rev’d on other\\ngrounds, 455 U.S. 72 (1982).\\n\\n Allowing “hybrid” claims to be brought in federal court\\nreads together the two statutes, the LMRA and the NLRA,\\ngiving effect to Congress’s provision of federal court\\njurisdiction for contractual claims and NLRB jurisdiction over\\nrepresentational claims. See Vaca, 386 U.S. at 179–80;\\n\\f 10\\nWachovia Bank v. Schmidt, 546 U.S. 303, 315–16 (2006)\\n(“under the in pari materia canon of statutory construction,\\nstatutes addressing the same subject matter generally should be\\nread as if they were one law”) (citations and quotation marks\\nomitted).\\n\\n Liberty continues to argue that if a case is both\\nrepresentational and contractual, a district court must place\\nthose claims on a sliding scale to determine if the case is\\nprimarily one or the other. Neither the Supreme Court nor this\\ncourt have required such an inquiry. In Liberty Maritime I, we\\ndescribed and discussed the practice in some circuits, which\\n“examine the major issues to be decided” and “determine\\nwhether they can be characterized as primarily representational\\nor primarily contractual” in order to dismiss “primarily\\nrepresentational” claims. 1 Liberty Maritime I, 815 F.3d at 840\\n(alterations, quotation marks, and citation omitted).\\n\\n Liberty’s reliance on the categories recognized by other\\ncircuits is misplaced. While the Liberty Maritime I court\\ndescribed the approaches from “several of our sister circuits,”\\nit did not adopt any of these competing decisions. Id. at 841.\\nBecause the categories identified by other circuits were not\\nnecessary to the decision, the Liberty Maritime I discussion of\\nthose cases “does not constitute a precedent to be followed with\\nrespect to that issue.” UC Health v. NLRB, 803 F.3d 669, 682\\n(D.C. Cir. 2015) (quotation marks and citations omitted). We\\nfollow Liberty Maritime I, which refused to define the\\n\\n1\\n Compare Iowa Elec. Light & Power Co., 668 F.2d at 419; Paper,\\nAllied–Indus., Chem. & Energy Workers Int’l Union v. Air Prods. &\\nChems., Inc., 300 F.3d 667, 675 (6th Cir. 2002); Pace v. Honolulu\\nDisposal Serv., Inc., 227 F.3d 1150, 1156 (9th Cir. 2000); Shoppers,\\n35 F.3d at 961; Copps Food Ctr., Inc. v. United Food & Commercial\\nWorkers Union, Local 73–A, No. 90–1905, 1991 WL 135508, at *2\\n(7th Cir. July 23, 1991) (unpublished).\\n\\f 11\\n“parameters of a claim that is ‘primarily representational’ as\\nopposed to ‘primarily contractual’” and declined “to shoehorn”\\na given “claim into one category or the other.” 815 F.3d at 840–\\n41.\\n\\n Liberty also contends that it cannot “be ruled out, based on\\nthe pleadings alone, that the major issues to be decided are\\nprimarily representational.” Appellant Br. 38. MEBA’s\\ngrievance letter, like its complaint, however, raised numerous\\ncontractual issues. See MEBA Compl. Exhibit G (“Please be\\nadvised that Liberty is in violation of the parties’ collective\\nbargaining agreement . . . . Consider this notice of an official\\ngrievance for violation of the parties’ labor contract . . . . We\\ndemand that the parties participate in expedited arbitration to\\nresolve this contractual dispute . . . .”). And Liberty does not\\nargue that the claim here is exclusively representational. See\\nLiberty Maritime I, 815 F.3d at 843 (“Garmon preemption is\\ndesigned to prevent a court from deciding a claim that can only\\nbe characterized as representational.”) (emphasis added). At\\nmost, then, Liberty has left open the possibility that this case\\ninvolves a hybrid claim raising both contractual and\\nrepresentational questions. As discussed, however, such hybrid\\nclaims are subject to the concurrent jurisdiction of the NLRB\\nand the federal courts. Id. at 840. To hold otherwise “would\\nfrustrate rather than serve the congressional policy expressed\\nin [Section 301].” Smith, 371 U.S. at 200.\\n\\n Finally, it may be true, as Liberty stresses, that “a party’s\\nmere assertion that a claim is contractual is not an automatic\\nticket to federal court” under the LMRA. Appellant Br. 35\\n(quoting Liberty Maritime I, 815 F.3d at 840). A plaintiff must\\nplausibly demonstrate the dispute falls within the terms of\\nSection 301, and is not an “end run around [the\\nNLRA] . . . under the guise of contract interpretation.” Pace,\\n227 F.3d at 1157 (citation and alteration omitted); accord\\n\\f 12\\nPaper, 300 F.3d at 675 (“[S]imply referring to the claim as a\\n‘breach of contract’ was insufficient for purposes of § 301\\nfederal courts’ jurisdiction.”). This proposition, however, goes\\nlittle further than the axiom that a plaintiff’s claim must invoke\\na proper basis for federal court jurisdiction within a well-\\npleaded complaint. Cf. Greenhill v. Spellings, 482 F.3d 569,\\n575 (D.C. Cir. 2007) (under the well-pleaded complaint rule,\\njurisdiction arising under federal law is established by looking\\nto the legal basis of plaintiff’s claim) (citing Louisville &\\nNashville R.R. Co. v. Mottley, 211 U.S. 149, 152–53 (1908));\\nsee also 13D Charles Alan Wright, Arthur R. Miller, Edward\\nH. Cooper, & Richard D. Freer, Federal Practice and\\nProcedure § 3566, at 261–62 (3d ed. 2008) (“The well-pleaded\\ncomplaint rule stands for the proposition that the court, in\\ndetermining whether the case arises under federal law, will\\nlook only to the claim itself.”).\\n\\n Put simply, Congress gave federal courts jurisdiction to\\nhear contractual claims between labor organizations and\\nemployers in Section 301 of the LMRA. The existence of\\nrepresentational issues does not divest the federal courts of\\njurisdiction. Here, MEBA’s suit alleges a breach of the parties’\\nlabor contract. The suit requires a judicial determination as to\\nwhether an arbitration clause in the agreements between\\nLiberty and MEBA covers the dispute over the Liberty Peace.\\nSee Gen. Elec. Co. v. Local 205, United Elec., Radio & Mach.\\nWorkers of Am., 353 U.S. 547, 548 (1957) (Section “301(a)\\nfurnishes a body of federal substantive law for the enforcement\\nof collective bargaining agreements” that provides for suits “to\\nenforce the obligation to arbitrate grievance disputes.”); see\\nalso Westinghouse, 375 U.S. at 267–68. As the dispute includes\\ncontractual claims, the district court properly concluded subject\\nmatter jurisdiction was established under Section 301 of the\\nLMRA.\\n\\f 13\\n III.\\n\\n Proceeding to the merits of Liberty’s appeal, we consider\\nnext whether the district court properly granted MEBA’s\\nmotion for judgment on the pleadings.\\n\\n A.\\n\\n This court reviews a Rule 12(c) judgment on the pleadings\\nde novo. Judicial Watch, Inc. v. United States Dep’t of\\nHomeland Sec., 895 F.3d 770, 777 (D.C. Cir. 2018); Mpoy v.\\nRhee, 758 F.3d 285, 287 (D.C. Cir. 2014). Federal Rule of Civil\\nProcedure 12(c) provides, “After the pleadings are closed—but\\nearly enough not to delay trial—a party may move for\\njudgment on the pleadings.” Pleadings include any “copy of a\\nwritten instrument that is an exhibit to a pleading,” Fed. R. Civ.\\nP. 10(c), such as relevant and authentic documents attached to\\nthe complaint. See, e.g., Philips v. Pitt Cty. Mem’l Hosp., 572\\nF.3d 176, 180 (4th Cir. 2009).\\n\\n Very few of our precedents discuss Rule 12(c), in part\\nbecause judgment on the pleadings is rare. As Wright & Miller\\nnotes, “Federal Rule 12(c) has its historical roots in common\\nlaw practice, which permitted either party, at any point in the\\nproceeding, to demur to his opponent’s pleading and secure a\\ndismissal or final judgment on the basis of the pleadings.” 5C\\nWright & Miller § 1367, at 205; see also Patel v. Contemp.\\nClassics of Beverly Hills, 259 F.3d 123, 126 (2d Cir. 2001)\\n(“[A] motion for judgment on the pleadings is the direct\\ndescendant of that ancient leper of the common law, the\\n‘speaking demurrer.’”); 5C Wright & Miller § 1369, at 265\\n(noting “the Rule 12(c) motion is little more than a relic of the\\ncommon law and code eras”).\\n\\n Because Rule 12(c) provides judicial resolution at an early\\nstage of a case, the party seeking judgment on the pleadings\\n\\f 14\\nshoulders a heavy burden of justification. A reviewing court\\n“will affirm the district court if the moving party demonstrates\\nthat no material fact is in dispute and that it is entitled to\\njudgment as a matter of law.” Peters v. Nat’l R.R. Passenger\\nCorp., 966 F.2d 1483, 1485 (D.C. Cir. 1992) (quotation marks\\nomitted). The moving party must demonstrate its entitlement to\\njudgment in its favor, even though the “court evaluating the\\n12(c) motion will accept as true the allegations in the\\nopponent’s pleadings, and as false all controverted assertions\\nof the movant.” Haynesworth v. Miller, 820 F.2d 1245, 1249\\nn.11 (D.C. Cir. 1987) (collecting cases), abrogated on other\\ngrounds by Hartman v. Moore, 547 U.S. 250 (2006); see also\\nBeal v. Missouri Pac. R.R. Corp., 312 U.S. 45, 51 (1941) (when\\nthe plaintiff moves for judgment on the pleadings, the\\ndefendant’s “denials and allegations of the answer which are\\nwell pleaded must be taken as true”). We must give “all\\nreasonable inferences to the opponent’s pleadings” before\\nentering a judgment on the pleadings. Wager v. Pro, 575 F.2d\\n882, 884 (D.C. Cir. 1976).\\n\\n Under this standard, “a judgment on the pleadings is not\\nappropriate” if there are “issues of fact which if proved would\\ndefeat recovery,” “even if the trial court is convinced that the\\nparty opposing the motion is unlikely to prevail at trial.” Id.\\n“[I]f material questions of fact are presented by the pleadings,\\nthe remedy by motion for judgment on the pleadings under\\nRule 12(c) is not available.” Noel v. Olds, 149 F.2d 13, 14 &\\nn.7 (D.C. Cir. 1945) (citing James A. Pike, Objections to\\nPleadings Under the New Federal Rules of Civil Procedure, 47\\nYale L.J. 50 (1937)).\\n\\n B.\\n\\n Applying these standards, MEBA’s pleadings and\\nattachments cannot carry the day. The district court erred in\\n\\f 15\\ngranting MEBA’s Rule 12(c) motion. The district court found\\nthat Liberty and MEBA agreed to arbitrate their disputes in the\\nMaster Agreements, which included clauses stating: “All\\ndisputes relating to the interpretation or performance of this\\nAgreement shall be determined in accordance with the\\nprovisions of this Section.” Dist. Ct. Op. at 13. The district\\ncourt acknowledged that Liberty contested “whether a\\ncollective bargaining agreement exists between the parties\\nunder which the union can assert its right to arbitrate” and that\\nLiberty argued the absence of such an agreement rendered\\njudgment on the pleadings inappropriate. Id. at 9–10 & n.7. But\\nthe district court construed Liberty’s assertion that no\\nagreement existed as a legal conclusion, not a factual dispute.\\nBecause legal conclusions about a collective bargaining\\nagreement are accorded “no special deference,” id. at n.7\\n(quoting Local Union No. 47, Int’l Bhd. Of Elec. Workers v.\\nNLRB, 927 F.2d 635, 640 (D.C. Cir. 1991)), the district court\\nfound no material fact about the contract to be in dispute.\\n\\n On appeal, Liberty argues that its denial was factual, not\\nlegal, and so failure to accept its denial as true was legal error;\\nthat the complete contractual terms and scope were not before\\nthe district court on the pleadings, as recognized by the district\\ncourt’s opinion and the pleadings; and that without a full\\ncontract to interpret, the order to arbitrate was error. Echoing\\nthe district court, MEBA simply responds that Liberty’s claim\\nthat it did not have an agreement with MEBA covering the\\nLiberty Peace is a legal conclusion, not a dispute of material\\nfact. MEBA also argues that Liberty never denied the existence\\nof a collective bargaining agreement that contains a broad\\narbitration clause, nor that those agreements are still in effect.\\nIt claims that Liberty’s “tactic” is “smoke and mirrors.”\\nAppellee Br. 10. MEBA maintains that Liberty contradicted its\\nappellate theory of the case in litigating the case below,\\nclaiming that Liberty conceded the existence of the contract\\n\\f 16\\nand disputed only whether the contract’s broad terms covered\\nthe Liberty Peace. Essentially, MEBA claims that Liberty’s\\nappeal is rooted in procedural technicalities, rather than a good\\nfaith factual dispute over whether an arbitration clause exists in\\nan authentic, extant contract.\\n\\n Giving all reasonable inferences to Liberty, Liberty raised\\nmaterial issues of fact that rendered judgment on the pleadings\\ninappropriate. First, Liberty properly disputed the existence of\\na contract to arbitrate in this case. Without a contract binding\\nthe parties to arbitrate, an order compelling arbitration is\\nimproper. “[A]rbitration is a matter of contract and a party\\ncannot be required to submit to arbitration any dispute which\\nhe has not agreed so to submit.” AT&T Techs., Inc. v.\\nCommc’ns Workers of Am., 475 U.S. 643, 648 (1986)\\n(quotation marks and citation omitted). Such contracts must be\\n“interpreted as a whole.” United States v. Hunt, 843 F.3d 1022,\\n1028 (D.C. Cir. 2016) (quoting Restatement (Second) of\\nContracts § 202(2)). And although “the determination that\\nparties have contractually bound themselves to arbitrate\\ndisputes . . . is a legal conclusion[,] . . . the findings upon which\\nthat conclusion is based are factual.” Bailey v. Fed. Nat’l\\nMortg. Ass’n, 209 F.3d 740, 744 (D.C. Cir. 2000) (quoting\\nChelsea Square Textiles, Inc. v. Bombay Dyeing & Mfg. Co.,\\n189 F.3d 289, 295 (2d Cir. 1999)); see also 11 Williston on\\nContracts § 30:3 (4th ed.) (“It is generally a question of fact . . .\\nwhether or not a contract . . . actually exists.”).\\n\\n The existence of a contract to arbitrate must first be\\nestablished through a factual determination of what constitutes\\nthe parties’ full agreement. As such, the district court erred\\nwhen it based its conclusion of arbitrability on a contract of\\ngenuinely disputed authenticity. Liberty denied the authenticity\\nof the copies of the Master Agreements that MEBA attached to\\nits complaint. Specifically, MEBA alleged in its complaint:\\n\\f 17\\n Liberty first became signatory to a collective bargaining\\n agreement with MEBA in 1988 when the Company signed\\n on to both the Union’s 1986-1990 Tanker Vessels Master\\n Agreement and the 1986-1990 Dry Cargo Vessels Master\\n Agreement. A copy of the 1986-1990 Tanker Vessels\\n Master Agreement and Liberty’s signature page is attached\\n hereto as Exhibit A. A copy of the 1986-1990 Dry Cargo\\n Vessels Master Agreement is attached hereto as Exhibit B.\\n These agreements cover all U.S. flagged vessels owned,\\n managed or operated by Liberty.\\n\\nMEBA Compl. ¶ 8. In its answer, Liberty admitted that Liberty\\nsigned the Master Agreements but denied all other allegations\\nin ¶ 8 of the complaint not expressly admitted. This included\\ndenying MEBA’s allegation that Exhibits A and B were copies\\nof the Master Agreements. Liberty also denied that even the\\nauthentic version of these agreements covered all U.S. flagged\\nvessels managed by Liberty.\\n\\n Accepting Liberty’s allegations as true and making all\\nreasonable inferences in Liberty’s favor, as the Rule 12(c)\\nstandards require, means that Exhibits A and B were not\\nauthenticated copies of the two Master Agreements. Without\\nauthenticated Master Agreements, the district court lacked\\nadequate factual support for its finding that the language of the\\nagreements contained an extant arbitration clause. This dispute\\nover authenticity created a material issue of fact that should\\nhave been enough to defeat MEBA’s Rule 12(c) motion. See\\nHorsley v. Feldt, 304 F.3d 1125, 1135 (11th Cir. 2002)\\n(“Because the authenticity of the [documents] attached to the\\namended answer is disputed, . . . they may not be considered in\\ndeciding the Rule 12(c) motion for judgment on the\\npleadings.”); see also Philips, 572 F.3d at 180.\\n\\f 18\\n Second, several disputed contract provisions were not\\nincluded in the pleadings and attachments. The pleadings\\nshowed that, at least in 2012, Liberty and MEBA were parties\\nto “a Memorandum of Understanding dated September 23,\\n2005, as amended; various Side Letters, dated June 8, 2005,\\nOctober 28, 2005, and July 14, 2010, respectively; and Letters\\nof Understanding, dated July 7, 2009, and February 21, 2010,\\nrespectively.” MEBA Compl. Exhibit C. Liberty conceded the\\nauthenticity of the MOU. But Liberty maintained that the six\\ndocuments referenced in the MOU, but not attached to\\nMEBA’s complaint, were relevant to whether any collective\\nbargaining agreement or arbitration clause applied to the new\\nvessel.\\n\\n Liberty stresses the pre-suit correspondence attached to\\nMEBA’s complaint referenced over a dozen other documents,\\nonly three of which were attached to the complaint, and only\\none of which was conceded to be authentic. Liberty also\\ndisputed that a contract covered the Liberty Peace on the basis\\nthat MEBA failed to define “this Agreement” in the arbitration\\nclause. Because the parties’ contractual relationship has been\\nmodified by numerous agreements over the decades since the\\nMaster Agreements were signed, the record does not make\\nclear to what “this Agreement” refers.\\n\\n The district court, however, assumed these numerous side\\nletters and letters of understanding did not modify the\\nagreement in any meaningful way, accepting that “the\\ngrievance and arbitration procedures contained in the Tanker\\nand Dry Cargo Master Agreements remain binding” without\\nhaving reviewed the “terms and conditions of employment of\\nthe [collective bargaining agreements], side letters, and letters\\nof understanding.” Dist. Ct. Op. at 3. These assumptions were\\ninappropriate in the face of controverted facts about the\\nAgreements’ content. The pleadings and limited attachments\\n\\f 19\\ndid not support the conclusion that “no material fact is in\\ndispute.” Peters, 966 F.2d at 1485.\\n\\n Third, the district court drew improper inferences against\\nLiberty, the opposing party, rather than against MEBA, the\\nmoving party. The district court concluded “Liberty failed to\\npoint to any evidence to rebut the presumption of arbitrability.”\\nDist. Ct. Op. at 14. Yet at this stage of the proceedings, it was\\nMEBA’s burden to demonstrate the existence of an applicable\\narbitration agreement, not Liberty’s burden to rebut it. See\\nBeal, 312 U.S. at 51. Although the district court correctly\\nidentified the federal policy in favor of arbitration agreements,\\nMoses H. Cone Mem’l Hosp. v. Mercury Constr. Corp., 460\\nU.S. 1, 24–25 (1983), that policy becomes a presumption only\\nwhen the factual existence of a contract for arbitration has been\\nestablished. See AT&T, 475 U.S. at 648; Bailey, 209 F.3d at\\n744. The district court also found that “no agreement between\\nthe parties contains language excluding this sort of dispute\\nfrom arbitration.” Dist. Ct. Op. at 14. But because the whole\\n“agreement” was not before the district court, it erred by\\ninferring, in a manner that favored MEBA, that the “cont[ent]”\\nand “language” of the agreement did not exclude the dispute\\nover the Liberty Peace from arbitration.\\n\\n Fourth, the district court erred by refusing to allow\\ndiscovery over Liberty’s objection that there was a material\\nfactual dispute about the existence of an agreement to arbitrate\\nissues relating to the Liberty Peace. See Dec. 14, 2017 Minute\\nOrder. Discovery would have required the district court to treat\\nMEBA’s motion as one for summary judgment. Fed. R. Civ. P.\\n12(d) (“If, on a motion under Rule 12(b)(6) or 12(c), matters\\noutside the pleadings are presented to and not excluded by the\\ncourt, the motion must be treated as one for summary judgment\\nunder Rule 56. All parties must be given a reasonable\\nopportunity to present all the material that is pertinent to the\\n\\f 20\\nmotion.”). Granting such discovery would have allowed\\nLiberty to attempt to substantiate, or MEBA to refute, the claim\\nthat the complete contractual record, when read as a whole,\\ndoes not require arbitration of this dispute.\\n\\n Despite MEBA’s claim that Liberty’s argument consists of\\n“smoke and mirrors,” MEBA has not shown that Liberty lacks\\na good faith basis for its appeal. Parties are presumed to have a\\ngood faith basis for denying allegations, even at the pleading\\nstage. See Amnesty Am. v. Town of W. Hartford, 361 F.3d 113,\\n131 (2d Cir. 2004) (“Because attorneys, as officers of the court,\\nare presumed not to offer in opposition” to dispositive motions\\n“evidence that they have no good faith basis to believe will be\\navailable or admissible at trial, the burden is on the moving\\nparty.”). For the reasons stated above and in light of the\\nincomplete contractual record, we assume there was a good\\nfaith basis for Liberty to challenge judgment on MEBA’s\\npleadings. We expect the parties, on remand, will act in good\\nfaith to narrow their dispute and avoid burying the district court\\nwith reams of contractual provisions not arguably relevant to\\nthe arbitrability issue in this case.\\n\\n ***\\n\\n This case involved disputed issues of material fact that\\nrendered the unusual remedy of judgment on the pleadings\\ninappropriate. Because the district court should have first\\ndetermined whether Liberty and MEBA had a valid contract\\nfor arbitration by looking at their whole agreement, we reverse\\nthe judgment on the pleadings and remand the case for further\\nproceedings consistent with this opinion.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4424306/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge RAO\"}]}"},"judge1":{"kind":"string","value":"MILLETT"},"judge2":{"kind":"string","value":"KATSAS"},"judge3":{"kind":"string","value":"RAO"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4647053/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: DISTRICT NO. 1, PACIFIC COAST DISTRICT, MARINE ENGINEERS BENEFICIAL ASSOCIATION, AFL-CIO, Appellee v. LIBERTY MARITIME CORPORATION, Appellant\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-09\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued May 10, 2019 Decided August 9, 2019\n\n No. 18-7148\n\n DISTRICT NO. 1, PACIFIC COAST DISTRICT, MARINE\n ENGINEERS BENEFICIAL ASSOCIATION, AFL-CIO,\n APPELLEE\n\n v.\n\n LIBERTY MARITIME CORPORATION,\n APPELLANT\n\n\n Appeal from the United States District Court\n for the District of Columbia\n (No. 1:17-cv-02173)\n\n\n Steffen N. Johnson argued the cause for appellant. With\nhim on the briefs were William G. Miossi and Paul N. Harold.\n\n Mark J. Murphy argued the cause and filed the brief for\nappellee.\n\n Before: MILLETT, KATSAS, and RAO, Circuit Judges.\n\n Opinion for the Court filed by Circuit Judge RAO.\n\n RAO, Circuit Judge: Liberty Maritime Corporation\n(Liberty) is a shipping company that has contracted over the\npast thirty years with District No. 1, Pacific Coast District,\n\f 2\nMarine Engineers Beneficial Association, AFL-CIO (MEBA),\na labor union representing supervisory employees in the\nmaritime industry. This case arises out of an underlying dispute\nabout whether Liberty was contractually required to hire\nMEBA employees on a new vessel managed by Liberty.\nMEBA sued in the United States District Court for the District\nof Columbia, claiming its contract with Liberty required the\nparties to submit the dispute to arbitration. The district court\nruled in favor of the union, granting judgment on the pleadings\nunder Federal Rule of Civil Procedure 12(c) and compelling\narbitration. Liberty timely appealed, arguing that the district\ncourt lacked subject matter jurisdiction, or in the alternative,\nerred in its application of the Rule 12(c) standards.\n\n For the reasons explained below, we agree that the district\ncourt had jurisdiction over MEBA’s claim under Section 301\nof the Labor Management Relations Act of 1947 (LMRA), 29\nU.S.C. §§ 141 et seq., which provides federal jurisdiction over\nsuits for “violation of contracts between an employer and a\nlabor organization.” Id. § 185(a). MEBA raised contractual\nissues regarding the arbitrability of the dispute and thus its\nclaim clearly falls within the district court’s statutory\njurisdiction. Although Liberty alleges that the dispute primarily\nraised representational issues and thus should be within the\nexclusive jurisdiction of the National Labor Relations Board\n(NLRB) under the doctrine of “Garmon preemption,” federal\ncourts retain jurisdiction over “hybrid” claims raising both\ncontractual and representational issues. Dist. No. 1, Pac. Coast\nDist., Marine Engineers’ Beneficial Ass’n, AFL-CIO v. Liberty\nMar. Corp., 815 F.3d 834, 840 (D.C. Cir. 2016) (“Liberty\nMaritime I”); see also William E. Arnold Co. v. Carpenters\nDist. Council of Jacksonville & Vicinity, 417 U.S. 12, 18\n(1974).\n\f 3\n Although jurisdiction here was proper, we reverse and\nremand because material facts remained in dispute regarding\nthe existence of an applicable arbitration clause, and therefore\nMEBA was not entitled to judgment on the pleadings under\nRule 12(c).\n\n I.\n\n Appellant Liberty is a shipping company that transports\ncommodities, vehicles, equipment, and other cargoes on the\nseagoing vessels it manages. Liberty’s clients include the U.S.\nGovernment, the United Nations, and commercial entities such\nas automobile manufacturers. Liberty manages vessels\ntransporting bulk cargo—including dry bulk, break bulk, and\nbagged commodities—and “roll on/roll-off” vessels, like car\nand truck carriers configured to transport vehicles that drive on\nand off the vessel. Many of these vessels are enrolled in the\nU.S. Maritime Security Program, a federal program that\nsubsidizes shipping companies for national security\npurposes—namely, to ensure a fleet of vessels is available in\nthe event of a war or national emergency. See generally 46\nU.S.C. §§ 53101 et seq. Appellee MEBA is a labor\norganization that represents supervisory employees in the U.S.\nmaritime industry at ports throughout the United States and on\noceangoing vessels. On car and truck carrier vessels operated\nby Liberty and enrolled in the U.S. Maritime Security Program,\nMEBA represents licensed officers and engineers.\n\n The parties’ relationship began in 1988 when they signed\ntwo agreements: the Tanker Vessels Master Agreement and the\nDry Cargo Vessels Master Agreement. Although the\nauthenticity of some of the documents attached to the pleadings\nis disputed, the documents that purport to be current copies of\nthese Master Agreements provide that “[a]ll disputes relating\nto the interpretation or performance of this Agreement shall be\n\f 4\ndetermined in accordance with the provisions of this Section.”\n“[T]his Section” states that grievances will be presented to a\nlicensed personnel board consisting of two persons appointed\nby the union and two persons appointed by the company; if the\nlicensed personnel board fails to resolve a grievance, an\narbitrator will assume jurisdiction over the grievance.\n\n Over the past three decades, the parties have modified their\ncontractual relationship on numerous occasions. At this stage\nof the proceedings, the record includes only a few of these\nagreements. Both parties agree, however, they were signatories\nto a 2012 Memorandum of Understanding (MOU). This MOU\nidentifies numerous prior agreements and states that prior\nagreements will remain in effect except as expressly modified,\nbut the MOU does not expressly modify any arbitration clause\nin a manner relevant to this case.\n\n This suit arises out of a dispute between Liberty and\nMEBA over a ship named the M/V Liberty Peace. On July 24,\n2017, Liberty sent MEBA a letter stating its intention to\ncommence managing this foreign flagged car and truck carrier\nvessel and operate it as a U.S. flagged vessel. In the letter,\nLiberty claimed the Liberty Peace would not fall under the\nparties’ collective bargaining agreements and the various\ncontractual modifications of those agreements because the\nvessel would not be enrolled in the U.S. Maritime Security\nProgram. MEBA disagreed, insisting the existing agreements\ncovered the new vessel. Although the parties met to discuss the\nmatter, they did not resolve their dispute. In the meantime,\nLiberty began managing the Liberty Peace as the agent of a\nthird party, and that third party entered into labor agreements\nwith a different union.\n\n MEBA sent Liberty a grievance letter on August 31, 2017,\nasserting Liberty was “in violation of the parties’ collective\n\f 5\nbargaining agreement by failing to apply the terms and\nconditions of the parties’ labor contract” to the Liberty Peace.\nLiberty did not submit MEBA’s grievance to arbitration.\n\n MEBA subsequently filed a “Complaint to Compel\nArbitration” in the United States District Court for the District\nof Columbia. MEBA requested the district court compel\nLiberty to participate in the arbitration process set forth in the\nparties’ collective bargaining agreement and grant any other\nappropriate relief, including attorneys’ fees and costs. MEBA\nattached as exhibits several documents purporting to be the two\noriginal Master Agreements, the MOU, MEBA’s August 31\ngrievance letter, and some additional correspondence between\nMEBA and Liberty.\n\n In its answer to MEBA’s complaint, Liberty admitted it\nhad signed the Master Agreements and the MOU. Liberty\nadmitted the authenticity of the MOU, but denied the\nauthenticity of the exhibits MEBA claimed were copies of the\nMaster Agreements. Liberty denied that the MOU incorporated\nthe terms of the Master Agreements and that the arbitration\nclauses covered the Liberty Peace. Liberty also denied that any\nlabor contract or arbitration agreement with MEBA covered the\nLiberty Peace. As an affirmative defense, Liberty alleged the\ndistrict court lacked subject matter jurisdiction because the suit\nconcerned representational rights and therefore was preempted\nby the jurisdiction of the NLRB under the terms of the National\nLabor Relations Act (NLRA).\n\n MEBA moved for judgment on the pleadings under\nFederal Rule of Civil Procedure 12(c), and the district court\ngranted the motion. The district court found that the Master\nAgreements stated, “[a]ll disputes relating to the interpretation\nor performance of this Agreement shall be determined in\naccordance with the provisions of this Section.” Dist. Ct. Op.\n\f 6\nat 3, 13. The district court concluded that this language created\na presumption of arbitrability; Liberty failed to rebut the\npresumption; and no agreement between the parties excluded\nthis sort of dispute from arbitration. Id. at 13–14. The district\ncourt also rejected Liberty’s preemption argument on the\ngrounds that federal courts have jurisdiction over contractual\nmatters and that MEBA’s suit “plainly requires deciding a\ncontractual matter: whether the arbitration clause covers the\ndispute at issue.” Id. at 10–11 n.7.\n\n Liberty timely appealed, challenging the district court’s\norder on jurisdictional grounds and arguing the district court\nviolated Rule 12(c) by making findings the pleadings did not\nadequately support.\n\n II.\n\n “The ‘first and fundamental question’ that we are ‘bound\nto ask and answer’ is whether the court has jurisdiction to\ndecide the case.” Bancoult v. McNamara, 445 F.3d 427, 432\n(D.C. Cir. 2006) (quoting Steel Co. v. Citizens for a Better\nEnv’t, 523 U.S. 83, 94 (1998)). The district court held that\nfederal courts have jurisdiction over contractual matters under\nSection 301 of the LMRA, which provides:\n\n Suits for violation of contracts between an employer and a\n labor organization representing employees in an industry\n affecting commerce as defined in this chapter, . . . may be\n brought in any district court of the United States having\n jurisdiction of the parties . . . .\n\n29 U.S.C. § 185(a). Section 301 confers federal court\njurisdiction over suits for breach of collective bargaining\nagreements, which are contractual. “Congress deliberately\nchose to leave the enforcement of collective agreements to the\n\f 7\nusual processes of the law.” Charles Dowd Box Co. v.\nCourtney, 368 U.S. 502, 513 (1962).\n\n Nevertheless, Liberty argues the district court lacked\nsubject matter jurisdiction over this case under the judicially\ncreated doctrine known as “Garmon preemption.” Washington\nServ. Contractors Coal. v. Dist. Columbia, 54 F.3d 811, 815\n(D.C. Cir. 1995) (citing San Diego Bldg. Trades Council v.\nGarmon, 359 U.S. 236 (1959)). This doctrine holds that\n“[w]hen an activity is arguably subject to § 7 or § 8 of the\n[NLRA], . . . the federal courts must defer to the exclusive\ncompetence of the [NLRB].” Garmon, 359 U.S. at 245. Suits\nimplicating § 7 or § 8 of the NLRA are often described as\n“representational.”\n\n Liberty attempts to rely on this court’s decision in a\nprevious suit between Liberty and MEBA, in which Liberty\nraised and lost a similar jurisdictional argument. Liberty cites\nthis case for the proposition that three categories of legal claims\nare preempted by the NLRA under Garmon: claims over which\nthe NLRB “has already exercised jurisdiction,” claims that call\nfor “an initial decision in the representation area,” and claims\n“in which the center of the dispute is a representational\nquestion.” Appellant Br. 35–36 (quoting Liberty Maritime I,\n815 F.3d at 841 (citations and quotation marks omitted)).\nLiberty urges this court to evaluate whether MEBA’s claim is\n“primarily representational or primarily contractual,” as\nseveral other circuits do. See, e.g., United Food & Commercial\nWorkers Union, Local 400 v. Shoppers Food Warehouse Corp.,\n35 F.3d 958, 961 (4th Cir. 1994); Local Union 204 of Int’l Bhd.\nof Elec. Workers, AFL-CIO v. Iowa Elec. Light & Power Co.,\n668 F.2d 413, 419 (8th Cir. 1982). Liberty claims such analysis\nwould show this suit may fall into the purported third category\nof claims preempted under Garmon: that the center of the\ndispute may concern a representational matter, such as whether\n\f 8\nMEBA or another union has representational rights over\ncrewmembers of the Liberty Peace. In order to determine the\ntrue center of this dispute, Liberty argues the district court\nshould have considered MEBA’s grievance letter instead of\nfocusing only on MEBA’s complaint. Liberty contends it was\nlegal error for the district court to have concluded that, as a\nmatter of law, it maintained jurisdiction over MEBA’s claims.\n\n MEBA responds that Liberty’s argument on appeal\n“conflates the type of claim with the effect of a claim’s\nenforcement.” Appellee Br. 26 (quoting Liberty Maritime I,\n815 F.3d at 843). While the possible outcome of its suit may\ntouch on representational issues, MEBA argues it has a contract\nthat requires Liberty to arbitrate, and MEBA asked the district\ncourt to compel compliance with that contract. MEBA\nmaintains this type of contractual dispute is squarely covered\nby Section 301 of the LMRA.\n\n The district court properly exercised jurisdiction over\nMEBA’s claim under the plain meaning of Section 301 as well\nas established Supreme Court and Circuit precedent. Section\n301 covers “[s]uits for violations of contracts between an\nemployer and a labor organization.” Liberty Maritime I, 815\nF.3d at 840 (quoting 29 U.S.C. § 185(a)). As the Supreme\nCourt has stated, Section 301 “permits suits for breach of a\ncollective bargaining agreement regardless of whether the\nparticular breach is also an unfair labor practice within the\njurisdiction of the Board.” Vaca v. Sipes, 386 U.S. 171, 179–\n80 (1967) (emphasis added); see also Carey v. Westinghouse\nCorp., 375 U.S. 261, 267–68 (1964) (holding that Section 301\ngives a federal court jurisdiction over a suit to enforce an\narbitration clause in a collective bargaining agreement even if\nthe case is “truly a representation case” that could also be heard\nby the NLRB under Section 9 of the NLRA). Thus, the\n“Garmon doctrine is ‘not relevant’ to actions within the\n\f 9\npurview of § 301” of the LMRA. Arnold, 417 U.S. at 16 (citing\nLocal 174, Teamsters v. Lucas Flour Co., 369 U.S. 95, 101 n.9\n(1962)). We held in Liberty Maritime I that federal courts and\nthe NLRB have concurrent jurisdiction over claims that are\n“both contractual and representational.” 815 F.3d at 840\n(emphasis original) (citing Arnold, 417 U.S. at 16, and Smith v.\nEvening News Ass’n, 371 U.S. 195, 197 (1962)).\n\n Thus, if a case is both representational and contractual, it\nis treated as a “hybrid” claim. See, e.g., United Parcel Serv.,\nInc. v. Mitchell, 451 U.S. 56, 66 (1981) (Stewart, J., concurring\nin the judgment) (describing “a hybrid ‘§ 301 and breach of\nduty suit’”); DelCostello v. Int’l Bhd. of Teamsters, 462 U.S.\n151, 165 (1983) (describing “hybrid § 301/fair representation\nlitigation”); Cephas v. MVM, Inc., 520 F.3d 480, 485 (D.C. Cir.\n2008) (same). Such “hybrid” claims create concurrent\njurisdiction for the federal courts and the NLRB, but they do\nnot divest courts of their statutory jurisdiction. Consistent with\nSupreme Court precedent, that is precisely what this court has\nheld: “Instead of forcing courts to shoehorn a hybrid claim into\none category or the other, the Supreme Court has held that they\nretain jurisdiction to hear a contractual claim even if the claim\nis also representational.” Liberty Maritime I, 815 F.3d at 840\n(citing Arnold, 417 U.S. at 16). “[F]ederal courts have\nindependent jurisdiction to decide cases alleging a breach of\ncollective bargaining agreements, even though that very breach\nmay also be an unfair labor practice.” Mullins v. Kaiser Steel\nCorp., 642 F.2d 1302, 1316 (D.C. Cir. 1980), rev’d on other\ngrounds, 455 U.S. 72 (1982).\n\n Allowing “hybrid” claims to be brought in federal court\nreads together the two statutes, the LMRA and the NLRA,\ngiving effect to Congress’s provision of federal court\njurisdiction for contractual claims and NLRB jurisdiction over\nrepresentational claims. See Vaca, 386 U.S. at 179–80;\n\f 10\nWachovia Bank v. Schmidt, 546 U.S. 303, 315–16 (2006)\n(“under the in pari materia canon of statutory construction,\nstatutes addressing the same subject matter generally should be\nread as if they were one law”) (citations and quotation marks\nomitted).\n\n Liberty continues to argue that if a case is both\nrepresentational and contractual, a district court must place\nthose claims on a sliding scale to determine if the case is\nprimarily one or the other. Neither the Supreme Court nor this\ncourt have required such an inquiry. In Liberty Maritime I, we\ndescribed and discussed the practice in some circuits, which\n“examine the major issues to be decided” and “determine\nwhether they can be characterized as primarily representational\nor primarily contractual” in order to dismiss “primarily\nrepresentational” claims. 1 Liberty Maritime I, 815 F.3d at 840\n(alterations, quotation marks, and citation omitted).\n\n Liberty’s reliance on the categories recognized by other\ncircuits is misplaced. While the Liberty Maritime I court\ndescribed the approaches from “several of our sister circuits,”\nit did not adopt any of these competing decisions. Id. at 841.\nBecause the categories identified by other circuits were not\nnecessary to the decision, the Liberty Maritime I discussion of\nthose cases “does not constitute a precedent to be followed with\nrespect to that issue.” UC Health v. NLRB, 803 F.3d 669, 682\n(D.C. Cir. 2015) (quotation marks and citations omitted). We\nfollow Liberty Maritime I, which refused to define the\n\n1\n Compare Iowa Elec. Light & Power Co., 668 F.2d at 419; Paper,\nAllied–Indus., Chem. & Energy Workers Int’l Union v. Air Prods. &\nChems., Inc., 300 F.3d 667, 675 (6th Cir. 2002); Pace v. Honolulu\nDisposal Serv., Inc., 227 F.3d 1150, 1156 (9th Cir. 2000); Shoppers,\n35 F.3d at 961; Copps Food Ctr., Inc. v. United Food & Commercial\nWorkers Union, Local 73–A, No. 90–1905, 1991 WL 135508, at *2\n(7th Cir. July 23, 1991) (unpublished).\n\f 11\n“parameters of a claim that is ‘primarily representational’ as\nopposed to ‘primarily contractual’” and declined “to shoehorn”\na given “claim into one category or the other.” 815 F.3d at 840–\n41.\n\n Liberty also contends that it cannot “be ruled out, based on\nthe pleadings alone, that the major issues to be decided are\nprimarily representational.” Appellant Br. 38. MEBA’s\ngrievance letter, like its complaint, however, raised numerous\ncontractual issues. See MEBA Compl. Exhibit G (“Please be\nadvised that Liberty is in violation of the parties’ collective\nbargaining agreement . . . . Consider this notice of an official\ngrievance for violation of the parties’ labor contract . . . . We\ndemand that the parties participate in expedited arbitration to\nresolve this contractual dispute . . . .”). And Liberty does not\nargue that the claim here is exclusively representational. See\nLiberty Maritime I, 815 F.3d at 843 (“Garmon preemption is\ndesigned to prevent a court from deciding a claim that can only\nbe characterized as representational.”) (emphasis added). At\nmost, then, Liberty has left open the possibility that this case\ninvolves a hybrid claim raising both contractual and\nrepresentational questions. As discussed, however, such hybrid\nclaims are subject to the concurrent jurisdiction of the NLRB\nand the federal courts. Id. at 840. To hold otherwise “would\nfrustrate rather than serve the congressional policy expressed\nin [Section 301].” Smith, 371 U.S. at 200.\n\n Finally, it may be true, as Liberty stresses, that “a party’s\nmere assertion that a claim is contractual is not an automatic\nticket to federal court” under the LMRA. Appellant Br. 35\n(quoting Liberty Maritime I, 815 F.3d at 840). A plaintiff must\nplausibly demonstrate the dispute falls within the terms of\nSection 301, and is not an “end run around [the\nNLRA] . . . under the guise of contract interpretation.” Pace,\n227 F.3d at 1157 (citation and alteration omitted); accord\n\f 12\nPaper, 300 F.3d at 675 (“[S]imply referring to the claim as a\n‘breach of contract’ was insufficient for purposes of § 301\nfederal courts’ jurisdiction.”). This proposition, however, goes\nlittle further than the axiom that a plaintiff’s claim must invoke\na proper basis for federal court jurisdiction within a well-\npleaded complaint. Cf. Greenhill v. Spellings, 482 F.3d 569,\n575 (D.C. Cir. 2007) (under the well-pleaded complaint rule,\njurisdiction arising under federal law is established by looking\nto the legal basis of plaintiff’s claim) (citing Louisville &\nNashville R.R. Co. v. Mottley, 211 U.S. 149, 152–53 (1908));\nsee also 13D Charles Alan Wright, Arthur R. Miller, Edward\nH. Cooper, & Richard D. Freer, Federal Practice and\nProcedure § 3566, at 261–62 (3d ed. 2008) (“The well-pleaded\ncomplaint rule stands for the proposition that the court, in\ndetermining whether the case arises under federal law, will\nlook only to the claim itself.”).\n\n Put simply, Congress gave federal courts jurisdiction to\nhear contractual claims between labor organizations and\nemployers in Section 301 of the LMRA. The existence of\nrepresentational issues does not divest the federal courts of\njurisdiction. Here, MEBA’s suit alleges a breach of the parties’\nlabor contract. The suit requires a judicial determination as to\nwhether an arbitration clause in the agreements between\nLiberty and MEBA covers the dispute over the Liberty Peace.\nSee Gen. Elec. Co. v. Local 205, United Elec., Radio & Mach.\nWorkers of Am., 353 U.S. 547, 548 (1957) (Section “301(a)\nfurnishes a body of federal substantive law for the enforcement\nof collective bargaining agreements” that provides for suits “to\nenforce the obligation to arbitrate grievance disputes.”); see\nalso Westinghouse, 375 U.S. at 267–68. As the dispute includes\ncontractual claims, the district court properly concluded subject\nmatter jurisdiction was established under Section 301 of the\nLMRA.\n\f 13\n III.\n\n Proceeding to the merits of Liberty’s appeal, we consider\nnext whether the district court properly granted MEBA’s\nmotion for judgment on the pleadings.\n\n A.\n\n This court reviews a Rule 12(c) judgment on the pleadings\nde novo. Judicial Watch, Inc. v. United States Dep’t of\nHomeland Sec., 895 F.3d 770, 777 (D.C. Cir. 2018); Mpoy v.\nRhee, 758 F.3d 285, 287 (D.C. Cir. 2014). Federal Rule of Civil\nProcedure 12(c) provides, “After the pleadings are closed—but\nearly enough not to delay trial—a party may move for\njudgment on the pleadings.” Pleadings include any “copy of a\nwritten instrument that is an exhibit to a pleading,” Fed. R. Civ.\nP. 10(c), such as relevant and authentic documents attached to\nthe complaint. See, e.g., Philips v. Pitt Cty. Mem’l Hosp., 572\nF.3d 176, 180 (4th Cir. 2009).\n\n Very few of our precedents discuss Rule 12(c), in part\nbecause judgment on the pleadings is rare. As Wright & Miller\nnotes, “Federal Rule 12(c) has its historical roots in common\nlaw practice, which permitted either party, at any point in the\nproceeding, to demur to his opponent’s pleading and secure a\ndismissal or final judgment on the basis of the pleadings.” 5C\nWright & Miller § 1367, at 205; see also Patel v. Contemp.\nClassics of Beverly Hills, 259 F.3d 123, 126 (2d Cir. 2001)\n(“[A] motion for judgment on the pleadings is the direct\ndescendant of that ancient leper of the common law, the\n‘speaking demurrer.’”); 5C Wright & Miller § 1369, at 265\n(noting “the Rule 12(c) motion is little more than a relic of the\ncommon law and code eras”).\n\n Because Rule 12(c) provides judicial resolution at an early\nstage of a case, the party seeking judgment on the pleadings\n\f 14\nshoulders a heavy burden of justification. A reviewing court\n“will affirm the district court if the moving party demonstrates\nthat no material fact is in dispute and that it is entitled to\njudgment as a matter of law.” Peters v. Nat’l R.R. Passenger\nCorp., 966 F.2d 1483, 1485 (D.C. Cir. 1992) (quotation marks\nomitted). The moving party must demonstrate its entitlement to\njudgment in its favor, even though the “court evaluating the\n12(c) motion will accept as true the allegations in the\nopponent’s pleadings, and as false all controverted assertions\nof the movant.” Haynesworth v. Miller, 820 F.2d 1245, 1249\nn.11 (D.C. Cir. 1987) (collecting cases), abrogated on other\ngrounds by Hartman v. Moore, 547 U.S. 250 (2006); see also\nBeal v. Missouri Pac. R.R. Corp., 312 U.S. 45, 51 (1941) (when\nthe plaintiff moves for judgment on the pleadings, the\ndefendant’s “denials and allegations of the answer which are\nwell pleaded must be taken as true”). We must give “all\nreasonable inferences to the opponent’s pleadings” before\nentering a judgment on the pleadings. Wager v. Pro, 575 F.2d\n882, 884 (D.C. Cir. 1976).\n\n Under this standard, “a judgment on the pleadings is not\nappropriate” if there are “issues of fact which if proved would\ndefeat recovery,” “even if the trial court is convinced that the\nparty opposing the motion is unlikely to prevail at trial.” Id.\n“[I]f material questions of fact are presented by the pleadings,\nthe remedy by motion for judgment on the pleadings under\nRule 12(c) is not available.” Noel v. Olds, 149 F.2d 13, 14 &\nn.7 (D.C. Cir. 1945) (citing James A. Pike, Objections to\nPleadings Under the New Federal Rules of Civil Procedure, 47\nYale L.J. 50 (1937)).\n\n B.\n\n Applying these standards, MEBA’s pleadings and\nattachments cannot carry the day. The district court erred in\n\f 15\ngranting MEBA’s Rule 12(c) motion. The district court found\nthat Liberty and MEBA agreed to arbitrate their disputes in the\nMaster Agreements, which included clauses stating: “All\ndisputes relating to the interpretation or performance of this\nAgreement shall be determined in accordance with the\nprovisions of this Section.” Dist. Ct. Op. at 13. The district\ncourt acknowledged that Liberty contested “whether a\ncollective bargaining agreement exists between the parties\nunder which the union can assert its right to arbitrate” and that\nLiberty argued the absence of such an agreement rendered\njudgment on the pleadings inappropriate. Id. at 9–10 & n.7. But\nthe district court construed Liberty’s assertion that no\nagreement existed as a legal conclusion, not a factual dispute.\nBecause legal conclusions about a collective bargaining\nagreement are accorded “no special deference,” id. at n.7\n(quoting Local Union No. 47, Int’l Bhd. Of Elec. Workers v.\nNLRB, 927 F.2d 635, 640 (D.C. Cir. 1991)), the district court\nfound no material fact about the contract to be in dispute.\n\n On appeal, Liberty argues that its denial was factual, not\nlegal, and so failure to accept its denial as true was legal error;\nthat the complete contractual terms and scope were not before\nthe district court on the pleadings, as recognized by the district\ncourt’s opinion and the pleadings; and that without a full\ncontract to interpret, the order to arbitrate was error. Echoing\nthe district court, MEBA simply responds that Liberty’s claim\nthat it did not have an agreement with MEBA covering the\nLiberty Peace is a legal conclusion, not a dispute of material\nfact. MEBA also argues that Liberty never denied the existence\nof a collective bargaining agreement that contains a broad\narbitration clause, nor that those agreements are still in effect.\nIt claims that Liberty’s “tactic” is “smoke and mirrors.”\nAppellee Br. 10. MEBA maintains that Liberty contradicted its\nappellate theory of the case in litigating the case below,\nclaiming that Liberty conceded the existence of the contract\n\f 16\nand disputed only whether the contract’s broad terms covered\nthe Liberty Peace. Essentially, MEBA claims that Liberty’s\nappeal is rooted in procedural technicalities, rather than a good\nfaith factual dispute over whether an arbitration clause exists in\nan authentic, extant contract.\n\n Giving all reasonable inferences to Liberty, Liberty raised\nmaterial issues of fact that rendered judgment on the pleadings\ninappropriate. First, Liberty properly disputed the existence of\na contract to arbitrate in this case. Without a contract binding\nthe parties to arbitrate, an order compelling arbitration is\nimproper. “[A]rbitration is a matter of contract and a party\ncannot be required to submit to arbitration any dispute which\nhe has not agreed so to submit.” AT&T Techs., Inc. v.\nCommc’ns Workers of Am., 475 U.S. 643, 648 (1986)\n(quotation marks and citation omitted). Such contracts must be\n“interpreted as a whole.” United States v. Hunt, 843 F.3d 1022,\n1028 (D.C. Cir. 2016) (quoting Restatement (Second) of\nContracts § 202(2)). And although “the determination that\nparties have contractually bound themselves to arbitrate\ndisputes . . . is a legal conclusion[,] . . . the findings upon which\nthat conclusion is based are factual.” Bailey v. Fed. Nat’l\nMortg. Ass’n, 209 F.3d 740, 744 (D.C. Cir. 2000) (quoting\nChelsea Square Textiles, Inc. v. Bombay Dyeing & Mfg. Co.,\n189 F.3d 289, 295 (2d Cir. 1999)); see also 11 Williston on\nContracts § 30:3 (4th ed.) (“It is generally a question of fact . . .\nwhether or not a contract . . . actually exists.”).\n\n The existence of a contract to arbitrate must first be\nestablished through a factual determination of what constitutes\nthe parties’ full agreement. As such, the district court erred\nwhen it based its conclusion of arbitrability on a contract of\ngenuinely disputed authenticity. Liberty denied the authenticity\nof the copies of the Master Agreements that MEBA attached to\nits complaint. Specifically, MEBA alleged in its complaint:\n\f 17\n Liberty first became signatory to a collective bargaining\n agreement with MEBA in 1988 when the Company signed\n on to both the Union’s 1986-1990 Tanker Vessels Master\n Agreement and the 1986-1990 Dry Cargo Vessels Master\n Agreement. A copy of the 1986-1990 Tanker Vessels\n Master Agreement and Liberty’s signature page is attached\n hereto as Exhibit A. A copy of the 1986-1990 Dry Cargo\n Vessels Master Agreement is attached hereto as Exhibit B.\n These agreements cover all U.S. flagged vessels owned,\n managed or operated by Liberty.\n\nMEBA Compl. ¶ 8. In its answer, Liberty admitted that Liberty\nsigned the Master Agreements but denied all other allegations\nin ¶ 8 of the complaint not expressly admitted. This included\ndenying MEBA’s allegation that Exhibits A and B were copies\nof the Master Agreements. Liberty also denied that even the\nauthentic version of these agreements covered all U.S. flagged\nvessels managed by Liberty.\n\n Accepting Liberty’s allegations as true and making all\nreasonable inferences in Liberty’s favor, as the Rule 12(c)\nstandards require, means that Exhibits A and B were not\nauthenticated copies of the two Master Agreements. Without\nauthenticated Master Agreements, the district court lacked\nadequate factual support for its finding that the language of the\nagreements contained an extant arbitration clause. This dispute\nover authenticity created a material issue of fact that should\nhave been enough to defeat MEBA’s Rule 12(c) motion. See\nHorsley v. Feldt, 304 F.3d 1125, 1135 (11th Cir. 2002)\n(“Because the authenticity of the [documents] attached to the\namended answer is disputed, . . . they may not be considered in\ndeciding the Rule 12(c) motion for judgment on the\npleadings.”); see also Philips, 572 F.3d at 180.\n\f 18\n Second, several disputed contract provisions were not\nincluded in the pleadings and attachments. The pleadings\nshowed that, at least in 2012, Liberty and MEBA were parties\nto “a Memorandum of Understanding dated September 23,\n2005, as amended; various Side Letters, dated June 8, 2005,\nOctober 28, 2005, and July 14, 2010, respectively; and Letters\nof Understanding, dated July 7, 2009, and February 21, 2010,\nrespectively.” MEBA Compl. Exhibit C. Liberty conceded the\nauthenticity of the MOU. But Liberty maintained that the six\ndocuments referenced in the MOU, but not attached to\nMEBA’s complaint, were relevant to whether any collective\nbargaining agreement or arbitration clause applied to the new\nvessel.\n\n Liberty stresses the pre-suit correspondence attached to\nMEBA’s complaint referenced over a dozen other documents,\nonly three of which were attached to the complaint, and only\none of which was conceded to be authentic. Liberty also\ndisputed that a contract covered the Liberty Peace on the basis\nthat MEBA failed to define “this Agreement” in the arbitration\nclause. Because the parties’ contractual relationship has been\nmodified by numerous agreements over the decades since the\nMaster Agreements were signed, the record does not make\nclear to what “this Agreement” refers.\n\n The district court, however, assumed these numerous side\nletters and letters of understanding did not modify the\nagreement in any meaningful way, accepting that “the\ngrievance and arbitration procedures contained in the Tanker\nand Dry Cargo Master Agreements remain binding” without\nhaving reviewed the “terms and conditions of employment of\nthe [collective bargaining agreements], side letters, and letters\nof understanding.” Dist. Ct. Op. at 3. These assumptions were\ninappropriate in the face of controverted facts about the\nAgreements’ content. The pleadings and limited attachments\n\f 19\ndid not support the conclusion that “no material fact is in\ndispute.” Peters, 966 F.2d at 1485.\n\n Third, the district court drew improper inferences against\nLiberty, the opposing party, rather than against MEBA, the\nmoving party. The district court concluded “Liberty failed to\npoint to any evidence to rebut the presumption of arbitrability.”\nDist. Ct. Op. at 14. Yet at this stage of the proceedings, it was\nMEBA’s burden to demonstrate the existence of an applicable\narbitration agreement, not Liberty’s burden to rebut it. See\nBeal, 312 U.S. at 51. Although the district court correctly\nidentified the federal policy in favor of arbitration agreements,\nMoses H. Cone Mem’l Hosp. v. Mercury Constr. Corp., 460\nU.S. 1, 24–25 (1983), that policy becomes a presumption only\nwhen the factual existence of a contract for arbitration has been\nestablished. See AT&T, 475 U.S. at 648; Bailey, 209 F.3d at\n744. The district court also found that “no agreement between\nthe parties contains language excluding this sort of dispute\nfrom arbitration.” Dist. Ct. Op. at 14. But because the whole\n“agreement” was not before the district court, it erred by\ninferring, in a manner that favored MEBA, that the “cont[ent]”\nand “language” of the agreement did not exclude the dispute\nover the Liberty Peace from arbitration.\n\n Fourth, the district court erred by refusing to allow\ndiscovery over Liberty’s objection that there was a material\nfactual dispute about the existence of an agreement to arbitrate\nissues relating to the Liberty Peace. See Dec. 14, 2017 Minute\nOrder. Discovery would have required the district court to treat\nMEBA’s motion as one for summary judgment. Fed. R. Civ. P.\n12(d) (“If, on a motion under Rule 12(b)(6) or 12(c), matters\noutside the pleadings are presented to and not excluded by the\ncourt, the motion must be treated as one for summary judgment\nunder Rule 56. All parties must be given a reasonable\nopportunity to present all the material that is pertinent to the\n\f 20\nmotion.”). Granting such discovery would have allowed\nLiberty to attempt to substantiate, or MEBA to refute, the claim\nthat the complete contractual record, when read as a whole,\ndoes not require arbitration of this dispute.\n\n Despite MEBA’s claim that Liberty’s argument consists of\n“smoke and mirrors,” MEBA has not shown that Liberty lacks\na good faith basis for its appeal. Parties are presumed to have a\ngood faith basis for denying allegations, even at the pleading\nstage. See Amnesty Am. v. Town of W. Hartford, 361 F.3d 113,\n131 (2d Cir. 2004) (“Because attorneys, as officers of the court,\nare presumed not to offer in opposition” to dispositive motions\n“evidence that they have no good faith basis to believe will be\navailable or admissible at trial, the burden is on the moving\nparty.”). For the reasons stated above and in light of the\nincomplete contractual record, we assume there was a good\nfaith basis for Liberty to challenge judgment on MEBA’s\npleadings. We expect the parties, on remand, will act in good\nfaith to narrow their dispute and avoid burying the district court\nwith reams of contractual provisions not arguably relevant to\nthe arbitrability issue in this case.\n\n ***\n\n This case involved disputed issues of material fact that\nrendered the unusual remedy of judgment on the pleadings\ninappropriate. Because the district court should have first\ndetermined whether Liberty and MEBA had a valid contract\nfor arbitration by looking at their whole agreement, we reverse\nthe judgment on the pleadings and remand the case for further\nproceedings consistent with this opinion.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: DISTRICT NO. 1, PACIFIC COAST DISTRICT, MARINE ENGINEERS BENEFICIAL ASSOCIATION, AFL-CIO, Appellee v. LIBERTY MARITIME CORPORATION, Appellant\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-09\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued May 10, 2019 Decided August 9, 2019\\n\\n No. 18-7148\\n\\n DISTRICT NO. 1, PACIFIC COAST DISTRICT, MARINE\\n ENGINEERS BENEFICIAL ASSOCIATION, AFL-CIO,\\n APPELLEE\\n\\n v.\\n\\n LIBERTY MARITIME CORPORATION,\\n APPELLANT\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:17-cv-02173)\\n\\n\\n Steffen N. Johnson argued the cause for appellant. With\\nhim on the briefs were William G. Miossi and Paul N. Harold.\\n\\n Mark J. Murphy argued the cause and filed the brief for\\nappellee.\\n\\n Before: MILLETT, KATSAS, and RAO, Circuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge RAO.\\n\\n RAO, Circuit Judge: Liberty Maritime Corporation\\n(Liberty) is a shipping company that has contracted over the\\npast thirty years with District No. 1, Pacific Coast District,\\n\\f 2\\nMarine Engineers Beneficial Association, AFL-CIO (MEBA),\\na labor union representing supervisory employees in the\\nmaritime industry. This case arises out of an underlying dispute\\nabout whether Liberty was contractually required to hire\\nMEBA employees on a new vessel managed by Liberty.\\nMEBA sued in the United States District Court for the District\\nof Columbia, claiming its contract with Liberty required the\\nparties to submit the dispute to arbitration. The district court\\nruled in favor of the union, granting judgment on the pleadings\\nunder Federal Rule of Civil Procedure 12(c) and compelling\\narbitration. Liberty timely appealed, arguing that the district\\ncourt lacked subject matter jurisdiction, or in the alternative,\\nerred in its application of the Rule 12(c) standards.\\n\\n For the reasons explained below, we agree that the district\\ncourt had jurisdiction over MEBA’s claim under Section 301\\nof the Labor Management Relations Act of 1947 (LMRA), 29\\nU.S.C. §§ 141 et seq., which provides federal jurisdiction over\\nsuits for “violation of contracts between an employer and a\\nlabor organization.” Id. § 185(a). MEBA raised contractual\\nissues regarding the arbitrability of the dispute and thus its\\nclaim clearly falls within the district court’s statutory\\njurisdiction. Although Liberty alleges that the dispute primarily\\nraised representational issues and thus should be within the\\nexclusive jurisdiction of the National Labor Relations Board\\n(NLRB) under the doctrine of “Garmon preemption,” federal\\ncourts retain jurisdiction over “hybrid” claims raising both\\ncontractual and representational issues. Dist. No. 1, Pac. Coast\\nDist., Marine Engineers’ Beneficial Ass’n, AFL-CIO v. Liberty\\nMar. Corp., 815 F.3d 834, 840 (D.C. Cir. 2016) (“Liberty\\nMaritime I”); see also William E. Arnold Co. v. Carpenters\\nDist. Council of Jacksonville & Vicinity, 417 U.S. 12, 18\\n(1974).\\n\\f 3\\n Although jurisdiction here was proper, we reverse and\\nremand because material facts remained in dispute regarding\\nthe existence of an applicable arbitration clause, and therefore\\nMEBA was not entitled to judgment on the pleadings under\\nRule 12(c).\\n\\n I.\\n\\n Appellant Liberty is a shipping company that transports\\ncommodities, vehicles, equipment, and other cargoes on the\\nseagoing vessels it manages. Liberty’s clients include the U.S.\\nGovernment, the United Nations, and commercial entities such\\nas automobile manufacturers. Liberty manages vessels\\ntransporting bulk cargo—including dry bulk, break bulk, and\\nbagged commodities—and “roll on/roll-off” vessels, like car\\nand truck carriers configured to transport vehicles that drive on\\nand off the vessel. Many of these vessels are enrolled in the\\nU.S. Maritime Security Program, a federal program that\\nsubsidizes shipping companies for national security\\npurposes—namely, to ensure a fleet of vessels is available in\\nthe event of a war or national emergency. See generally 46\\nU.S.C. §§ 53101 et seq. Appellee MEBA is a labor\\norganization that represents supervisory employees in the U.S.\\nmaritime industry at ports throughout the United States and on\\noceangoing vessels. On car and truck carrier vessels operated\\nby Liberty and enrolled in the U.S. Maritime Security Program,\\nMEBA represents licensed officers and engineers.\\n\\n The parties’ relationship began in 1988 when they signed\\ntwo agreements: the Tanker Vessels Master Agreement and the\\nDry Cargo Vessels Master Agreement. Although the\\nauthenticity of some of the documents attached to the pleadings\\nis disputed, the documents that purport to be current copies of\\nthese Master Agreements provide that “[a]ll disputes relating\\nto the interpretation or performance of this Agreement shall be\\n\\f 4\\ndetermined in accordance with the provisions of this Section.”\\n“[T]his Section” states that grievances will be presented to a\\nlicensed personnel board consisting of two persons appointed\\nby the union and two persons appointed by the company; if the\\nlicensed personnel board fails to resolve a grievance, an\\narbitrator will assume jurisdiction over the grievance.\\n\\n Over the past three decades, the parties have modified their\\ncontractual relationship on numerous occasions. At this stage\\nof the proceedings, the record includes only a few of these\\nagreements. Both parties agree, however, they were signatories\\nto a 2012 Memorandum of Understanding (MOU). This MOU\\nidentifies numerous prior agreements and states that prior\\nagreements will remain in effect except as expressly modified,\\nbut the MOU does not expressly modify any arbitration clause\\nin a manner relevant to this case.\\n\\n This suit arises out of a dispute between Liberty and\\nMEBA over a ship named the M/V Liberty Peace. On July 24,\\n2017, Liberty sent MEBA a letter stating its intention to\\ncommence managing this foreign flagged car and truck carrier\\nvessel and operate it as a U.S. flagged vessel. In the letter,\\nLiberty claimed the Liberty Peace would not fall under the\\nparties’ collective bargaining agreements and the various\\ncontractual modifications of those agreements because the\\nvessel would not be enrolled in the U.S. Maritime Security\\nProgram. MEBA disagreed, insisting the existing agreements\\ncovered the new vessel. Although the parties met to discuss the\\nmatter, they did not resolve their dispute. In the meantime,\\nLiberty began managing the Liberty Peace as the agent of a\\nthird party, and that third party entered into labor agreements\\nwith a different union.\\n\\n MEBA sent Liberty a grievance letter on August 31, 2017,\\nasserting Liberty was “in violation of the parties’ collective\\n\\f 5\\nbargaining agreement by failing to apply the terms and\\nconditions of the parties’ labor contract” to the Liberty Peace.\\nLiberty did not submit MEBA’s grievance to arbitration.\\n\\n MEBA subsequently filed a “Complaint to Compel\\nArbitration” in the United States District Court for the District\\nof Columbia. MEBA requested the district court compel\\nLiberty to participate in the arbitration process set forth in the\\nparties’ collective bargaining agreement and grant any other\\nappropriate relief, including attorneys’ fees and costs. MEBA\\nattached as exhibits several documents purporting to be the two\\noriginal Master Agreements, the MOU, MEBA’s August 31\\ngrievance letter, and some additional correspondence between\\nMEBA and Liberty.\\n\\n In its answer to MEBA’s complaint, Liberty admitted it\\nhad signed the Master Agreements and the MOU. Liberty\\nadmitted the authenticity of the MOU, but denied the\\nauthenticity of the exhibits MEBA claimed were copies of the\\nMaster Agreements. Liberty denied that the MOU incorporated\\nthe terms of the Master Agreements and that the arbitration\\nclauses covered the Liberty Peace. Liberty also denied that any\\nlabor contract or arbitration agreement with MEBA covered the\\nLiberty Peace. As an affirmative defense, Liberty alleged the\\ndistrict court lacked subject matter jurisdiction because the suit\\nconcerned representational rights and therefore was preempted\\nby the jurisdiction of the NLRB under the terms of the National\\nLabor Relations Act (NLRA).\\n\\n MEBA moved for judgment on the pleadings under\\nFederal Rule of Civil Procedure 12(c), and the district court\\ngranted the motion. The district court found that the Master\\nAgreements stated, “[a]ll disputes relating to the interpretation\\nor performance of this Agreement shall be determined in\\naccordance with the provisions of this Section.” Dist. Ct. Op.\\n\\f 6\\nat 3, 13. The district court concluded that this language created\\na presumption of arbitrability; Liberty failed to rebut the\\npresumption; and no agreement between the parties excluded\\nthis sort of dispute from arbitration. Id. at 13–14. The district\\ncourt also rejected Liberty’s preemption argument on the\\ngrounds that federal courts have jurisdiction over contractual\\nmatters and that MEBA’s suit “plainly requires deciding a\\ncontractual matter: whether the arbitration clause covers the\\ndispute at issue.” Id. at 10–11 n.7.\\n\\n Liberty timely appealed, challenging the district court’s\\norder on jurisdictional grounds and arguing the district court\\nviolated Rule 12(c) by making findings the pleadings did not\\nadequately support.\\n\\n II.\\n\\n “The ‘first and fundamental question’ that we are ‘bound\\nto ask and answer’ is whether the court has jurisdiction to\\ndecide the case.” Bancoult v. McNamara, 445 F.3d 427, 432\\n(D.C. Cir. 2006) (quoting Steel Co. v. Citizens for a Better\\nEnv’t, 523 U.S. 83, 94 (1998)). The district court held that\\nfederal courts have jurisdiction over contractual matters under\\nSection 301 of the LMRA, which provides:\\n\\n Suits for violation of contracts between an employer and a\\n labor organization representing employees in an industry\\n affecting commerce as defined in this chapter, . . . may be\\n brought in any district court of the United States having\\n jurisdiction of the parties . . . .\\n\\n29 U.S.C. § 185(a). Section 301 confers federal court\\njurisdiction over suits for breach of collective bargaining\\nagreements, which are contractual. “Congress deliberately\\nchose to leave the enforcement of collective agreements to the\\n\\f 7\\nusual processes of the law.” Charles Dowd Box Co. v.\\nCourtney, 368 U.S. 502, 513 (1962).\\n\\n Nevertheless, Liberty argues the district court lacked\\nsubject matter jurisdiction over this case under the judicially\\ncreated doctrine known as “Garmon preemption.” Washington\\nServ. Contractors Coal. v. Dist. Columbia, 54 F.3d 811, 815\\n(D.C. Cir. 1995) (citing San Diego Bldg. Trades Council v.\\nGarmon, 359 U.S. 236 (1959)). This doctrine holds that\\n“[w]hen an activity is arguably subject to § 7 or § 8 of the\\n[NLRA], . . . the federal courts must defer to the exclusive\\ncompetence of the [NLRB].” Garmon, 359 U.S. at 245. Suits\\nimplicating § 7 or § 8 of the NLRA are often described as\\n“representational.”\\n\\n Liberty attempts to rely on this court’s decision in a\\nprevious suit between Liberty and MEBA, in which Liberty\\nraised and lost a similar jurisdictional argument. Liberty cites\\nthis case for the proposition that three categories of legal claims\\nare preempted by the NLRA under Garmon: claims over which\\nthe NLRB “has already exercised jurisdiction,” claims that call\\nfor “an initial decision in the representation area,” and claims\\n“in which the center of the dispute is a representational\\nquestion.” Appellant Br. 35–36 (quoting Liberty Maritime I,\\n815 F.3d at 841 (citations and quotation marks omitted)).\\nLiberty urges this court to evaluate whether MEBA’s claim is\\n“primarily representational or primarily contractual,” as\\nseveral other circuits do. See, e.g., United Food & Commercial\\nWorkers Union, Local 400 v. Shoppers Food Warehouse Corp.,\\n35 F.3d 958, 961 (4th Cir. 1994); Local Union 204 of Int’l Bhd.\\nof Elec. Workers, AFL-CIO v. Iowa Elec. Light & Power Co.,\\n668 F.2d 413, 419 (8th Cir. 1982). Liberty claims such analysis\\nwould show this suit may fall into the purported third category\\nof claims preempted under Garmon: that the center of the\\ndispute may concern a representational matter, such as whether\\n\\f 8\\nMEBA or another union has representational rights over\\ncrewmembers of the Liberty Peace. In order to determine the\\ntrue center of this dispute, Liberty argues the district court\\nshould have considered MEBA’s grievance letter instead of\\nfocusing only on MEBA’s complaint. Liberty contends it was\\nlegal error for the district court to have concluded that, as a\\nmatter of law, it maintained jurisdiction over MEBA’s claims.\\n\\n MEBA responds that Liberty’s argument on appeal\\n“conflates the type of claim with the effect of a claim’s\\nenforcement.” Appellee Br. 26 (quoting Liberty Maritime I,\\n815 F.3d at 843). While the possible outcome of its suit may\\ntouch on representational issues, MEBA argues it has a contract\\nthat requires Liberty to arbitrate, and MEBA asked the district\\ncourt to compel compliance with that contract. MEBA\\nmaintains this type of contractual dispute is squarely covered\\nby Section 301 of the LMRA.\\n\\n The district court properly exercised jurisdiction over\\nMEBA’s claim under the plain meaning of Section 301 as well\\nas established Supreme Court and Circuit precedent. Section\\n301 covers “[s]uits for violations of contracts between an\\nemployer and a labor organization.” Liberty Maritime I, 815\\nF.3d at 840 (quoting 29 U.S.C. § 185(a)). As the Supreme\\nCourt has stated, Section 301 “permits suits for breach of a\\ncollective bargaining agreement regardless of whether the\\nparticular breach is also an unfair labor practice within the\\njurisdiction of the Board.” Vaca v. Sipes, 386 U.S. 171, 179–\\n80 (1967) (emphasis added); see also Carey v. Westinghouse\\nCorp., 375 U.S. 261, 267–68 (1964) (holding that Section 301\\ngives a federal court jurisdiction over a suit to enforce an\\narbitration clause in a collective bargaining agreement even if\\nthe case is “truly a representation case” that could also be heard\\nby the NLRB under Section 9 of the NLRA). Thus, the\\n“Garmon doctrine is ‘not relevant’ to actions within the\\n\\f 9\\npurview of § 301” of the LMRA. Arnold, 417 U.S. at 16 (citing\\nLocal 174, Teamsters v. Lucas Flour Co., 369 U.S. 95, 101 n.9\\n(1962)). We held in Liberty Maritime I that federal courts and\\nthe NLRB have concurrent jurisdiction over claims that are\\n“both contractual and representational.” 815 F.3d at 840\\n(emphasis original) (citing Arnold, 417 U.S. at 16, and Smith v.\\nEvening News Ass’n, 371 U.S. 195, 197 (1962)).\\n\\n Thus, if a case is both representational and contractual, it\\nis treated as a “hybrid” claim. See, e.g., United Parcel Serv.,\\nInc. v. Mitchell, 451 U.S. 56, 66 (1981) (Stewart, J., concurring\\nin the judgment) (describing “a hybrid ‘§ 301 and breach of\\nduty suit’”); DelCostello v. Int’l Bhd. of Teamsters, 462 U.S.\\n151, 165 (1983) (describing “hybrid § 301/fair representation\\nlitigation”); Cephas v. MVM, Inc., 520 F.3d 480, 485 (D.C. Cir.\\n2008) (same). Such “hybrid” claims create concurrent\\njurisdiction for the federal courts and the NLRB, but they do\\nnot divest courts of their statutory jurisdiction. Consistent with\\nSupreme Court precedent, that is precisely what this court has\\nheld: “Instead of forcing courts to shoehorn a hybrid claim into\\none category or the other, the Supreme Court has held that they\\nretain jurisdiction to hear a contractual claim even if the claim\\nis also representational.” Liberty Maritime I, 815 F.3d at 840\\n(citing Arnold, 417 U.S. at 16). “[F]ederal courts have\\nindependent jurisdiction to decide cases alleging a breach of\\ncollective bargaining agreements, even though that very breach\\nmay also be an unfair labor practice.” Mullins v. Kaiser Steel\\nCorp., 642 F.2d 1302, 1316 (D.C. Cir. 1980), rev’d on other\\ngrounds, 455 U.S. 72 (1982).\\n\\n Allowing “hybrid” claims to be brought in federal court\\nreads together the two statutes, the LMRA and the NLRA,\\ngiving effect to Congress’s provision of federal court\\njurisdiction for contractual claims and NLRB jurisdiction over\\nrepresentational claims. See Vaca, 386 U.S. at 179–80;\\n\\f 10\\nWachovia Bank v. Schmidt, 546 U.S. 303, 315–16 (2006)\\n(“under the in pari materia canon of statutory construction,\\nstatutes addressing the same subject matter generally should be\\nread as if they were one law”) (citations and quotation marks\\nomitted).\\n\\n Liberty continues to argue that if a case is both\\nrepresentational and contractual, a district court must place\\nthose claims on a sliding scale to determine if the case is\\nprimarily one or the other. Neither the Supreme Court nor this\\ncourt have required such an inquiry. In Liberty Maritime I, we\\ndescribed and discussed the practice in some circuits, which\\n“examine the major issues to be decided” and “determine\\nwhether they can be characterized as primarily representational\\nor primarily contractual” in order to dismiss “primarily\\nrepresentational” claims. 1 Liberty Maritime I, 815 F.3d at 840\\n(alterations, quotation marks, and citation omitted).\\n\\n Liberty’s reliance on the categories recognized by other\\ncircuits is misplaced. While the Liberty Maritime I court\\ndescribed the approaches from “several of our sister circuits,”\\nit did not adopt any of these competing decisions. Id. at 841.\\nBecause the categories identified by other circuits were not\\nnecessary to the decision, the Liberty Maritime I discussion of\\nthose cases “does not constitute a precedent to be followed with\\nrespect to that issue.” UC Health v. NLRB, 803 F.3d 669, 682\\n(D.C. Cir. 2015) (quotation marks and citations omitted). We\\nfollow Liberty Maritime I, which refused to define the\\n\\n1\\n Compare Iowa Elec. Light & Power Co., 668 F.2d at 419; Paper,\\nAllied–Indus., Chem. & Energy Workers Int’l Union v. Air Prods. &\\nChems., Inc., 300 F.3d 667, 675 (6th Cir. 2002); Pace v. Honolulu\\nDisposal Serv., Inc., 227 F.3d 1150, 1156 (9th Cir. 2000); Shoppers,\\n35 F.3d at 961; Copps Food Ctr., Inc. v. United Food & Commercial\\nWorkers Union, Local 73–A, No. 90–1905, 1991 WL 135508, at *2\\n(7th Cir. July 23, 1991) (unpublished).\\n\\f 11\\n“parameters of a claim that is ‘primarily representational’ as\\nopposed to ‘primarily contractual’” and declined “to shoehorn”\\na given “claim into one category or the other.” 815 F.3d at 840–\\n41.\\n\\n Liberty also contends that it cannot “be ruled out, based on\\nthe pleadings alone, that the major issues to be decided are\\nprimarily representational.” Appellant Br. 38. MEBA’s\\ngrievance letter, like its complaint, however, raised numerous\\ncontractual issues. See MEBA Compl. Exhibit G (“Please be\\nadvised that Liberty is in violation of the parties’ collective\\nbargaining agreement . . . . Consider this notice of an official\\ngrievance for violation of the parties’ labor contract . . . . We\\ndemand that the parties participate in expedited arbitration to\\nresolve this contractual dispute . . . .”). And Liberty does not\\nargue that the claim here is exclusively representational. See\\nLiberty Maritime I, 815 F.3d at 843 (“Garmon preemption is\\ndesigned to prevent a court from deciding a claim that can only\\nbe characterized as representational.”) (emphasis added). At\\nmost, then, Liberty has left open the possibility that this case\\ninvolves a hybrid claim raising both contractual and\\nrepresentational questions. As discussed, however, such hybrid\\nclaims are subject to the concurrent jurisdiction of the NLRB\\nand the federal courts. Id. at 840. To hold otherwise “would\\nfrustrate rather than serve the congressional policy expressed\\nin [Section 301].” Smith, 371 U.S. at 200.\\n\\n Finally, it may be true, as Liberty stresses, that “a party’s\\nmere assertion that a claim is contractual is not an automatic\\nticket to federal court” under the LMRA. Appellant Br. 35\\n(quoting Liberty Maritime I, 815 F.3d at 840). A plaintiff must\\nplausibly demonstrate the dispute falls within the terms of\\nSection 301, and is not an “end run around [the\\nNLRA] . . . under the guise of contract interpretation.” Pace,\\n227 F.3d at 1157 (citation and alteration omitted); accord\\n\\f 12\\nPaper, 300 F.3d at 675 (“[S]imply referring to the claim as a\\n‘breach of contract’ was insufficient for purposes of § 301\\nfederal courts’ jurisdiction.”). This proposition, however, goes\\nlittle further than the axiom that a plaintiff’s claim must invoke\\na proper basis for federal court jurisdiction within a well-\\npleaded complaint. Cf. Greenhill v. Spellings, 482 F.3d 569,\\n575 (D.C. Cir. 2007) (under the well-pleaded complaint rule,\\njurisdiction arising under federal law is established by looking\\nto the legal basis of plaintiff’s claim) (citing Louisville &\\nNashville R.R. Co. v. Mottley, 211 U.S. 149, 152–53 (1908));\\nsee also 13D Charles Alan Wright, Arthur R. Miller, Edward\\nH. Cooper, & Richard D. Freer, Federal Practice and\\nProcedure § 3566, at 261–62 (3d ed. 2008) (“The well-pleaded\\ncomplaint rule stands for the proposition that the court, in\\ndetermining whether the case arises under federal law, will\\nlook only to the claim itself.”).\\n\\n Put simply, Congress gave federal courts jurisdiction to\\nhear contractual claims between labor organizations and\\nemployers in Section 301 of the LMRA. The existence of\\nrepresentational issues does not divest the federal courts of\\njurisdiction. Here, MEBA’s suit alleges a breach of the parties’\\nlabor contract. The suit requires a judicial determination as to\\nwhether an arbitration clause in the agreements between\\nLiberty and MEBA covers the dispute over the Liberty Peace.\\nSee Gen. Elec. Co. v. Local 205, United Elec., Radio & Mach.\\nWorkers of Am., 353 U.S. 547, 548 (1957) (Section “301(a)\\nfurnishes a body of federal substantive law for the enforcement\\nof collective bargaining agreements” that provides for suits “to\\nenforce the obligation to arbitrate grievance disputes.”); see\\nalso Westinghouse, 375 U.S. at 267–68. As the dispute includes\\ncontractual claims, the district court properly concluded subject\\nmatter jurisdiction was established under Section 301 of the\\nLMRA.\\n\\f 13\\n III.\\n\\n Proceeding to the merits of Liberty’s appeal, we consider\\nnext whether the district court properly granted MEBA’s\\nmotion for judgment on the pleadings.\\n\\n A.\\n\\n This court reviews a Rule 12(c) judgment on the pleadings\\nde novo. Judicial Watch, Inc. v. United States Dep’t of\\nHomeland Sec., 895 F.3d 770, 777 (D.C. Cir. 2018); Mpoy v.\\nRhee, 758 F.3d 285, 287 (D.C. Cir. 2014). Federal Rule of Civil\\nProcedure 12(c) provides, “After the pleadings are closed—but\\nearly enough not to delay trial—a party may move for\\njudgment on the pleadings.” Pleadings include any “copy of a\\nwritten instrument that is an exhibit to a pleading,” Fed. R. Civ.\\nP. 10(c), such as relevant and authentic documents attached to\\nthe complaint. See, e.g., Philips v. Pitt Cty. Mem’l Hosp., 572\\nF.3d 176, 180 (4th Cir. 2009).\\n\\n Very few of our precedents discuss Rule 12(c), in part\\nbecause judgment on the pleadings is rare. As Wright & Miller\\nnotes, “Federal Rule 12(c) has its historical roots in common\\nlaw practice, which permitted either party, at any point in the\\nproceeding, to demur to his opponent’s pleading and secure a\\ndismissal or final judgment on the basis of the pleadings.” 5C\\nWright & Miller § 1367, at 205; see also Patel v. Contemp.\\nClassics of Beverly Hills, 259 F.3d 123, 126 (2d Cir. 2001)\\n(“[A] motion for judgment on the pleadings is the direct\\ndescendant of that ancient leper of the common law, the\\n‘speaking demurrer.’”); 5C Wright & Miller § 1369, at 265\\n(noting “the Rule 12(c) motion is little more than a relic of the\\ncommon law and code eras”).\\n\\n Because Rule 12(c) provides judicial resolution at an early\\nstage of a case, the party seeking judgment on the pleadings\\n\\f 14\\nshoulders a heavy burden of justification. A reviewing court\\n“will affirm the district court if the moving party demonstrates\\nthat no material fact is in dispute and that it is entitled to\\njudgment as a matter of law.” Peters v. Nat’l R.R. Passenger\\nCorp., 966 F.2d 1483, 1485 (D.C. Cir. 1992) (quotation marks\\nomitted). The moving party must demonstrate its entitlement to\\njudgment in its favor, even though the “court evaluating the\\n12(c) motion will accept as true the allegations in the\\nopponent’s pleadings, and as false all controverted assertions\\nof the movant.” Haynesworth v. Miller, 820 F.2d 1245, 1249\\nn.11 (D.C. Cir. 1987) (collecting cases), abrogated on other\\ngrounds by Hartman v. Moore, 547 U.S. 250 (2006); see also\\nBeal v. Missouri Pac. R.R. Corp., 312 U.S. 45, 51 (1941) (when\\nthe plaintiff moves for judgment on the pleadings, the\\ndefendant’s “denials and allegations of the answer which are\\nwell pleaded must be taken as true”). We must give “all\\nreasonable inferences to the opponent’s pleadings” before\\nentering a judgment on the pleadings. Wager v. Pro, 575 F.2d\\n882, 884 (D.C. Cir. 1976).\\n\\n Under this standard, “a judgment on the pleadings is not\\nappropriate” if there are “issues of fact which if proved would\\ndefeat recovery,” “even if the trial court is convinced that the\\nparty opposing the motion is unlikely to prevail at trial.” Id.\\n“[I]f material questions of fact are presented by the pleadings,\\nthe remedy by motion for judgment on the pleadings under\\nRule 12(c) is not available.” Noel v. Olds, 149 F.2d 13, 14 &\\nn.7 (D.C. Cir. 1945) (citing James A. Pike, Objections to\\nPleadings Under the New Federal Rules of Civil Procedure, 47\\nYale L.J. 50 (1937)).\\n\\n B.\\n\\n Applying these standards, MEBA’s pleadings and\\nattachments cannot carry the day. The district court erred in\\n\\f 15\\ngranting MEBA’s Rule 12(c) motion. The district court found\\nthat Liberty and MEBA agreed to arbitrate their disputes in the\\nMaster Agreements, which included clauses stating: “All\\ndisputes relating to the interpretation or performance of this\\nAgreement shall be determined in accordance with the\\nprovisions of this Section.” Dist. Ct. Op. at 13. The district\\ncourt acknowledged that Liberty contested “whether a\\ncollective bargaining agreement exists between the parties\\nunder which the union can assert its right to arbitrate” and that\\nLiberty argued the absence of such an agreement rendered\\njudgment on the pleadings inappropriate. Id. at 9–10 & n.7. But\\nthe district court construed Liberty’s assertion that no\\nagreement existed as a legal conclusion, not a factual dispute.\\nBecause legal conclusions about a collective bargaining\\nagreement are accorded “no special deference,” id. at n.7\\n(quoting Local Union No. 47, Int’l Bhd. Of Elec. Workers v.\\nNLRB, 927 F.2d 635, 640 (D.C. Cir. 1991)), the district court\\nfound no material fact about the contract to be in dispute.\\n\\n On appeal, Liberty argues that its denial was factual, not\\nlegal, and so failure to accept its denial as true was legal error;\\nthat the complete contractual terms and scope were not before\\nthe district court on the pleadings, as recognized by the district\\ncourt’s opinion and the pleadings; and that without a full\\ncontract to interpret, the order to arbitrate was error. Echoing\\nthe district court, MEBA simply responds that Liberty’s claim\\nthat it did not have an agreement with MEBA covering the\\nLiberty Peace is a legal conclusion, not a dispute of material\\nfact. MEBA also argues that Liberty never denied the existence\\nof a collective bargaining agreement that contains a broad\\narbitration clause, nor that those agreements are still in effect.\\nIt claims that Liberty’s “tactic” is “smoke and mirrors.”\\nAppellee Br. 10. MEBA maintains that Liberty contradicted its\\nappellate theory of the case in litigating the case below,\\nclaiming that Liberty conceded the existence of the contract\\n\\f 16\\nand disputed only whether the contract’s broad terms covered\\nthe Liberty Peace. Essentially, MEBA claims that Liberty’s\\nappeal is rooted in procedural technicalities, rather than a good\\nfaith factual dispute over whether an arbitration clause exists in\\nan authentic, extant contract.\\n\\n Giving all reasonable inferences to Liberty, Liberty raised\\nmaterial issues of fact that rendered judgment on the pleadings\\ninappropriate. First, Liberty properly disputed the existence of\\na contract to arbitrate in this case. Without a contract binding\\nthe parties to arbitrate, an order compelling arbitration is\\nimproper. “[A]rbitration is a matter of contract and a party\\ncannot be required to submit to arbitration any dispute which\\nhe has not agreed so to submit.” AT&T Techs., Inc. v.\\nCommc’ns Workers of Am., 475 U.S. 643, 648 (1986)\\n(quotation marks and citation omitted). Such contracts must be\\n“interpreted as a whole.” United States v. Hunt, 843 F.3d 1022,\\n1028 (D.C. Cir. 2016) (quoting Restatement (Second) of\\nContracts § 202(2)). And although “the determination that\\nparties have contractually bound themselves to arbitrate\\ndisputes . . . is a legal conclusion[,] . . . the findings upon which\\nthat conclusion is based are factual.” Bailey v. Fed. Nat’l\\nMortg. Ass’n, 209 F.3d 740, 744 (D.C. Cir. 2000) (quoting\\nChelsea Square Textiles, Inc. v. Bombay Dyeing & Mfg. Co.,\\n189 F.3d 289, 295 (2d Cir. 1999)); see also 11 Williston on\\nContracts § 30:3 (4th ed.) (“It is generally a question of fact . . .\\nwhether or not a contract . . . actually exists.”).\\n\\n The existence of a contract to arbitrate must first be\\nestablished through a factual determination of what constitutes\\nthe parties’ full agreement. As such, the district court erred\\nwhen it based its conclusion of arbitrability on a contract of\\ngenuinely disputed authenticity. Liberty denied the authenticity\\nof the copies of the Master Agreements that MEBA attached to\\nits complaint. Specifically, MEBA alleged in its complaint:\\n\\f 17\\n Liberty first became signatory to a collective bargaining\\n agreement with MEBA in 1988 when the Company signed\\n on to both the Union’s 1986-1990 Tanker Vessels Master\\n Agreement and the 1986-1990 Dry Cargo Vessels Master\\n Agreement. A copy of the 1986-1990 Tanker Vessels\\n Master Agreement and Liberty’s signature page is attached\\n hereto as Exhibit A. A copy of the 1986-1990 Dry Cargo\\n Vessels Master Agreement is attached hereto as Exhibit B.\\n These agreements cover all U.S. flagged vessels owned,\\n managed or operated by Liberty.\\n\\nMEBA Compl. ¶ 8. In its answer, Liberty admitted that Liberty\\nsigned the Master Agreements but denied all other allegations\\nin ¶ 8 of the complaint not expressly admitted. This included\\ndenying MEBA’s allegation that Exhibits A and B were copies\\nof the Master Agreements. Liberty also denied that even the\\nauthentic version of these agreements covered all U.S. flagged\\nvessels managed by Liberty.\\n\\n Accepting Liberty’s allegations as true and making all\\nreasonable inferences in Liberty’s favor, as the Rule 12(c)\\nstandards require, means that Exhibits A and B were not\\nauthenticated copies of the two Master Agreements. Without\\nauthenticated Master Agreements, the district court lacked\\nadequate factual support for its finding that the language of the\\nagreements contained an extant arbitration clause. This dispute\\nover authenticity created a material issue of fact that should\\nhave been enough to defeat MEBA’s Rule 12(c) motion. See\\nHorsley v. Feldt, 304 F.3d 1125, 1135 (11th Cir. 2002)\\n(“Because the authenticity of the [documents] attached to the\\namended answer is disputed, . . . they may not be considered in\\ndeciding the Rule 12(c) motion for judgment on the\\npleadings.”); see also Philips, 572 F.3d at 180.\\n\\f 18\\n Second, several disputed contract provisions were not\\nincluded in the pleadings and attachments. The pleadings\\nshowed that, at least in 2012, Liberty and MEBA were parties\\nto “a Memorandum of Understanding dated September 23,\\n2005, as amended; various Side Letters, dated June 8, 2005,\\nOctober 28, 2005, and July 14, 2010, respectively; and Letters\\nof Understanding, dated July 7, 2009, and February 21, 2010,\\nrespectively.” MEBA Compl. Exhibit C. Liberty conceded the\\nauthenticity of the MOU. But Liberty maintained that the six\\ndocuments referenced in the MOU, but not attached to\\nMEBA’s complaint, were relevant to whether any collective\\nbargaining agreement or arbitration clause applied to the new\\nvessel.\\n\\n Liberty stresses the pre-suit correspondence attached to\\nMEBA’s complaint referenced over a dozen other documents,\\nonly three of which were attached to the complaint, and only\\none of which was conceded to be authentic. Liberty also\\ndisputed that a contract covered the Liberty Peace on the basis\\nthat MEBA failed to define “this Agreement” in the arbitration\\nclause. Because the parties’ contractual relationship has been\\nmodified by numerous agreements over the decades since the\\nMaster Agreements were signed, the record does not make\\nclear to what “this Agreement” refers.\\n\\n The district court, however, assumed these numerous side\\nletters and letters of understanding did not modify the\\nagreement in any meaningful way, accepting that “the\\ngrievance and arbitration procedures contained in the Tanker\\nand Dry Cargo Master Agreements remain binding” without\\nhaving reviewed the “terms and conditions of employment of\\nthe [collective bargaining agreements], side letters, and letters\\nof understanding.” Dist. Ct. Op. at 3. These assumptions were\\ninappropriate in the face of controverted facts about the\\nAgreements’ content. The pleadings and limited attachments\\n\\f 19\\ndid not support the conclusion that “no material fact is in\\ndispute.” Peters, 966 F.2d at 1485.\\n\\n Third, the district court drew improper inferences against\\nLiberty, the opposing party, rather than against MEBA, the\\nmoving party. The district court concluded “Liberty failed to\\npoint to any evidence to rebut the presumption of arbitrability.”\\nDist. Ct. Op. at 14. Yet at this stage of the proceedings, it was\\nMEBA’s burden to demonstrate the existence of an applicable\\narbitration agreement, not Liberty’s burden to rebut it. See\\nBeal, 312 U.S. at 51. Although the district court correctly\\nidentified the federal policy in favor of arbitration agreements,\\nMoses H. Cone Mem’l Hosp. v. Mercury Constr. Corp., 460\\nU.S. 1, 24–25 (1983), that policy becomes a presumption only\\nwhen the factual existence of a contract for arbitration has been\\nestablished. See AT&T, 475 U.S. at 648; Bailey, 209 F.3d at\\n744. The district court also found that “no agreement between\\nthe parties contains language excluding this sort of dispute\\nfrom arbitration.” Dist. Ct. Op. at 14. But because the whole\\n“agreement” was not before the district court, it erred by\\ninferring, in a manner that favored MEBA, that the “cont[ent]”\\nand “language” of the agreement did not exclude the dispute\\nover the Liberty Peace from arbitration.\\n\\n Fourth, the district court erred by refusing to allow\\ndiscovery over Liberty’s objection that there was a material\\nfactual dispute about the existence of an agreement to arbitrate\\nissues relating to the Liberty Peace. See Dec. 14, 2017 Minute\\nOrder. Discovery would have required the district court to treat\\nMEBA’s motion as one for summary judgment. Fed. R. Civ. P.\\n12(d) (“If, on a motion under Rule 12(b)(6) or 12(c), matters\\noutside the pleadings are presented to and not excluded by the\\ncourt, the motion must be treated as one for summary judgment\\nunder Rule 56. All parties must be given a reasonable\\nopportunity to present all the material that is pertinent to the\\n\\f 20\\nmotion.”). Granting such discovery would have allowed\\nLiberty to attempt to substantiate, or MEBA to refute, the claim\\nthat the complete contractual record, when read as a whole,\\ndoes not require arbitration of this dispute.\\n\\n Despite MEBA’s claim that Liberty’s argument consists of\\n“smoke and mirrors,” MEBA has not shown that Liberty lacks\\na good faith basis for its appeal. Parties are presumed to have a\\ngood faith basis for denying allegations, even at the pleading\\nstage. See Amnesty Am. v. Town of W. Hartford, 361 F.3d 113,\\n131 (2d Cir. 2004) (“Because attorneys, as officers of the court,\\nare presumed not to offer in opposition” to dispositive motions\\n“evidence that they have no good faith basis to believe will be\\navailable or admissible at trial, the burden is on the moving\\nparty.”). For the reasons stated above and in light of the\\nincomplete contractual record, we assume there was a good\\nfaith basis for Liberty to challenge judgment on MEBA’s\\npleadings. We expect the parties, on remand, will act in good\\nfaith to narrow their dispute and avoid burying the district court\\nwith reams of contractual provisions not arguably relevant to\\nthe arbitrability issue in this case.\\n\\n ***\\n\\n This case involved disputed issues of material fact that\\nrendered the unusual remedy of judgment on the pleadings\\ninappropriate. Because the district court should have first\\ndetermined whether Liberty and MEBA had a valid contract\\nfor arbitration by looking at their whole agreement, we reverse\\nthe judgment on the pleadings and remand the case for further\\nproceedings consistent with this opinion.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4173,"cells":{"id":{"kind":"number","value":4647783,"string":"4,647,783"},"name":{"kind":"string","value":"Hamed Sufyan Othman ALMAQRAMI, Et Al., Appellants v. Michael R. POMPEO, in His Official Capacity as Secretary of State and John Does, #1-#50, in Their Official Capacity as the Consular Officials Responsible for Issuing Diversity Visas, Appellees"},"name_abbreviation":{"kind":"string","value":"Hamed Almaqrami v. Michael Pompeo"},"decision_date":{"kind":"string","value":"2019-08-13"},"docket_number":{"kind":"string","value":"18-5156"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Tatel, Griffith, Williams\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge GRIFFITH\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued December 14, 2018 Decided August 13, 2019\\n\\n No. 18-5156\\n\\n HAMED SUFYAN OTHMAN ALMAQRAMI, ET AL.,\\n APPELLANTS\\n\\n v.\\n\\n MICHAEL R. POMPEO, IN HIS OFFICIAL CAPACITY AS\\n SECRETARY OF STATE AND JOHN DOES, #1-#50, IN THEIR\\nOFFICIAL CAPACITY AS THE CONSULAR OFFICIALS RESPONSIBLE\\n FOR ISSUING DIVERSITY VISAS,\\n APPELLEES\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:17-cv-01533)\\n\\n\\n Benjamin M. Eidelson argued the cause for appellants.\\nWith him on the briefs were Matthew E. Price, Max J. Minzner,\\nYolanda Rondon, Arthur B. Spitzer, and Scott Michelman.\\n\\n Scott G. Stewart, Attorney, U.S. Department of Justice,\\nargued the cause for appellees. On the brief were Gisela A.\\nWestwater and Erez Reuveni, Assistant Directors, and Joshua\\nS. Press, Trial Attorney.\\n\\n Before: TATEL and GRIFFITH, Circuit Judges, and\\nWILLIAMS, Senior Circuit Judge.\\n\\f 2\\n\\n Opinion for the Court filed by Circuit Judge GRIFFITH.\\n\\n GRIFFITH, Circuit Judge: Plaintiffs were offered the\\nchance to apply for a select number of “diversity visas.” The\\ngovernment never granted them visas, and the statutory\\ndeadline to do so has now passed. But this case is not moot\\nbecause whether the district court retains the authority to award\\nplaintiffs relief is a merits question. We reverse the district\\ncourt’s decision to the contrary.\\n\\n I\\n\\n A\\n\\n In general, a citizen of a foreign country who wishes to\\ncome to the United States must first obtain a U.S. visa, which\\nis placed in the traveler’s passport. A visa does not guarantee\\nentry into the United States; it only confers the right to travel\\nto a port of entry and apply for admission to enter the country.\\n8 U.S.C. § 1201(h). Failure to satisfy certain requirements in\\nthe Immigration and Nationality Act (INA) will render an alien\\nineligible for a visa and ineligible for entry. For example, an\\nalien cannot receive a visa if she has “a communicable disease\\nof public health significance,” and if she contracts such a\\ndisease after receiving her visa, she will be denied entry. 8\\nU.S.C. § 1182(a)(1)(A). Other parts of the INA apply to only\\nvisas or entry. The Secretary of State may, for instance, decline\\nto issue a visa to an alien who abused a position of power to\\nexpropriate American property, id. § 1182d, and the Attorney\\nGeneral may decide that certain aliens must pay bonds before\\nentering the country, id. § 1183.\\n\\n Each fiscal year, the State Department grants\\napproximately 50,000 diversity immigrant visas to individuals\\n\\f 3\\nfrom countries underrepresented in the immigration process,\\nwhich allow recipients who are granted admission to enter the\\ncountry as lawful permanent residents who may live and work\\nhere indefinitely. See 8 U.S.C. §§ 1151(e), 1153(c)(1); Pls.’ Br.\\n4 & n.1. 1 The process by which the State Department awards\\ndiversity visas is competitive and complicated. An applicant\\nmust first apply for and win the diversity visa “lottery.” Pls.’\\nBr. 5; see 8 U.S.C. § 1153(e); 22 C.F.R. § 42.33(b)-(c). A\\nlottery winner or “selectee” must submit an application and\\nvarious documents to be eligible for a visa number—an\\nadministrative device used by the State Department to ensure\\nthat it does not grant more than 50,000 visas per year. Gov’t\\nBr. 9-10; see 8 U.S.C. § 1202(b); 22 C.F.R. §§ 42.33(f)-(g),\\n42.61-67. A selectee is eligible to receive a visa number only\\nduring the fiscal year in which he applied and was selected\\n(“selection FY”). 8 U.S.C. § 1153(e)(2); 22 C.F.R. § 42.33(f).\\nVisa number in hand, the selectee may schedule a consular\\ninterview, and if he meets the criteria to obtain one, the State\\nDepartment “shall” issue him a diversity visa. 8 U.S.C.\\n§ 1153(c), (e)(1); 22 C.F.R. §§ 40.6, 42.81(a); see 8 U.S.C.\\n§ 1202(h). Consulates return unused visa numbers to the State\\nDepartment at the end of each month so that they may be\\nreassigned, but the State Department stops granting visa\\nnumbers altogether once it projects that it will issue all\\navailable visas to existing visa number holders. Gov’t Br. 9-10.\\nBecause the diversity visa program restarts each fiscal year,\\nconsular officers may not issue diversity visas after midnight\\non September 30 of the selection FY. 8 U.S.C.\\n\\n\\n 1\\n A diversity visa applicant who is already living in the United\\nStates in another legal status would instead apply to U.S. Citizenship\\nand Immigration Services (USCIS) for adjustment of status. If\\nsuccessful, he receives a lawful permanent resident identification\\ncard. There is no need to satisfy the entry requirements. We use the\\nterm “diversity visas” to describe both processes.\\n\\f 4\\n§§ 1153(c)(1), 1154(a)(1)(I)(ii)(II); 22 C.F.R. § 42.33(a)(1),\\n(d); see 31 U.S.C. § 1102.\\n\\n B\\n\\n In March 2017, President Trump invoked his authority\\nunder 8 U.S.C. § 1182(f) to “suspend the entry of all aliens or\\nany class of aliens as . . . he may deem to be appropriate” and\\nissued the second iteration of his “travel ban”—an Executive\\nOrder that temporarily prohibited nationals of specific\\ncountries from entering the United States, subject to\\nexemptions and waivers. Exec. Order No. 13,780 (“EO-2”),\\nProtecting the Nation From Foreign Terrorist Entry Into the\\nUnited States §§ 2(c), 3(c), 12(e), 82 Fed. Reg. 13,209, 13,213-\\n15, 13,218 (2017); see Trump v. Hawaii, 138 S. Ct. 2392, 2437\\n(2018). Several district courts issued preliminary injunctions\\npreventing the government from enforcing EO-2’s entry\\nrestriction, which were largely affirmed by the courts of\\nappeals. See Hawaii, 138 S. Ct. at 2437. But in June 2017, the\\nSupreme Court held that EO-2’s entry restriction could take\\neffect while the Court considered the appeals of the preliminary\\ninjunctions, except as to foreign nationals with “a credible\\nclaim of a bona fide relationship with a person or entity in the\\nUnited States.” Trump v. Int’l Refugee Assistance Project\\n(IRAP I), 137 S. Ct. 2080, 2088 (2017) (per curiam).\\n\\n Two days after the Supreme Court’s ruling, the State\\nDepartment issued a “Guidance Memo” instructing consular\\nofficers reviewing diversity visa applications about how EO-\\n2’s entry ban affected visa eligibility: A consular officer should\\nfirst determine whether the selectee “is eligible for the [visa],\\nwithout regard to [EO-2].” J.A. 17. If so, and if he is from a\\ncountry subject to EO-2, the officer must evaluate whether the\\nselectee qualifies for an exemption or waiver, or can establish\\n\\f 5\\na bona fide relationship with the United States. If he cannot, his\\nvisa will be refused.\\n\\n C\\n\\n Plaintiffs won the 2017 diversity visa lottery but were\\ndenied visas pursuant to the Guidance Memo. Consular officers\\ninterviewed plaintiffs and, in accordance with the Memo,\\ndetermined that they would have been eligible for diversity\\nvisas but for the issuance of EO-2. However, because plaintiffs\\nwere from Iran and Yemen—countries subject to the entry\\nban—and could not qualify for exemptions or waivers or\\nsatisfy IRAP I’s bona fide relationship requirement, the\\nconsular officers determined that plaintiffs were “not exempt\\nfrom [EO-2’s] suspension of entry” and denied them visas. See\\nJ.A. 17.\\n\\n Plaintiffs filed suit in the U.S. District Court for the\\nDistrict of Columbia in August 2017. The amended complaint\\nalleges that the Guidance Memo directed consular officers to\\nmake visa determinations on the basis of nationality in\\nviolation of 8 U.S.C. § 1152(a)(1)(A). It further alleges that\\n§ 1182(f) only authorized the President to restrict entry, so by\\nrelying on EO-2 to deny plaintiffs visas, consular officers\\nviolated their duty under the INA to issue visas to all statutorily\\neligible applicants. Among other relief, plaintiffs asked the\\ndistrict court to “enjoin[] the State Department from\\nimplementing the policy set forth in [the Guidance Memo]”\\nand order “consular officers to process [p]laintiffs’ applications\\npursuant to the [INA].” P.K. v. Tillerson, 302 F. Supp. 3d 1, 6\\n(D.D.C. 2017).\\n\\n The next month, the State Department informed the district\\ncourt that it was on track to issue all 50,000 visas allocated for\\nFY 2017 prior to October 1 and would no longer process\\n\\f 6\\nadditional requests for visa numbers made during FY 2017.\\nThe government argued that this development “counsel[ed]\\nagainst” an order that it give plaintiffs visa numbers or process\\ntheir applications. J.A. 33-34. Plaintiffs presented the district\\ncourt with several options. From past practice, it appeared that\\nthe State Department would not reallocate visa numbers\\nreturned in September. Instead of “wast[ing]” those numbers,\\nplaintiffs suggested that the court could order the government\\nto reassign them to plaintiffs. J.A. 46. Recognizing that the\\ndistrict court might have reservations about issuing such a\\nruling while the Supreme Court was reviewing the orders\\nenjoining EO-2’s entry restriction as unlawful, plaintiffs\\nexplained that the district court could instead “maintain the\\nstatus quo” by ordering the State Department to “reserve any\\nunused visa numbers until” IRAP I was “resolved.” Id.\\n\\n On September 24, 2017, EO-2 expired and was replaced\\nby the third iteration of President Trump’s travel ban, the\\n“Proclamation.” Proclamation No. 9,645, Enhancing Vetting\\nCapabilities and Processes for Detecting Attempted Entry Into\\nthe United States by Terrorists or Other Public-Safety Threats,\\n82 Fed. Reg. 45,161 (2017); see P.K., 302 F. Supp. 3d at 3-4.\\nAlthough the Proclamation modified the scope and duration of\\nEO-2’s entry restrictions, the restrictions on Iranians and\\nYemenis remained largely the same.\\n\\n Five days later, on September 29—one day before the end\\nof FY 2017—the district court issued its first ruling. Plaintiffs\\nhad argued that the litigation over EO-2’s entry restrictions was\\n“irrelevant” to their case about visas, J.A. 46, but the district\\ncourt concluded otherwise. Relying on language equating visas\\nand entry in the courts of appeals decisions affirming the\\ninjunctions of EO-2 and the nature of the relief sought in those\\ncases, the district court determined that the Supreme Court’s\\norder staying challenges to EO-2’s entry ban also necessarily\\n\\f 7\\nstayed challenges to EO-2’s effect on visas. P.K., 302 F. Supp.\\n3d at 7-8 & n.7. But in the midst of the uncertainty about the\\nstrength and status of the legal challenges before the Supreme\\nCourt, and with only one day before the end of plaintiffs’\\nselection FY, the district court sought to preserve the status quo\\nand thus keep alive the possibility that plaintiffs could yet\\nreceive their visas. The court therefore ordered the State\\nDepartment to report the number of unused visa numbers for\\nFY 2017 and “hold those visa numbers to process [p]laintiffs’\\nvisa applications in the event the Supreme Court finds [EO-2]\\nto be unlawful.” J.A. 112 (“September 29 Order”). The\\ngovernment has since advised that 27,241 diversity visa\\nnumbers were returned unused and that it issued 49,976\\ndiversity visas in FY 2017—24 shy of the statutory target. This\\nappeal primarily concerns whether the district court may order\\nthe government to keep those unused visas available in the\\nevent these plaintiffs eventually prevail on their claims.\\n\\n In October 2017, the Supreme Court explained that\\nchallenges to the expired entry restrictions of EO-2 were moot.\\nTrump v. Hawaii, 138 S. Ct. 377, 377 (2017); Trump v. IRAP,\\n138 S. Ct. 353, 353 (2017). That same month, the government\\nfiled a motion to dismiss this case as moot, arguing that once\\nFY 2017 ended, the court lacked the power to order the\\ngovernment to issue plaintiffs FY 2017 diversity visas. The\\ndistrict court rejected that theory, but agreed that the case was\\nmoot because the Guidance Memo was issued to implement the\\nnow-expired EO-2 and the September 29 Order had “expressly\\npredicated” any future order requiring the government to\\nprocess plaintiffs’ applications upon “the Supreme Court\\nfind[ing] [EO-2] to be unlawful,” which it had not done.\\nAlmaqrami v. Tillerson, 304 F. Supp. 3d 1, 7-8 (D.D.C. 2018)\\n(first alteration in original) (internal quotation marks omitted).\\nPlaintiffs appeal.\\n\\f 8\\n While this litigation was ongoing, other plaintiffs\\nchallenged the Proclamation’s entry ban on grounds similar to\\nthose argued to enjoin EO-2. Once again, district courts\\nenjoined that ban, and the courts of appeals affirmed. In June\\n2018, the Supreme Court held that those plaintiffs were not\\nlikely to show that the Proclamation was unlawful. Hawaii, 138\\nS. Ct. 2392.\\n\\n II\\n\\n We have jurisdiction pursuant to 28 U.S.C. § 1291. Our\\nreview is de novo, though we accept as true the facts plaintiffs\\nhave alleged. Schnitzer v. Harvey, 389 F.3d 200, 202 (D.C. Cir.\\n2004).\\n\\n A federal court’s jurisdiction is limited to “Cases or\\nControversies.” U.S. Const. art. III, § 2, cl. 1. A lawsuit\\nbecomes moot—and is therefore no longer a “Case” or\\n“Controversy”—“when the issues presented are no longer\\n‘live’ or the parties lack a legally cognizable interest in the\\noutcome.” Chafin v. Chafin, 568 U.S. 165, 172 (2013) (quoting\\nAlready, LLC v. Nike, Inc., 568 U.S. 85, 91 (2013)). But a case\\nbecomes moot only if, assuming the plaintiff prevails, “it is\\nimpossible for a court to grant [her] any effectual relief\\nwhatever.” Id. (quoting Knox v. Serv. Emps. Int’l Union, 567\\nU.S. 298, 307 (2012)). We must assume that the plaintiff will\\n“prevail” unless her argument that the relief sought is legally\\navailable and that she is entitled to it is “so implausible that it\\nis insufficient to preserve jurisdiction.” Id. at 174. And if there\\nis “any chance” that relief will be effective in securing what\\nshe seeks, she has an interest in obtaining it. Mission Prod.\\nHoldings, Inc. v. Tempnology, LLC, 139 S. Ct. 1652, 1660\\n(2019); see Chafin, 568 U.S. at 175-76 (“[T]he fact that a\\ndefendant is insolvent does not moot a claim for damages.”).\\n\\f 9\\n Under this “demanding standard,” plaintiffs’ claims are\\nnot moot. See Tempnology, 139 S. Ct. at 1660. They seek a\\ncourt order instructing the government to stop implementing\\nthe Guidance Memo, process their visa applications, and issue\\nthem diversity visas. Neither their claim that such relief is\\nlegally available nor their claim that they are entitled to that\\nrelief is so implausible as to deprive the district court of\\njurisdiction. And there is some chance that this relief will be\\neffective at securing their immigration to the United States.\\n\\n A\\n\\n The parties dispute whether the district court may lawfully\\ntake steps to grant plaintiffs relief, notwithstanding the fact that\\nFY 2017 is over. This question goes to the merits, and because\\nplaintiffs’ argument that the district court may do so is not “so\\nimplausible that it is insufficient to preserve jurisdiction,” the\\ncase is not moot. Chafin, 568 U.S. at 174 (explaining that an\\nargument about “the legal availability of a certain kind of\\nrelief” is a merits question).\\n\\n Courts are often asked to intervene in disputes over\\ndiversity visas, and the end of the selection FY does often\\nrender those cases moot. In a straightforward case, a plaintiff\\nwho believed the government had erroneously denied her visa\\nor was not processing her application quickly enough would\\nfile suit well in advance of the end of the selection FY. If the\\ncourt agreed, it would order the government to correct the error,\\nthe government would timely comply, and the plaintiff would\\nreceive her visa before the selection FY ended. More often, the\\nplaintiff files suit after the selection FY has ended. Because\\ndiversity visas expire when the selection FY ends, 8 U.S.C.\\n§§ 1153(c)(1), 1154(a)(1)(I)(ii)(II), that plaintiff does not have\\na statutory right to the requested visa and the government does\\nnot have a duty to issue her one. Courts faced with this situation\\n\\f 10\\nhave dismissed these lawsuits as moot. Zixiang Li v. Kerry, 710\\nF.3d 995 (9th Cir. 2013); Mohamed v. Gonzales, 436 F.3d 79\\n(2d Cir. 2006) (per curiam); Nyaga v. Ashcroft, 323 F.3d 906\\n(11th Cir. 2003) (per curiam). But see Coraggioso v. Ashcroft,\\n355 F.3d 730 (3d Cir. 2004) (dismissing suit on the merits);\\nIddir v. INS, 301 F.3d 492 (7th Cir. 2002) (same).\\n\\n Sometimes a plaintiff files suit before the selection FY\\nends but the court fails to act on that request until after\\nSeptember 30, at which point the State Department lacks\\nauthority to issue a diversity visa sought in the prior fiscal year.\\nCourts have likewise dismissed these cases as moot. See\\nMwasaru v. Napolitano, 619 F.3d 545 (6th Cir. 2010); Zapata\\nv. INS, 93 F. Supp. 2d 355 (S.D.N.Y. 2000).\\n\\n Other cases involve a different twist. The plaintiff files suit\\nand the court grants some relief—but not the visa—before\\nOctober 1. In such a case, after the selection FY has ended, the\\ncourt might lawfully take steps to compel the government to\\nprocess the plaintiff’s application and issue her a diversity visa\\nanyway.\\n\\n That is what happened here. On September 29, the district\\ncourt ordered the government to reserve unused FY 2017 visa\\nnumbers so that, if it turned out that plaintiffs had erroneously\\nbeen denied their diversity visas, the court could order the\\ngovernment “to process visas [for plaintiffs] past the statutory\\ndeadline.” P.K., 302 F. Supp. 3d at 10-11. The district court\\ncited two cases in which courts had done something similar:\\nPrzhebelskaya v. USCIS, 338 F. Supp. 2d 399 (E.D.N.Y. 2004)\\nand Paunescu v. INS, 76 F. Supp. 2d 896 (N.D. Ill. 1999). In\\nboth, before the selection FY ended, the district court ordered\\nthe government to timely process plaintiffs’ diversity visa\\napplications. Had the government complied, plaintiffs would\\nhave received visas. But the government did not, so the court\\n\\f 11\\ninvoked its equitable power to enforce prior orders and\\ninstructed the government to issue the plaintiffs visas even\\nthough the selection FY had ended. See also Marcetic v. INS,\\nNo. 97 C 7018, 1998 WL 173129, at *2 (N.D. Ill. Apr. 6, 1998)\\n(ordering government to comply with prior order entered by an\\nimmigration judge requiring it to do what was necessary to\\nissue plaintiff’s green card).\\n\\n The question is whether, now that the selection FY has\\nended, the district court can order the State Department to do\\nanything with the unused visa numbers held in reserve pursuant\\nto the September 29 Order. Following Chafin, we hold that this\\nis a merits question in the context of this case. There, a father\\nsought a court order directing the mother of his daughter to\\nreturn the child to the United States after taking her to live\\noverseas. Chafin, 568 U.S. at 173. The mother argued that the\\nfather did not have a legally cognizable interest in obtaining\\nthat order because the court “lack[ed] the authority to issue” it\\n“pursuant to its inherent equitable powers.” Id. at 174. But that\\nargument—“which goes to the . . . legal availability of a certain\\nkind of relief—confuse[d] mootness with the merits.” Id.\\nBecause the husband’s claim that the district court possessed\\nthe equitable power to issue such an order was not “so\\nimplausible that it is insufficient to preserve jurisdiction,” the\\ncourt had to assume that claim would prevail, meaning the case\\nwas not moot. Id.\\n\\n It is likewise not “implausible” that the district court here\\ncould rely on equity to take steps to compel the issuance of\\ndiversity visas, notwithstanding the end of FY 2017. Indeed,\\nthe government acknowledges that courts have that power, but\\nin its view, that power is limited to cases like Przhebelskaya\\nand Paunescu in which the court orders the government to\\nprocess a visa application. By contrast, the district court here\\nordered the government to hold available visa numbers to\\n\\f 12\\npotentially process plaintiffs’ applications. Because there is no\\n“prior court order that was not complied with,” Gov’t Br. 37,\\nthe court cannot even “arguabl[y]” invoke equity to provide\\nplaintiffs relief, id. at 31, or so the argument goes, see id. at 28-\\n37.\\n\\n This argument assigns more determinacy to the meaning\\nof the September 29 Order than it can bear, in our view, for\\npurposes of our mootness analysis. That Order instructed the\\ngovernment to “hold [unused] visa numbers to process\\n[p]laintiffs’ visa applications in the event the Supreme Court\\nfinds [EO-2] to be unlawful.” J.A. 112. We hold that the\\nSeptember 29 Order need not be read to limit the authority of\\nthe district court to grant additional relief to a scenario in which\\nthe Supreme Court finds EO-2 unlawful. 2 Rather, it is at least\\npossible to read that Order as doing one or both of two other\\nthings.\\n\\n First, the Order might simply have preserved the “status\\nquo . . . while the legality of [EO-2]” was pending before the\\nSupreme Court, meaning it preserved the district court’s\\nability, as of September 29, to rectify the erroneous denial of\\nplaintiffs’ visas based on a legally questionable Guidance\\nMemo or erroneous interpretation of the INA. See P.K., 302 F.\\nSupp. 3d at 7. Second, it may have told the government that if\\none specific eventuality arose—the Supreme Court found EO-\\n2 unlawful—the State Department must process plaintiffs’ visa\\n\\n 2\\n In the district court’s view, that is all it did, and because the\\nSupreme Court never found EO-2 unlawful, the district court\\nconcluded that it could not grant plaintiffs further relief. Almaqrami,\\n304 F. Supp. 3d at 8. Although the government does not defend this\\nreasoning on appeal, and plaintiffs argue this was not “an\\nindependent alternative holding,” Pls.’ Br. 24, we are obligated to\\naddress potential jurisdictional defects, Floyd v. District of\\nColumbia, 129 F.3d 152, 155 (D.C. Cir. 1997).\\n\\f 13\\napplications. These readings may be combined. That is, the\\nOrder could be read as (1) instructing the State Department to\\n“hold” these unused visa numbers for the purpose of enabling\\na later judicial judgment, which might require the government\\n“to process” plaintiffs’ applications after any salient obstacles\\nwere removed, and (2) identifying the specific judgment that\\nwould issue if the Supreme Court took a certain action. Cf.\\nCrystal Clear Commc’ns, Inc. v. Sw. Bell Tel. Co., 415 F.3d\\n1171, 1177-78 (10th Cir. 2005) (concluding from context that\\nan order stating plaintiffs must submit “claims” to the FCC\\nonly intended to refer some claims, but not others, to the FCC);\\nGjertsen v. Bd. of Election Comm’rs of City of Chi., 751 F.2d\\n199 (7th Cir. 1984) (Posner, J.) (context and district court’s\\nintent demonstrated that an order styled as a “final judgment”\\nwas actually an order granting a preliminary injunction).\\n\\n On that reading, the September 29 Order left open whether\\na later judgment would issue and, if so, what it would look like\\nin the event the Supreme Court took any of myriad other\\ntacks—for example, holding that the President could rely on\\n§ 1182(f) to restrict entry but not visas, that the IRAP I\\nplaintiffs lacked standing, that their claims were nonjusticiable,\\nor that the case must be dismissed as moot, which the district\\ncourt knew was a possibility on September 29. See P.K., 302\\nF. Supp. 3d at 4 n.4 (explaining that the Court had vacated oral\\nargument in IRAP I and ordered the parties to brief whether the\\nProclamation and expiration of EO-2 mooted that case). We\\nneed not decide the merit of these various readings; we merely\\nnote that they are possible constructions of the September 29\\nOrder.\\n\\n That means plaintiffs’ claim that further relief is legally\\navailable is not “so implausible” as to be “insufficient to\\npreserve jurisdiction.” Chafin, 568 U.S. at 174. For example, if\\nwe were to read the Order as leaving open the possibility that a\\n\\f 14\\nlater judgment would issue, now that IRAP I—the primary\\nobstacle the district court identified to granting plaintiffs\\nadditional relief—has been dismissed as moot, it is not\\n“implausible” that the district court could grant plaintiffs\\nadditional relief. As the district court explained, if it “were to\\nnow order the State Department to use the unused visa numbers\\nto process [p]laintiffs’ visa applications, it would [arguably] be\\nrequiring the State Department to fulfill its obligations under”\\nthe September 29 Order, which instructed the State Department\\nto “reserve the unused visa numbers . . . for a specific purpose:\\nthe future processing of [p]laintiffs’ visa applications.”\\nAlmaqrami, 304 F. Supp. 3d at 6. Like in Paunescu and\\nPrzhebelskaya, such an order would give effect to the district\\ncourt’s prior directive, entered before the end of the selection\\nFY, to preserve an essential (and otherwise expiring) ingredient\\nof relief. To be sure, those cases required the government to\\ncomply with a prior order to process applications, see Gov’t\\nBr. 37, but they offer useful examples, not binding models, and\\nneither of those courts limited their holdings to the precise\\nscenario they confronted, see Almaqrami, 304 F. Supp. 3d at 7.\\nAnd this case is more similar to Paunescu and Przhebelskaya\\nthan the cases dismissed as moot because the plaintiff filed too\\nlate or the court did not act in time. All told, this is enough to\\nsuggest that plaintiffs’ argument that the district court could\\ngrant them additional relief, despite the end of the selection FY,\\nis not so “completely devoid of merit as not to involve a federal\\ncontroversy.” Steel Co. v. Citizens for a Better Env’t, 523 U.S.\\n83, 89 (1998) (quoting Oneida Indian Nation of N.Y. v. County\\nof Oneida, 414 U.S. 661, 666 (1974)).\\n\\n B\\n\\n Three other issues warrant brief discussion. First, the\\ndistrict court concluded that because the expiration of EO-2\\nrendered challenges to that order moot, the same was true for\\n\\f 15\\nplaintiffs’ challenges to the Guidance Memo that implemented\\nEO-2, and therefore “this case” was moot. Almaqrami, 304 F.\\nSupp. 3d at 8. But plaintiffs asked the district court to declare\\nthe policy in the Guidance Memo unlawful and enjoin the State\\nDepartment from implementing it and to order consular\\nofficers to process their visa applications in accordance with\\nthe INA. See P.K., 302 F. Supp. 3d at 7. If the Guidance Memo\\nis no longer in force, that arguably removes an obstacle to the\\nplaintiffs’ ability to obtain an order instructing the government\\nto process their applications and issue them visas pursuant to\\nthe INA separate and apart from anything provided in the\\nMemo. As for their challenges to the Memo itself, although\\ncourts generally cannot declare unlawful or enjoin policies that\\nare no longer in force, the district court never actually found\\nthat the Guidance Memo “expired by its own terms” along with\\nEO-2. See Almaqrami, 304 F. Supp. 3d at 7 (quoting IRAP, 138\\nS. Ct. at 353; and citing Hawaii, 138 S. Ct. at 377). Indeed, the\\npresent record suggests that the Memo did not. See Tr. of Oral\\nArg. at 5:10-12 (plaintiffs’ counsel explaining without\\ncontradiction that “there is nothing in the record suggesting\\nthat” the Guidance Memo or the policy it announced “was ever\\nrescinded or modified”). Without making such a finding, the\\ndistrict court was too quick to conclude that plaintiffs’\\nchallenges to the Memo were moot. Compare Nat’l Ass’n of\\nMfrs. v. Dep’t of Def., 138 S. Ct. 617, 627 n.5 (2018) (“Because\\nthe [challenged rule] remains on the books for now, the parties\\nretain ‘a concrete interest’ in the outcome of this litigation, and\\nit is not ‘impossible for a court to grant any effectual relief . . .\\nto the prevailing party.’” (quoting Chafin, 568 U.S. at 172)),\\nwith Safari Club Int’l v. Jewell, 842 F.3d 1280, 1288 (D.C. Cir.\\n2016) (“expiration of a government policy ordinarily moots a\\nchallenge to it”). See also Rembert v. Sheahan, 62 F.3d 937,\\n938 (7th Cir. 1995) (declining to address mootness in the face\\nof an insufficient record).\\n\\f 16\\n Furthermore, whether the Memo has expired has no effect\\non the potential viability of plaintiffs’ theories of relief, which\\nwe must assume are valid unless they are “wholly insubstantial\\nand frivolous.” Steel Co., 523 U.S. at 89 (quoting Bell v. Hood,\\n327 U.S. 678, 682-83 (1946)). EO-2 prevented nationals of\\nspecific countries from entering the United States; it said\\nnothing about diversity visas. Although “aliens who are\\ninadmissible” to the United States for reasons described in\\ncertain provisions of the INA “are ineligible to receive visas,”\\n8 U.S.C. § 1182(a), plaintiffs argue that § 1182(f), which the\\nPresident relied on to issue EO-2 and the Proclamation, is not\\none of those designated provisions. Plaintiffs also argue that by\\ndenying them visas on the basis of nationality, the State\\nDepartment and its consular officers violated 8 U.S.C.\\n§ 1152(a)(1)(A). They admit that some language in the\\nSupreme Court’s Hawaii opinion “is in tension with [their]\\nclaim[s],” but they contend that language is neither binding nor\\npersuasive, and that other language in Hawaii is “profoundly\\nhelpful.” Tr. of Oral Arg. at 6:14-17; see id. at 7:10-9:2, 23:17-\\n25:20. We take no position on the merits of these arguments,\\nor whether plaintiffs have “state[d] a claim to relief that is\\nplausible on its face,” as required by Federal Rule of Civil\\nProcedure 12(b)(6). Bell Atl. Corp. v. Twombly, 550 U.S. 544,\\n570 (2007). We merely note that they are sufficient to preserve\\nthe district court’s subject matter jurisdiction. See Chafin, 568\\nU.S. at 174; Steel Co., 523 U.S. at 89.\\n\\n Finally, because there is some chance that this relief would\\nbe effective at securing plaintiffs’ immigration to the United\\nStates, their “suit remains live.” Tempnology, 139 S. Ct. at\\n1660; see Chafin, 568 U.S. at 172, 174-75 (explaining that,\\nunless it is impossible to grant relief that would be effective in\\nsecuring the plaintiff’s goal, the case is not moot). Although\\nthe Proclamation currently prevents nationals of Iran and\\nYemen from entering the country, plaintiffs could qualify for\\n\\f 17\\nan exemption or waiver. Or the President might lift these\\nparticular entry restrictions, as he has done for nationals of\\nChad. Proclamation No. 9,723, Maintaining Enhanced Vetting\\nCapabilities and Processes for Detecting Attempted Entry Into\\nthe United States by Terrorists or Other Public-Safety Threats,\\n83 Fed. Reg. 15,937, 15,938-39 (2018); see also Hawaii, 138\\nS. Ct. at 2422 (explaining other changes made since the initial\\nentry restrictions took effect). These possibilities “may be\\nuncertain or even unlikely,” Tempnology, 139 S. Ct. at 1660,\\nbut that “does not typically render cases moot,” Chafin, 568 at\\n175; see Del. Riverkeeper Network v. FERC, 857 F.3d 388, 397\\n(D.C. Cir. 2017) (case was not moot even though the court’s\\norder “would almost certainly have no real world\\nconsequences”). This case is no different.\\n\\n III\\n\\n On the present record, this case is not moot. We reverse\\nthe order dismissing this case for lack of subject matter\\njurisdiction and remand it to the district court for further\\nproceedings. 3\\n\\n 3\\n The government also moved to dismiss plaintiffs’ amended\\ncomplaint as nonjusticiable based on the doctrine of consular\\nnonreviewability and for failure to state a claim under Federal Rule\\nof Civil Procedure 12(b)(6). Mem. of Supporting Points and\\nAuthorities at 26-31, 34-39, No. 17-cv-01533 (D.D.C. Oct. 20,\\n2017), Dkt. No. 53-1. The district court had already rejected the\\ngovernment’s consular nonreviewability argument. P.K., 302 F.\\nSupp. 3d at 11-12. Because the court dismissed the case as moot, it\\ndid not reconsider that consular nonreviewability determination or\\nreach the government’s 12(b)(6) argument. Almaqrami, 304 F. Supp.\\n3d at 9 & n.3.\\n\\n When we reverse the dismissal of a case as moot, our usual\\npractice is to remand for the district court to consider arguments\\n\\f 18\\n\\n So ordered.\\n\\n\\n\\n\\nabout the merits in the first instance, assuming no other threshold\\nissues exist. E.g., Judicial Watch, Inc. v. Kerry, 844 F.3d 952, 956\\n(D.C. Cir. 2016). Because courts “may assume without deciding that\\nplaintiffs’ statutory claims are reviewable” and proceed to the merits\\n“notwithstanding consular nonreviewability,” we see no need to\\naddress consular nonreviewability here. Hawaii, 138 S. Ct. at 2407;\\nsee Gov’t Br. 38-43 (not arguing this doctrine is jurisdictional).\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4425036/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge GRIFFITH\"}]}"},"judge1":{"kind":"string","value":"TATEL"},"judge2":{"kind":"string","value":"GRIFFITH"},"judge3":{"kind":"string","value":"WILLIAMS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4647783/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Hamed Sufyan Othman ALMAQRAMI, Et Al., Appellants v. Michael R. POMPEO, in His Official Capacity as Secretary of State and John Does, #1-#50, in Their Official Capacity as the Consular Officials Responsible for Issuing Diversity Visas, Appellees\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-13\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued December 14, 2018 Decided August 13, 2019\n\n No. 18-5156\n\n HAMED SUFYAN OTHMAN ALMAQRAMI, ET AL.,\n APPELLANTS\n\n v.\n\n MICHAEL R. POMPEO, IN HIS OFFICIAL CAPACITY AS\n SECRETARY OF STATE AND JOHN DOES, #1-#50, IN THEIR\nOFFICIAL CAPACITY AS THE CONSULAR OFFICIALS RESPONSIBLE\n FOR ISSUING DIVERSITY VISAS,\n APPELLEES\n\n\n Appeal from the United States District Court\n for the District of Columbia\n (No. 1:17-cv-01533)\n\n\n Benjamin M. Eidelson argued the cause for appellants.\nWith him on the briefs were Matthew E. Price, Max J. Minzner,\nYolanda Rondon, Arthur B. Spitzer, and Scott Michelman.\n\n Scott G. Stewart, Attorney, U.S. Department of Justice,\nargued the cause for appellees. On the brief were Gisela A.\nWestwater and Erez Reuveni, Assistant Directors, and Joshua\nS. Press, Trial Attorney.\n\n Before: TATEL and GRIFFITH, Circuit Judges, and\nWILLIAMS, Senior Circuit Judge.\n\f 2\n\n Opinion for the Court filed by Circuit Judge GRIFFITH.\n\n GRIFFITH, Circuit Judge: Plaintiffs were offered the\nchance to apply for a select number of “diversity visas.” The\ngovernment never granted them visas, and the statutory\ndeadline to do so has now passed. But this case is not moot\nbecause whether the district court retains the authority to award\nplaintiffs relief is a merits question. We reverse the district\ncourt’s decision to the contrary.\n\n I\n\n A\n\n In general, a citizen of a foreign country who wishes to\ncome to the United States must first obtain a U.S. visa, which\nis placed in the traveler’s passport. A visa does not guarantee\nentry into the United States; it only confers the right to travel\nto a port of entry and apply for admission to enter the country.\n8 U.S.C. § 1201(h). Failure to satisfy certain requirements in\nthe Immigration and Nationality Act (INA) will render an alien\nineligible for a visa and ineligible for entry. For example, an\nalien cannot receive a visa if she has “a communicable disease\nof public health significance,” and if she contracts such a\ndisease after receiving her visa, she will be denied entry. 8\nU.S.C. § 1182(a)(1)(A). Other parts of the INA apply to only\nvisas or entry. The Secretary of State may, for instance, decline\nto issue a visa to an alien who abused a position of power to\nexpropriate American property, id. § 1182d, and the Attorney\nGeneral may decide that certain aliens must pay bonds before\nentering the country, id. § 1183.\n\n Each fiscal year, the State Department grants\napproximately 50,000 diversity immigrant visas to individuals\n\f 3\nfrom countries underrepresented in the immigration process,\nwhich allow recipients who are granted admission to enter the\ncountry as lawful permanent residents who may live and work\nhere indefinitely. See 8 U.S.C. §§ 1151(e), 1153(c)(1); Pls.’ Br.\n4 & n.1. 1 The process by which the State Department awards\ndiversity visas is competitive and complicated. An applicant\nmust first apply for and win the diversity visa “lottery.” Pls.’\nBr. 5; see 8 U.S.C. § 1153(e); 22 C.F.R. § 42.33(b)-(c). A\nlottery winner or “selectee” must submit an application and\nvarious documents to be eligible for a visa number—an\nadministrative device used by the State Department to ensure\nthat it does not grant more than 50,000 visas per year. Gov’t\nBr. 9-10; see 8 U.S.C. § 1202(b); 22 C.F.R. §§ 42.33(f)-(g),\n42.61-67. A selectee is eligible to receive a visa number only\nduring the fiscal year in which he applied and was selected\n(“selection FY”). 8 U.S.C. § 1153(e)(2); 22 C.F.R. § 42.33(f).\nVisa number in hand, the selectee may schedule a consular\ninterview, and if he meets the criteria to obtain one, the State\nDepartment “shall” issue him a diversity visa. 8 U.S.C.\n§ 1153(c), (e)(1); 22 C.F.R. §§ 40.6, 42.81(a); see 8 U.S.C.\n§ 1202(h). Consulates return unused visa numbers to the State\nDepartment at the end of each month so that they may be\nreassigned, but the State Department stops granting visa\nnumbers altogether once it projects that it will issue all\navailable visas to existing visa number holders. Gov’t Br. 9-10.\nBecause the diversity visa program restarts each fiscal year,\nconsular officers may not issue diversity visas after midnight\non September 30 of the selection FY. 8 U.S.C.\n\n\n 1\n A diversity visa applicant who is already living in the United\nStates in another legal status would instead apply to U.S. Citizenship\nand Immigration Services (USCIS) for adjustment of status. If\nsuccessful, he receives a lawful permanent resident identification\ncard. There is no need to satisfy the entry requirements. We use the\nterm “diversity visas” to describe both processes.\n\f 4\n§§ 1153(c)(1), 1154(a)(1)(I)(ii)(II); 22 C.F.R. § 42.33(a)(1),\n(d); see 31 U.S.C. § 1102.\n\n B\n\n In March 2017, President Trump invoked his authority\nunder 8 U.S.C. § 1182(f) to “suspend the entry of all aliens or\nany class of aliens as . . . he may deem to be appropriate” and\nissued the second iteration of his “travel ban”—an Executive\nOrder that temporarily prohibited nationals of specific\ncountries from entering the United States, subject to\nexemptions and waivers. Exec. Order No. 13,780 (“EO-2”),\nProtecting the Nation From Foreign Terrorist Entry Into the\nUnited States §§ 2(c), 3(c), 12(e), 82 Fed. Reg. 13,209, 13,213-\n15, 13,218 (2017); see Trump v. Hawaii, 138 S. Ct. 2392, 2437\n(2018). Several district courts issued preliminary injunctions\npreventing the government from enforcing EO-2’s entry\nrestriction, which were largely affirmed by the courts of\nappeals. See Hawaii, 138 S. Ct. at 2437. But in June 2017, the\nSupreme Court held that EO-2’s entry restriction could take\neffect while the Court considered the appeals of the preliminary\ninjunctions, except as to foreign nationals with “a credible\nclaim of a bona fide relationship with a person or entity in the\nUnited States.” Trump v. Int’l Refugee Assistance Project\n(IRAP I), 137 S. Ct. 2080, 2088 (2017) (per curiam).\n\n Two days after the Supreme Court’s ruling, the State\nDepartment issued a “Guidance Memo” instructing consular\nofficers reviewing diversity visa applications about how EO-\n2’s entry ban affected visa eligibility: A consular officer should\nfirst determine whether the selectee “is eligible for the [visa],\nwithout regard to [EO-2].” J.A. 17. If so, and if he is from a\ncountry subject to EO-2, the officer must evaluate whether the\nselectee qualifies for an exemption or waiver, or can establish\n\f 5\na bona fide relationship with the United States. If he cannot, his\nvisa will be refused.\n\n C\n\n Plaintiffs won the 2017 diversity visa lottery but were\ndenied visas pursuant to the Guidance Memo. Consular officers\ninterviewed plaintiffs and, in accordance with the Memo,\ndetermined that they would have been eligible for diversity\nvisas but for the issuance of EO-2. However, because plaintiffs\nwere from Iran and Yemen—countries subject to the entry\nban—and could not qualify for exemptions or waivers or\nsatisfy IRAP I’s bona fide relationship requirement, the\nconsular officers determined that plaintiffs were “not exempt\nfrom [EO-2’s] suspension of entry” and denied them visas. See\nJ.A. 17.\n\n Plaintiffs filed suit in the U.S. District Court for the\nDistrict of Columbia in August 2017. The amended complaint\nalleges that the Guidance Memo directed consular officers to\nmake visa determinations on the basis of nationality in\nviolation of 8 U.S.C. § 1152(a)(1)(A). It further alleges that\n§ 1182(f) only authorized the President to restrict entry, so by\nrelying on EO-2 to deny plaintiffs visas, consular officers\nviolated their duty under the INA to issue visas to all statutorily\neligible applicants. Among other relief, plaintiffs asked the\ndistrict court to “enjoin[] the State Department from\nimplementing the policy set forth in [the Guidance Memo]”\nand order “consular officers to process [p]laintiffs’ applications\npursuant to the [INA].” P.K. v. Tillerson, 302 F. Supp. 3d 1, 6\n(D.D.C. 2017).\n\n The next month, the State Department informed the district\ncourt that it was on track to issue all 50,000 visas allocated for\nFY 2017 prior to October 1 and would no longer process\n\f 6\nadditional requests for visa numbers made during FY 2017.\nThe government argued that this development “counsel[ed]\nagainst” an order that it give plaintiffs visa numbers or process\ntheir applications. J.A. 33-34. Plaintiffs presented the district\ncourt with several options. From past practice, it appeared that\nthe State Department would not reallocate visa numbers\nreturned in September. Instead of “wast[ing]” those numbers,\nplaintiffs suggested that the court could order the government\nto reassign them to plaintiffs. J.A. 46. Recognizing that the\ndistrict court might have reservations about issuing such a\nruling while the Supreme Court was reviewing the orders\nenjoining EO-2’s entry restriction as unlawful, plaintiffs\nexplained that the district court could instead “maintain the\nstatus quo” by ordering the State Department to “reserve any\nunused visa numbers until” IRAP I was “resolved.” Id.\n\n On September 24, 2017, EO-2 expired and was replaced\nby the third iteration of President Trump’s travel ban, the\n“Proclamation.” Proclamation No. 9,645, Enhancing Vetting\nCapabilities and Processes for Detecting Attempted Entry Into\nthe United States by Terrorists or Other Public-Safety Threats,\n82 Fed. Reg. 45,161 (2017); see P.K., 302 F. Supp. 3d at 3-4.\nAlthough the Proclamation modified the scope and duration of\nEO-2’s entry restrictions, the restrictions on Iranians and\nYemenis remained largely the same.\n\n Five days later, on September 29—one day before the end\nof FY 2017—the district court issued its first ruling. Plaintiffs\nhad argued that the litigation over EO-2’s entry restrictions was\n“irrelevant” to their case about visas, J.A. 46, but the district\ncourt concluded otherwise. Relying on language equating visas\nand entry in the courts of appeals decisions affirming the\ninjunctions of EO-2 and the nature of the relief sought in those\ncases, the district court determined that the Supreme Court’s\norder staying challenges to EO-2’s entry ban also necessarily\n\f 7\nstayed challenges to EO-2’s effect on visas. P.K., 302 F. Supp.\n3d at 7-8 & n.7. But in the midst of the uncertainty about the\nstrength and status of the legal challenges before the Supreme\nCourt, and with only one day before the end of plaintiffs’\nselection FY, the district court sought to preserve the status quo\nand thus keep alive the possibility that plaintiffs could yet\nreceive their visas. The court therefore ordered the State\nDepartment to report the number of unused visa numbers for\nFY 2017 and “hold those visa numbers to process [p]laintiffs’\nvisa applications in the event the Supreme Court finds [EO-2]\nto be unlawful.” J.A. 112 (“September 29 Order”). The\ngovernment has since advised that 27,241 diversity visa\nnumbers were returned unused and that it issued 49,976\ndiversity visas in FY 2017—24 shy of the statutory target. This\nappeal primarily concerns whether the district court may order\nthe government to keep those unused visas available in the\nevent these plaintiffs eventually prevail on their claims.\n\n In October 2017, the Supreme Court explained that\nchallenges to the expired entry restrictions of EO-2 were moot.\nTrump v. Hawaii, 138 S. Ct. 377, 377 (2017); Trump v. IRAP,\n138 S. Ct. 353, 353 (2017). That same month, the government\nfiled a motion to dismiss this case as moot, arguing that once\nFY 2017 ended, the court lacked the power to order the\ngovernment to issue plaintiffs FY 2017 diversity visas. The\ndistrict court rejected that theory, but agreed that the case was\nmoot because the Guidance Memo was issued to implement the\nnow-expired EO-2 and the September 29 Order had “expressly\npredicated” any future order requiring the government to\nprocess plaintiffs’ applications upon “the Supreme Court\nfind[ing] [EO-2] to be unlawful,” which it had not done.\nAlmaqrami v. Tillerson, 304 F. Supp. 3d 1, 7-8 (D.D.C. 2018)\n(first alteration in original) (internal quotation marks omitted).\nPlaintiffs appeal.\n\f 8\n While this litigation was ongoing, other plaintiffs\nchallenged the Proclamation’s entry ban on grounds similar to\nthose argued to enjoin EO-2. Once again, district courts\nenjoined that ban, and the courts of appeals affirmed. In June\n2018, the Supreme Court held that those plaintiffs were not\nlikely to show that the Proclamation was unlawful. Hawaii, 138\nS. Ct. 2392.\n\n II\n\n We have jurisdiction pursuant to 28 U.S.C. § 1291. Our\nreview is de novo, though we accept as true the facts plaintiffs\nhave alleged. Schnitzer v. Harvey, 389 F.3d 200, 202 (D.C. Cir.\n2004).\n\n A federal court’s jurisdiction is limited to “Cases or\nControversies.” U.S. Const. art. III, § 2, cl. 1. A lawsuit\nbecomes moot—and is therefore no longer a “Case” or\n“Controversy”—“when the issues presented are no longer\n‘live’ or the parties lack a legally cognizable interest in the\noutcome.” Chafin v. Chafin, 568 U.S. 165, 172 (2013) (quoting\nAlready, LLC v. Nike, Inc., 568 U.S. 85, 91 (2013)). But a case\nbecomes moot only if, assuming the plaintiff prevails, “it is\nimpossible for a court to grant [her] any effectual relief\nwhatever.” Id. (quoting Knox v. Serv. Emps. Int’l Union, 567\nU.S. 298, 307 (2012)). We must assume that the plaintiff will\n“prevail” unless her argument that the relief sought is legally\navailable and that she is entitled to it is “so implausible that it\nis insufficient to preserve jurisdiction.” Id. at 174. And if there\nis “any chance” that relief will be effective in securing what\nshe seeks, she has an interest in obtaining it. Mission Prod.\nHoldings, Inc. v. Tempnology, LLC, 139 S. Ct. 1652, 1660\n(2019); see Chafin, 568 U.S. at 175-76 (“[T]he fact that a\ndefendant is insolvent does not moot a claim for damages.”).\n\f 9\n Under this “demanding standard,” plaintiffs’ claims are\nnot moot. See Tempnology, 139 S. Ct. at 1660. They seek a\ncourt order instructing the government to stop implementing\nthe Guidance Memo, process their visa applications, and issue\nthem diversity visas. Neither their claim that such relief is\nlegally available nor their claim that they are entitled to that\nrelief is so implausible as to deprive the district court of\njurisdiction. And there is some chance that this relief will be\neffective at securing their immigration to the United States.\n\n A\n\n The parties dispute whether the district court may lawfully\ntake steps to grant plaintiffs relief, notwithstanding the fact that\nFY 2017 is over. This question goes to the merits, and because\nplaintiffs’ argument that the district court may do so is not “so\nimplausible that it is insufficient to preserve jurisdiction,” the\ncase is not moot. Chafin, 568 U.S. at 174 (explaining that an\nargument about “the legal availability of a certain kind of\nrelief” is a merits question).\n\n Courts are often asked to intervene in disputes over\ndiversity visas, and the end of the selection FY does often\nrender those cases moot. In a straightforward case, a plaintiff\nwho believed the government had erroneously denied her visa\nor was not processing her application quickly enough would\nfile suit well in advance of the end of the selection FY. If the\ncourt agreed, it would order the government to correct the error,\nthe government would timely comply, and the plaintiff would\nreceive her visa before the selection FY ended. More often, the\nplaintiff files suit after the selection FY has ended. Because\ndiversity visas expire when the selection FY ends, 8 U.S.C.\n§§ 1153(c)(1), 1154(a)(1)(I)(ii)(II), that plaintiff does not have\na statutory right to the requested visa and the government does\nnot have a duty to issue her one. Courts faced with this situation\n\f 10\nhave dismissed these lawsuits as moot. Zixiang Li v. Kerry, 710\nF.3d 995 (9th Cir. 2013); Mohamed v. Gonzales, 436 F.3d 79\n(2d Cir. 2006) (per curiam); Nyaga v. Ashcroft, 323 F.3d 906\n(11th Cir. 2003) (per curiam). But see Coraggioso v. Ashcroft,\n355 F.3d 730 (3d Cir. 2004) (dismissing suit on the merits);\nIddir v. INS, 301 F.3d 492 (7th Cir. 2002) (same).\n\n Sometimes a plaintiff files suit before the selection FY\nends but the court fails to act on that request until after\nSeptember 30, at which point the State Department lacks\nauthority to issue a diversity visa sought in the prior fiscal year.\nCourts have likewise dismissed these cases as moot. See\nMwasaru v. Napolitano, 619 F.3d 545 (6th Cir. 2010); Zapata\nv. INS, 93 F. Supp. 2d 355 (S.D.N.Y. 2000).\n\n Other cases involve a different twist. The plaintiff files suit\nand the court grants some relief—but not the visa—before\nOctober 1. In such a case, after the selection FY has ended, the\ncourt might lawfully take steps to compel the government to\nprocess the plaintiff’s application and issue her a diversity visa\nanyway.\n\n That is what happened here. On September 29, the district\ncourt ordered the government to reserve unused FY 2017 visa\nnumbers so that, if it turned out that plaintiffs had erroneously\nbeen denied their diversity visas, the court could order the\ngovernment “to process visas [for plaintiffs] past the statutory\ndeadline.” P.K., 302 F. Supp. 3d at 10-11. The district court\ncited two cases in which courts had done something similar:\nPrzhebelskaya v. USCIS, 338 F. Supp. 2d 399 (E.D.N.Y. 2004)\nand Paunescu v. INS, 76 F. Supp. 2d 896 (N.D. Ill. 1999). In\nboth, before the selection FY ended, the district court ordered\nthe government to timely process plaintiffs’ diversity visa\napplications. Had the government complied, plaintiffs would\nhave received visas. But the government did not, so the court\n\f 11\ninvoked its equitable power to enforce prior orders and\ninstructed the government to issue the plaintiffs visas even\nthough the selection FY had ended. See also Marcetic v. INS,\nNo. 97 C 7018, 1998 WL 173129, at *2 (N.D. Ill. Apr. 6, 1998)\n(ordering government to comply with prior order entered by an\nimmigration judge requiring it to do what was necessary to\nissue plaintiff’s green card).\n\n The question is whether, now that the selection FY has\nended, the district court can order the State Department to do\nanything with the unused visa numbers held in reserve pursuant\nto the September 29 Order. Following Chafin, we hold that this\nis a merits question in the context of this case. There, a father\nsought a court order directing the mother of his daughter to\nreturn the child to the United States after taking her to live\noverseas. Chafin, 568 U.S. at 173. The mother argued that the\nfather did not have a legally cognizable interest in obtaining\nthat order because the court “lack[ed] the authority to issue” it\n“pursuant to its inherent equitable powers.” Id. at 174. But that\nargument—“which goes to the . . . legal availability of a certain\nkind of relief—confuse[d] mootness with the merits.” Id.\nBecause the husband’s claim that the district court possessed\nthe equitable power to issue such an order was not “so\nimplausible that it is insufficient to preserve jurisdiction,” the\ncourt had to assume that claim would prevail, meaning the case\nwas not moot. Id.\n\n It is likewise not “implausible” that the district court here\ncould rely on equity to take steps to compel the issuance of\ndiversity visas, notwithstanding the end of FY 2017. Indeed,\nthe government acknowledges that courts have that power, but\nin its view, that power is limited to cases like Przhebelskaya\nand Paunescu in which the court orders the government to\nprocess a visa application. By contrast, the district court here\nordered the government to hold available visa numbers to\n\f 12\npotentially process plaintiffs’ applications. Because there is no\n“prior court order that was not complied with,” Gov’t Br. 37,\nthe court cannot even “arguabl[y]” invoke equity to provide\nplaintiffs relief, id. at 31, or so the argument goes, see id. at 28-\n37.\n\n This argument assigns more determinacy to the meaning\nof the September 29 Order than it can bear, in our view, for\npurposes of our mootness analysis. That Order instructed the\ngovernment to “hold [unused] visa numbers to process\n[p]laintiffs’ visa applications in the event the Supreme Court\nfinds [EO-2] to be unlawful.” J.A. 112. We hold that the\nSeptember 29 Order need not be read to limit the authority of\nthe district court to grant additional relief to a scenario in which\nthe Supreme Court finds EO-2 unlawful. 2 Rather, it is at least\npossible to read that Order as doing one or both of two other\nthings.\n\n First, the Order might simply have preserved the “status\nquo . . . while the legality of [EO-2]” was pending before the\nSupreme Court, meaning it preserved the district court’s\nability, as of September 29, to rectify the erroneous denial of\nplaintiffs’ visas based on a legally questionable Guidance\nMemo or erroneous interpretation of the INA. See P.K., 302 F.\nSupp. 3d at 7. Second, it may have told the government that if\none specific eventuality arose—the Supreme Court found EO-\n2 unlawful—the State Department must process plaintiffs’ visa\n\n 2\n In the district court’s view, that is all it did, and because the\nSupreme Court never found EO-2 unlawful, the district court\nconcluded that it could not grant plaintiffs further relief. Almaqrami,\n304 F. Supp. 3d at 8. Although the government does not defend this\nreasoning on appeal, and plaintiffs argue this was not “an\nindependent alternative holding,” Pls.’ Br. 24, we are obligated to\naddress potential jurisdictional defects, Floyd v. District of\nColumbia, 129 F.3d 152, 155 (D.C. Cir. 1997).\n\f 13\napplications. These readings may be combined. That is, the\nOrder could be read as (1) instructing the State Department to\n“hold” these unused visa numbers for the purpose of enabling\na later judicial judgment, which might require the government\n“to process” plaintiffs’ applications after any salient obstacles\nwere removed, and (2) identifying the specific judgment that\nwould issue if the Supreme Court took a certain action. Cf.\nCrystal Clear Commc’ns, Inc. v. Sw. Bell Tel. Co., 415 F.3d\n1171, 1177-78 (10th Cir. 2005) (concluding from context that\nan order stating plaintiffs must submit “claims” to the FCC\nonly intended to refer some claims, but not others, to the FCC);\nGjertsen v. Bd. of Election Comm’rs of City of Chi., 751 F.2d\n199 (7th Cir. 1984) (Posner, J.) (context and district court’s\nintent demonstrated that an order styled as a “final judgment”\nwas actually an order granting a preliminary injunction).\n\n On that reading, the September 29 Order left open whether\na later judgment would issue and, if so, what it would look like\nin the event the Supreme Court took any of myriad other\ntacks—for example, holding that the President could rely on\n§ 1182(f) to restrict entry but not visas, that the IRAP I\nplaintiffs lacked standing, that their claims were nonjusticiable,\nor that the case must be dismissed as moot, which the district\ncourt knew was a possibility on September 29. See P.K., 302\nF. Supp. 3d at 4 n.4 (explaining that the Court had vacated oral\nargument in IRAP I and ordered the parties to brief whether the\nProclamation and expiration of EO-2 mooted that case). We\nneed not decide the merit of these various readings; we merely\nnote that they are possible constructions of the September 29\nOrder.\n\n That means plaintiffs’ claim that further relief is legally\navailable is not “so implausible” as to be “insufficient to\npreserve jurisdiction.” Chafin, 568 U.S. at 174. For example, if\nwe were to read the Order as leaving open the possibility that a\n\f 14\nlater judgment would issue, now that IRAP I—the primary\nobstacle the district court identified to granting plaintiffs\nadditional relief—has been dismissed as moot, it is not\n“implausible” that the district court could grant plaintiffs\nadditional relief. As the district court explained, if it “were to\nnow order the State Department to use the unused visa numbers\nto process [p]laintiffs’ visa applications, it would [arguably] be\nrequiring the State Department to fulfill its obligations under”\nthe September 29 Order, which instructed the State Department\nto “reserve the unused visa numbers . . . for a specific purpose:\nthe future processing of [p]laintiffs’ visa applications.”\nAlmaqrami, 304 F. Supp. 3d at 6. Like in Paunescu and\nPrzhebelskaya, such an order would give effect to the district\ncourt’s prior directive, entered before the end of the selection\nFY, to preserve an essential (and otherwise expiring) ingredient\nof relief. To be sure, those cases required the government to\ncomply with a prior order to process applications, see Gov’t\nBr. 37, but they offer useful examples, not binding models, and\nneither of those courts limited their holdings to the precise\nscenario they confronted, see Almaqrami, 304 F. Supp. 3d at 7.\nAnd this case is more similar to Paunescu and Przhebelskaya\nthan the cases dismissed as moot because the plaintiff filed too\nlate or the court did not act in time. All told, this is enough to\nsuggest that plaintiffs’ argument that the district court could\ngrant them additional relief, despite the end of the selection FY,\nis not so “completely devoid of merit as not to involve a federal\ncontroversy.” Steel Co. v. Citizens for a Better Env’t, 523 U.S.\n83, 89 (1998) (quoting Oneida Indian Nation of N.Y. v. County\nof Oneida, 414 U.S. 661, 666 (1974)).\n\n B\n\n Three other issues warrant brief discussion. First, the\ndistrict court concluded that because the expiration of EO-2\nrendered challenges to that order moot, the same was true for\n\f 15\nplaintiffs’ challenges to the Guidance Memo that implemented\nEO-2, and therefore “this case” was moot. Almaqrami, 304 F.\nSupp. 3d at 8. But plaintiffs asked the district court to declare\nthe policy in the Guidance Memo unlawful and enjoin the State\nDepartment from implementing it and to order consular\nofficers to process their visa applications in accordance with\nthe INA. See P.K., 302 F. Supp. 3d at 7. If the Guidance Memo\nis no longer in force, that arguably removes an obstacle to the\nplaintiffs’ ability to obtain an order instructing the government\nto process their applications and issue them visas pursuant to\nthe INA separate and apart from anything provided in the\nMemo. As for their challenges to the Memo itself, although\ncourts generally cannot declare unlawful or enjoin policies that\nare no longer in force, the district court never actually found\nthat the Guidance Memo “expired by its own terms” along with\nEO-2. See Almaqrami, 304 F. Supp. 3d at 7 (quoting IRAP, 138\nS. Ct. at 353; and citing Hawaii, 138 S. Ct. at 377). Indeed, the\npresent record suggests that the Memo did not. See Tr. of Oral\nArg. at 5:10-12 (plaintiffs’ counsel explaining without\ncontradiction that “there is nothing in the record suggesting\nthat” the Guidance Memo or the policy it announced “was ever\nrescinded or modified”). Without making such a finding, the\ndistrict court was too quick to conclude that plaintiffs’\nchallenges to the Memo were moot. Compare Nat’l Ass’n of\nMfrs. v. Dep’t of Def., 138 S. Ct. 617, 627 n.5 (2018) (“Because\nthe [challenged rule] remains on the books for now, the parties\nretain ‘a concrete interest’ in the outcome of this litigation, and\nit is not ‘impossible for a court to grant any effectual relief . . .\nto the prevailing party.’” (quoting Chafin, 568 U.S. at 172)),\nwith Safari Club Int’l v. Jewell, 842 F.3d 1280, 1288 (D.C. Cir.\n2016) (“expiration of a government policy ordinarily moots a\nchallenge to it”). See also Rembert v. Sheahan, 62 F.3d 937,\n938 (7th Cir. 1995) (declining to address mootness in the face\nof an insufficient record).\n\f 16\n Furthermore, whether the Memo has expired has no effect\non the potential viability of plaintiffs’ theories of relief, which\nwe must assume are valid unless they are “wholly insubstantial\nand frivolous.” Steel Co., 523 U.S. at 89 (quoting Bell v. Hood,\n327 U.S. 678, 682-83 (1946)). EO-2 prevented nationals of\nspecific countries from entering the United States; it said\nnothing about diversity visas. Although “aliens who are\ninadmissible” to the United States for reasons described in\ncertain provisions of the INA “are ineligible to receive visas,”\n8 U.S.C. § 1182(a), plaintiffs argue that § 1182(f), which the\nPresident relied on to issue EO-2 and the Proclamation, is not\none of those designated provisions. Plaintiffs also argue that by\ndenying them visas on the basis of nationality, the State\nDepartment and its consular officers violated 8 U.S.C.\n§ 1152(a)(1)(A). They admit that some language in the\nSupreme Court’s Hawaii opinion “is in tension with [their]\nclaim[s],” but they contend that language is neither binding nor\npersuasive, and that other language in Hawaii is “profoundly\nhelpful.” Tr. of Oral Arg. at 6:14-17; see id. at 7:10-9:2, 23:17-\n25:20. We take no position on the merits of these arguments,\nor whether plaintiffs have “state[d] a claim to relief that is\nplausible on its face,” as required by Federal Rule of Civil\nProcedure 12(b)(6). Bell Atl. Corp. v. Twombly, 550 U.S. 544,\n570 (2007). We merely note that they are sufficient to preserve\nthe district court’s subject matter jurisdiction. See Chafin, 568\nU.S. at 174; Steel Co., 523 U.S. at 89.\n\n Finally, because there is some chance that this relief would\nbe effective at securing plaintiffs’ immigration to the United\nStates, their “suit remains live.” Tempnology, 139 S. Ct. at\n1660; see Chafin, 568 U.S. at 172, 174-75 (explaining that,\nunless it is impossible to grant relief that would be effective in\nsecuring the plaintiff’s goal, the case is not moot). Although\nthe Proclamation currently prevents nationals of Iran and\nYemen from entering the country, plaintiffs could qualify for\n\f 17\nan exemption or waiver. Or the President might lift these\nparticular entry restrictions, as he has done for nationals of\nChad. Proclamation No. 9,723, Maintaining Enhanced Vetting\nCapabilities and Processes for Detecting Attempted Entry Into\nthe United States by Terrorists or Other Public-Safety Threats,\n83 Fed. Reg. 15,937, 15,938-39 (2018); see also Hawaii, 138\nS. Ct. at 2422 (explaining other changes made since the initial\nentry restrictions took effect). These possibilities “may be\nuncertain or even unlikely,” Tempnology, 139 S. Ct. at 1660,\nbut that “does not typically render cases moot,” Chafin, 568 at\n175; see Del. Riverkeeper Network v. FERC, 857 F.3d 388, 397\n(D.C. Cir. 2017) (case was not moot even though the court’s\norder “would almost certainly have no real world\nconsequences”). This case is no different.\n\n III\n\n On the present record, this case is not moot. We reverse\nthe order dismissing this case for lack of subject matter\njurisdiction and remand it to the district court for further\nproceedings. 3\n\n 3\n The government also moved to dismiss plaintiffs’ amended\ncomplaint as nonjusticiable based on the doctrine of consular\nnonreviewability and for failure to state a claim under Federal Rule\nof Civil Procedure 12(b)(6). Mem. of Supporting Points and\nAuthorities at 26-31, 34-39, No. 17-cv-01533 (D.D.C. Oct. 20,\n2017), Dkt. No. 53-1. The district court had already rejected the\ngovernment’s consular nonreviewability argument. P.K., 302 F.\nSupp. 3d at 11-12. Because the court dismissed the case as moot, it\ndid not reconsider that consular nonreviewability determination or\nreach the government’s 12(b)(6) argument. Almaqrami, 304 F. Supp.\n3d at 9 & n.3.\n\n When we reverse the dismissal of a case as moot, our usual\npractice is to remand for the district court to consider arguments\n\f 18\n\n So ordered.\n\n\n\n\nabout the merits in the first instance, assuming no other threshold\nissues exist. E.g., Judicial Watch, Inc. v. Kerry, 844 F.3d 952, 956\n(D.C. Cir. 2016). Because courts “may assume without deciding that\nplaintiffs’ statutory claims are reviewable” and proceed to the merits\n“notwithstanding consular nonreviewability,” we see no need to\naddress consular nonreviewability here. Hawaii, 138 S. Ct. at 2407;\nsee Gov’t Br. 38-43 (not arguing this doctrine is jurisdictional).\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Hamed Sufyan Othman ALMAQRAMI, Et Al., Appellants v. Michael R. POMPEO, in His Official Capacity as Secretary of State and John Does, #1-#50, in Their Official Capacity as the Consular Officials Responsible for Issuing Diversity Visas, Appellees\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-13\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued December 14, 2018 Decided August 13, 2019\\n\\n No. 18-5156\\n\\n HAMED SUFYAN OTHMAN ALMAQRAMI, ET AL.,\\n APPELLANTS\\n\\n v.\\n\\n MICHAEL R. POMPEO, IN HIS OFFICIAL CAPACITY AS\\n SECRETARY OF STATE AND JOHN DOES, #1-#50, IN THEIR\\nOFFICIAL CAPACITY AS THE CONSULAR OFFICIALS RESPONSIBLE\\n FOR ISSUING DIVERSITY VISAS,\\n APPELLEES\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:17-cv-01533)\\n\\n\\n Benjamin M. Eidelson argued the cause for appellants.\\nWith him on the briefs were Matthew E. Price, Max J. Minzner,\\nYolanda Rondon, Arthur B. Spitzer, and Scott Michelman.\\n\\n Scott G. Stewart, Attorney, U.S. Department of Justice,\\nargued the cause for appellees. On the brief were Gisela A.\\nWestwater and Erez Reuveni, Assistant Directors, and Joshua\\nS. Press, Trial Attorney.\\n\\n Before: TATEL and GRIFFITH, Circuit Judges, and\\nWILLIAMS, Senior Circuit Judge.\\n\\f 2\\n\\n Opinion for the Court filed by Circuit Judge GRIFFITH.\\n\\n GRIFFITH, Circuit Judge: Plaintiffs were offered the\\nchance to apply for a select number of “diversity visas.” The\\ngovernment never granted them visas, and the statutory\\ndeadline to do so has now passed. But this case is not moot\\nbecause whether the district court retains the authority to award\\nplaintiffs relief is a merits question. We reverse the district\\ncourt’s decision to the contrary.\\n\\n I\\n\\n A\\n\\n In general, a citizen of a foreign country who wishes to\\ncome to the United States must first obtain a U.S. visa, which\\nis placed in the traveler’s passport. A visa does not guarantee\\nentry into the United States; it only confers the right to travel\\nto a port of entry and apply for admission to enter the country.\\n8 U.S.C. § 1201(h). Failure to satisfy certain requirements in\\nthe Immigration and Nationality Act (INA) will render an alien\\nineligible for a visa and ineligible for entry. For example, an\\nalien cannot receive a visa if she has “a communicable disease\\nof public health significance,” and if she contracts such a\\ndisease after receiving her visa, she will be denied entry. 8\\nU.S.C. § 1182(a)(1)(A). Other parts of the INA apply to only\\nvisas or entry. The Secretary of State may, for instance, decline\\nto issue a visa to an alien who abused a position of power to\\nexpropriate American property, id. § 1182d, and the Attorney\\nGeneral may decide that certain aliens must pay bonds before\\nentering the country, id. § 1183.\\n\\n Each fiscal year, the State Department grants\\napproximately 50,000 diversity immigrant visas to individuals\\n\\f 3\\nfrom countries underrepresented in the immigration process,\\nwhich allow recipients who are granted admission to enter the\\ncountry as lawful permanent residents who may live and work\\nhere indefinitely. See 8 U.S.C. §§ 1151(e), 1153(c)(1); Pls.’ Br.\\n4 & n.1. 1 The process by which the State Department awards\\ndiversity visas is competitive and complicated. An applicant\\nmust first apply for and win the diversity visa “lottery.” Pls.’\\nBr. 5; see 8 U.S.C. § 1153(e); 22 C.F.R. § 42.33(b)-(c). A\\nlottery winner or “selectee” must submit an application and\\nvarious documents to be eligible for a visa number—an\\nadministrative device used by the State Department to ensure\\nthat it does not grant more than 50,000 visas per year. Gov’t\\nBr. 9-10; see 8 U.S.C. § 1202(b); 22 C.F.R. §§ 42.33(f)-(g),\\n42.61-67. A selectee is eligible to receive a visa number only\\nduring the fiscal year in which he applied and was selected\\n(“selection FY”). 8 U.S.C. § 1153(e)(2); 22 C.F.R. § 42.33(f).\\nVisa number in hand, the selectee may schedule a consular\\ninterview, and if he meets the criteria to obtain one, the State\\nDepartment “shall” issue him a diversity visa. 8 U.S.C.\\n§ 1153(c), (e)(1); 22 C.F.R. §§ 40.6, 42.81(a); see 8 U.S.C.\\n§ 1202(h). Consulates return unused visa numbers to the State\\nDepartment at the end of each month so that they may be\\nreassigned, but the State Department stops granting visa\\nnumbers altogether once it projects that it will issue all\\navailable visas to existing visa number holders. Gov’t Br. 9-10.\\nBecause the diversity visa program restarts each fiscal year,\\nconsular officers may not issue diversity visas after midnight\\non September 30 of the selection FY. 8 U.S.C.\\n\\n\\n 1\\n A diversity visa applicant who is already living in the United\\nStates in another legal status would instead apply to U.S. Citizenship\\nand Immigration Services (USCIS) for adjustment of status. If\\nsuccessful, he receives a lawful permanent resident identification\\ncard. There is no need to satisfy the entry requirements. We use the\\nterm “diversity visas” to describe both processes.\\n\\f 4\\n§§ 1153(c)(1), 1154(a)(1)(I)(ii)(II); 22 C.F.R. § 42.33(a)(1),\\n(d); see 31 U.S.C. § 1102.\\n\\n B\\n\\n In March 2017, President Trump invoked his authority\\nunder 8 U.S.C. § 1182(f) to “suspend the entry of all aliens or\\nany class of aliens as . . . he may deem to be appropriate” and\\nissued the second iteration of his “travel ban”—an Executive\\nOrder that temporarily prohibited nationals of specific\\ncountries from entering the United States, subject to\\nexemptions and waivers. Exec. Order No. 13,780 (“EO-2”),\\nProtecting the Nation From Foreign Terrorist Entry Into the\\nUnited States §§ 2(c), 3(c), 12(e), 82 Fed. Reg. 13,209, 13,213-\\n15, 13,218 (2017); see Trump v. Hawaii, 138 S. Ct. 2392, 2437\\n(2018). Several district courts issued preliminary injunctions\\npreventing the government from enforcing EO-2’s entry\\nrestriction, which were largely affirmed by the courts of\\nappeals. See Hawaii, 138 S. Ct. at 2437. But in June 2017, the\\nSupreme Court held that EO-2’s entry restriction could take\\neffect while the Court considered the appeals of the preliminary\\ninjunctions, except as to foreign nationals with “a credible\\nclaim of a bona fide relationship with a person or entity in the\\nUnited States.” Trump v. Int’l Refugee Assistance Project\\n(IRAP I), 137 S. Ct. 2080, 2088 (2017) (per curiam).\\n\\n Two days after the Supreme Court’s ruling, the State\\nDepartment issued a “Guidance Memo” instructing consular\\nofficers reviewing diversity visa applications about how EO-\\n2’s entry ban affected visa eligibility: A consular officer should\\nfirst determine whether the selectee “is eligible for the [visa],\\nwithout regard to [EO-2].” J.A. 17. If so, and if he is from a\\ncountry subject to EO-2, the officer must evaluate whether the\\nselectee qualifies for an exemption or waiver, or can establish\\n\\f 5\\na bona fide relationship with the United States. If he cannot, his\\nvisa will be refused.\\n\\n C\\n\\n Plaintiffs won the 2017 diversity visa lottery but were\\ndenied visas pursuant to the Guidance Memo. Consular officers\\ninterviewed plaintiffs and, in accordance with the Memo,\\ndetermined that they would have been eligible for diversity\\nvisas but for the issuance of EO-2. However, because plaintiffs\\nwere from Iran and Yemen—countries subject to the entry\\nban—and could not qualify for exemptions or waivers or\\nsatisfy IRAP I’s bona fide relationship requirement, the\\nconsular officers determined that plaintiffs were “not exempt\\nfrom [EO-2’s] suspension of entry” and denied them visas. See\\nJ.A. 17.\\n\\n Plaintiffs filed suit in the U.S. District Court for the\\nDistrict of Columbia in August 2017. The amended complaint\\nalleges that the Guidance Memo directed consular officers to\\nmake visa determinations on the basis of nationality in\\nviolation of 8 U.S.C. § 1152(a)(1)(A). It further alleges that\\n§ 1182(f) only authorized the President to restrict entry, so by\\nrelying on EO-2 to deny plaintiffs visas, consular officers\\nviolated their duty under the INA to issue visas to all statutorily\\neligible applicants. Among other relief, plaintiffs asked the\\ndistrict court to “enjoin[] the State Department from\\nimplementing the policy set forth in [the Guidance Memo]”\\nand order “consular officers to process [p]laintiffs’ applications\\npursuant to the [INA].” P.K. v. Tillerson, 302 F. Supp. 3d 1, 6\\n(D.D.C. 2017).\\n\\n The next month, the State Department informed the district\\ncourt that it was on track to issue all 50,000 visas allocated for\\nFY 2017 prior to October 1 and would no longer process\\n\\f 6\\nadditional requests for visa numbers made during FY 2017.\\nThe government argued that this development “counsel[ed]\\nagainst” an order that it give plaintiffs visa numbers or process\\ntheir applications. J.A. 33-34. Plaintiffs presented the district\\ncourt with several options. From past practice, it appeared that\\nthe State Department would not reallocate visa numbers\\nreturned in September. Instead of “wast[ing]” those numbers,\\nplaintiffs suggested that the court could order the government\\nto reassign them to plaintiffs. J.A. 46. Recognizing that the\\ndistrict court might have reservations about issuing such a\\nruling while the Supreme Court was reviewing the orders\\nenjoining EO-2’s entry restriction as unlawful, plaintiffs\\nexplained that the district court could instead “maintain the\\nstatus quo” by ordering the State Department to “reserve any\\nunused visa numbers until” IRAP I was “resolved.” Id.\\n\\n On September 24, 2017, EO-2 expired and was replaced\\nby the third iteration of President Trump’s travel ban, the\\n“Proclamation.” Proclamation No. 9,645, Enhancing Vetting\\nCapabilities and Processes for Detecting Attempted Entry Into\\nthe United States by Terrorists or Other Public-Safety Threats,\\n82 Fed. Reg. 45,161 (2017); see P.K., 302 F. Supp. 3d at 3-4.\\nAlthough the Proclamation modified the scope and duration of\\nEO-2’s entry restrictions, the restrictions on Iranians and\\nYemenis remained largely the same.\\n\\n Five days later, on September 29—one day before the end\\nof FY 2017—the district court issued its first ruling. Plaintiffs\\nhad argued that the litigation over EO-2’s entry restrictions was\\n“irrelevant” to their case about visas, J.A. 46, but the district\\ncourt concluded otherwise. Relying on language equating visas\\nand entry in the courts of appeals decisions affirming the\\ninjunctions of EO-2 and the nature of the relief sought in those\\ncases, the district court determined that the Supreme Court’s\\norder staying challenges to EO-2’s entry ban also necessarily\\n\\f 7\\nstayed challenges to EO-2’s effect on visas. P.K., 302 F. Supp.\\n3d at 7-8 & n.7. But in the midst of the uncertainty about the\\nstrength and status of the legal challenges before the Supreme\\nCourt, and with only one day before the end of plaintiffs’\\nselection FY, the district court sought to preserve the status quo\\nand thus keep alive the possibility that plaintiffs could yet\\nreceive their visas. The court therefore ordered the State\\nDepartment to report the number of unused visa numbers for\\nFY 2017 and “hold those visa numbers to process [p]laintiffs’\\nvisa applications in the event the Supreme Court finds [EO-2]\\nto be unlawful.” J.A. 112 (“September 29 Order”). The\\ngovernment has since advised that 27,241 diversity visa\\nnumbers were returned unused and that it issued 49,976\\ndiversity visas in FY 2017—24 shy of the statutory target. This\\nappeal primarily concerns whether the district court may order\\nthe government to keep those unused visas available in the\\nevent these plaintiffs eventually prevail on their claims.\\n\\n In October 2017, the Supreme Court explained that\\nchallenges to the expired entry restrictions of EO-2 were moot.\\nTrump v. Hawaii, 138 S. Ct. 377, 377 (2017); Trump v. IRAP,\\n138 S. Ct. 353, 353 (2017). That same month, the government\\nfiled a motion to dismiss this case as moot, arguing that once\\nFY 2017 ended, the court lacked the power to order the\\ngovernment to issue plaintiffs FY 2017 diversity visas. The\\ndistrict court rejected that theory, but agreed that the case was\\nmoot because the Guidance Memo was issued to implement the\\nnow-expired EO-2 and the September 29 Order had “expressly\\npredicated” any future order requiring the government to\\nprocess plaintiffs’ applications upon “the Supreme Court\\nfind[ing] [EO-2] to be unlawful,” which it had not done.\\nAlmaqrami v. Tillerson, 304 F. Supp. 3d 1, 7-8 (D.D.C. 2018)\\n(first alteration in original) (internal quotation marks omitted).\\nPlaintiffs appeal.\\n\\f 8\\n While this litigation was ongoing, other plaintiffs\\nchallenged the Proclamation’s entry ban on grounds similar to\\nthose argued to enjoin EO-2. Once again, district courts\\nenjoined that ban, and the courts of appeals affirmed. In June\\n2018, the Supreme Court held that those plaintiffs were not\\nlikely to show that the Proclamation was unlawful. Hawaii, 138\\nS. Ct. 2392.\\n\\n II\\n\\n We have jurisdiction pursuant to 28 U.S.C. § 1291. Our\\nreview is de novo, though we accept as true the facts plaintiffs\\nhave alleged. Schnitzer v. Harvey, 389 F.3d 200, 202 (D.C. Cir.\\n2004).\\n\\n A federal court’s jurisdiction is limited to “Cases or\\nControversies.” U.S. Const. art. III, § 2, cl. 1. A lawsuit\\nbecomes moot—and is therefore no longer a “Case” or\\n“Controversy”—“when the issues presented are no longer\\n‘live’ or the parties lack a legally cognizable interest in the\\noutcome.” Chafin v. Chafin, 568 U.S. 165, 172 (2013) (quoting\\nAlready, LLC v. Nike, Inc., 568 U.S. 85, 91 (2013)). But a case\\nbecomes moot only if, assuming the plaintiff prevails, “it is\\nimpossible for a court to grant [her] any effectual relief\\nwhatever.” Id. (quoting Knox v. Serv. Emps. Int’l Union, 567\\nU.S. 298, 307 (2012)). We must assume that the plaintiff will\\n“prevail” unless her argument that the relief sought is legally\\navailable and that she is entitled to it is “so implausible that it\\nis insufficient to preserve jurisdiction.” Id. at 174. And if there\\nis “any chance” that relief will be effective in securing what\\nshe seeks, she has an interest in obtaining it. Mission Prod.\\nHoldings, Inc. v. Tempnology, LLC, 139 S. Ct. 1652, 1660\\n(2019); see Chafin, 568 U.S. at 175-76 (“[T]he fact that a\\ndefendant is insolvent does not moot a claim for damages.”).\\n\\f 9\\n Under this “demanding standard,” plaintiffs’ claims are\\nnot moot. See Tempnology, 139 S. Ct. at 1660. They seek a\\ncourt order instructing the government to stop implementing\\nthe Guidance Memo, process their visa applications, and issue\\nthem diversity visas. Neither their claim that such relief is\\nlegally available nor their claim that they are entitled to that\\nrelief is so implausible as to deprive the district court of\\njurisdiction. And there is some chance that this relief will be\\neffective at securing their immigration to the United States.\\n\\n A\\n\\n The parties dispute whether the district court may lawfully\\ntake steps to grant plaintiffs relief, notwithstanding the fact that\\nFY 2017 is over. This question goes to the merits, and because\\nplaintiffs’ argument that the district court may do so is not “so\\nimplausible that it is insufficient to preserve jurisdiction,” the\\ncase is not moot. Chafin, 568 U.S. at 174 (explaining that an\\nargument about “the legal availability of a certain kind of\\nrelief” is a merits question).\\n\\n Courts are often asked to intervene in disputes over\\ndiversity visas, and the end of the selection FY does often\\nrender those cases moot. In a straightforward case, a plaintiff\\nwho believed the government had erroneously denied her visa\\nor was not processing her application quickly enough would\\nfile suit well in advance of the end of the selection FY. If the\\ncourt agreed, it would order the government to correct the error,\\nthe government would timely comply, and the plaintiff would\\nreceive her visa before the selection FY ended. More often, the\\nplaintiff files suit after the selection FY has ended. Because\\ndiversity visas expire when the selection FY ends, 8 U.S.C.\\n§§ 1153(c)(1), 1154(a)(1)(I)(ii)(II), that plaintiff does not have\\na statutory right to the requested visa and the government does\\nnot have a duty to issue her one. Courts faced with this situation\\n\\f 10\\nhave dismissed these lawsuits as moot. Zixiang Li v. Kerry, 710\\nF.3d 995 (9th Cir. 2013); Mohamed v. Gonzales, 436 F.3d 79\\n(2d Cir. 2006) (per curiam); Nyaga v. Ashcroft, 323 F.3d 906\\n(11th Cir. 2003) (per curiam). But see Coraggioso v. Ashcroft,\\n355 F.3d 730 (3d Cir. 2004) (dismissing suit on the merits);\\nIddir v. INS, 301 F.3d 492 (7th Cir. 2002) (same).\\n\\n Sometimes a plaintiff files suit before the selection FY\\nends but the court fails to act on that request until after\\nSeptember 30, at which point the State Department lacks\\nauthority to issue a diversity visa sought in the prior fiscal year.\\nCourts have likewise dismissed these cases as moot. See\\nMwasaru v. Napolitano, 619 F.3d 545 (6th Cir. 2010); Zapata\\nv. INS, 93 F. Supp. 2d 355 (S.D.N.Y. 2000).\\n\\n Other cases involve a different twist. The plaintiff files suit\\nand the court grants some relief—but not the visa—before\\nOctober 1. In such a case, after the selection FY has ended, the\\ncourt might lawfully take steps to compel the government to\\nprocess the plaintiff’s application and issue her a diversity visa\\nanyway.\\n\\n That is what happened here. On September 29, the district\\ncourt ordered the government to reserve unused FY 2017 visa\\nnumbers so that, if it turned out that plaintiffs had erroneously\\nbeen denied their diversity visas, the court could order the\\ngovernment “to process visas [for plaintiffs] past the statutory\\ndeadline.” P.K., 302 F. Supp. 3d at 10-11. The district court\\ncited two cases in which courts had done something similar:\\nPrzhebelskaya v. USCIS, 338 F. Supp. 2d 399 (E.D.N.Y. 2004)\\nand Paunescu v. INS, 76 F. Supp. 2d 896 (N.D. Ill. 1999). In\\nboth, before the selection FY ended, the district court ordered\\nthe government to timely process plaintiffs’ diversity visa\\napplications. Had the government complied, plaintiffs would\\nhave received visas. But the government did not, so the court\\n\\f 11\\ninvoked its equitable power to enforce prior orders and\\ninstructed the government to issue the plaintiffs visas even\\nthough the selection FY had ended. See also Marcetic v. INS,\\nNo. 97 C 7018, 1998 WL 173129, at *2 (N.D. Ill. Apr. 6, 1998)\\n(ordering government to comply with prior order entered by an\\nimmigration judge requiring it to do what was necessary to\\nissue plaintiff’s green card).\\n\\n The question is whether, now that the selection FY has\\nended, the district court can order the State Department to do\\nanything with the unused visa numbers held in reserve pursuant\\nto the September 29 Order. Following Chafin, we hold that this\\nis a merits question in the context of this case. There, a father\\nsought a court order directing the mother of his daughter to\\nreturn the child to the United States after taking her to live\\noverseas. Chafin, 568 U.S. at 173. The mother argued that the\\nfather did not have a legally cognizable interest in obtaining\\nthat order because the court “lack[ed] the authority to issue” it\\n“pursuant to its inherent equitable powers.” Id. at 174. But that\\nargument—“which goes to the . . . legal availability of a certain\\nkind of relief—confuse[d] mootness with the merits.” Id.\\nBecause the husband’s claim that the district court possessed\\nthe equitable power to issue such an order was not “so\\nimplausible that it is insufficient to preserve jurisdiction,” the\\ncourt had to assume that claim would prevail, meaning the case\\nwas not moot. Id.\\n\\n It is likewise not “implausible” that the district court here\\ncould rely on equity to take steps to compel the issuance of\\ndiversity visas, notwithstanding the end of FY 2017. Indeed,\\nthe government acknowledges that courts have that power, but\\nin its view, that power is limited to cases like Przhebelskaya\\nand Paunescu in which the court orders the government to\\nprocess a visa application. By contrast, the district court here\\nordered the government to hold available visa numbers to\\n\\f 12\\npotentially process plaintiffs’ applications. Because there is no\\n“prior court order that was not complied with,” Gov’t Br. 37,\\nthe court cannot even “arguabl[y]” invoke equity to provide\\nplaintiffs relief, id. at 31, or so the argument goes, see id. at 28-\\n37.\\n\\n This argument assigns more determinacy to the meaning\\nof the September 29 Order than it can bear, in our view, for\\npurposes of our mootness analysis. That Order instructed the\\ngovernment to “hold [unused] visa numbers to process\\n[p]laintiffs’ visa applications in the event the Supreme Court\\nfinds [EO-2] to be unlawful.” J.A. 112. We hold that the\\nSeptember 29 Order need not be read to limit the authority of\\nthe district court to grant additional relief to a scenario in which\\nthe Supreme Court finds EO-2 unlawful. 2 Rather, it is at least\\npossible to read that Order as doing one or both of two other\\nthings.\\n\\n First, the Order might simply have preserved the “status\\nquo . . . while the legality of [EO-2]” was pending before the\\nSupreme Court, meaning it preserved the district court’s\\nability, as of September 29, to rectify the erroneous denial of\\nplaintiffs’ visas based on a legally questionable Guidance\\nMemo or erroneous interpretation of the INA. See P.K., 302 F.\\nSupp. 3d at 7. Second, it may have told the government that if\\none specific eventuality arose—the Supreme Court found EO-\\n2 unlawful—the State Department must process plaintiffs’ visa\\n\\n 2\\n In the district court’s view, that is all it did, and because the\\nSupreme Court never found EO-2 unlawful, the district court\\nconcluded that it could not grant plaintiffs further relief. Almaqrami,\\n304 F. Supp. 3d at 8. Although the government does not defend this\\nreasoning on appeal, and plaintiffs argue this was not “an\\nindependent alternative holding,” Pls.’ Br. 24, we are obligated to\\naddress potential jurisdictional defects, Floyd v. District of\\nColumbia, 129 F.3d 152, 155 (D.C. Cir. 1997).\\n\\f 13\\napplications. These readings may be combined. That is, the\\nOrder could be read as (1) instructing the State Department to\\n“hold” these unused visa numbers for the purpose of enabling\\na later judicial judgment, which might require the government\\n“to process” plaintiffs’ applications after any salient obstacles\\nwere removed, and (2) identifying the specific judgment that\\nwould issue if the Supreme Court took a certain action. Cf.\\nCrystal Clear Commc’ns, Inc. v. Sw. Bell Tel. Co., 415 F.3d\\n1171, 1177-78 (10th Cir. 2005) (concluding from context that\\nan order stating plaintiffs must submit “claims” to the FCC\\nonly intended to refer some claims, but not others, to the FCC);\\nGjertsen v. Bd. of Election Comm’rs of City of Chi., 751 F.2d\\n199 (7th Cir. 1984) (Posner, J.) (context and district court’s\\nintent demonstrated that an order styled as a “final judgment”\\nwas actually an order granting a preliminary injunction).\\n\\n On that reading, the September 29 Order left open whether\\na later judgment would issue and, if so, what it would look like\\nin the event the Supreme Court took any of myriad other\\ntacks—for example, holding that the President could rely on\\n§ 1182(f) to restrict entry but not visas, that the IRAP I\\nplaintiffs lacked standing, that their claims were nonjusticiable,\\nor that the case must be dismissed as moot, which the district\\ncourt knew was a possibility on September 29. See P.K., 302\\nF. Supp. 3d at 4 n.4 (explaining that the Court had vacated oral\\nargument in IRAP I and ordered the parties to brief whether the\\nProclamation and expiration of EO-2 mooted that case). We\\nneed not decide the merit of these various readings; we merely\\nnote that they are possible constructions of the September 29\\nOrder.\\n\\n That means plaintiffs’ claim that further relief is legally\\navailable is not “so implausible” as to be “insufficient to\\npreserve jurisdiction.” Chafin, 568 U.S. at 174. For example, if\\nwe were to read the Order as leaving open the possibility that a\\n\\f 14\\nlater judgment would issue, now that IRAP I—the primary\\nobstacle the district court identified to granting plaintiffs\\nadditional relief—has been dismissed as moot, it is not\\n“implausible” that the district court could grant plaintiffs\\nadditional relief. As the district court explained, if it “were to\\nnow order the State Department to use the unused visa numbers\\nto process [p]laintiffs’ visa applications, it would [arguably] be\\nrequiring the State Department to fulfill its obligations under”\\nthe September 29 Order, which instructed the State Department\\nto “reserve the unused visa numbers . . . for a specific purpose:\\nthe future processing of [p]laintiffs’ visa applications.”\\nAlmaqrami, 304 F. Supp. 3d at 6. Like in Paunescu and\\nPrzhebelskaya, such an order would give effect to the district\\ncourt’s prior directive, entered before the end of the selection\\nFY, to preserve an essential (and otherwise expiring) ingredient\\nof relief. To be sure, those cases required the government to\\ncomply with a prior order to process applications, see Gov’t\\nBr. 37, but they offer useful examples, not binding models, and\\nneither of those courts limited their holdings to the precise\\nscenario they confronted, see Almaqrami, 304 F. Supp. 3d at 7.\\nAnd this case is more similar to Paunescu and Przhebelskaya\\nthan the cases dismissed as moot because the plaintiff filed too\\nlate or the court did not act in time. All told, this is enough to\\nsuggest that plaintiffs’ argument that the district court could\\ngrant them additional relief, despite the end of the selection FY,\\nis not so “completely devoid of merit as not to involve a federal\\ncontroversy.” Steel Co. v. Citizens for a Better Env’t, 523 U.S.\\n83, 89 (1998) (quoting Oneida Indian Nation of N.Y. v. County\\nof Oneida, 414 U.S. 661, 666 (1974)).\\n\\n B\\n\\n Three other issues warrant brief discussion. First, the\\ndistrict court concluded that because the expiration of EO-2\\nrendered challenges to that order moot, the same was true for\\n\\f 15\\nplaintiffs’ challenges to the Guidance Memo that implemented\\nEO-2, and therefore “this case” was moot. Almaqrami, 304 F.\\nSupp. 3d at 8. But plaintiffs asked the district court to declare\\nthe policy in the Guidance Memo unlawful and enjoin the State\\nDepartment from implementing it and to order consular\\nofficers to process their visa applications in accordance with\\nthe INA. See P.K., 302 F. Supp. 3d at 7. If the Guidance Memo\\nis no longer in force, that arguably removes an obstacle to the\\nplaintiffs’ ability to obtain an order instructing the government\\nto process their applications and issue them visas pursuant to\\nthe INA separate and apart from anything provided in the\\nMemo. As for their challenges to the Memo itself, although\\ncourts generally cannot declare unlawful or enjoin policies that\\nare no longer in force, the district court never actually found\\nthat the Guidance Memo “expired by its own terms” along with\\nEO-2. See Almaqrami, 304 F. Supp. 3d at 7 (quoting IRAP, 138\\nS. Ct. at 353; and citing Hawaii, 138 S. Ct. at 377). Indeed, the\\npresent record suggests that the Memo did not. See Tr. of Oral\\nArg. at 5:10-12 (plaintiffs’ counsel explaining without\\ncontradiction that “there is nothing in the record suggesting\\nthat” the Guidance Memo or the policy it announced “was ever\\nrescinded or modified”). Without making such a finding, the\\ndistrict court was too quick to conclude that plaintiffs’\\nchallenges to the Memo were moot. Compare Nat’l Ass’n of\\nMfrs. v. Dep’t of Def., 138 S. Ct. 617, 627 n.5 (2018) (“Because\\nthe [challenged rule] remains on the books for now, the parties\\nretain ‘a concrete interest’ in the outcome of this litigation, and\\nit is not ‘impossible for a court to grant any effectual relief . . .\\nto the prevailing party.’” (quoting Chafin, 568 U.S. at 172)),\\nwith Safari Club Int’l v. Jewell, 842 F.3d 1280, 1288 (D.C. Cir.\\n2016) (“expiration of a government policy ordinarily moots a\\nchallenge to it”). See also Rembert v. Sheahan, 62 F.3d 937,\\n938 (7th Cir. 1995) (declining to address mootness in the face\\nof an insufficient record).\\n\\f 16\\n Furthermore, whether the Memo has expired has no effect\\non the potential viability of plaintiffs’ theories of relief, which\\nwe must assume are valid unless they are “wholly insubstantial\\nand frivolous.” Steel Co., 523 U.S. at 89 (quoting Bell v. Hood,\\n327 U.S. 678, 682-83 (1946)). EO-2 prevented nationals of\\nspecific countries from entering the United States; it said\\nnothing about diversity visas. Although “aliens who are\\ninadmissible” to the United States for reasons described in\\ncertain provisions of the INA “are ineligible to receive visas,”\\n8 U.S.C. § 1182(a), plaintiffs argue that § 1182(f), which the\\nPresident relied on to issue EO-2 and the Proclamation, is not\\none of those designated provisions. Plaintiffs also argue that by\\ndenying them visas on the basis of nationality, the State\\nDepartment and its consular officers violated 8 U.S.C.\\n§ 1152(a)(1)(A). They admit that some language in the\\nSupreme Court’s Hawaii opinion “is in tension with [their]\\nclaim[s],” but they contend that language is neither binding nor\\npersuasive, and that other language in Hawaii is “profoundly\\nhelpful.” Tr. of Oral Arg. at 6:14-17; see id. at 7:10-9:2, 23:17-\\n25:20. We take no position on the merits of these arguments,\\nor whether plaintiffs have “state[d] a claim to relief that is\\nplausible on its face,” as required by Federal Rule of Civil\\nProcedure 12(b)(6). Bell Atl. Corp. v. Twombly, 550 U.S. 544,\\n570 (2007). We merely note that they are sufficient to preserve\\nthe district court’s subject matter jurisdiction. See Chafin, 568\\nU.S. at 174; Steel Co., 523 U.S. at 89.\\n\\n Finally, because there is some chance that this relief would\\nbe effective at securing plaintiffs’ immigration to the United\\nStates, their “suit remains live.” Tempnology, 139 S. Ct. at\\n1660; see Chafin, 568 U.S. at 172, 174-75 (explaining that,\\nunless it is impossible to grant relief that would be effective in\\nsecuring the plaintiff’s goal, the case is not moot). Although\\nthe Proclamation currently prevents nationals of Iran and\\nYemen from entering the country, plaintiffs could qualify for\\n\\f 17\\nan exemption or waiver. Or the President might lift these\\nparticular entry restrictions, as he has done for nationals of\\nChad. Proclamation No. 9,723, Maintaining Enhanced Vetting\\nCapabilities and Processes for Detecting Attempted Entry Into\\nthe United States by Terrorists or Other Public-Safety Threats,\\n83 Fed. Reg. 15,937, 15,938-39 (2018); see also Hawaii, 138\\nS. Ct. at 2422 (explaining other changes made since the initial\\nentry restrictions took effect). These possibilities “may be\\nuncertain or even unlikely,” Tempnology, 139 S. Ct. at 1660,\\nbut that “does not typically render cases moot,” Chafin, 568 at\\n175; see Del. Riverkeeper Network v. FERC, 857 F.3d 388, 397\\n(D.C. Cir. 2017) (case was not moot even though the court’s\\norder “would almost certainly have no real world\\nconsequences”). This case is no different.\\n\\n III\\n\\n On the present record, this case is not moot. We reverse\\nthe order dismissing this case for lack of subject matter\\njurisdiction and remand it to the district court for further\\nproceedings. 3\\n\\n 3\\n The government also moved to dismiss plaintiffs’ amended\\ncomplaint as nonjusticiable based on the doctrine of consular\\nnonreviewability and for failure to state a claim under Federal Rule\\nof Civil Procedure 12(b)(6). Mem. of Supporting Points and\\nAuthorities at 26-31, 34-39, No. 17-cv-01533 (D.D.C. Oct. 20,\\n2017), Dkt. No. 53-1. The district court had already rejected the\\ngovernment’s consular nonreviewability argument. P.K., 302 F.\\nSupp. 3d at 11-12. Because the court dismissed the case as moot, it\\ndid not reconsider that consular nonreviewability determination or\\nreach the government’s 12(b)(6) argument. Almaqrami, 304 F. Supp.\\n3d at 9 & n.3.\\n\\n When we reverse the dismissal of a case as moot, our usual\\npractice is to remand for the district court to consider arguments\\n\\f 18\\n\\n So ordered.\\n\\n\\n\\n\\nabout the merits in the first instance, assuming no other threshold\\nissues exist. E.g., Judicial Watch, Inc. v. Kerry, 844 F.3d 952, 956\\n(D.C. Cir. 2016). Because courts “may assume without deciding that\\nplaintiffs’ statutory claims are reviewable” and proceed to the merits\\n“notwithstanding consular nonreviewability,” we see no need to\\naddress consular nonreviewability here. Hawaii, 138 S. Ct. at 2407;\\nsee Gov’t Br. 38-43 (not arguing this doctrine is jurisdictional).\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4174,"cells":{"id":{"kind":"number","value":4647784,"string":"4,647,784"},"name":{"kind":"string","value":"CHILDREN'S HOSPITAL ASSOCIATION OF TEXAS, Et Al., Appellees v. Alex Michael AZAR, II, in His Official Capacity, Secretary, Department of Health and Human Services, Et Al., Appellants"},"name_abbreviation":{"kind":"string","value":"Children's Hospital Ass'n v. Alex Azar"},"decision_date":{"kind":"string","value":"2019-08-13"},"docket_number":{"kind":"string","value":"18-5135"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Henderson, Rogers, Sentelle\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge HENDERSON\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 9, 2019 Decided August 13, 2019\\n\\n No. 18-5135\\n\\n CHILDREN’S HOSPITAL ASSOCIATION OF TEXAS, ET AL.,\\n APPELLEES\\n\\n v.\\n\\n ALEX MICHAEL AZAR, II, IN HIS OFFICIAL CAPACITY,\\n SECRETARY,\\n DEPARTMENT OF HEALTH AND HUMAN SERVICES, ET AL.,\\n APPELLANTS\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:17-cv-00844)\\n\\n\\n Tara S. Morrissey, Attorney, United States Department of\\nJustice, argued the cause for the appellants. Mark B. Stern and\\nSamantha L. Chaifetz, Attorneys, Robert P. Charrow, General\\nCounsel, United States Department of Health & Human\\nServices, Janice L. Hoffman, Associate General Counsel,\\nSusan M. Lyons, Deputy Associate General Counsel, and\\nDavid L. Hoskins, Attorney, were with her on brief.\\n\\n David B. Salmons argued the cause for the appellees.\\nGeraldine E. Edens, Michael E. Kenneally, and Susan Feigin\\nHarris were with him on brief. Christopher H. Marraro\\nentered an appearance.\\n\\f 2\\n\\n Daniel G. Jarcho and Michael H. Park were on brief for\\nthe amicus curiae Children’s Hospital Association in support\\nof the appellees and in support of affirmance.\\n\\n Before: HENDERSON and ROGERS, Circuit Judges, and\\nSENTELLE, Senior Circuit Judge.\\n\\n Opinion for the Court filed by Circuit Judge HENDERSON.\\n\\n KAREN LECRAFT HENDERSON, Circuit Judge: Under the\\nMedicaid Act (Act), the federal government provides each\\nstate funds for distribution to hospitals that serve a\\ndisproportionate number of low-income patients. See 42\\nU.S.C. § 1396-1. A state distributes the funds through\\nDisproportionate Share Hospital (DSH) payments. See id.\\n§ 1396r-4(b), (c). A hospital may not receive a DSH payment\\nthat exceeds its “costs incurred” in furnishing hospital services\\nto low-income patients. Id. § 1396r-4(g)(1)(A). “Costs\\nincurred” are, inter alia, “determined by the Secretary” of the\\nUnited States Department of Health and Human Services\\n(Secretary). Id. In 2017, the Secretary promulgated a regulation\\ndefining “costs incurred.” Medicaid Program;\\nDisproportionate Share Hospital Payments–Treatment of Third\\nParty Payers in Calculating Uncompensated Care Costs, 82\\nFed. Reg. 16,114, 16,122 (Apr. 3, 2017) (“2017 Rule”). The\\nplaintiffs, a group of children’s hospitals that receive DSH\\npayments, argue that the regulatory definition is contrary to the\\nMedicaid Act and otherwise arbitrary and capricious. The\\ndistrict court agreed that the definition is inconsistent with the\\nAct and vacated the 2017 Rule. Children’s Hosp. Ass’n of Tex.\\nv. Azar, 300 F. Supp. 3d 190 (D.D.C. 2018). We now reverse.\\n\\f 3\\n I. Background\\n\\n “Medicaid is a cooperative federal-state program through\\nwhich the Federal Government provides financial assistance to\\nStates so that they may furnish medical care to needy\\nindividuals.” Wilder v. Va. Hosp. Ass’n, 496 U.S. 498, 502\\n(1990). States implement their own Medicaid plans, subject to\\nthe federal government’s review and approval. See 42 U.S.C.\\n§ 1396a. Treating the indigent proves costly even for hospitals\\nthat receive Medicaid payments. Indeed, not all hospital\\nservices are covered by Medicaid; not all costs associated with\\ncovered services are allowed by Medicaid; and Medicaid does\\nnot fully reimburse hospitals for all allowable costs associated\\nwith covered services. Recognizing this, the Congress\\nauthorizes supplemental payments (“DSH payments”) to\\nhospitals that serve a disproportionate share of low-income\\npatients (“DSH hospitals”). 42 U.S.C. § 1396a(13)(A)(iv)\\n(requiring that Medicaid payment rates “take into account (in a\\nmanner consistent with section 1396r-4 of this title) the\\nsituation of hospitals which serve a disproportionate number of\\nlow-income patients with special needs”); 42 U.S.C. § 1396r-\\n4 (entitled “Adjustment in payment for inpatient hospital\\nservices furnished by disproportionate share hospitals”). There\\nis both a state-specific and a hospital-specific limit on DSH\\npayments. The state-specific limit—not at issue in this case—\\ndictates that all DSH payments to DSH hospitals within a\\nsingle state must be drawn from the same pool of federal funds.\\nSee 42 U.S.C. § 1396r-4(f). The hospital-specific limit, which\\nis at issue in this case, dictates that a DSH payment to a single\\nhospital cannot exceed:\\n\\n [T]he costs incurred during the year of\\n furnishing hospital services (as determined by\\n the Secretary and net of payments under this\\n subchapter, other than under this section, and by\\n\\f 4\\n uninsured patients) by the hospital to\\n individuals who either are eligible for medical\\n assistance under the State plan or have no health\\n insurance (or other source of third party\\n coverage) for services provided during the year.\\n\\n42 U.S.C. § 1396r-4(g)(1)(A). This sentence—although not the\\npicture of clarity—establishes a few matters clearly. A DSH\\nhospital cannot receive a DSH payment that exceeds its “costs\\nincurred during the year of furnishing hospital services” to\\nMedicaid-eligible and uninsured individuals. The Secretary is\\nassigned the task of determining “costs incurred.” And “costs\\nincurred” are “net of payments under this subchapter, other\\nthan under this section, and by uninsured patients”; in other\\nwords, payments made by Medicaid and uninsured individuals\\nmust be subtracted out when calculating a hospital’s “costs\\nincurred.” The dispute here is about whether payments made\\nby Medicare and private insurers should also be subtracted out.\\n\\n In 2003, the Congress enacted legislation requiring states\\nto submit annual reports and independent certified audits\\nregarding their DSH programs. See 42 U.S.C. § 1396r-4(j). The\\nreports must identify which hospitals receive DSH payments\\nand the audits must verify that the DSH payments comply with\\nthe statutory requirements. Id.\\n\\n In 2008, the Centers for Medicare & Medicaid Services\\n(CMS), using the authority delegated it by the Secretary,\\npromulgated a regulation implementing the reporting and\\nauditing requirements. Medicaid Program; Disproportionate\\nShare Hospital Payments, 73 Fed. Reg. 77,904 (Dec. 19, 2008)\\n(“2008 Rule”). The 2008 Rule provided that each state must\\nreport to CMS the cost of each DSH hospital’s “Total Medicaid\\nUncompensated Care.” Id. at 77,950 (codified at 42 C.F.R.\\n§ 447.299(c)(11)). The 2008 Rule did not state whether third-\\nparty payments, including payments by Medicare and private\\n\\f 5\\ninsurers, were meant to be included in calculating the amount.\\nSee id. Three courts of appeals concluded from this silence that\\nthe 2008 Rule left uncertain whether these payments should be\\nconsidered. See Children’s Health Care v. Ctrs. for Medicare\\n& Medicaid Servs., 900 F.3d 1022, 1025 (8th Cir. 2018);\\nChildren’s Hosp. of the King’s Daughters, Inc. v. Azar, 896\\nF.3d 615, 621 (4th Cir. 2018); N.H. Hosp. Ass’n v. Azar, 887\\nF.3d 62, 75 (1st Cir. 2018). One court of appeals concluded that\\nthe 2008 Rule made clear that these payments should not be\\nconsidered. See Tenn. Hosp. Ass’n v. Azar, 908 F.3d 1029,\\n1043–44 (6th Cir. 2018).\\n\\n In 2010, CMS posted a Frequently Asked Questions\\ndocument on its website clarifying that payments made by\\nMedicare and private insurers should be included. See CMS,\\nAdditional Information on the DSH Reporting and Audit\\nRequirements, FAQs 33 and 34 (2010),\\nhttps://www.medicaid.gov/medicaid/finance/downloads/part-\\n1-additional-info-on-dsh-reporting-and-auditing.pdf. A\\nnumber of hospitals brought suit, arguing that the FAQs\\nposting was invalid because it represented a substantive policy\\nchange without notice and an opportunity for public comment.\\nIn response, CMS issued a notice of proposed rulemaking and\\nsubsequently promulgated the 2017 Rule. The 2017 Rule\\nestablishes that payments by Medicare and private insurers are\\nto be included in calculating a hospital’s “costs incurred.” 82\\nFed. Reg. at 16,122 (codified at 42 C.F.R. § 447.299(c)(10)).\\nIt provides, inter alia, “costs . . . [a]re defined as costs net of\\nthird-party payments, including, but not limited to, payments\\nby Medicare and private insurance.” Id. The Secretary explains\\nthat considering payments by Medicare and private insurers\\n“best fulfills the purpose of the DSH statute,” is “necessary to\\nensure that limited DSH resources are allocated to hospitals\\nthat have a net financial shortfall in serving Medicaid patients”\\nand “is necessary to facilitate the Congressional directive . . .\\n\\f 6\\nof limiting the DSH payment to a hospital’s uncompensated\\ncare costs.” Id. at 16,116, 16,118. He maintains that the 2017\\nRule did not effect a legal change but instead continued the\\npreexisting policy. Id. at 16,119.\\n\\n The plaintiffs are four children’s hospitals in Minnesota,\\nVirginia and Washington and an association representing eight\\nchildren’s hospitals in Texas. They claim the 2017 Rule\\nviolates the Administrative Procedure Act because it exceeds\\nthe Secretary’s authority under the Medicaid Act and is the\\nproduct of arbitrary and capricious reasoning. See 5 U.S.C.\\n§ 706(2)(A), (C). The district court entered summary judgment\\nfor the plaintiffs, holding that the Rule “is inconsistent with the\\nplain language of the Medicaid Act,” which “clearly indicates\\nwhich payments can be subtracted from the total costs incurred\\nduring the year by hospitals” and “nowhere mentions\\nsubtracting other third-party payments made on behalf of\\nMedicaid-eligible patients from the total costs incurred.”\\nChildren’s Hosp. Ass’n of Tex., 300 F. Supp. 3d at 205, 207.\\nHaving held the 2017 Rule invalid under § 706(2)(C) (“ultra\\nvires” prohibition), the district court did not reach the\\nplaintiffs’ § 706(2)(A) challenge (“arbitrary and capricious”\\nprohibition). Id. at 205. The district court ultimately vacated\\nthe 2017 Rule and the Secretary timely appealed. Id. at 210–\\n11. Our review is de novo. See Ark Initiative v. Tidwell, 816\\nF.3d 119, 126–27 (D.C. Cir. 2016).\\n\\n II. Analysis\\n\\n A. Exceeds Statutory Authority\\n\\n The plaintiffs first challenge the Rule as exceeding the\\nSecretary’s authority under the Medicaid Act, in violation of 5\\nU.S.C. § 706(2)(C). The familiar Chevron framework guides\\nour review. See Ass’n of Private Sector Colls. & Univs. v.\\nDuncan, 681 F.3d 427, 441 (D.C. Cir. 2012) (“Appellant’s\\n\\f 7\\nclaims that various provisions of the challenged regulations are\\n‘in excess of statutory jurisdiction, authority, or limitations, or\\nshort of statutory right,’ are reviewed under the well-known\\nChevron framework.” (citation omitted) (quoting 5 U.S.C.\\n§ 706(2)(C))). “Under that framework, we ask whether the\\nstatute is ambiguous and, if so, whether the agency’s\\ninterpretation is reasonable.” King v. Burwell, 135 S. Ct. 2480,\\n2488 (2015). “This approach ‘is premised on the theory that a\\nstatute’s ambiguity constitutes an implicit delegation from\\nCongress to the agency to fill in the statutory gaps.’” Id.\\n(quoting FDA v. Brown & Williamson Tobacco Corp., 529\\nU.S. 120, 159 (2000)). Because the delegation at issue here is\\nexpress rather than implied, see 42 U.S.C. § 1396r-4(g)(1)(A)\\n(“[T]he costs incurred during the year of furnishing hospital\\nservices (as determined by the Secretary . . .) . . . .” (emphasis\\nadded)); see also Transitional Hosps. Corp. of La. v. Shalala,\\n222 F.3d 1019, 1025 (D.C. Cir. 2000) (“as determined by the\\nSecretary” is “express delegation”), we have no need to search\\nfor statutory ambiguity. We skip straight to asking whether the\\nRule is reasonable. See Chevron U.S.A. Inc. v. Nat. Res. Def.\\nCouncil, Inc., 467 U.S. 837, 843–44 (1984) (“If Congress has\\nexplicitly left a gap for the agency to fill, there is an express\\ndelegation of authority to the agency to elucidate a specific\\nprovision of the statute by regulation. Such legislative\\nregulations are given controlling weight unless they are\\narbitrary, capricious, or manifestly contrary to the statute.”);\\nsee also Transitional Hosps. Corp. of La., 222 F.3d at 1025 (if\\ndelegation is express “we are bound to uphold [the Secretary’s]\\ndetermination as long as she exercises [her] discretion in a\\nreasonable way”).\\n\\n The plaintiffs offer four principal reasons the statute does\\nnot grant the Secretary authority to require that payments by\\nMedicare and private insurers be considered in calculating a\\nhospital’s “costs incurred.” First, the statute exclusively\\n\\f 8\\nspecifies which payments can be considered. Second, the Rule\\nrenders superfluous the statute’s specification that certain\\npayments must be considered. Third, the Congress required\\nconsideration of third-party payments in a different statutory\\nprovision but not in the relevant provision. Fourth, the statute\\nplainly distinguishes costs and payments. We reject all four\\narguments.1\\n\\n First, we disagree with the plaintiffs’ argument that the\\nstatute exclusively specifies which payments can be considered\\nin calculating “costs incurred.” See Plaintiffs’ Br. at 58;\\nChildren’s Hosp. Ass’n of Tex., 300 F. Supp. 3d at 207 (“On its\\nface, the statute clearly indicates which payments can be\\nsubtracted from the total costs incurred during the year by\\nhospitals: (1) ‘payments under this subchapter,’ i.e., payments\\nmade by Medicaid; and (2) payments made by uninsured\\npatients. The statute nowhere mentions subtracting other third-\\nparty payments made on behalf of Medicaid-eligible patients\\nfrom the total costs incurred.”). Although the statute\\nestablishes that payments by Medicaid and the uninsured must\\nbe considered, it nowhere states that those are the only\\npayments that may be considered. To conclude otherwise, we\\nwould have to rely on the interpretive canon expressio unius\\nest exclusio alterius, which means “expressing one item of [an]\\nassociated group or series excludes another left unmentioned.”\\nChevron U.S.A. Inc. v. Echazabal, 536 U.S. 73, 80 (2002)\\n(alteration in original) (quoting United States v. Vonn, 535 U.S.\\n55, 65 (2002)). But that canon has been called a “feeble helper\\nin an administrative setting.” Adirondack Med. Ctr. v. Sebelius,\\n740 F.3d 692, 697 (D.C. Cir. 2014) (quoting Cheney R.R. Co.\\nv. ICC, 902 F.2d 66, 69 (1990)). And, in any setting, it “applies\\nonly when ‘circumstances support[] a sensible inference that\\n\\n 1\\n We have also considered and reject the plaintiffs’ other\\narguments.\\n\\f 9\\nthe term left out must have been meant to be excluded.’” NLRB\\nv. SW Gen., Inc., 137 S. Ct. 929, 940 (2017) (quoting\\nEchazabal, 536 U.S. at 81); see also Barnhart v. Peabody Coal\\nCo., 537 U.S. 149, 168 (2003) (“[W]e do not read the\\nenumeration of one . . . to exclude the other unless it is fair to\\nsuppose that Congress considered the unnamed possibility and\\nmeant to say no to it.”). There is reason to believe the Congress\\ndid not intend to exclude Medicare and private insurance\\npayments from consideration. Indeed, the parties agree that the\\nmost common sources of payment for treating Medicaid-\\neligible and uninsured individuals are Medicaid and the\\nuninsured. The Congress may have wanted to ensure that the\\nmost common sources of payment must be considered but at\\nthe same time allow the Secretary to decide whether less-\\ncommon sources of payment should be as well. Especially in\\nlight of this plausible alternative explanation, we will not rely\\non the expressio unius canon to find that the statute exclusively\\nspecifies which payments can be considered in calculating\\n“costs incurred.” See Indep. Ins. Agents of Am., Inc. v. Hawke,\\n211 F.3d 638, 644 (D.C. Cir. 2000) (“[I]f there are other\\nreasonable explanations for an omission in a statute, expressio\\nunius may not be a useful tool.”).\\n\\n Second, we disagree with the plaintiffs’ argument that the\\nRule renders superfluous the statute’s specification that\\npayments by Medicaid and the uninsured must be considered.\\nSee Plaintiffs’ Br. 41–42; Children’s Hosp. Ass’n of Tex., 300\\nF. Supp. 3d at 207 (“To allow the Secretary to redefine ‘costs’\\nto net out a third category of payments—i.e., ‘third-party\\npayments, including but not limited to, payments by Medicare\\nand private insurance’ . . .—would ‘render the Congressional\\ndefinition of payments in the very same clause superfluous.’”\\n(quoting Children’s Hosp. of the King’s Daughters, Inc. v.\\nPrice, 258 F. Supp. 3d 672, 687 (E.D. Va. 2017))); id.\\n(“[D]efendants’ interpretation of the statute would render\\n\\f 10\\nportions of the statutory language superfluous.”). The statute’s\\nspecification that two forms of payment must be considered\\nremoves the Secretary’s discretion as to those two forms of\\npayment. But it does nothing to disturb the Secretary’s\\ndiscretion as to other forms of payment, which may be\\nconsidered. See Tenn. Hosp. Ass’n, 908 F.3d at 1038 (“[T]he\\nfact that certain payments must be deducted from costs does\\nnot mean that other payments cannot be.”); N.H. Hosp. Ass’n,\\n887 F.3d at 66 (“Congress identified two specific sources of\\npayment that must be offset against total costs, but otherwise\\nsimply stated that ‘costs incurred’ are ‘as determined by the\\nSecretary.’”).\\n\\n Third, we reject the plaintiffs’ argument that we should\\ninfer from the Congress’s requiring consideration of third party\\npayments under 42 U.S.C. § 1396r-4(g)(2)(A)—a provision\\nthat, before 1995, allowed certain hospitals to receive\\npayments that exceeded their uncompensated costs—that it\\nmeant to prohibit consideration of third party payments under\\n§ 1396r-4(g)(1)(A). Plaintiffs’ Br. 34–35. This argument is\\nbased on the so-called “Russello presumption—that the\\npresence of a phrase in one provision and its absence in another\\nreveals Congress’ design.” City of Columbus v. Ours Garage\\n& Wrecker Serv., Inc., 536 U.S. 424, 435–36 (2002) (citing\\nRussello v. United States, 464 U.S. 16 (1983)). But that\\npresumption “grows weaker with each difference in the\\nformulation of the provisions under inspection.” Id. at 436.\\nBecause sections (g)(1)(A) and (g)(2)(A) are fundamentally\\ndifferent, we find the plaintiffs’ argument unpersuasive. See\\nTenn. Hosp. Ass’n, 908 F.3d at 1039 (“There is no tension,\\nhowever, in Congress requiring third-party payment\\ndeductions in subsection (g)(2)(A) and allowing third-party\\npayment deductions in subsection (g)(1)(A). The DSH\\npayments provided for in (g)(2)(A) are above and beyond those\\nmandated by (g)(1)(A); it therefore makes sense for Congress\\n\\f 11\\nto impose a hard limit on the ceiling of (g)(2)(A) funds—i.e.,\\nno more than 200% of the costs of serving Medicaid-eligible\\npatients, less payments from Medicaid, uninsured patients, and\\n‘third party payors’—while giving CMS more discretion to\\ncalibrate the appropriate cap on the ‘standard’ DSH payments\\ndiscussed in (g)(1)(A).”).\\n\\n Fourth, we disagree with the plaintiffs’ argument that the\\nstatute plainly distinguishes between costs and payments such\\nthat payments can never be considered in calculating “costs\\nincurred.” See Plaintiffs’ Br. at 26, 30, 33, 40, 57. The statute\\nestablishes that a hospital’s DSH payment cannot exceed its\\n“costs incurred during the year of furnishing hospital services\\n(as determined by the Secretary and net of payments under this\\nsubchapter, other than under this section, and by uninsured\\npatients).” 42 U.S.C. § 1396r-4(g)(1)(A). Both of the\\nparenthetical’s adjoining phrases modify “costs incurred”—\\nthat is, “costs incurred” are both “as determined by the\\nSecretary” and “net of payments under [Medicaid] and by\\nuninsured patients.” Id. In other words, the statute requires that\\nsome payments be considered in calculating a hospital’s “costs\\nincurred.” The argument that the statute separates costs and\\npayments therefore flies in the teeth of the statutory text.\\n\\n Contrary to the plaintiffs’ contention, we believe the 2017\\nRule is consistent with the statute’s context and purpose, both\\nof which suggest DSH payments are meant to assist those\\nhospitals that need them most by covering only those costs for\\nwhich DSH hospitals are in fact uncompensated. See 42 U.S.C.\\n§ 1396r-4(g)(1) (heading of provision at issue: “Amount of\\nadjustment subject to uncompensated costs”); 2 42 U.S.C.\\n§ 1396r-4(j)(2)(C) (requiring states to certify that “[o]nly the\\n\\n 2\\n Headings, although “not commanding,” “supply clues” about\\nCongressional intent. Yates v. United States, 135 S. Ct. 1074, 1083\\n(2015).\\n\\f 12\\nuncompensated care costs of providing inpatient hospital and\\noutpatient hospital services to individuals described in\\n[(g)(1)(A)] are included in the calculation of the hospital-\\nspecific limits.”); H.R. Rep. No. 103-111, at 211 (in enacting\\nDSH payment limit, Congress noted “some States have made\\nDSH payment[s] . . . to State psychiatric or university hospitals\\nin amounts that exceed the net costs, and in some instances the\\ntotal costs, of operating the facilities”); Tenn. Hosp. Ass’n, 908\\nF.3d at 1040 (“essence” of Congressional concern in enacting\\nstatute was “that hospitals were double dipping by collecting\\nDSH payments to cover costs that had already been\\nreimbursed”). By requiring the inclusion of payments by\\nMedicare and private insurers, the 2017 Rule ensures that DSH\\npayments will go to hospitals that have been compensated least\\nand are thus most in need. Because the 2017 Rule is consistent\\nwith the statute, it does not violate § 706(2)(C).\\n\\n B. Arbitrary and Capricious\\n\\n The plaintiffs next challenge the 2017 Rule as the product\\nof arbitrary and capricious reasoning, in violation of 5 U.S.C.\\n§ 706(2)(A). A reviewing court “shall . . . hold unlawful and\\nset aside agency action, findings, and conclusions found to be\\n. . . arbitrary, capricious, an abuse of discretion, or otherwise\\nnot in accordance with law.” 5 U.S.C. § 706(2)(A). An agency\\nrulemaking is arbitrary and capricious “if the agency has relied\\non factors which Congress has not intended it to consider,\\nentirely failed to consider an important aspect of the problem,\\noffered an explanation for its decision that runs counter to the\\nevidence before the agency, or is so implausible that it could\\nnot be ascribed to a difference in view or the product of agency\\nexpertise.” Motor Vehicles Mfrs. Ass’n of U.S. v. State Farm\\nMut. Auto. Ins. Co., 463 U.S. 29, 43 (1983).\\n\\n The plaintiffs first contend that “CMS has never\\nacknowledged, let alone justified, its new Rule’s departure\\n\\f 13\\nfrom the 2008 rule.” Plaintiffs’ Br. 27–28. We disagree.\\n“Agencies are free to change their existing policies as long as\\nthey provide a reasoned explanation for the change.” Encino\\nMotorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125 (2016). An\\nagency need not provide a more detailed justification for a\\nchanged policy than it would for a brand-new policy. Id. But it\\nmust provide “a reasoned explanation . . . for disregarding facts\\nand circumstances that underlay or were engendered by the\\nprior policy.” Id. at 2126 (quoting FCC v. Fox Television\\nStations, Inc., 556 U.S. 502, 516 (2009)). An “unexplained\\ninconsistency” with an earlier position renders a changed\\npolicy arbitrary and capricious. Id. (quoting Nat. Cable &\\nTelecomm. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 981\\n(2005)).\\n\\n There is no unexplained inconsistency with an earlier\\nposition here. To be clear, we agree with the plaintiffs that the\\n2017 Rule and the 2008 Rule establish different policies.3 But\\nit makes no difference. CMS explained why the statute’s\\npurposes are better fulfilled by a policy that requires\\nconsideration of payments by Medicare and private insurers\\n\\n 3\\n The Secretary maintains that the 2017 Rule is consistent with\\nthe 2008 Rule and so does not establish a new policy. That argument\\nhas been rejected by four courts of appeals, all of which found the\\n2010 FAQs procedurally invalid because the policy established\\ntherein, which is the same policy established by the 2017 Rule,\\nmarked a departure from the policy established by the 2008 Rule\\nwithout notice and an opportunity for public comment. See Tenn.\\nHosp. Ass’n, 908 F.3d at 1043 (“As three circuit courts and several\\ndistrict courts have now held, the payment-deduction policy\\nelucidated in the FAQs and hinted at in the preamble to the 2008 rule\\nseeks to amend, rather than merely clarify, the 2008 regulations.”\\n(citing Children’s Health Care, 900 F.3d at 1026–27; Children’s\\nHosp. of the King’s Daughters, 896 F.3d at 623; N.H. Hosp. Ass’n,\\n887 F.3d at 74)). We agree with our sister circuits.\\n\\f 14\\n(the 2017 Rule) than one that does not (the 2008 Rule, as we\\ninterpret it):\\n\\n In light of the statutory requirement limiting\\n DSH payments on a hospital-specific basis to\\n uncompensated care costs, it is inconsistent\\n with the statute to assist hospitals with costs that\\n have already been compensated by third party\\n payments. [The 2017] rule is designed to\\n reiterate the policy and make explicit within the\\n terms of the regulation that all costs and\\n payments associated with dual eligible and\\n individuals with a source of third party\\n coverage must be included in calculating the\\n hospital-specific DSH limit. This policy is\\n necessary to ensure that only actual\\n uncompensated care costs are included in the\\n Medicaid hospital-specific DSH limit. And,\\n because state DSH payments are limited to an\\n annual federal allotment, this policy is also\\n necessary to ensure that limited DSH resources\\n are allocated to hospitals that have a net\\n financial shortfall in serving Medicaid patients.\\n\\n82 Fed. Reg. at 16,117. This explanation is more than sufficient\\nto survive review under § 706(2)(A).\\n\\n The plaintiffs also claim that the Secretary has not tied the\\n2017 Rule to the administrative record. According to their\\nreading, the record shows that CMS reduces DSH payments to\\nthe plaintiff hospitals when it considers private insurance\\npayments, notwithstanding “they have among the highest\\nMedicaid inpatient utilization rates in their respective states\\nand the highest net financial shortfalls in serving Medicaid\\npatients.” Plaintiffs Br. 65. The plaintiffs claim this outcome is\\ninconsistent with the purpose of the 2017 Rule, which is “to\\n\\f 15\\nensure that limited DSH resources are allocated to hospitals\\nthat have a net financial shortfall in serving Medicaid patients.”\\n82 Fed. Reg. at 16,117.\\n\\n Their argument is doubly flawed. For starters, “Medicaid\\ninpatient utilization rates” are not mentioned in § 1396r-\\n4(g)(1)(A). See 42 U.S.C. § 1396r-4(g)(1)(A). More\\nimportantly, the plaintiffs misstate which hospitals suffer a\\n“net financial shortfall.” Programs and services a hospital\\nprovides that are not paid for by Medicaid are not relevant to\\nthe shortfall calculation. See 42 U.S.C. § 1396r-4(j)(2)(C)\\n(“Only the uncompensated care costs of providing inpatient\\nhospital and outpatient hospital services to individuals\\ndescribed in paragraph [(g)(1)(A)] are included in the\\ncalculation of the hospital-specific limits under such\\nsubsection.”). Indeed, the statute does not consider a hospital’s\\nactual costs; it considers only those costs that Medicaid pays\\nfor. See 82 Fed. Reg. at 16,118 (“Ancillary programs and\\nservices that hospitals provide to patients may be laudable, but\\nthey are not paid for by Medicaid because they are not costs\\nassociated with furnishing inpatient and outpatient hospital\\nservices.”). Calculating “net financial shortfall” using only\\nthose costs that Medicaid pays for, no hospital that suffers a\\n“net financial shortfall” will be denied a DSH payment. Thus,\\nwe disagree with the plaintiffs’ argument that the Secretary has\\nfailed to tie the Rule to the record. Like their § 706(2)(C)\\nchallenge, their § 706(2)(A) challenge fails.\\n\\n For the foregoing reasons, we reverse the judgment of the\\ndistrict court, reinstate the 2017 Rule and remand the case for\\nfurther proceedings consistent with this opinion.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4425037/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge HENDERSON\"}]}"},"judge1":{"kind":"string","value":"HENDERSON"},"judge2":{"kind":"string","value":"ROGERS"},"judge3":{"kind":"string","value":"SENTELLE"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4647784/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: CHILDREN'S HOSPITAL ASSOCIATION OF TEXAS, Et Al., Appellees v. Alex Michael AZAR, II, in His Official Capacity, Secretary, Department of Health and Human Services, Et Al., Appellants\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-13\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued April 9, 2019 Decided August 13, 2019\n\n No. 18-5135\n\n CHILDREN’S HOSPITAL ASSOCIATION OF TEXAS, ET AL.,\n APPELLEES\n\n v.\n\n ALEX MICHAEL AZAR, II, IN HIS OFFICIAL CAPACITY,\n SECRETARY,\n DEPARTMENT OF HEALTH AND HUMAN SERVICES, ET AL.,\n APPELLANTS\n\n\n Appeal from the United States District Court\n for the District of Columbia\n (No. 1:17-cv-00844)\n\n\n Tara S. Morrissey, Attorney, United States Department of\nJustice, argued the cause for the appellants. Mark B. Stern and\nSamantha L. Chaifetz, Attorneys, Robert P. Charrow, General\nCounsel, United States Department of Health & Human\nServices, Janice L. Hoffman, Associate General Counsel,\nSusan M. Lyons, Deputy Associate General Counsel, and\nDavid L. Hoskins, Attorney, were with her on brief.\n\n David B. Salmons argued the cause for the appellees.\nGeraldine E. Edens, Michael E. Kenneally, and Susan Feigin\nHarris were with him on brief. Christopher H. Marraro\nentered an appearance.\n\f 2\n\n Daniel G. Jarcho and Michael H. Park were on brief for\nthe amicus curiae Children’s Hospital Association in support\nof the appellees and in support of affirmance.\n\n Before: HENDERSON and ROGERS, Circuit Judges, and\nSENTELLE, Senior Circuit Judge.\n\n Opinion for the Court filed by Circuit Judge HENDERSON.\n\n KAREN LECRAFT HENDERSON, Circuit Judge: Under the\nMedicaid Act (Act), the federal government provides each\nstate funds for distribution to hospitals that serve a\ndisproportionate number of low-income patients. See 42\nU.S.C. § 1396-1. A state distributes the funds through\nDisproportionate Share Hospital (DSH) payments. See id.\n§ 1396r-4(b), (c). A hospital may not receive a DSH payment\nthat exceeds its “costs incurred” in furnishing hospital services\nto low-income patients. Id. § 1396r-4(g)(1)(A). “Costs\nincurred” are, inter alia, “determined by the Secretary” of the\nUnited States Department of Health and Human Services\n(Secretary). Id. In 2017, the Secretary promulgated a regulation\ndefining “costs incurred.” Medicaid Program;\nDisproportionate Share Hospital Payments–Treatment of Third\nParty Payers in Calculating Uncompensated Care Costs, 82\nFed. Reg. 16,114, 16,122 (Apr. 3, 2017) (“2017 Rule”). The\nplaintiffs, a group of children’s hospitals that receive DSH\npayments, argue that the regulatory definition is contrary to the\nMedicaid Act and otherwise arbitrary and capricious. The\ndistrict court agreed that the definition is inconsistent with the\nAct and vacated the 2017 Rule. Children’s Hosp. Ass’n of Tex.\nv. Azar, 300 F. Supp. 3d 190 (D.D.C. 2018). We now reverse.\n\f 3\n I. Background\n\n “Medicaid is a cooperative federal-state program through\nwhich the Federal Government provides financial assistance to\nStates so that they may furnish medical care to needy\nindividuals.” Wilder v. Va. Hosp. Ass’n, 496 U.S. 498, 502\n(1990). States implement their own Medicaid plans, subject to\nthe federal government’s review and approval. See 42 U.S.C.\n§ 1396a. Treating the indigent proves costly even for hospitals\nthat receive Medicaid payments. Indeed, not all hospital\nservices are covered by Medicaid; not all costs associated with\ncovered services are allowed by Medicaid; and Medicaid does\nnot fully reimburse hospitals for all allowable costs associated\nwith covered services. Recognizing this, the Congress\nauthorizes supplemental payments (“DSH payments”) to\nhospitals that serve a disproportionate share of low-income\npatients (“DSH hospitals”). 42 U.S.C. § 1396a(13)(A)(iv)\n(requiring that Medicaid payment rates “take into account (in a\nmanner consistent with section 1396r-4 of this title) the\nsituation of hospitals which serve a disproportionate number of\nlow-income patients with special needs”); 42 U.S.C. § 1396r-\n4 (entitled “Adjustment in payment for inpatient hospital\nservices furnished by disproportionate share hospitals”). There\nis both a state-specific and a hospital-specific limit on DSH\npayments. The state-specific limit—not at issue in this case—\ndictates that all DSH payments to DSH hospitals within a\nsingle state must be drawn from the same pool of federal funds.\nSee 42 U.S.C. § 1396r-4(f). The hospital-specific limit, which\nis at issue in this case, dictates that a DSH payment to a single\nhospital cannot exceed:\n\n [T]he costs incurred during the year of\n furnishing hospital services (as determined by\n the Secretary and net of payments under this\n subchapter, other than under this section, and by\n\f 4\n uninsured patients) by the hospital to\n individuals who either are eligible for medical\n assistance under the State plan or have no health\n insurance (or other source of third party\n coverage) for services provided during the year.\n\n42 U.S.C. § 1396r-4(g)(1)(A). This sentence—although not the\npicture of clarity—establishes a few matters clearly. A DSH\nhospital cannot receive a DSH payment that exceeds its “costs\nincurred during the year of furnishing hospital services” to\nMedicaid-eligible and uninsured individuals. The Secretary is\nassigned the task of determining “costs incurred.” And “costs\nincurred” are “net of payments under this subchapter, other\nthan under this section, and by uninsured patients”; in other\nwords, payments made by Medicaid and uninsured individuals\nmust be subtracted out when calculating a hospital’s “costs\nincurred.” The dispute here is about whether payments made\nby Medicare and private insurers should also be subtracted out.\n\n In 2003, the Congress enacted legislation requiring states\nto submit annual reports and independent certified audits\nregarding their DSH programs. See 42 U.S.C. § 1396r-4(j). The\nreports must identify which hospitals receive DSH payments\nand the audits must verify that the DSH payments comply with\nthe statutory requirements. Id.\n\n In 2008, the Centers for Medicare & Medicaid Services\n(CMS), using the authority delegated it by the Secretary,\npromulgated a regulation implementing the reporting and\nauditing requirements. Medicaid Program; Disproportionate\nShare Hospital Payments, 73 Fed. Reg. 77,904 (Dec. 19, 2008)\n(“2008 Rule”). The 2008 Rule provided that each state must\nreport to CMS the cost of each DSH hospital’s “Total Medicaid\nUncompensated Care.” Id. at 77,950 (codified at 42 C.F.R.\n§ 447.299(c)(11)). The 2008 Rule did not state whether third-\nparty payments, including payments by Medicare and private\n\f 5\ninsurers, were meant to be included in calculating the amount.\nSee id. Three courts of appeals concluded from this silence that\nthe 2008 Rule left uncertain whether these payments should be\nconsidered. See Children’s Health Care v. Ctrs. for Medicare\n& Medicaid Servs., 900 F.3d 1022, 1025 (8th Cir. 2018);\nChildren’s Hosp. of the King’s Daughters, Inc. v. Azar, 896\nF.3d 615, 621 (4th Cir. 2018); N.H. Hosp. Ass’n v. Azar, 887\nF.3d 62, 75 (1st Cir. 2018). One court of appeals concluded that\nthe 2008 Rule made clear that these payments should not be\nconsidered. See Tenn. Hosp. Ass’n v. Azar, 908 F.3d 1029,\n1043–44 (6th Cir. 2018).\n\n In 2010, CMS posted a Frequently Asked Questions\ndocument on its website clarifying that payments made by\nMedicare and private insurers should be included. See CMS,\nAdditional Information on the DSH Reporting and Audit\nRequirements, FAQs 33 and 34 (2010),\nhttps://www.medicaid.gov/medicaid/finance/downloads/part-\n1-additional-info-on-dsh-reporting-and-auditing.pdf. A\nnumber of hospitals brought suit, arguing that the FAQs\nposting was invalid because it represented a substantive policy\nchange without notice and an opportunity for public comment.\nIn response, CMS issued a notice of proposed rulemaking and\nsubsequently promulgated the 2017 Rule. The 2017 Rule\nestablishes that payments by Medicare and private insurers are\nto be included in calculating a hospital’s “costs incurred.” 82\nFed. Reg. at 16,122 (codified at 42 C.F.R. § 447.299(c)(10)).\nIt provides, inter alia, “costs . . . [a]re defined as costs net of\nthird-party payments, including, but not limited to, payments\nby Medicare and private insurance.” Id. The Secretary explains\nthat considering payments by Medicare and private insurers\n“best fulfills the purpose of the DSH statute,” is “necessary to\nensure that limited DSH resources are allocated to hospitals\nthat have a net financial shortfall in serving Medicaid patients”\nand “is necessary to facilitate the Congressional directive . . .\n\f 6\nof limiting the DSH payment to a hospital’s uncompensated\ncare costs.” Id. at 16,116, 16,118. He maintains that the 2017\nRule did not effect a legal change but instead continued the\npreexisting policy. Id. at 16,119.\n\n The plaintiffs are four children’s hospitals in Minnesota,\nVirginia and Washington and an association representing eight\nchildren’s hospitals in Texas. They claim the 2017 Rule\nviolates the Administrative Procedure Act because it exceeds\nthe Secretary’s authority under the Medicaid Act and is the\nproduct of arbitrary and capricious reasoning. See 5 U.S.C.\n§ 706(2)(A), (C). The district court entered summary judgment\nfor the plaintiffs, holding that the Rule “is inconsistent with the\nplain language of the Medicaid Act,” which “clearly indicates\nwhich payments can be subtracted from the total costs incurred\nduring the year by hospitals” and “nowhere mentions\nsubtracting other third-party payments made on behalf of\nMedicaid-eligible patients from the total costs incurred.”\nChildren’s Hosp. Ass’n of Tex., 300 F. Supp. 3d at 205, 207.\nHaving held the 2017 Rule invalid under § 706(2)(C) (“ultra\nvires” prohibition), the district court did not reach the\nplaintiffs’ § 706(2)(A) challenge (“arbitrary and capricious”\nprohibition). Id. at 205. The district court ultimately vacated\nthe 2017 Rule and the Secretary timely appealed. Id. at 210–\n11. Our review is de novo. See Ark Initiative v. Tidwell, 816\nF.3d 119, 126–27 (D.C. Cir. 2016).\n\n II. Analysis\n\n A. Exceeds Statutory Authority\n\n The plaintiffs first challenge the Rule as exceeding the\nSecretary’s authority under the Medicaid Act, in violation of 5\nU.S.C. § 706(2)(C). The familiar Chevron framework guides\nour review. See Ass’n of Private Sector Colls. & Univs. v.\nDuncan, 681 F.3d 427, 441 (D.C. Cir. 2012) (“Appellant’s\n\f 7\nclaims that various provisions of the challenged regulations are\n‘in excess of statutory jurisdiction, authority, or limitations, or\nshort of statutory right,’ are reviewed under the well-known\nChevron framework.” (citation omitted) (quoting 5 U.S.C.\n§ 706(2)(C))). “Under that framework, we ask whether the\nstatute is ambiguous and, if so, whether the agency’s\ninterpretation is reasonable.” King v. Burwell, 135 S. Ct. 2480,\n2488 (2015). “This approach ‘is premised on the theory that a\nstatute’s ambiguity constitutes an implicit delegation from\nCongress to the agency to fill in the statutory gaps.’” Id.\n(quoting FDA v. Brown & Williamson Tobacco Corp., 529\nU.S. 120, 159 (2000)). Because the delegation at issue here is\nexpress rather than implied, see 42 U.S.C. § 1396r-4(g)(1)(A)\n(“[T]he costs incurred during the year of furnishing hospital\nservices (as determined by the Secretary . . .) . . . .” (emphasis\nadded)); see also Transitional Hosps. Corp. of La. v. Shalala,\n222 F.3d 1019, 1025 (D.C. Cir. 2000) (“as determined by the\nSecretary” is “express delegation”), we have no need to search\nfor statutory ambiguity. We skip straight to asking whether the\nRule is reasonable. See Chevron U.S.A. Inc. v. Nat. Res. Def.\nCouncil, Inc., 467 U.S. 837, 843–44 (1984) (“If Congress has\nexplicitly left a gap for the agency to fill, there is an express\ndelegation of authority to the agency to elucidate a specific\nprovision of the statute by regulation. Such legislative\nregulations are given controlling weight unless they are\narbitrary, capricious, or manifestly contrary to the statute.”);\nsee also Transitional Hosps. Corp. of La., 222 F.3d at 1025 (if\ndelegation is express “we are bound to uphold [the Secretary’s]\ndetermination as long as she exercises [her] discretion in a\nreasonable way”).\n\n The plaintiffs offer four principal reasons the statute does\nnot grant the Secretary authority to require that payments by\nMedicare and private insurers be considered in calculating a\nhospital’s “costs incurred.” First, the statute exclusively\n\f 8\nspecifies which payments can be considered. Second, the Rule\nrenders superfluous the statute’s specification that certain\npayments must be considered. Third, the Congress required\nconsideration of third-party payments in a different statutory\nprovision but not in the relevant provision. Fourth, the statute\nplainly distinguishes costs and payments. We reject all four\narguments.1\n\n First, we disagree with the plaintiffs’ argument that the\nstatute exclusively specifies which payments can be considered\nin calculating “costs incurred.” See Plaintiffs’ Br. at 58;\nChildren’s Hosp. Ass’n of Tex., 300 F. Supp. 3d at 207 (“On its\nface, the statute clearly indicates which payments can be\nsubtracted from the total costs incurred during the year by\nhospitals: (1) ‘payments under this subchapter,’ i.e., payments\nmade by Medicaid; and (2) payments made by uninsured\npatients. The statute nowhere mentions subtracting other third-\nparty payments made on behalf of Medicaid-eligible patients\nfrom the total costs incurred.”). Although the statute\nestablishes that payments by Medicaid and the uninsured must\nbe considered, it nowhere states that those are the only\npayments that may be considered. To conclude otherwise, we\nwould have to rely on the interpretive canon expressio unius\nest exclusio alterius, which means “expressing one item of [an]\nassociated group or series excludes another left unmentioned.”\nChevron U.S.A. Inc. v. Echazabal, 536 U.S. 73, 80 (2002)\n(alteration in original) (quoting United States v. Vonn, 535 U.S.\n55, 65 (2002)). But that canon has been called a “feeble helper\nin an administrative setting.” Adirondack Med. Ctr. v. Sebelius,\n740 F.3d 692, 697 (D.C. Cir. 2014) (quoting Cheney R.R. Co.\nv. ICC, 902 F.2d 66, 69 (1990)). And, in any setting, it “applies\nonly when ‘circumstances support[] a sensible inference that\n\n 1\n We have also considered and reject the plaintiffs’ other\narguments.\n\f 9\nthe term left out must have been meant to be excluded.’” NLRB\nv. SW Gen., Inc., 137 S. Ct. 929, 940 (2017) (quoting\nEchazabal, 536 U.S. at 81); see also Barnhart v. Peabody Coal\nCo., 537 U.S. 149, 168 (2003) (“[W]e do not read the\nenumeration of one . . . to exclude the other unless it is fair to\nsuppose that Congress considered the unnamed possibility and\nmeant to say no to it.”). There is reason to believe the Congress\ndid not intend to exclude Medicare and private insurance\npayments from consideration. Indeed, the parties agree that the\nmost common sources of payment for treating Medicaid-\neligible and uninsured individuals are Medicaid and the\nuninsured. The Congress may have wanted to ensure that the\nmost common sources of payment must be considered but at\nthe same time allow the Secretary to decide whether less-\ncommon sources of payment should be as well. Especially in\nlight of this plausible alternative explanation, we will not rely\non the expressio unius canon to find that the statute exclusively\nspecifies which payments can be considered in calculating\n“costs incurred.” See Indep. Ins. Agents of Am., Inc. v. Hawke,\n211 F.3d 638, 644 (D.C. Cir. 2000) (“[I]f there are other\nreasonable explanations for an omission in a statute, expressio\nunius may not be a useful tool.”).\n\n Second, we disagree with the plaintiffs’ argument that the\nRule renders superfluous the statute’s specification that\npayments by Medicaid and the uninsured must be considered.\nSee Plaintiffs’ Br. 41–42; Children’s Hosp. Ass’n of Tex., 300\nF. Supp. 3d at 207 (“To allow the Secretary to redefine ‘costs’\nto net out a third category of payments—i.e., ‘third-party\npayments, including but not limited to, payments by Medicare\nand private insurance’ . . .—would ‘render the Congressional\ndefinition of payments in the very same clause superfluous.’”\n(quoting Children’s Hosp. of the King’s Daughters, Inc. v.\nPrice, 258 F. Supp. 3d 672, 687 (E.D. Va. 2017))); id.\n(“[D]efendants’ interpretation of the statute would render\n\f 10\nportions of the statutory language superfluous.”). The statute’s\nspecification that two forms of payment must be considered\nremoves the Secretary’s discretion as to those two forms of\npayment. But it does nothing to disturb the Secretary’s\ndiscretion as to other forms of payment, which may be\nconsidered. See Tenn. Hosp. Ass’n, 908 F.3d at 1038 (“[T]he\nfact that certain payments must be deducted from costs does\nnot mean that other payments cannot be.”); N.H. Hosp. Ass’n,\n887 F.3d at 66 (“Congress identified two specific sources of\npayment that must be offset against total costs, but otherwise\nsimply stated that ‘costs incurred’ are ‘as determined by the\nSecretary.’”).\n\n Third, we reject the plaintiffs’ argument that we should\ninfer from the Congress’s requiring consideration of third party\npayments under 42 U.S.C. § 1396r-4(g)(2)(A)—a provision\nthat, before 1995, allowed certain hospitals to receive\npayments that exceeded their uncompensated costs—that it\nmeant to prohibit consideration of third party payments under\n§ 1396r-4(g)(1)(A). Plaintiffs’ Br. 34–35. This argument is\nbased on the so-called “Russello presumption—that the\npresence of a phrase in one provision and its absence in another\nreveals Congress’ design.” City of Columbus v. Ours Garage\n& Wrecker Serv., Inc., 536 U.S. 424, 435–36 (2002) (citing\nRussello v. United States, 464 U.S. 16 (1983)). But that\npresumption “grows weaker with each difference in the\nformulation of the provisions under inspection.” Id. at 436.\nBecause sections (g)(1)(A) and (g)(2)(A) are fundamentally\ndifferent, we find the plaintiffs’ argument unpersuasive. See\nTenn. Hosp. Ass’n, 908 F.3d at 1039 (“There is no tension,\nhowever, in Congress requiring third-party payment\ndeductions in subsection (g)(2)(A) and allowing third-party\npayment deductions in subsection (g)(1)(A). The DSH\npayments provided for in (g)(2)(A) are above and beyond those\nmandated by (g)(1)(A); it therefore makes sense for Congress\n\f 11\nto impose a hard limit on the ceiling of (g)(2)(A) funds—i.e.,\nno more than 200% of the costs of serving Medicaid-eligible\npatients, less payments from Medicaid, uninsured patients, and\n‘third party payors’—while giving CMS more discretion to\ncalibrate the appropriate cap on the ‘standard’ DSH payments\ndiscussed in (g)(1)(A).”).\n\n Fourth, we disagree with the plaintiffs’ argument that the\nstatute plainly distinguishes between costs and payments such\nthat payments can never be considered in calculating “costs\nincurred.” See Plaintiffs’ Br. at 26, 30, 33, 40, 57. The statute\nestablishes that a hospital’s DSH payment cannot exceed its\n“costs incurred during the year of furnishing hospital services\n(as determined by the Secretary and net of payments under this\nsubchapter, other than under this section, and by uninsured\npatients).” 42 U.S.C. § 1396r-4(g)(1)(A). Both of the\nparenthetical’s adjoining phrases modify “costs incurred”—\nthat is, “costs incurred” are both “as determined by the\nSecretary” and “net of payments under [Medicaid] and by\nuninsured patients.” Id. In other words, the statute requires that\nsome payments be considered in calculating a hospital’s “costs\nincurred.” The argument that the statute separates costs and\npayments therefore flies in the teeth of the statutory text.\n\n Contrary to the plaintiffs’ contention, we believe the 2017\nRule is consistent with the statute’s context and purpose, both\nof which suggest DSH payments are meant to assist those\nhospitals that need them most by covering only those costs for\nwhich DSH hospitals are in fact uncompensated. See 42 U.S.C.\n§ 1396r-4(g)(1) (heading of provision at issue: “Amount of\nadjustment subject to uncompensated costs”); 2 42 U.S.C.\n§ 1396r-4(j)(2)(C) (requiring states to certify that “[o]nly the\n\n 2\n Headings, although “not commanding,” “supply clues” about\nCongressional intent. Yates v. United States, 135 S. Ct. 1074, 1083\n(2015).\n\f 12\nuncompensated care costs of providing inpatient hospital and\noutpatient hospital services to individuals described in\n[(g)(1)(A)] are included in the calculation of the hospital-\nspecific limits.”); H.R. Rep. No. 103-111, at 211 (in enacting\nDSH payment limit, Congress noted “some States have made\nDSH payment[s] . . . to State psychiatric or university hospitals\nin amounts that exceed the net costs, and in some instances the\ntotal costs, of operating the facilities”); Tenn. Hosp. Ass’n, 908\nF.3d at 1040 (“essence” of Congressional concern in enacting\nstatute was “that hospitals were double dipping by collecting\nDSH payments to cover costs that had already been\nreimbursed”). By requiring the inclusion of payments by\nMedicare and private insurers, the 2017 Rule ensures that DSH\npayments will go to hospitals that have been compensated least\nand are thus most in need. Because the 2017 Rule is consistent\nwith the statute, it does not violate § 706(2)(C).\n\n B. Arbitrary and Capricious\n\n The plaintiffs next challenge the 2017 Rule as the product\nof arbitrary and capricious reasoning, in violation of 5 U.S.C.\n§ 706(2)(A). A reviewing court “shall . . . hold unlawful and\nset aside agency action, findings, and conclusions found to be\n. . . arbitrary, capricious, an abuse of discretion, or otherwise\nnot in accordance with law.” 5 U.S.C. § 706(2)(A). An agency\nrulemaking is arbitrary and capricious “if the agency has relied\non factors which Congress has not intended it to consider,\nentirely failed to consider an important aspect of the problem,\noffered an explanation for its decision that runs counter to the\nevidence before the agency, or is so implausible that it could\nnot be ascribed to a difference in view or the product of agency\nexpertise.” Motor Vehicles Mfrs. Ass’n of U.S. v. State Farm\nMut. Auto. Ins. Co., 463 U.S. 29, 43 (1983).\n\n The plaintiffs first contend that “CMS has never\nacknowledged, let alone justified, its new Rule’s departure\n\f 13\nfrom the 2008 rule.” Plaintiffs’ Br. 27–28. We disagree.\n“Agencies are free to change their existing policies as long as\nthey provide a reasoned explanation for the change.” Encino\nMotorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125 (2016). An\nagency need not provide a more detailed justification for a\nchanged policy than it would for a brand-new policy. Id. But it\nmust provide “a reasoned explanation . . . for disregarding facts\nand circumstances that underlay or were engendered by the\nprior policy.” Id. at 2126 (quoting FCC v. Fox Television\nStations, Inc., 556 U.S. 502, 516 (2009)). An “unexplained\ninconsistency” with an earlier position renders a changed\npolicy arbitrary and capricious. Id. (quoting Nat. Cable &\nTelecomm. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 981\n(2005)).\n\n There is no unexplained inconsistency with an earlier\nposition here. To be clear, we agree with the plaintiffs that the\n2017 Rule and the 2008 Rule establish different policies.3 But\nit makes no difference. CMS explained why the statute’s\npurposes are better fulfilled by a policy that requires\nconsideration of payments by Medicare and private insurers\n\n 3\n The Secretary maintains that the 2017 Rule is consistent with\nthe 2008 Rule and so does not establish a new policy. That argument\nhas been rejected by four courts of appeals, all of which found the\n2010 FAQs procedurally invalid because the policy established\ntherein, which is the same policy established by the 2017 Rule,\nmarked a departure from the policy established by the 2008 Rule\nwithout notice and an opportunity for public comment. See Tenn.\nHosp. Ass’n, 908 F.3d at 1043 (“As three circuit courts and several\ndistrict courts have now held, the payment-deduction policy\nelucidated in the FAQs and hinted at in the preamble to the 2008 rule\nseeks to amend, rather than merely clarify, the 2008 regulations.”\n(citing Children’s Health Care, 900 F.3d at 1026–27; Children’s\nHosp. of the King’s Daughters, 896 F.3d at 623; N.H. Hosp. Ass’n,\n887 F.3d at 74)). We agree with our sister circuits.\n\f 14\n(the 2017 Rule) than one that does not (the 2008 Rule, as we\ninterpret it):\n\n In light of the statutory requirement limiting\n DSH payments on a hospital-specific basis to\n uncompensated care costs, it is inconsistent\n with the statute to assist hospitals with costs that\n have already been compensated by third party\n payments. [The 2017] rule is designed to\n reiterate the policy and make explicit within the\n terms of the regulation that all costs and\n payments associated with dual eligible and\n individuals with a source of third party\n coverage must be included in calculating the\n hospital-specific DSH limit. This policy is\n necessary to ensure that only actual\n uncompensated care costs are included in the\n Medicaid hospital-specific DSH limit. And,\n because state DSH payments are limited to an\n annual federal allotment, this policy is also\n necessary to ensure that limited DSH resources\n are allocated to hospitals that have a net\n financial shortfall in serving Medicaid patients.\n\n82 Fed. Reg. at 16,117. This explanation is more than sufficient\nto survive review under § 706(2)(A).\n\n The plaintiffs also claim that the Secretary has not tied the\n2017 Rule to the administrative record. According to their\nreading, the record shows that CMS reduces DSH payments to\nthe plaintiff hospitals when it considers private insurance\npayments, notwithstanding “they have among the highest\nMedicaid inpatient utilization rates in their respective states\nand the highest net financial shortfalls in serving Medicaid\npatients.” Plaintiffs Br. 65. The plaintiffs claim this outcome is\ninconsistent with the purpose of the 2017 Rule, which is “to\n\f 15\nensure that limited DSH resources are allocated to hospitals\nthat have a net financial shortfall in serving Medicaid patients.”\n82 Fed. Reg. at 16,117.\n\n Their argument is doubly flawed. For starters, “Medicaid\ninpatient utilization rates” are not mentioned in § 1396r-\n4(g)(1)(A). See 42 U.S.C. § 1396r-4(g)(1)(A). More\nimportantly, the plaintiffs misstate which hospitals suffer a\n“net financial shortfall.” Programs and services a hospital\nprovides that are not paid for by Medicaid are not relevant to\nthe shortfall calculation. See 42 U.S.C. § 1396r-4(j)(2)(C)\n(“Only the uncompensated care costs of providing inpatient\nhospital and outpatient hospital services to individuals\ndescribed in paragraph [(g)(1)(A)] are included in the\ncalculation of the hospital-specific limits under such\nsubsection.”). Indeed, the statute does not consider a hospital’s\nactual costs; it considers only those costs that Medicaid pays\nfor. See 82 Fed. Reg. at 16,118 (“Ancillary programs and\nservices that hospitals provide to patients may be laudable, but\nthey are not paid for by Medicaid because they are not costs\nassociated with furnishing inpatient and outpatient hospital\nservices.”). Calculating “net financial shortfall” using only\nthose costs that Medicaid pays for, no hospital that suffers a\n“net financial shortfall” will be denied a DSH payment. Thus,\nwe disagree with the plaintiffs’ argument that the Secretary has\nfailed to tie the Rule to the record. Like their § 706(2)(C)\nchallenge, their § 706(2)(A) challenge fails.\n\n For the foregoing reasons, we reverse the judgment of the\ndistrict court, reinstate the 2017 Rule and remand the case for\nfurther proceedings consistent with this opinion.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: CHILDREN'S HOSPITAL ASSOCIATION OF TEXAS, Et Al., Appellees v. Alex Michael AZAR, II, in His Official Capacity, Secretary, Department of Health and Human Services, Et Al., Appellants\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-13\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 9, 2019 Decided August 13, 2019\\n\\n No. 18-5135\\n\\n CHILDREN’S HOSPITAL ASSOCIATION OF TEXAS, ET AL.,\\n APPELLEES\\n\\n v.\\n\\n ALEX MICHAEL AZAR, II, IN HIS OFFICIAL CAPACITY,\\n SECRETARY,\\n DEPARTMENT OF HEALTH AND HUMAN SERVICES, ET AL.,\\n APPELLANTS\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:17-cv-00844)\\n\\n\\n Tara S. Morrissey, Attorney, United States Department of\\nJustice, argued the cause for the appellants. Mark B. Stern and\\nSamantha L. Chaifetz, Attorneys, Robert P. Charrow, General\\nCounsel, United States Department of Health & Human\\nServices, Janice L. Hoffman, Associate General Counsel,\\nSusan M. Lyons, Deputy Associate General Counsel, and\\nDavid L. Hoskins, Attorney, were with her on brief.\\n\\n David B. Salmons argued the cause for the appellees.\\nGeraldine E. Edens, Michael E. Kenneally, and Susan Feigin\\nHarris were with him on brief. Christopher H. Marraro\\nentered an appearance.\\n\\f 2\\n\\n Daniel G. Jarcho and Michael H. Park were on brief for\\nthe amicus curiae Children’s Hospital Association in support\\nof the appellees and in support of affirmance.\\n\\n Before: HENDERSON and ROGERS, Circuit Judges, and\\nSENTELLE, Senior Circuit Judge.\\n\\n Opinion for the Court filed by Circuit Judge HENDERSON.\\n\\n KAREN LECRAFT HENDERSON, Circuit Judge: Under the\\nMedicaid Act (Act), the federal government provides each\\nstate funds for distribution to hospitals that serve a\\ndisproportionate number of low-income patients. See 42\\nU.S.C. § 1396-1. A state distributes the funds through\\nDisproportionate Share Hospital (DSH) payments. See id.\\n§ 1396r-4(b), (c). A hospital may not receive a DSH payment\\nthat exceeds its “costs incurred” in furnishing hospital services\\nto low-income patients. Id. § 1396r-4(g)(1)(A). “Costs\\nincurred” are, inter alia, “determined by the Secretary” of the\\nUnited States Department of Health and Human Services\\n(Secretary). Id. In 2017, the Secretary promulgated a regulation\\ndefining “costs incurred.” Medicaid Program;\\nDisproportionate Share Hospital Payments–Treatment of Third\\nParty Payers in Calculating Uncompensated Care Costs, 82\\nFed. Reg. 16,114, 16,122 (Apr. 3, 2017) (“2017 Rule”). The\\nplaintiffs, a group of children’s hospitals that receive DSH\\npayments, argue that the regulatory definition is contrary to the\\nMedicaid Act and otherwise arbitrary and capricious. The\\ndistrict court agreed that the definition is inconsistent with the\\nAct and vacated the 2017 Rule. Children’s Hosp. Ass’n of Tex.\\nv. Azar, 300 F. Supp. 3d 190 (D.D.C. 2018). We now reverse.\\n\\f 3\\n I. Background\\n\\n “Medicaid is a cooperative federal-state program through\\nwhich the Federal Government provides financial assistance to\\nStates so that they may furnish medical care to needy\\nindividuals.” Wilder v. Va. Hosp. Ass’n, 496 U.S. 498, 502\\n(1990). States implement their own Medicaid plans, subject to\\nthe federal government’s review and approval. See 42 U.S.C.\\n§ 1396a. Treating the indigent proves costly even for hospitals\\nthat receive Medicaid payments. Indeed, not all hospital\\nservices are covered by Medicaid; not all costs associated with\\ncovered services are allowed by Medicaid; and Medicaid does\\nnot fully reimburse hospitals for all allowable costs associated\\nwith covered services. Recognizing this, the Congress\\nauthorizes supplemental payments (“DSH payments”) to\\nhospitals that serve a disproportionate share of low-income\\npatients (“DSH hospitals”). 42 U.S.C. § 1396a(13)(A)(iv)\\n(requiring that Medicaid payment rates “take into account (in a\\nmanner consistent with section 1396r-4 of this title) the\\nsituation of hospitals which serve a disproportionate number of\\nlow-income patients with special needs”); 42 U.S.C. § 1396r-\\n4 (entitled “Adjustment in payment for inpatient hospital\\nservices furnished by disproportionate share hospitals”). There\\nis both a state-specific and a hospital-specific limit on DSH\\npayments. The state-specific limit—not at issue in this case—\\ndictates that all DSH payments to DSH hospitals within a\\nsingle state must be drawn from the same pool of federal funds.\\nSee 42 U.S.C. § 1396r-4(f). The hospital-specific limit, which\\nis at issue in this case, dictates that a DSH payment to a single\\nhospital cannot exceed:\\n\\n [T]he costs incurred during the year of\\n furnishing hospital services (as determined by\\n the Secretary and net of payments under this\\n subchapter, other than under this section, and by\\n\\f 4\\n uninsured patients) by the hospital to\\n individuals who either are eligible for medical\\n assistance under the State plan or have no health\\n insurance (or other source of third party\\n coverage) for services provided during the year.\\n\\n42 U.S.C. § 1396r-4(g)(1)(A). This sentence—although not the\\npicture of clarity—establishes a few matters clearly. A DSH\\nhospital cannot receive a DSH payment that exceeds its “costs\\nincurred during the year of furnishing hospital services” to\\nMedicaid-eligible and uninsured individuals. The Secretary is\\nassigned the task of determining “costs incurred.” And “costs\\nincurred” are “net of payments under this subchapter, other\\nthan under this section, and by uninsured patients”; in other\\nwords, payments made by Medicaid and uninsured individuals\\nmust be subtracted out when calculating a hospital’s “costs\\nincurred.” The dispute here is about whether payments made\\nby Medicare and private insurers should also be subtracted out.\\n\\n In 2003, the Congress enacted legislation requiring states\\nto submit annual reports and independent certified audits\\nregarding their DSH programs. See 42 U.S.C. § 1396r-4(j). The\\nreports must identify which hospitals receive DSH payments\\nand the audits must verify that the DSH payments comply with\\nthe statutory requirements. Id.\\n\\n In 2008, the Centers for Medicare & Medicaid Services\\n(CMS), using the authority delegated it by the Secretary,\\npromulgated a regulation implementing the reporting and\\nauditing requirements. Medicaid Program; Disproportionate\\nShare Hospital Payments, 73 Fed. Reg. 77,904 (Dec. 19, 2008)\\n(“2008 Rule”). The 2008 Rule provided that each state must\\nreport to CMS the cost of each DSH hospital’s “Total Medicaid\\nUncompensated Care.” Id. at 77,950 (codified at 42 C.F.R.\\n§ 447.299(c)(11)). The 2008 Rule did not state whether third-\\nparty payments, including payments by Medicare and private\\n\\f 5\\ninsurers, were meant to be included in calculating the amount.\\nSee id. Three courts of appeals concluded from this silence that\\nthe 2008 Rule left uncertain whether these payments should be\\nconsidered. See Children’s Health Care v. Ctrs. for Medicare\\n& Medicaid Servs., 900 F.3d 1022, 1025 (8th Cir. 2018);\\nChildren’s Hosp. of the King’s Daughters, Inc. v. Azar, 896\\nF.3d 615, 621 (4th Cir. 2018); N.H. Hosp. Ass’n v. Azar, 887\\nF.3d 62, 75 (1st Cir. 2018). One court of appeals concluded that\\nthe 2008 Rule made clear that these payments should not be\\nconsidered. See Tenn. Hosp. Ass’n v. Azar, 908 F.3d 1029,\\n1043–44 (6th Cir. 2018).\\n\\n In 2010, CMS posted a Frequently Asked Questions\\ndocument on its website clarifying that payments made by\\nMedicare and private insurers should be included. See CMS,\\nAdditional Information on the DSH Reporting and Audit\\nRequirements, FAQs 33 and 34 (2010),\\nhttps://www.medicaid.gov/medicaid/finance/downloads/part-\\n1-additional-info-on-dsh-reporting-and-auditing.pdf. A\\nnumber of hospitals brought suit, arguing that the FAQs\\nposting was invalid because it represented a substantive policy\\nchange without notice and an opportunity for public comment.\\nIn response, CMS issued a notice of proposed rulemaking and\\nsubsequently promulgated the 2017 Rule. The 2017 Rule\\nestablishes that payments by Medicare and private insurers are\\nto be included in calculating a hospital’s “costs incurred.” 82\\nFed. Reg. at 16,122 (codified at 42 C.F.R. § 447.299(c)(10)).\\nIt provides, inter alia, “costs . . . [a]re defined as costs net of\\nthird-party payments, including, but not limited to, payments\\nby Medicare and private insurance.” Id. The Secretary explains\\nthat considering payments by Medicare and private insurers\\n“best fulfills the purpose of the DSH statute,” is “necessary to\\nensure that limited DSH resources are allocated to hospitals\\nthat have a net financial shortfall in serving Medicaid patients”\\nand “is necessary to facilitate the Congressional directive . . .\\n\\f 6\\nof limiting the DSH payment to a hospital’s uncompensated\\ncare costs.” Id. at 16,116, 16,118. He maintains that the 2017\\nRule did not effect a legal change but instead continued the\\npreexisting policy. Id. at 16,119.\\n\\n The plaintiffs are four children’s hospitals in Minnesota,\\nVirginia and Washington and an association representing eight\\nchildren’s hospitals in Texas. They claim the 2017 Rule\\nviolates the Administrative Procedure Act because it exceeds\\nthe Secretary’s authority under the Medicaid Act and is the\\nproduct of arbitrary and capricious reasoning. See 5 U.S.C.\\n§ 706(2)(A), (C). The district court entered summary judgment\\nfor the plaintiffs, holding that the Rule “is inconsistent with the\\nplain language of the Medicaid Act,” which “clearly indicates\\nwhich payments can be subtracted from the total costs incurred\\nduring the year by hospitals” and “nowhere mentions\\nsubtracting other third-party payments made on behalf of\\nMedicaid-eligible patients from the total costs incurred.”\\nChildren’s Hosp. Ass’n of Tex., 300 F. Supp. 3d at 205, 207.\\nHaving held the 2017 Rule invalid under § 706(2)(C) (“ultra\\nvires” prohibition), the district court did not reach the\\nplaintiffs’ § 706(2)(A) challenge (“arbitrary and capricious”\\nprohibition). Id. at 205. The district court ultimately vacated\\nthe 2017 Rule and the Secretary timely appealed. Id. at 210–\\n11. Our review is de novo. See Ark Initiative v. Tidwell, 816\\nF.3d 119, 126–27 (D.C. Cir. 2016).\\n\\n II. Analysis\\n\\n A. Exceeds Statutory Authority\\n\\n The plaintiffs first challenge the Rule as exceeding the\\nSecretary’s authority under the Medicaid Act, in violation of 5\\nU.S.C. § 706(2)(C). The familiar Chevron framework guides\\nour review. See Ass’n of Private Sector Colls. & Univs. v.\\nDuncan, 681 F.3d 427, 441 (D.C. Cir. 2012) (“Appellant’s\\n\\f 7\\nclaims that various provisions of the challenged regulations are\\n‘in excess of statutory jurisdiction, authority, or limitations, or\\nshort of statutory right,’ are reviewed under the well-known\\nChevron framework.” (citation omitted) (quoting 5 U.S.C.\\n§ 706(2)(C))). “Under that framework, we ask whether the\\nstatute is ambiguous and, if so, whether the agency’s\\ninterpretation is reasonable.” King v. Burwell, 135 S. Ct. 2480,\\n2488 (2015). “This approach ‘is premised on the theory that a\\nstatute’s ambiguity constitutes an implicit delegation from\\nCongress to the agency to fill in the statutory gaps.’” Id.\\n(quoting FDA v. Brown & Williamson Tobacco Corp., 529\\nU.S. 120, 159 (2000)). Because the delegation at issue here is\\nexpress rather than implied, see 42 U.S.C. § 1396r-4(g)(1)(A)\\n(“[T]he costs incurred during the year of furnishing hospital\\nservices (as determined by the Secretary . . .) . . . .” (emphasis\\nadded)); see also Transitional Hosps. Corp. of La. v. Shalala,\\n222 F.3d 1019, 1025 (D.C. Cir. 2000) (“as determined by the\\nSecretary” is “express delegation”), we have no need to search\\nfor statutory ambiguity. We skip straight to asking whether the\\nRule is reasonable. See Chevron U.S.A. Inc. v. Nat. Res. Def.\\nCouncil, Inc., 467 U.S. 837, 843–44 (1984) (“If Congress has\\nexplicitly left a gap for the agency to fill, there is an express\\ndelegation of authority to the agency to elucidate a specific\\nprovision of the statute by regulation. Such legislative\\nregulations are given controlling weight unless they are\\narbitrary, capricious, or manifestly contrary to the statute.”);\\nsee also Transitional Hosps. Corp. of La., 222 F.3d at 1025 (if\\ndelegation is express “we are bound to uphold [the Secretary’s]\\ndetermination as long as she exercises [her] discretion in a\\nreasonable way”).\\n\\n The plaintiffs offer four principal reasons the statute does\\nnot grant the Secretary authority to require that payments by\\nMedicare and private insurers be considered in calculating a\\nhospital’s “costs incurred.” First, the statute exclusively\\n\\f 8\\nspecifies which payments can be considered. Second, the Rule\\nrenders superfluous the statute’s specification that certain\\npayments must be considered. Third, the Congress required\\nconsideration of third-party payments in a different statutory\\nprovision but not in the relevant provision. Fourth, the statute\\nplainly distinguishes costs and payments. We reject all four\\narguments.1\\n\\n First, we disagree with the plaintiffs’ argument that the\\nstatute exclusively specifies which payments can be considered\\nin calculating “costs incurred.” See Plaintiffs’ Br. at 58;\\nChildren’s Hosp. Ass’n of Tex., 300 F. Supp. 3d at 207 (“On its\\nface, the statute clearly indicates which payments can be\\nsubtracted from the total costs incurred during the year by\\nhospitals: (1) ‘payments under this subchapter,’ i.e., payments\\nmade by Medicaid; and (2) payments made by uninsured\\npatients. The statute nowhere mentions subtracting other third-\\nparty payments made on behalf of Medicaid-eligible patients\\nfrom the total costs incurred.”). Although the statute\\nestablishes that payments by Medicaid and the uninsured must\\nbe considered, it nowhere states that those are the only\\npayments that may be considered. To conclude otherwise, we\\nwould have to rely on the interpretive canon expressio unius\\nest exclusio alterius, which means “expressing one item of [an]\\nassociated group or series excludes another left unmentioned.”\\nChevron U.S.A. Inc. v. Echazabal, 536 U.S. 73, 80 (2002)\\n(alteration in original) (quoting United States v. Vonn, 535 U.S.\\n55, 65 (2002)). But that canon has been called a “feeble helper\\nin an administrative setting.” Adirondack Med. Ctr. v. Sebelius,\\n740 F.3d 692, 697 (D.C. Cir. 2014) (quoting Cheney R.R. Co.\\nv. ICC, 902 F.2d 66, 69 (1990)). And, in any setting, it “applies\\nonly when ‘circumstances support[] a sensible inference that\\n\\n 1\\n We have also considered and reject the plaintiffs’ other\\narguments.\\n\\f 9\\nthe term left out must have been meant to be excluded.’” NLRB\\nv. SW Gen., Inc., 137 S. Ct. 929, 940 (2017) (quoting\\nEchazabal, 536 U.S. at 81); see also Barnhart v. Peabody Coal\\nCo., 537 U.S. 149, 168 (2003) (“[W]e do not read the\\nenumeration of one . . . to exclude the other unless it is fair to\\nsuppose that Congress considered the unnamed possibility and\\nmeant to say no to it.”). There is reason to believe the Congress\\ndid not intend to exclude Medicare and private insurance\\npayments from consideration. Indeed, the parties agree that the\\nmost common sources of payment for treating Medicaid-\\neligible and uninsured individuals are Medicaid and the\\nuninsured. The Congress may have wanted to ensure that the\\nmost common sources of payment must be considered but at\\nthe same time allow the Secretary to decide whether less-\\ncommon sources of payment should be as well. Especially in\\nlight of this plausible alternative explanation, we will not rely\\non the expressio unius canon to find that the statute exclusively\\nspecifies which payments can be considered in calculating\\n“costs incurred.” See Indep. Ins. Agents of Am., Inc. v. Hawke,\\n211 F.3d 638, 644 (D.C. Cir. 2000) (“[I]f there are other\\nreasonable explanations for an omission in a statute, expressio\\nunius may not be a useful tool.”).\\n\\n Second, we disagree with the plaintiffs’ argument that the\\nRule renders superfluous the statute’s specification that\\npayments by Medicaid and the uninsured must be considered.\\nSee Plaintiffs’ Br. 41–42; Children’s Hosp. Ass’n of Tex., 300\\nF. Supp. 3d at 207 (“To allow the Secretary to redefine ‘costs’\\nto net out a third category of payments—i.e., ‘third-party\\npayments, including but not limited to, payments by Medicare\\nand private insurance’ . . .—would ‘render the Congressional\\ndefinition of payments in the very same clause superfluous.’”\\n(quoting Children’s Hosp. of the King’s Daughters, Inc. v.\\nPrice, 258 F. Supp. 3d 672, 687 (E.D. Va. 2017))); id.\\n(“[D]efendants’ interpretation of the statute would render\\n\\f 10\\nportions of the statutory language superfluous.”). The statute’s\\nspecification that two forms of payment must be considered\\nremoves the Secretary’s discretion as to those two forms of\\npayment. But it does nothing to disturb the Secretary’s\\ndiscretion as to other forms of payment, which may be\\nconsidered. See Tenn. Hosp. Ass’n, 908 F.3d at 1038 (“[T]he\\nfact that certain payments must be deducted from costs does\\nnot mean that other payments cannot be.”); N.H. Hosp. Ass’n,\\n887 F.3d at 66 (“Congress identified two specific sources of\\npayment that must be offset against total costs, but otherwise\\nsimply stated that ‘costs incurred’ are ‘as determined by the\\nSecretary.’”).\\n\\n Third, we reject the plaintiffs’ argument that we should\\ninfer from the Congress’s requiring consideration of third party\\npayments under 42 U.S.C. § 1396r-4(g)(2)(A)—a provision\\nthat, before 1995, allowed certain hospitals to receive\\npayments that exceeded their uncompensated costs—that it\\nmeant to prohibit consideration of third party payments under\\n§ 1396r-4(g)(1)(A). Plaintiffs’ Br. 34–35. This argument is\\nbased on the so-called “Russello presumption—that the\\npresence of a phrase in one provision and its absence in another\\nreveals Congress’ design.” City of Columbus v. Ours Garage\\n& Wrecker Serv., Inc., 536 U.S. 424, 435–36 (2002) (citing\\nRussello v. United States, 464 U.S. 16 (1983)). But that\\npresumption “grows weaker with each difference in the\\nformulation of the provisions under inspection.” Id. at 436.\\nBecause sections (g)(1)(A) and (g)(2)(A) are fundamentally\\ndifferent, we find the plaintiffs’ argument unpersuasive. See\\nTenn. Hosp. Ass’n, 908 F.3d at 1039 (“There is no tension,\\nhowever, in Congress requiring third-party payment\\ndeductions in subsection (g)(2)(A) and allowing third-party\\npayment deductions in subsection (g)(1)(A). The DSH\\npayments provided for in (g)(2)(A) are above and beyond those\\nmandated by (g)(1)(A); it therefore makes sense for Congress\\n\\f 11\\nto impose a hard limit on the ceiling of (g)(2)(A) funds—i.e.,\\nno more than 200% of the costs of serving Medicaid-eligible\\npatients, less payments from Medicaid, uninsured patients, and\\n‘third party payors’—while giving CMS more discretion to\\ncalibrate the appropriate cap on the ‘standard’ DSH payments\\ndiscussed in (g)(1)(A).”).\\n\\n Fourth, we disagree with the plaintiffs’ argument that the\\nstatute plainly distinguishes between costs and payments such\\nthat payments can never be considered in calculating “costs\\nincurred.” See Plaintiffs’ Br. at 26, 30, 33, 40, 57. The statute\\nestablishes that a hospital’s DSH payment cannot exceed its\\n“costs incurred during the year of furnishing hospital services\\n(as determined by the Secretary and net of payments under this\\nsubchapter, other than under this section, and by uninsured\\npatients).” 42 U.S.C. § 1396r-4(g)(1)(A). Both of the\\nparenthetical’s adjoining phrases modify “costs incurred”—\\nthat is, “costs incurred” are both “as determined by the\\nSecretary” and “net of payments under [Medicaid] and by\\nuninsured patients.” Id. In other words, the statute requires that\\nsome payments be considered in calculating a hospital’s “costs\\nincurred.” The argument that the statute separates costs and\\npayments therefore flies in the teeth of the statutory text.\\n\\n Contrary to the plaintiffs’ contention, we believe the 2017\\nRule is consistent with the statute’s context and purpose, both\\nof which suggest DSH payments are meant to assist those\\nhospitals that need them most by covering only those costs for\\nwhich DSH hospitals are in fact uncompensated. See 42 U.S.C.\\n§ 1396r-4(g)(1) (heading of provision at issue: “Amount of\\nadjustment subject to uncompensated costs”); 2 42 U.S.C.\\n§ 1396r-4(j)(2)(C) (requiring states to certify that “[o]nly the\\n\\n 2\\n Headings, although “not commanding,” “supply clues” about\\nCongressional intent. Yates v. United States, 135 S. Ct. 1074, 1083\\n(2015).\\n\\f 12\\nuncompensated care costs of providing inpatient hospital and\\noutpatient hospital services to individuals described in\\n[(g)(1)(A)] are included in the calculation of the hospital-\\nspecific limits.”); H.R. Rep. No. 103-111, at 211 (in enacting\\nDSH payment limit, Congress noted “some States have made\\nDSH payment[s] . . . to State psychiatric or university hospitals\\nin amounts that exceed the net costs, and in some instances the\\ntotal costs, of operating the facilities”); Tenn. Hosp. Ass’n, 908\\nF.3d at 1040 (“essence” of Congressional concern in enacting\\nstatute was “that hospitals were double dipping by collecting\\nDSH payments to cover costs that had already been\\nreimbursed”). By requiring the inclusion of payments by\\nMedicare and private insurers, the 2017 Rule ensures that DSH\\npayments will go to hospitals that have been compensated least\\nand are thus most in need. Because the 2017 Rule is consistent\\nwith the statute, it does not violate § 706(2)(C).\\n\\n B. Arbitrary and Capricious\\n\\n The plaintiffs next challenge the 2017 Rule as the product\\nof arbitrary and capricious reasoning, in violation of 5 U.S.C.\\n§ 706(2)(A). A reviewing court “shall . . . hold unlawful and\\nset aside agency action, findings, and conclusions found to be\\n. . . arbitrary, capricious, an abuse of discretion, or otherwise\\nnot in accordance with law.” 5 U.S.C. § 706(2)(A). An agency\\nrulemaking is arbitrary and capricious “if the agency has relied\\non factors which Congress has not intended it to consider,\\nentirely failed to consider an important aspect of the problem,\\noffered an explanation for its decision that runs counter to the\\nevidence before the agency, or is so implausible that it could\\nnot be ascribed to a difference in view or the product of agency\\nexpertise.” Motor Vehicles Mfrs. Ass’n of U.S. v. State Farm\\nMut. Auto. Ins. Co., 463 U.S. 29, 43 (1983).\\n\\n The plaintiffs first contend that “CMS has never\\nacknowledged, let alone justified, its new Rule’s departure\\n\\f 13\\nfrom the 2008 rule.” Plaintiffs’ Br. 27–28. We disagree.\\n“Agencies are free to change their existing policies as long as\\nthey provide a reasoned explanation for the change.” Encino\\nMotorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125 (2016). An\\nagency need not provide a more detailed justification for a\\nchanged policy than it would for a brand-new policy. Id. But it\\nmust provide “a reasoned explanation . . . for disregarding facts\\nand circumstances that underlay or were engendered by the\\nprior policy.” Id. at 2126 (quoting FCC v. Fox Television\\nStations, Inc., 556 U.S. 502, 516 (2009)). An “unexplained\\ninconsistency” with an earlier position renders a changed\\npolicy arbitrary and capricious. Id. (quoting Nat. Cable &\\nTelecomm. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 981\\n(2005)).\\n\\n There is no unexplained inconsistency with an earlier\\nposition here. To be clear, we agree with the plaintiffs that the\\n2017 Rule and the 2008 Rule establish different policies.3 But\\nit makes no difference. CMS explained why the statute’s\\npurposes are better fulfilled by a policy that requires\\nconsideration of payments by Medicare and private insurers\\n\\n 3\\n The Secretary maintains that the 2017 Rule is consistent with\\nthe 2008 Rule and so does not establish a new policy. That argument\\nhas been rejected by four courts of appeals, all of which found the\\n2010 FAQs procedurally invalid because the policy established\\ntherein, which is the same policy established by the 2017 Rule,\\nmarked a departure from the policy established by the 2008 Rule\\nwithout notice and an opportunity for public comment. See Tenn.\\nHosp. Ass’n, 908 F.3d at 1043 (“As three circuit courts and several\\ndistrict courts have now held, the payment-deduction policy\\nelucidated in the FAQs and hinted at in the preamble to the 2008 rule\\nseeks to amend, rather than merely clarify, the 2008 regulations.”\\n(citing Children’s Health Care, 900 F.3d at 1026–27; Children’s\\nHosp. of the King’s Daughters, 896 F.3d at 623; N.H. Hosp. Ass’n,\\n887 F.3d at 74)). We agree with our sister circuits.\\n\\f 14\\n(the 2017 Rule) than one that does not (the 2008 Rule, as we\\ninterpret it):\\n\\n In light of the statutory requirement limiting\\n DSH payments on a hospital-specific basis to\\n uncompensated care costs, it is inconsistent\\n with the statute to assist hospitals with costs that\\n have already been compensated by third party\\n payments. [The 2017] rule is designed to\\n reiterate the policy and make explicit within the\\n terms of the regulation that all costs and\\n payments associated with dual eligible and\\n individuals with a source of third party\\n coverage must be included in calculating the\\n hospital-specific DSH limit. This policy is\\n necessary to ensure that only actual\\n uncompensated care costs are included in the\\n Medicaid hospital-specific DSH limit. And,\\n because state DSH payments are limited to an\\n annual federal allotment, this policy is also\\n necessary to ensure that limited DSH resources\\n are allocated to hospitals that have a net\\n financial shortfall in serving Medicaid patients.\\n\\n82 Fed. Reg. at 16,117. This explanation is more than sufficient\\nto survive review under § 706(2)(A).\\n\\n The plaintiffs also claim that the Secretary has not tied the\\n2017 Rule to the administrative record. According to their\\nreading, the record shows that CMS reduces DSH payments to\\nthe plaintiff hospitals when it considers private insurance\\npayments, notwithstanding “they have among the highest\\nMedicaid inpatient utilization rates in their respective states\\nand the highest net financial shortfalls in serving Medicaid\\npatients.” Plaintiffs Br. 65. The plaintiffs claim this outcome is\\ninconsistent with the purpose of the 2017 Rule, which is “to\\n\\f 15\\nensure that limited DSH resources are allocated to hospitals\\nthat have a net financial shortfall in serving Medicaid patients.”\\n82 Fed. Reg. at 16,117.\\n\\n Their argument is doubly flawed. For starters, “Medicaid\\ninpatient utilization rates” are not mentioned in § 1396r-\\n4(g)(1)(A). See 42 U.S.C. § 1396r-4(g)(1)(A). More\\nimportantly, the plaintiffs misstate which hospitals suffer a\\n“net financial shortfall.” Programs and services a hospital\\nprovides that are not paid for by Medicaid are not relevant to\\nthe shortfall calculation. See 42 U.S.C. § 1396r-4(j)(2)(C)\\n(“Only the uncompensated care costs of providing inpatient\\nhospital and outpatient hospital services to individuals\\ndescribed in paragraph [(g)(1)(A)] are included in the\\ncalculation of the hospital-specific limits under such\\nsubsection.”). Indeed, the statute does not consider a hospital’s\\nactual costs; it considers only those costs that Medicaid pays\\nfor. See 82 Fed. Reg. at 16,118 (“Ancillary programs and\\nservices that hospitals provide to patients may be laudable, but\\nthey are not paid for by Medicaid because they are not costs\\nassociated with furnishing inpatient and outpatient hospital\\nservices.”). Calculating “net financial shortfall” using only\\nthose costs that Medicaid pays for, no hospital that suffers a\\n“net financial shortfall” will be denied a DSH payment. Thus,\\nwe disagree with the plaintiffs’ argument that the Secretary has\\nfailed to tie the Rule to the record. Like their § 706(2)(C)\\nchallenge, their § 706(2)(A) challenge fails.\\n\\n For the foregoing reasons, we reverse the judgment of the\\ndistrict court, reinstate the 2017 Rule and remand the case for\\nfurther proceedings consistent with this opinion.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4175,"cells":{"id":{"kind":"number","value":4647785,"string":"4,647,785"},"name":{"kind":"string","value":"CAPITOL SERVICES MANAGEMENT, INC., Appellant v. VESTA CORPORATION, Appellee"},"name_abbreviation":{"kind":"string","value":"Capitol Services Management v. Vesta Corporation"},"decision_date":{"kind":"string","value":"2019-08-13"},"docket_number":{"kind":"string","value":"18-7135"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Tatel, Millett, Katsas\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge MILLETT\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 12, 2019 Decided August 13, 2019\\n\\n No. 18-7135\\n\\n CAPITOL SERVICES MANAGEMENT, INC.,\\n APPELLANT\\n\\n v.\\n\\n VESTA CORPORATION,\\n APPELLEE\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:17-cv-01756)\\n\\n\\n Donald M. Temple argued the cause and filed the briefs for\\nappellant.\\n\\n Lindsay D. Breedlove argued the cause and filed the brief\\nfor appellee.\\n\\n Before: TATEL, MILLETT, and KATSAS, Circuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge MILLETT.\\n\\n MILLETT, Circuit Judge: This case concerns the proper\\napplication of the discovery rule to tortious interference claims\\nunder District of Columbia law. The district court determined\\nthat the lawsuit filed by plaintiff Capitol Services Management,\\n\\f 2\\nInc. was barred by the statute of limitations because Capitol\\nServices was on “inquiry notice” of the defendant’s alleged\\ninterference with its contract long before the limitations period\\nexpired. But at the motion-to-dismiss stage, dismissal on\\nstatute-of-limitations grounds is permissible only if a plaintiff’s\\nclaims are conclusively time-barred on the face of the\\ncomplaint. That strict standard was not met here, so we reverse\\nand remand the case for further proceedings.\\n\\n I\\n\\n A\\n\\n The Park Southern building (the “Property”) is an\\napartment complex located in Southeast Washington, D.C. In\\n2006, the Park Southern Neighborhood Corporation (“Park\\nSouthern”) acquired the Property from its prior owner, the\\nDistrict of Columbia, subject to a Deed of Trust. Under the\\nterms of their agreement, if Park Southern defaulted on the\\nterms of the Deed, the District could resume control of the\\nProperty.\\n\\n Park Southern contracted with Vesta Corporation to serve\\nas the Property’s manager. However, Park Southern eventually\\nbecame dissatisfied with Vesta’s performance and, in March\\n2014, terminated its contract. Park Southern and Capitol\\nServices then entered into a one-year management agreement\\nthat would then continue on a “year-to-year basis,” unless\\neither party terminated it with three months’ notice. Complaint\\n¶¶ 9–10, J.A. 6.\\n\\n District officials began talking almost immediately about\\nPark Southern’s management reshuffle. Milton Bailey, a\\nDistrict employee, emailed the Mayor’s chief of staff advising\\nthat “‘[w]e already have enough to send default and foreclosure\\n\\f 3\\nnotices’ to effectively takeover [sic] the property.” Complaint\\n¶ 13, J.A. 7. In a telephone call in late March, Vesta relayed to\\nBailey its intent to “continue to manage the property and to\\nprovide [the District with] whatever assistance [it] need[s] with\\nrespect to the ongoing condition of the property.” Id. ¶ 16, J.A.\\n7. Vesta also communicated with the District by private email,\\nand participated in at least five other conference calls regarding\\nCapitol Services, the Property’s “management issues,” and\\nVesta’s “interests.” Id. ¶¶ 17–18, J.A. 8.\\n\\n Things came to a head on May 2, 2014, when the District\\nexercised its default remedy under the Deed of Trust and\\nimmediately took over the Property. That same day, without\\nproviding any notice to Capitol Services, the District entered\\ninto an “emergency contract” with Vesta, authorizing it to take\\nover property management duties from Capitol Services the\\nvery next day.\\n\\n Capitol Services filed suit against the District in Superior\\nCourt in July 2014. It filed an amended complaint against the\\nDistrict in October 2014, asserting (as relevant here) claims of\\ntortious interference with contract and with business\\nopportunity. In July 2016, as part of discovery in that action,\\nCapitol Services deposed Milton Bailey. In Capitol Services’\\nview, Bailey’s testimony “revealed considerable evidence” of\\nVesta’s alleged tortious interference. J.A. 57. Yet with respect\\nto its claims against the District, Capitol Services came up\\nshort. In a May 2017 opinion, the Superior Court entered\\njudgment for the District, finding that the court lacked\\njurisdiction because the District was entitled to sovereign\\nimmunity. The court also ruled that Capitol Services had failed\\nto establish a prima facie case of tortious interference by the\\nDistrict.\\n\\f 4\\n B\\n\\n On August 28, 2017, Capitol Services filed suit against\\nVesta in the United States District Court for the District of\\nColumbia. Its complaint asserted claims for tortious\\ninterference with business relations and tortious interference\\nwith reasonable expectation of prospective economic\\nadvantage. Vesta moved to dismiss, asserting that (i) Capitol\\nServices’ claims were time-barred; (ii) Capitol Services was\\ncollaterally estopped from relitigating issues decided in the\\nSuperior Court action; and (iii) Capitol Services failed to state\\nclaims for which relief could be granted.\\n\\n Reaching only the first issue, the district court agreed that\\nthe statute of limitations barred the suit. The court ruled that\\nthe District’s three-year statute of limitations for tortious\\ninterference claims began to run on May 3, 2014, when the\\nDistrict terminated Capitol Services’ contract and substituted\\nVesta in its place. Because Capitol Services did not file suit\\nuntil August 28, 2017, the court dismissed the lawsuit as\\nuntimely.\\n\\n In so ruling, the district court rejected Capitol Services’\\nrejoinder that the discovery rule delayed the start of the\\nlimitations period until July 2016, when Capitol Services\\ndeposed Milton Bailey as part of the Superior Court action. In\\nthe district court’s view, Capitol Services was on inquiry notice\\nof its claims against Vesta as soon as the District ended the\\ncontract because, at that point, Capitol Services had “reason to\\nsuspect that [Vesta] did some wrong.” J.A. 111. The court\\nadded that Capitol Services surely was on inquiry notice far\\nearlier than Bailey’s deposition because Capitol Services’\\nSuperior Court amended complaint against the District alleged\\nVesta’s involvement in terminating its management agreement.\\n\\f 5\\n Capitol Services timely appealed.\\n\\n II\\n\\n The district court had diversity jurisdiction under 28\\nU.S.C. § 1332(a)(1). This court’s jurisdiction rests on 28\\nU.S.C. § 1291.\\n\\n We review de novo a Rule 12(b)(6) dismissal on statute-\\nof-limitations grounds, accepting plaintiff’s well-pleaded\\nfactual allegations as true and drawing all reasonable\\ninferences in plaintiff’s favor. Momenian v. Davison, 878 F.3d\\n381, 387 (D.C. Cir. 2017). At the motion to dismiss stage,\\ndismissal on statute-of-limitations grounds is proper “only if\\nthe complaint on its face is conclusively time-barred.”\\nCommonwealth Land Title Ins. Co. v. KCI Techs., Inc., 922\\nF.3d 459, 464 (D.C. Cir. 2019); accord Firestone v. Firestone,\\n76 F.3d 1205, 1209 (D.C. Cir. 1996).\\n\\n Capitol Services argues that the district court effectively\\nconverted the case to summary judgment by considering its\\npleadings in the Superior Court action, and so the summary-\\njudgment standard of review should govern. But no such\\nconversion occurred. District courts may, at the Rule 12(b)(6)\\nstage, take judicial notice of publicly filed pleadings in related\\nactions as evidence of what was alleged in the other actions,\\nalthough they may not treat as true the matter alleged. Hurd v.\\nDistrict of Columbia, 864 F.3d 671, 686 (D.C. Cir. 2017); see\\nalso FED. R. EVID. 201(b).\\n\\n That is all the district court did here. The district court\\nproperly took notice of Capitol Services’ allegations in the\\nSuperior Court action not for their truth, but for the fact that\\nCapitol Services believed those allegations to be true and\\n\\f 6\\nviable as legal claims at the time they were made. See Watkins\\nv. United States, 854 F.3d 947, 949 (7th Cir. 2017).\\n\\n So this remains an appeal of a motion to dismiss under\\nFederal Rule of Civil Procedure 12(b)(6).\\n\\n III\\n\\n On appeal, Capitol Services asserts that its complaint\\nagainst Vesta was timely because the three-year statute of\\nlimitations did not begin to run until it had inquiry notice of\\n“all of the essential elements of its cause[s] of action” against\\nVesta. Capitol Servs. Br. 32. In Capitol Services’ view, that\\ndid not happen until it deposed Milton Bailey in July 2016.\\nCapitol Servs. Br. 33–39. From that deposition, reports Capitol\\nServices, it learned that Vesta had the motive or purpose to\\n“intentional[ly] interfere[]” with its business dealings\\nmanaging the Property. Capitol Servs. Br. 34–39. Vesta, for\\nits part, insists that Capitol Services had sufficient notice to\\ntrigger the statute of limitations at the earliest when it was\\nousted from its contract in May 2014 and Vesta took over, and\\nat the latest in July 2014 when it filed suit against the District.\\nSee Vesta Br. 23–33.\\n\\n As it happens, neither party is correct. As explained\\nbelow, inquiry notice was triggered sometime after the filing of\\nthe original July 2014 complaint against the District, but before\\nthe filing of the October 2014 amended complaint, which\\nreferenced Vesta’s interference. When exactly during that\\nthree-month period Capitol Services was on inquiry notice of\\nVesta’s alleged role is an unresolved factual question on this\\nrecord, making dismissal under Rule 12(b)(6) erroneous.\\n\\f 7\\n A\\n\\n Before reaching the merits, we first address Vesta’s\\nargument that Capitol Services did not properly preserve its\\nstatute-of-limitations argument for appellate review. We find\\nno merit to this objection.\\n\\n As an appellate court, “we are a court of review, not of first\\nview[.]” Cutter v. Wilkinson, 544 U.S. 709, 718 n.7 (2005).\\nFor that reason, absent “exceptional circumstances,” this court\\ndoes not entertain issues first raised on appeal. Woodruff v.\\nPeters, 482 F.3d 521, 525 (D.C. Cir. 2007).\\n\\n Vesta objects that, before the district court, Capitol\\nServices argued that it had inquiry notice only upon learning of\\nthe “depth and breadth of Vesta’s involvement and\\ncommunications with the District.” Vesta Br. 20 (quoting J.A.\\n57). According to Vesta, Capitol Services has shifted gears on\\nappeal in arguing that it lacked any knowledge of Vesta’s\\n“motive” or “intent” to interfere. Vesta Br. 20.\\n\\n That is a distinction without a difference. Both before the\\ndistrict court and this court, Capitol Services has argued that it\\nwas not on inquiry notice of Vesta’s alleged tortious\\ninterference at the time its contract was terminated in May 2014\\nand, indeed, until Bailey’s July 2016 deposition revealed\\ndetails about Vesta’s role. Although its emphasis has changed,\\nCapitol Services has argued consistently that the facts it learned\\nat that deposition were necessary for it to have inquiry notice\\nof its causes of action. See Yee v. City of Escondido, 503 U.S.\\n519, 534 (1992) (“Once a * * * claim is properly presented, a\\nparty can make any argument in support of that claim; parties\\nare not limited to the precise arguments they made below.”); In\\nre Harman Int’l Indus., Inc. Sec. Litig., 791 F.3d 90, 100 (D.C.\\nCir. 2015) (“On appeal, a party may refine and clarify its\\n\\f 8\\nanalysis in light of the district court’s ruling, including citing\\nadditional support.”) (formatting modified); Koch v. Cox, 489\\nF.3d 384, 391 (D.C. Cir. 2007) (A party may cite “additional\\nsupport for his side of an issue upon which the district court did\\nrule, much like citing a case for the first time on appeal.”).\\n\\n On to the merits.\\n\\n B\\n\\n We take as given the parties’ agreement that District of\\nColumbia law governs this tort dispute. See Patton Boggs LLP\\nv. Chevron Corp., 683 F.3d 397, 403 (D.C. Cir. 2012)\\n(applying District of Columbia law “which both parties assume\\napplies”). Under District law, Capitol Services’ claims for\\ntortious interference with business relations and tortious\\ninterference with reasonable expectation of prospective\\neconomic advantage are subject to a three-year statute of\\nlimitations. D.C. CODE § 12-301(8); see Beard v. Edmondson\\n& Gallagher, 790 A.2d 541, 546 (D.C. 2002); Carr v. Brown,\\n395 A.2d 79, 83 (D.C. 1978). See generally Hartford Accident\\n& Indem. Co. v. Pro-Football, Inc., 127 F.3d 1111, 1118 (D.C.\\nCir. 1997) (“Since [plaintiff’s] claims arise under District law,\\nthe applicable statute of limitations is also that of the\\nDistrict.”).\\n\\n Ordinarily under District law, the statute of limitations for\\na tort claim starts to run when the plaintiff is injured. See\\nAmobi v. District of Columbia Dep’t of Corr., 755 F.3d 980,\\n994 (D.C. Cir. 2014). All agree that Capitol Services was\\ninjured on May 3, 2014, when the District terminated its\\nmanagement contract. If that were the end of the story, basic\\nmath would render Capitol Services’ August 28, 2017 filing of\\nthis lawsuit untimely.\\n\\f 9\\n But Capitol Services contends, and Vesta does not\\nmeaningfully dispute, that the District of Columbia’s discovery\\nrule delays the start of the statute of limitations until Capitol\\nServices had actual or inquiry notice of its potential causes of\\naction against Vesta. See Momenian, 878 F.3d at 388. 1\\n\\n Where Capitol Services and Vesta part ways is over (i)\\nwhat “inquiry notice” demands, and (ii) whether it is\\n“conclusive” on the face of the complaint that Capitol Services\\nwas on inquiry notice of its claims against Vesta more than\\nthree years before it filed suit.\\n\\n 1\\n\\n A claim accrues under the discovery rule when the plaintiff\\n“either has actual knowledge of a cause of action” or is\\notherwise “charged with knowledge of that cause of action.”\\nCevenini v. Archbishop of Wash., 707 A.2d 768, 771 (D.C.\\n1998). The latter is known as “inquiry notice.” Inquiry notice\\nis “that notice which a plaintiff would have possessed after due\\ninvestigation.” Diamond v. Davis, 680 A.2d 364, 372 (D.C.\\n1996). It does not refer to “the amount of information that\\ntriggers a duty to investigate,” but rather the “quantum of\\nknowledge required to” “charge[] [a plaintiff] with knowledge\\n\\n 1\\n Vesta dedicated a measly three parenthetical words in its brief\\nto the applicability of the discovery rule. See Vesta Br. 27 (“In the\\ncase at bar, the district court * * * concluded that even if the\\ndiscovery rule applied (it does not), Capitol Services’ claim\\nnonetheless accrued in March 2014.”) (emphasis added). That was\\nnot enough to put the issue in controversy. See Davis v. Pension\\nBenefit Guar. Corp., 734 F.3d 1161, 1166–1167 (D.C. Cir. 2013) (A\\nparty may not “mention a possible argument in the most skeletal way,\\nleaving the court to do counsel’s work, create the ossature for the\\nargument, and put flesh on its bones.”) (internal quotation marks\\nomitted).\\n\\f 10\\nof [his] cause of action[.]” Id.; see BDO Seidman v. Morgan,\\nLewis & Bockius LLP, 89 A.3d 492, 500 (D.C. 2014). Under\\nDistrict of Columbia law, the time at which a plaintiff acquires\\ninquiry notice “is a question of fact.” Diamond, 680 A.2d at\\n372.\\n\\n Following multiple decisions of the District of Columbia\\nCourt of Appeals, this Court has repeatedly applied the\\ndiscovery rule to postpone the running of a statute of\\nlimitations until the plaintiff “knows or by the exercise of\\nreasonable diligence should know: (1) of the injury; (2) the\\ninjury’s cause in fact; and (3) of some evidence of\\nwrongdoing.” Commonwealth Land, 922 F.3d at 464 (quoting\\nCapital Place I Assocs. L.P. v. George Hyman Constr. Co., 673\\nA.2d 194, 199 (D.C. 1996)). 2\\n\\n Under that test, Capitol Services’ claims accrued when,\\nafter due investigation, it would have learned of its injury, the\\ninjury’s cause in fact, and “some evidence” of Vesta’s\\nwrongdoing. Commonwealth Land, 922 F.3d at 464. Yet as\\nCapitol Services notes, the District’s Court of Appeals has\\nsometimes articulated its discovery rule differently, stating that\\nclaims do not accrue until the plaintiff has “discovered or\\nreasonably should have discovered all of the essential elements\\nof his possible cause of action.” Price v. Washington Metro.\\n\\n 2\\n Accord Hancock v. HomEq Servicing Corp., 526 F.3d 785,\\n786 (D.C. Cir. 2008) (quoting Diamond, 680 A.2d at 381); Bradley\\nv. National Ass’n of Sec. Dealers Dispute Resolution, Inc., 433 F.3d\\n846, 849 (D.C. Cir. 2005) (quoting Bussineau v. President & Dirs.\\nof Georgetown College, 518 A.2d 423, 425 (D.C. 1986)); Jung v.\\nMundy, Holt & Mance, P.C., 372 F.3d 429, 433 (D.C. Cir. 2004)\\n(quoting Knight v. Furlow, 553 A.2d 1232, 1234 (D.C. 1989));\\nGoldman v. Bequai, 19 F.3d 666, 671–672 (D.C. Cir. 1994) (quoting\\nKnight, 553 A.2d at 1234); Williams v. Mordofsky, 901 F.2d 158, 162\\n(D.C. Cir. 1990) (quoting Knight, 553 A.2d at 1234).\\n\\f 11\\nArea Transit Auth., 41 A.3d 526, 533 (D.C. 2012) (formatting\\nmodified); accord Owens-Corning Fiberglass Corp. v. Henkel,\\n689 A.2d 1224, 1231 (D.C. 1997); Arrington v. District of\\nColumbia, 673 A.2d 674, 678 (D.C. 1996); Colbert v.\\nGeorgetown Univ., 641 A.2d 469, 473 (D.C. 1994) (en banc).\\nAnd because a prima facie case of tortious interference with\\nbusiness relations or with reasonable expectation of\\nprospective economic advantage requires that a plaintiff plead\\n“intentional interference,” Jankovic v. International Crisis\\nGroup, 593 F.3d 22, 29 (D.C. Cir. 2010); NCRIC, Inc. v.\\nColumbia Hosp. for Women Med. Ctr., Inc., 957 A.2d 890, 900\\n(D.C. 2008), Capitol Services reasons that the statute of\\nlimitations did not begin running until it “received notice that\\nVesta intentionally interfered with [its] expectation that it\\nwould continue to perform property management services” at\\nthe Property, Capitol Servs. Br. 34.\\n\\n Predicting whether the District’s Court of Appeals would\\nfind light between the “some evidence of wrongdoing” and “all\\nelements” tests would be no easy task. See Earle v. District of\\nColumbia, 707 F.3d 299, 310 (D.C. Cir. 2012). There is\\ncaselaw on both sides of the ledger. Compare Hardi v.\\nMezzanotte, 818 A.2d 974, 979 (D.C. 2003), and East v.\\nGraphic Arts Indus. Joint Pension Trust, 718 A.2d 153, 157\\n(D.C. 1998), with Morton v. National Med. Enters., Inc., 725\\nA.2d 462, 469 (D.C. 1999), and Cevenini, 707 A.2d at 771.\\nUltimately, that is a job for another day. Even assuming the\\n“some evidence of wrongdoing” standard is stricter for\\nplaintiffs, Capitol Services’ lawsuit is timely.\\n\\n 2\\n\\n There is no dispute that, as of May 3, 2014, Capitol\\nServices had “actual notice” of its injury—the cancellation of\\nits property management contract. And Capitol Services knew\\n\\f 12\\nthen that the injury was caused by the District’s decision to\\naward an “emergency contract” to Vesta. See Vesta Br. 27–28;\\nCapitol Servs. Br. 33–34. There is also no meaningful dispute\\nthat Capitol Services lacked “actual notice” of Vesta’s\\nwrongdoing as of that date. See Vesta Br. 28–31; Capitol\\nServs. Br. 30, 36–37.\\n\\n The district court, instead, ruled that Capitol Services was\\non inquiry notice as of May 3rd. Specifically, the district court\\nreasoned that Capitol Services had “reason to suspect that\\nVesta did some wrong” because it knew that (i) the District had\\nissued an “emergency contract” to Vesta, (ii) the District\\n“inappropriately terminated” Capitol Services’ agreement, and\\n(iii) Capitol Services’ relationship with Park Southern “ended.”\\nJ.A. 110–111.\\n\\n While those facts certainly put Capitol Services on notice\\nof its claims against the District, there is nothing in the\\ndescription of the District’s actions that hinted at Vesta’s role\\nin the contract’s termination. “[P]laintiff’s knowledge of\\nwrongdoing on the part of” the District itself “did not cause\\naccrual of his action against another[.]” Diamond, 680 A.2d at\\n380; see also Cevenini, 707 A.2d at 773 (“[K]nowledge of\\nmisconduct on the part of one defendant will not automatically\\ncreate inquiry notice of claims against a potential co-\\ndefendant[.]”).\\n\\n Nor did the district court identify what evidence a “diligent\\ninvestigation” at that point would have uncovered about\\nVesta’s participation in the District’s decisionmaking.\\nDiamond, 680 A.2d at 372; see Commonwealth Land, 922 F.3d\\nat 464. Because Capitol Services had reported to Park\\nSouthern, not the District, and because the District had declined\\nto provide notice to Capitol Services of its ouster, nothing in\\nthe initial complaint suggests that Capitol Services had any\\n\\f 13\\nknowledge or even suspicions of background relationships or\\nconnections between Vesta and the District government. Even\\nthe written communication between the District and Vesta were\\nby “private e-mails, which effectively circumvented FOIA\\nscrutiny.” Complaint ¶ 17, J.A. 8. The complaint does not\\n“conclusively establish” any facts uncoverable by Capitol\\nServices on May 3, 2014, or even by August 28, 2014 (that is,\\nthree years before the complaint against Vesta was first filed)\\nthat would have implicated Vesta in its injury within the three-\\nyear limitations period.\\n\\n Vesta argues, in the alternative, that Capitol Services was\\non inquiry notice in July 2014, when it filed its complaint\\nagainst the District in Superior Court. This is so, Vesta\\nexplains, because the complaint repeatedly alleges Vesta’s role\\nin inducing Capitol Services’ contract termination. See, e.g.,\\nJ.A. 32 ¶ 18 (The District “terminated the contract and awarded\\nit immediately to Vesta because of political and personal\\nconnections on the pretense of an ‘emergency contract’ award.\\nUpon information and belief, certain [District] representatives\\njointly engaged in this decision. Said officials also\\ncollaborated with Vesta regarding the government’s actions in\\nadvance of [the District’s] May 3 * * * takeover of the\\nproperty.”); J.A. 32 ¶ 19 (The District’s “purported\\nqualification of the contract award to Vesta as an ‘Emergency’\\nwas pretextual. It was intended to cause the removal of\\n[Capitol Services] and the substitution of Vesta as the\\nproperty’s property manager.”); J.A. 34 ¶ 29 (“Upon\\ninformation and belief, * * * one or more [District] officials\\nconsulted improperly and illegally with yet known [sic] Vesta\\nofficials to achieve the termination of [Capitol Services’]\\nAgreement and the award of an ‘emergency’ contract to Vesta\\nin order to block [Capitol Services’] business opportunities.”);\\nJ.A. 38 ¶ 48 (The District “impeded [Capitol Services’]\\nbusiness relationship and prospective advantages/opportunities\\n\\f 14\\nin order to advantage the interests of [Vesta], declared a non-\\nexisting pretextual ‘emergency’ situation and awarded an\\nillegal ‘emergency [contract]’ to [Vesta].”); J.A. 39 (asserting\\na claim against the District for civil conspiracy); see also D.C.\\nR. CIV. P. 11(b)(3) (imposing an obligation on counsel to assert\\nonly factual contentions having evidentiary support).\\n\\n That would be a really good argument if Vesta were\\ndiscussing the right version of the complaint. But those\\nallegations appear nowhere in Capitol Services’ July 2014\\ncomplaint against the District. The allegations in the July 2014\\ncomplaint reference Vesta only in innocuous and fleeting\\nfactual recitations. See Complaint ¶ 17, Capitol Servs. Mgmt.\\nInc. v. District of Columbia, No. 2014 CA 004551 B (D.C.\\nSuper. Ct. July 24, 2014) (The District “entered into an\\n‘emergency contract’ with Vesta * * * on May 2, 2014, to\\nauthorize Vesta to take over the management of Park Southern\\nApartments on May 3, 2014.”); id. ¶ 18 (“In executing said\\ncontract with Vesta, [the District] immediately terminated the\\nmanagement agreement between [Park Southern] and [Capitol\\nServices].”); id. ¶ 31.\\n\\n The allegations that Vesta cites and on which Vesta’s\\nargument relies come from Capitol Services’ amended\\nSuperior Court complaint. That was filed in October 2014—\\nsquarely within the three-year period preceding the lawsuit\\nagainst Vesta.\\n\\n Finally, in light of the allegations in its amended\\ncomplaint, Capitol Services’ insistence that it had no notice\\nuntil its deposition of Milton Bailey in July 2016 pushes things\\ntoo far. Inquiry notice commences when, with due diligence, a\\nplaintiff would have been on notice that a particular\\ndefendant’s wrongdoing caused or contributed to a potential\\ninjury. See, e.g., Diamond, 680 A.2d at 372. It does not wait\\n\\f 15\\nuntil a plaintiff has actually assembled the evidence in\\ndeposition form.\\n\\n So all that the limited record shows at this Rule 12(b)(6)\\nstage is that Capitol Services was not on actual or inquiry\\nnotice in either May or July of 2014, but was on inquiry notice\\nat least by October 2014. The record is inconclusive as to\\nCapitol Services’ knowledge about Vesta’s role prior to August\\n28, 2014—three years before this suit against Vesta was filed.\\nWhen during that intervening period Capitol Services had\\ninquiry notice of Vesta’s potential role thus is an open factual\\nquestion that cannot be resolved at this early juncture.\\n\\n IV\\n\\n This court can affirm a district court’s judgment on any\\nbasis supported by the record. See United States v. Hicks, 911\\nF.3d 623, 626 (D.C. Cir. 2018). Vesta asks this court to do just\\nthat, arguing that even if the complaint is timely, Capitol\\nServices nonetheless is collaterally estopped from relitigating\\nthe question of alleged tortious interference with its contract.\\nVesta points to the prior Superior Court judgment dismissing\\nCapitol Services’ case against the District arising from\\ntermination of the Property contract. See J.A. 43–49.\\n\\n Vesta fares no better on this front. The Superior Court held\\nthat it lacked jurisdiction to adjudicate the suit against the\\nDistrict, and so a central element for collateral estoppel is\\nmissing.\\n\\n Collateral estoppel, also known as issue preclusion,\\ngenerally confines plaintiffs to one bite at the litigation apple.\\nSee B&B Hardware, Inc. v. Hargis Indus., Inc., 135 S. Ct.\\n1293, 1302–1303 (2015). Specifically, collateral estoppel bars\\nsuccessive litigation of an issue of fact or law when\\n\\f 16\\n\\n (1) the issue is actually litigated; (2) determined by a\\n valid, final judgment on the merits; (3) after a full and\\n fair opportunity for litigation by the parties or their\\n privies; and (4) under circumstances where the\\n determination was essential to the judgment, and not\\n merely dictum.\\n\\nWalker v. FedEx Office & Print Servs., Inc., 123 A.3d 160, 164\\n(D.C. 2015) (quoting Hogue v. Hopper, 728 A.2d 611, 614\\n(D.C. 1999)). See generally Bode & Grenier, LLP v. Knight,\\n808 F.3d 852, 857–858 (D.C. Cir. 2015) (applying preclusion\\nrules of the jurisdiction in which the judgment was rendered).\\n\\n Vesta directs our attention to language in the Superior\\nCourt’s 2017 decision stating that it found “no evidence” “that\\nthe District of Columbia actually interfered with [Capitol\\nServices’] contract or its business relationship with [Park\\nSouthern] because neither the District of Columbia nor [Park\\nSouthern] took any action to terminate it,” and instead “it\\nappear[ed] to the court that [Capitol Services] voluntarily\\ndecided not to enforce its claimed rights[.]” J.A. 48. Because\\nCapitol Services failed to show any interference in that case,\\nVesta argues, Capitol Services should be precluded from\\nrelitigating whether it suffered tortious interference at Vesta’s\\nhands.\\n\\n The Achilles heel of that argument is that, immediately\\nbefore remarking on contract interference, the Superior Court\\nruled that it could not adjudicate the lawsuit because the\\nDistrict enjoyed sovereign immunity from suit, and so was\\n“entitled to summary judgment” on that basis. J.A. 48. The\\nDistrict’s sovereign immunity is a jurisdictional bar to\\nlitigation—if it is not waived, the sovereign’s entitlement to\\nimmunity deprives the court of subject matter jurisdiction over\\n\\f 17\\nthe action. See Washington Metro. Area Transit Auth. v.\\nBarksdale-Showell, 965 A.2d 16, 21 n.4 (D.C. 2009). So once\\nthe Superior Court concluded that it was without jurisdiction,\\nit could not “reach beyond that determination to consider\\nwhether th[e plaintiff’s] claims would be meritorious if the\\ndefect in [jurisdiction] were cured[.]” UMC Dev., LLC v.\\nDistrict of Columbia, 120 A.3d 37, 49 (D.C. 2015).\\n\\n The absence of jurisdiction meant the Superior Court\\ncould not and did not actually resolve “on the merits” Capitol\\nServices’ tortious interference claim against the District, and so\\nthe court’s comments on the subject were “merely dictum[,]”\\nin no way “essential to the judgment[.]” Walker, 123 A.3d at\\n164. 3 As a result, the critical elements of collateral estoppel\\nhave not been established in this case.\\n\\n V\\n\\n Lastly, Vesta asks that we affirm on the ground that\\nCapitol Services failed to state a claim on the merits for either\\ntortious interference with business relations, or tortious\\ninterference with reasonable expectation of prospective\\neconomic advantage. Because the district court did not address\\nthose questions and the parties have dedicated scant appellate\\n\\n 3\\n See also United States ex rel. Bledsoe v. Community Health\\nSys., Inc., 501 F.3d 493, 507 (6th Cir. 2007) (holding that the\\nconclusion in a prior court’s opinion was “more properly\\ncharacterized as dicta than an alternative holding[]” because the\\ncourt had “concluded that it lacked subject matter jurisdiction”);\\nMoreland v. Federal Bureau of Prisons, 431 F.3d 180, 185 (5th Cir.\\n2005) (holding that the conclusion in a prior court’s opinion was\\n“dicta because the petition in [that opinion] was dismissed for lack\\nof subject-matter jurisdiction,” and was not an “alternative holding\\nbecause it could not support the actual judgment in that case, which\\nwas dismissal for lack of subject-matter jurisdiction”).\\n\\f 18\\nbriefing to the issues, it would be unwise for us to wade in at\\nthis time and on this record. The issue, of course, remains open\\nfor the district court to address on remand.\\n\\n VI\\n\\n For the foregoing reasons, we reverse the district court’s\\njudgment dismissing the case under Federal Rule of Civil\\nProcedure 12(b)(6) on statute-of-limitations grounds and\\nremand for further proceedings consistent with this opinion.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4425038/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge MILLETT\"}]}"},"judge1":{"kind":"string","value":"TATEL"},"judge2":{"kind":"string","value":"MILLETT"},"judge3":{"kind":"string","value":"KATSAS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4647785/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: CAPITOL SERVICES MANAGEMENT, INC., Appellant v. VESTA CORPORATION, Appellee\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-13\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued April 12, 2019 Decided August 13, 2019\n\n No. 18-7135\n\n CAPITOL SERVICES MANAGEMENT, INC.,\n APPELLANT\n\n v.\n\n VESTA CORPORATION,\n APPELLEE\n\n\n Appeal from the United States District Court\n for the District of Columbia\n (No. 1:17-cv-01756)\n\n\n Donald M. Temple argued the cause and filed the briefs for\nappellant.\n\n Lindsay D. Breedlove argued the cause and filed the brief\nfor appellee.\n\n Before: TATEL, MILLETT, and KATSAS, Circuit Judges.\n\n Opinion for the Court filed by Circuit Judge MILLETT.\n\n MILLETT, Circuit Judge: This case concerns the proper\napplication of the discovery rule to tortious interference claims\nunder District of Columbia law. The district court determined\nthat the lawsuit filed by plaintiff Capitol Services Management,\n\f 2\nInc. was barred by the statute of limitations because Capitol\nServices was on “inquiry notice” of the defendant’s alleged\ninterference with its contract long before the limitations period\nexpired. But at the motion-to-dismiss stage, dismissal on\nstatute-of-limitations grounds is permissible only if a plaintiff’s\nclaims are conclusively time-barred on the face of the\ncomplaint. That strict standard was not met here, so we reverse\nand remand the case for further proceedings.\n\n I\n\n A\n\n The Park Southern building (the “Property”) is an\napartment complex located in Southeast Washington, D.C. In\n2006, the Park Southern Neighborhood Corporation (“Park\nSouthern”) acquired the Property from its prior owner, the\nDistrict of Columbia, subject to a Deed of Trust. Under the\nterms of their agreement, if Park Southern defaulted on the\nterms of the Deed, the District could resume control of the\nProperty.\n\n Park Southern contracted with Vesta Corporation to serve\nas the Property’s manager. However, Park Southern eventually\nbecame dissatisfied with Vesta’s performance and, in March\n2014, terminated its contract. Park Southern and Capitol\nServices then entered into a one-year management agreement\nthat would then continue on a “year-to-year basis,” unless\neither party terminated it with three months’ notice. Complaint\n¶¶ 9–10, J.A. 6.\n\n District officials began talking almost immediately about\nPark Southern’s management reshuffle. Milton Bailey, a\nDistrict employee, emailed the Mayor’s chief of staff advising\nthat “‘[w]e already have enough to send default and foreclosure\n\f 3\nnotices’ to effectively takeover [sic] the property.” Complaint\n¶ 13, J.A. 7. In a telephone call in late March, Vesta relayed to\nBailey its intent to “continue to manage the property and to\nprovide [the District with] whatever assistance [it] need[s] with\nrespect to the ongoing condition of the property.” Id. ¶ 16, J.A.\n7. Vesta also communicated with the District by private email,\nand participated in at least five other conference calls regarding\nCapitol Services, the Property’s “management issues,” and\nVesta’s “interests.” Id. ¶¶ 17–18, J.A. 8.\n\n Things came to a head on May 2, 2014, when the District\nexercised its default remedy under the Deed of Trust and\nimmediately took over the Property. That same day, without\nproviding any notice to Capitol Services, the District entered\ninto an “emergency contract” with Vesta, authorizing it to take\nover property management duties from Capitol Services the\nvery next day.\n\n Capitol Services filed suit against the District in Superior\nCourt in July 2014. It filed an amended complaint against the\nDistrict in October 2014, asserting (as relevant here) claims of\ntortious interference with contract and with business\nopportunity. In July 2016, as part of discovery in that action,\nCapitol Services deposed Milton Bailey. In Capitol Services’\nview, Bailey’s testimony “revealed considerable evidence” of\nVesta’s alleged tortious interference. J.A. 57. Yet with respect\nto its claims against the District, Capitol Services came up\nshort. In a May 2017 opinion, the Superior Court entered\njudgment for the District, finding that the court lacked\njurisdiction because the District was entitled to sovereign\nimmunity. The court also ruled that Capitol Services had failed\nto establish a prima facie case of tortious interference by the\nDistrict.\n\f 4\n B\n\n On August 28, 2017, Capitol Services filed suit against\nVesta in the United States District Court for the District of\nColumbia. Its complaint asserted claims for tortious\ninterference with business relations and tortious interference\nwith reasonable expectation of prospective economic\nadvantage. Vesta moved to dismiss, asserting that (i) Capitol\nServices’ claims were time-barred; (ii) Capitol Services was\ncollaterally estopped from relitigating issues decided in the\nSuperior Court action; and (iii) Capitol Services failed to state\nclaims for which relief could be granted.\n\n Reaching only the first issue, the district court agreed that\nthe statute of limitations barred the suit. The court ruled that\nthe District’s three-year statute of limitations for tortious\ninterference claims began to run on May 3, 2014, when the\nDistrict terminated Capitol Services’ contract and substituted\nVesta in its place. Because Capitol Services did not file suit\nuntil August 28, 2017, the court dismissed the lawsuit as\nuntimely.\n\n In so ruling, the district court rejected Capitol Services’\nrejoinder that the discovery rule delayed the start of the\nlimitations period until July 2016, when Capitol Services\ndeposed Milton Bailey as part of the Superior Court action. In\nthe district court’s view, Capitol Services was on inquiry notice\nof its claims against Vesta as soon as the District ended the\ncontract because, at that point, Capitol Services had “reason to\nsuspect that [Vesta] did some wrong.” J.A. 111. The court\nadded that Capitol Services surely was on inquiry notice far\nearlier than Bailey’s deposition because Capitol Services’\nSuperior Court amended complaint against the District alleged\nVesta’s involvement in terminating its management agreement.\n\f 5\n Capitol Services timely appealed.\n\n II\n\n The district court had diversity jurisdiction under 28\nU.S.C. § 1332(a)(1). This court’s jurisdiction rests on 28\nU.S.C. § 1291.\n\n We review de novo a Rule 12(b)(6) dismissal on statute-\nof-limitations grounds, accepting plaintiff’s well-pleaded\nfactual allegations as true and drawing all reasonable\ninferences in plaintiff’s favor. Momenian v. Davison, 878 F.3d\n381, 387 (D.C. Cir. 2017). At the motion to dismiss stage,\ndismissal on statute-of-limitations grounds is proper “only if\nthe complaint on its face is conclusively time-barred.”\nCommonwealth Land Title Ins. Co. v. KCI Techs., Inc., 922\nF.3d 459, 464 (D.C. Cir. 2019); accord Firestone v. Firestone,\n76 F.3d 1205, 1209 (D.C. Cir. 1996).\n\n Capitol Services argues that the district court effectively\nconverted the case to summary judgment by considering its\npleadings in the Superior Court action, and so the summary-\njudgment standard of review should govern. But no such\nconversion occurred. District courts may, at the Rule 12(b)(6)\nstage, take judicial notice of publicly filed pleadings in related\nactions as evidence of what was alleged in the other actions,\nalthough they may not treat as true the matter alleged. Hurd v.\nDistrict of Columbia, 864 F.3d 671, 686 (D.C. Cir. 2017); see\nalso FED. R. EVID. 201(b).\n\n That is all the district court did here. The district court\nproperly took notice of Capitol Services’ allegations in the\nSuperior Court action not for their truth, but for the fact that\nCapitol Services believed those allegations to be true and\n\f 6\nviable as legal claims at the time they were made. See Watkins\nv. United States, 854 F.3d 947, 949 (7th Cir. 2017).\n\n So this remains an appeal of a motion to dismiss under\nFederal Rule of Civil Procedure 12(b)(6).\n\n III\n\n On appeal, Capitol Services asserts that its complaint\nagainst Vesta was timely because the three-year statute of\nlimitations did not begin to run until it had inquiry notice of\n“all of the essential elements of its cause[s] of action” against\nVesta. Capitol Servs. Br. 32. In Capitol Services’ view, that\ndid not happen until it deposed Milton Bailey in July 2016.\nCapitol Servs. Br. 33–39. From that deposition, reports Capitol\nServices, it learned that Vesta had the motive or purpose to\n“intentional[ly] interfere[]” with its business dealings\nmanaging the Property. Capitol Servs. Br. 34–39. Vesta, for\nits part, insists that Capitol Services had sufficient notice to\ntrigger the statute of limitations at the earliest when it was\nousted from its contract in May 2014 and Vesta took over, and\nat the latest in July 2014 when it filed suit against the District.\nSee Vesta Br. 23–33.\n\n As it happens, neither party is correct. As explained\nbelow, inquiry notice was triggered sometime after the filing of\nthe original July 2014 complaint against the District, but before\nthe filing of the October 2014 amended complaint, which\nreferenced Vesta’s interference. When exactly during that\nthree-month period Capitol Services was on inquiry notice of\nVesta’s alleged role is an unresolved factual question on this\nrecord, making dismissal under Rule 12(b)(6) erroneous.\n\f 7\n A\n\n Before reaching the merits, we first address Vesta’s\nargument that Capitol Services did not properly preserve its\nstatute-of-limitations argument for appellate review. We find\nno merit to this objection.\n\n As an appellate court, “we are a court of review, not of first\nview[.]” Cutter v. Wilkinson, 544 U.S. 709, 718 n.7 (2005).\nFor that reason, absent “exceptional circumstances,” this court\ndoes not entertain issues first raised on appeal. Woodruff v.\nPeters, 482 F.3d 521, 525 (D.C. Cir. 2007).\n\n Vesta objects that, before the district court, Capitol\nServices argued that it had inquiry notice only upon learning of\nthe “depth and breadth of Vesta’s involvement and\ncommunications with the District.” Vesta Br. 20 (quoting J.A.\n57). According to Vesta, Capitol Services has shifted gears on\nappeal in arguing that it lacked any knowledge of Vesta’s\n“motive” or “intent” to interfere. Vesta Br. 20.\n\n That is a distinction without a difference. Both before the\ndistrict court and this court, Capitol Services has argued that it\nwas not on inquiry notice of Vesta’s alleged tortious\ninterference at the time its contract was terminated in May 2014\nand, indeed, until Bailey’s July 2016 deposition revealed\ndetails about Vesta’s role. Although its emphasis has changed,\nCapitol Services has argued consistently that the facts it learned\nat that deposition were necessary for it to have inquiry notice\nof its causes of action. See Yee v. City of Escondido, 503 U.S.\n519, 534 (1992) (“Once a * * * claim is properly presented, a\nparty can make any argument in support of that claim; parties\nare not limited to the precise arguments they made below.”); In\nre Harman Int’l Indus., Inc. Sec. Litig., 791 F.3d 90, 100 (D.C.\nCir. 2015) (“On appeal, a party may refine and clarify its\n\f 8\nanalysis in light of the district court’s ruling, including citing\nadditional support.”) (formatting modified); Koch v. Cox, 489\nF.3d 384, 391 (D.C. Cir. 2007) (A party may cite “additional\nsupport for his side of an issue upon which the district court did\nrule, much like citing a case for the first time on appeal.”).\n\n On to the merits.\n\n B\n\n We take as given the parties’ agreement that District of\nColumbia law governs this tort dispute. See Patton Boggs LLP\nv. Chevron Corp., 683 F.3d 397, 403 (D.C. Cir. 2012)\n(applying District of Columbia law “which both parties assume\napplies”). Under District law, Capitol Services’ claims for\ntortious interference with business relations and tortious\ninterference with reasonable expectation of prospective\neconomic advantage are subject to a three-year statute of\nlimitations. D.C. CODE § 12-301(8); see Beard v. Edmondson\n& Gallagher, 790 A.2d 541, 546 (D.C. 2002); Carr v. Brown,\n395 A.2d 79, 83 (D.C. 1978). See generally Hartford Accident\n& Indem. Co. v. Pro-Football, Inc., 127 F.3d 1111, 1118 (D.C.\nCir. 1997) (“Since [plaintiff’s] claims arise under District law,\nthe applicable statute of limitations is also that of the\nDistrict.”).\n\n Ordinarily under District law, the statute of limitations for\na tort claim starts to run when the plaintiff is injured. See\nAmobi v. District of Columbia Dep’t of Corr., 755 F.3d 980,\n994 (D.C. Cir. 2014). All agree that Capitol Services was\ninjured on May 3, 2014, when the District terminated its\nmanagement contract. If that were the end of the story, basic\nmath would render Capitol Services’ August 28, 2017 filing of\nthis lawsuit untimely.\n\f 9\n But Capitol Services contends, and Vesta does not\nmeaningfully dispute, that the District of Columbia’s discovery\nrule delays the start of the statute of limitations until Capitol\nServices had actual or inquiry notice of its potential causes of\naction against Vesta. See Momenian, 878 F.3d at 388. 1\n\n Where Capitol Services and Vesta part ways is over (i)\nwhat “inquiry notice” demands, and (ii) whether it is\n“conclusive” on the face of the complaint that Capitol Services\nwas on inquiry notice of its claims against Vesta more than\nthree years before it filed suit.\n\n 1\n\n A claim accrues under the discovery rule when the plaintiff\n“either has actual knowledge of a cause of action” or is\notherwise “charged with knowledge of that cause of action.”\nCevenini v. Archbishop of Wash., 707 A.2d 768, 771 (D.C.\n1998). The latter is known as “inquiry notice.” Inquiry notice\nis “that notice which a plaintiff would have possessed after due\ninvestigation.” Diamond v. Davis, 680 A.2d 364, 372 (D.C.\n1996). It does not refer to “the amount of information that\ntriggers a duty to investigate,” but rather the “quantum of\nknowledge required to” “charge[] [a plaintiff] with knowledge\n\n 1\n Vesta dedicated a measly three parenthetical words in its brief\nto the applicability of the discovery rule. See Vesta Br. 27 (“In the\ncase at bar, the district court * * * concluded that even if the\ndiscovery rule applied (it does not), Capitol Services’ claim\nnonetheless accrued in March 2014.”) (emphasis added). That was\nnot enough to put the issue in controversy. See Davis v. Pension\nBenefit Guar. Corp., 734 F.3d 1161, 1166–1167 (D.C. Cir. 2013) (A\nparty may not “mention a possible argument in the most skeletal way,\nleaving the court to do counsel’s work, create the ossature for the\nargument, and put flesh on its bones.”) (internal quotation marks\nomitted).\n\f 10\nof [his] cause of action[.]” Id.; see BDO Seidman v. Morgan,\nLewis & Bockius LLP, 89 A.3d 492, 500 (D.C. 2014). Under\nDistrict of Columbia law, the time at which a plaintiff acquires\ninquiry notice “is a question of fact.” Diamond, 680 A.2d at\n372.\n\n Following multiple decisions of the District of Columbia\nCourt of Appeals, this Court has repeatedly applied the\ndiscovery rule to postpone the running of a statute of\nlimitations until the plaintiff “knows or by the exercise of\nreasonable diligence should know: (1) of the injury; (2) the\ninjury’s cause in fact; and (3) of some evidence of\nwrongdoing.” Commonwealth Land, 922 F.3d at 464 (quoting\nCapital Place I Assocs. L.P. v. George Hyman Constr. Co., 673\nA.2d 194, 199 (D.C. 1996)). 2\n\n Under that test, Capitol Services’ claims accrued when,\nafter due investigation, it would have learned of its injury, the\ninjury’s cause in fact, and “some evidence” of Vesta’s\nwrongdoing. Commonwealth Land, 922 F.3d at 464. Yet as\nCapitol Services notes, the District’s Court of Appeals has\nsometimes articulated its discovery rule differently, stating that\nclaims do not accrue until the plaintiff has “discovered or\nreasonably should have discovered all of the essential elements\nof his possible cause of action.” Price v. Washington Metro.\n\n 2\n Accord Hancock v. HomEq Servicing Corp., 526 F.3d 785,\n786 (D.C. Cir. 2008) (quoting Diamond, 680 A.2d at 381); Bradley\nv. National Ass’n of Sec. Dealers Dispute Resolution, Inc., 433 F.3d\n846, 849 (D.C. Cir. 2005) (quoting Bussineau v. President & Dirs.\nof Georgetown College, 518 A.2d 423, 425 (D.C. 1986)); Jung v.\nMundy, Holt & Mance, P.C., 372 F.3d 429, 433 (D.C. Cir. 2004)\n(quoting Knight v. Furlow, 553 A.2d 1232, 1234 (D.C. 1989));\nGoldman v. Bequai, 19 F.3d 666, 671–672 (D.C. Cir. 1994) (quoting\nKnight, 553 A.2d at 1234); Williams v. Mordofsky, 901 F.2d 158, 162\n(D.C. Cir. 1990) (quoting Knight, 553 A.2d at 1234).\n\f 11\nArea Transit Auth., 41 A.3d 526, 533 (D.C. 2012) (formatting\nmodified); accord Owens-Corning Fiberglass Corp. v. Henkel,\n689 A.2d 1224, 1231 (D.C. 1997); Arrington v. District of\nColumbia, 673 A.2d 674, 678 (D.C. 1996); Colbert v.\nGeorgetown Univ., 641 A.2d 469, 473 (D.C. 1994) (en banc).\nAnd because a prima facie case of tortious interference with\nbusiness relations or with reasonable expectation of\nprospective economic advantage requires that a plaintiff plead\n“intentional interference,” Jankovic v. International Crisis\nGroup, 593 F.3d 22, 29 (D.C. Cir. 2010); NCRIC, Inc. v.\nColumbia Hosp. for Women Med. Ctr., Inc., 957 A.2d 890, 900\n(D.C. 2008), Capitol Services reasons that the statute of\nlimitations did not begin running until it “received notice that\nVesta intentionally interfered with [its] expectation that it\nwould continue to perform property management services” at\nthe Property, Capitol Servs. Br. 34.\n\n Predicting whether the District’s Court of Appeals would\nfind light between the “some evidence of wrongdoing” and “all\nelements” tests would be no easy task. See Earle v. District of\nColumbia, 707 F.3d 299, 310 (D.C. Cir. 2012). There is\ncaselaw on both sides of the ledger. Compare Hardi v.\nMezzanotte, 818 A.2d 974, 979 (D.C. 2003), and East v.\nGraphic Arts Indus. Joint Pension Trust, 718 A.2d 153, 157\n(D.C. 1998), with Morton v. National Med. Enters., Inc., 725\nA.2d 462, 469 (D.C. 1999), and Cevenini, 707 A.2d at 771.\nUltimately, that is a job for another day. Even assuming the\n“some evidence of wrongdoing” standard is stricter for\nplaintiffs, Capitol Services’ lawsuit is timely.\n\n 2\n\n There is no dispute that, as of May 3, 2014, Capitol\nServices had “actual notice” of its injury—the cancellation of\nits property management contract. And Capitol Services knew\n\f 12\nthen that the injury was caused by the District’s decision to\naward an “emergency contract” to Vesta. See Vesta Br. 27–28;\nCapitol Servs. Br. 33–34. There is also no meaningful dispute\nthat Capitol Services lacked “actual notice” of Vesta’s\nwrongdoing as of that date. See Vesta Br. 28–31; Capitol\nServs. Br. 30, 36–37.\n\n The district court, instead, ruled that Capitol Services was\non inquiry notice as of May 3rd. Specifically, the district court\nreasoned that Capitol Services had “reason to suspect that\nVesta did some wrong” because it knew that (i) the District had\nissued an “emergency contract” to Vesta, (ii) the District\n“inappropriately terminated” Capitol Services’ agreement, and\n(iii) Capitol Services’ relationship with Park Southern “ended.”\nJ.A. 110–111.\n\n While those facts certainly put Capitol Services on notice\nof its claims against the District, there is nothing in the\ndescription of the District’s actions that hinted at Vesta’s role\nin the contract’s termination. “[P]laintiff’s knowledge of\nwrongdoing on the part of” the District itself “did not cause\naccrual of his action against another[.]” Diamond, 680 A.2d at\n380; see also Cevenini, 707 A.2d at 773 (“[K]nowledge of\nmisconduct on the part of one defendant will not automatically\ncreate inquiry notice of claims against a potential co-\ndefendant[.]”).\n\n Nor did the district court identify what evidence a “diligent\ninvestigation” at that point would have uncovered about\nVesta’s participation in the District’s decisionmaking.\nDiamond, 680 A.2d at 372; see Commonwealth Land, 922 F.3d\nat 464. Because Capitol Services had reported to Park\nSouthern, not the District, and because the District had declined\nto provide notice to Capitol Services of its ouster, nothing in\nthe initial complaint suggests that Capitol Services had any\n\f 13\nknowledge or even suspicions of background relationships or\nconnections between Vesta and the District government. Even\nthe written communication between the District and Vesta were\nby “private e-mails, which effectively circumvented FOIA\nscrutiny.” Complaint ¶ 17, J.A. 8. The complaint does not\n“conclusively establish” any facts uncoverable by Capitol\nServices on May 3, 2014, or even by August 28, 2014 (that is,\nthree years before the complaint against Vesta was first filed)\nthat would have implicated Vesta in its injury within the three-\nyear limitations period.\n\n Vesta argues, in the alternative, that Capitol Services was\non inquiry notice in July 2014, when it filed its complaint\nagainst the District in Superior Court. This is so, Vesta\nexplains, because the complaint repeatedly alleges Vesta’s role\nin inducing Capitol Services’ contract termination. See, e.g.,\nJ.A. 32 ¶ 18 (The District “terminated the contract and awarded\nit immediately to Vesta because of political and personal\nconnections on the pretense of an ‘emergency contract’ award.\nUpon information and belief, certain [District] representatives\njointly engaged in this decision. Said officials also\ncollaborated with Vesta regarding the government’s actions in\nadvance of [the District’s] May 3 * * * takeover of the\nproperty.”); J.A. 32 ¶ 19 (The District’s “purported\nqualification of the contract award to Vesta as an ‘Emergency’\nwas pretextual. It was intended to cause the removal of\n[Capitol Services] and the substitution of Vesta as the\nproperty’s property manager.”); J.A. 34 ¶ 29 (“Upon\ninformation and belief, * * * one or more [District] officials\nconsulted improperly and illegally with yet known [sic] Vesta\nofficials to achieve the termination of [Capitol Services’]\nAgreement and the award of an ‘emergency’ contract to Vesta\nin order to block [Capitol Services’] business opportunities.”);\nJ.A. 38 ¶ 48 (The District “impeded [Capitol Services’]\nbusiness relationship and prospective advantages/opportunities\n\f 14\nin order to advantage the interests of [Vesta], declared a non-\nexisting pretextual ‘emergency’ situation and awarded an\nillegal ‘emergency [contract]’ to [Vesta].”); J.A. 39 (asserting\na claim against the District for civil conspiracy); see also D.C.\nR. CIV. P. 11(b)(3) (imposing an obligation on counsel to assert\nonly factual contentions having evidentiary support).\n\n That would be a really good argument if Vesta were\ndiscussing the right version of the complaint. But those\nallegations appear nowhere in Capitol Services’ July 2014\ncomplaint against the District. The allegations in the July 2014\ncomplaint reference Vesta only in innocuous and fleeting\nfactual recitations. See Complaint ¶ 17, Capitol Servs. Mgmt.\nInc. v. District of Columbia, No. 2014 CA 004551 B (D.C.\nSuper. Ct. July 24, 2014) (The District “entered into an\n‘emergency contract’ with Vesta * * * on May 2, 2014, to\nauthorize Vesta to take over the management of Park Southern\nApartments on May 3, 2014.”); id. ¶ 18 (“In executing said\ncontract with Vesta, [the District] immediately terminated the\nmanagement agreement between [Park Southern] and [Capitol\nServices].”); id. ¶ 31.\n\n The allegations that Vesta cites and on which Vesta’s\nargument relies come from Capitol Services’ amended\nSuperior Court complaint. That was filed in October 2014—\nsquarely within the three-year period preceding the lawsuit\nagainst Vesta.\n\n Finally, in light of the allegations in its amended\ncomplaint, Capitol Services’ insistence that it had no notice\nuntil its deposition of Milton Bailey in July 2016 pushes things\ntoo far. Inquiry notice commences when, with due diligence, a\nplaintiff would have been on notice that a particular\ndefendant’s wrongdoing caused or contributed to a potential\ninjury. See, e.g., Diamond, 680 A.2d at 372. It does not wait\n\f 15\nuntil a plaintiff has actually assembled the evidence in\ndeposition form.\n\n So all that the limited record shows at this Rule 12(b)(6)\nstage is that Capitol Services was not on actual or inquiry\nnotice in either May or July of 2014, but was on inquiry notice\nat least by October 2014. The record is inconclusive as to\nCapitol Services’ knowledge about Vesta’s role prior to August\n28, 2014—three years before this suit against Vesta was filed.\nWhen during that intervening period Capitol Services had\ninquiry notice of Vesta’s potential role thus is an open factual\nquestion that cannot be resolved at this early juncture.\n\n IV\n\n This court can affirm a district court’s judgment on any\nbasis supported by the record. See United States v. Hicks, 911\nF.3d 623, 626 (D.C. Cir. 2018). Vesta asks this court to do just\nthat, arguing that even if the complaint is timely, Capitol\nServices nonetheless is collaterally estopped from relitigating\nthe question of alleged tortious interference with its contract.\nVesta points to the prior Superior Court judgment dismissing\nCapitol Services’ case against the District arising from\ntermination of the Property contract. See J.A. 43–49.\n\n Vesta fares no better on this front. The Superior Court held\nthat it lacked jurisdiction to adjudicate the suit against the\nDistrict, and so a central element for collateral estoppel is\nmissing.\n\n Collateral estoppel, also known as issue preclusion,\ngenerally confines plaintiffs to one bite at the litigation apple.\nSee B&B Hardware, Inc. v. Hargis Indus., Inc., 135 S. Ct.\n1293, 1302–1303 (2015). Specifically, collateral estoppel bars\nsuccessive litigation of an issue of fact or law when\n\f 16\n\n (1) the issue is actually litigated; (2) determined by a\n valid, final judgment on the merits; (3) after a full and\n fair opportunity for litigation by the parties or their\n privies; and (4) under circumstances where the\n determination was essential to the judgment, and not\n merely dictum.\n\nWalker v. FedEx Office & Print Servs., Inc., 123 A.3d 160, 164\n(D.C. 2015) (quoting Hogue v. Hopper, 728 A.2d 611, 614\n(D.C. 1999)). See generally Bode & Grenier, LLP v. Knight,\n808 F.3d 852, 857–858 (D.C. Cir. 2015) (applying preclusion\nrules of the jurisdiction in which the judgment was rendered).\n\n Vesta directs our attention to language in the Superior\nCourt’s 2017 decision stating that it found “no evidence” “that\nthe District of Columbia actually interfered with [Capitol\nServices’] contract or its business relationship with [Park\nSouthern] because neither the District of Columbia nor [Park\nSouthern] took any action to terminate it,” and instead “it\nappear[ed] to the court that [Capitol Services] voluntarily\ndecided not to enforce its claimed rights[.]” J.A. 48. Because\nCapitol Services failed to show any interference in that case,\nVesta argues, Capitol Services should be precluded from\nrelitigating whether it suffered tortious interference at Vesta’s\nhands.\n\n The Achilles heel of that argument is that, immediately\nbefore remarking on contract interference, the Superior Court\nruled that it could not adjudicate the lawsuit because the\nDistrict enjoyed sovereign immunity from suit, and so was\n“entitled to summary judgment” on that basis. J.A. 48. The\nDistrict’s sovereign immunity is a jurisdictional bar to\nlitigation—if it is not waived, the sovereign’s entitlement to\nimmunity deprives the court of subject matter jurisdiction over\n\f 17\nthe action. See Washington Metro. Area Transit Auth. v.\nBarksdale-Showell, 965 A.2d 16, 21 n.4 (D.C. 2009). So once\nthe Superior Court concluded that it was without jurisdiction,\nit could not “reach beyond that determination to consider\nwhether th[e plaintiff’s] claims would be meritorious if the\ndefect in [jurisdiction] were cured[.]” UMC Dev., LLC v.\nDistrict of Columbia, 120 A.3d 37, 49 (D.C. 2015).\n\n The absence of jurisdiction meant the Superior Court\ncould not and did not actually resolve “on the merits” Capitol\nServices’ tortious interference claim against the District, and so\nthe court’s comments on the subject were “merely dictum[,]”\nin no way “essential to the judgment[.]” Walker, 123 A.3d at\n164. 3 As a result, the critical elements of collateral estoppel\nhave not been established in this case.\n\n V\n\n Lastly, Vesta asks that we affirm on the ground that\nCapitol Services failed to state a claim on the merits for either\ntortious interference with business relations, or tortious\ninterference with reasonable expectation of prospective\neconomic advantage. Because the district court did not address\nthose questions and the parties have dedicated scant appellate\n\n 3\n See also United States ex rel. Bledsoe v. Community Health\nSys., Inc., 501 F.3d 493, 507 (6th Cir. 2007) (holding that the\nconclusion in a prior court’s opinion was “more properly\ncharacterized as dicta than an alternative holding[]” because the\ncourt had “concluded that it lacked subject matter jurisdiction”);\nMoreland v. Federal Bureau of Prisons, 431 F.3d 180, 185 (5th Cir.\n2005) (holding that the conclusion in a prior court’s opinion was\n“dicta because the petition in [that opinion] was dismissed for lack\nof subject-matter jurisdiction,” and was not an “alternative holding\nbecause it could not support the actual judgment in that case, which\nwas dismissal for lack of subject-matter jurisdiction”).\n\f 18\nbriefing to the issues, it would be unwise for us to wade in at\nthis time and on this record. The issue, of course, remains open\nfor the district court to address on remand.\n\n VI\n\n For the foregoing reasons, we reverse the district court’s\njudgment dismissing the case under Federal Rule of Civil\nProcedure 12(b)(6) on statute-of-limitations grounds and\nremand for further proceedings consistent with this opinion.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: CAPITOL SERVICES MANAGEMENT, INC., Appellant v. VESTA CORPORATION, Appellee\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-13\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 12, 2019 Decided August 13, 2019\\n\\n No. 18-7135\\n\\n CAPITOL SERVICES MANAGEMENT, INC.,\\n APPELLANT\\n\\n v.\\n\\n VESTA CORPORATION,\\n APPELLEE\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:17-cv-01756)\\n\\n\\n Donald M. Temple argued the cause and filed the briefs for\\nappellant.\\n\\n Lindsay D. Breedlove argued the cause and filed the brief\\nfor appellee.\\n\\n Before: TATEL, MILLETT, and KATSAS, Circuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge MILLETT.\\n\\n MILLETT, Circuit Judge: This case concerns the proper\\napplication of the discovery rule to tortious interference claims\\nunder District of Columbia law. The district court determined\\nthat the lawsuit filed by plaintiff Capitol Services Management,\\n\\f 2\\nInc. was barred by the statute of limitations because Capitol\\nServices was on “inquiry notice” of the defendant’s alleged\\ninterference with its contract long before the limitations period\\nexpired. But at the motion-to-dismiss stage, dismissal on\\nstatute-of-limitations grounds is permissible only if a plaintiff’s\\nclaims are conclusively time-barred on the face of the\\ncomplaint. That strict standard was not met here, so we reverse\\nand remand the case for further proceedings.\\n\\n I\\n\\n A\\n\\n The Park Southern building (the “Property”) is an\\napartment complex located in Southeast Washington, D.C. In\\n2006, the Park Southern Neighborhood Corporation (“Park\\nSouthern”) acquired the Property from its prior owner, the\\nDistrict of Columbia, subject to a Deed of Trust. Under the\\nterms of their agreement, if Park Southern defaulted on the\\nterms of the Deed, the District could resume control of the\\nProperty.\\n\\n Park Southern contracted with Vesta Corporation to serve\\nas the Property’s manager. However, Park Southern eventually\\nbecame dissatisfied with Vesta’s performance and, in March\\n2014, terminated its contract. Park Southern and Capitol\\nServices then entered into a one-year management agreement\\nthat would then continue on a “year-to-year basis,” unless\\neither party terminated it with three months’ notice. Complaint\\n¶¶ 9–10, J.A. 6.\\n\\n District officials began talking almost immediately about\\nPark Southern’s management reshuffle. Milton Bailey, a\\nDistrict employee, emailed the Mayor’s chief of staff advising\\nthat “‘[w]e already have enough to send default and foreclosure\\n\\f 3\\nnotices’ to effectively takeover [sic] the property.” Complaint\\n¶ 13, J.A. 7. In a telephone call in late March, Vesta relayed to\\nBailey its intent to “continue to manage the property and to\\nprovide [the District with] whatever assistance [it] need[s] with\\nrespect to the ongoing condition of the property.” Id. ¶ 16, J.A.\\n7. Vesta also communicated with the District by private email,\\nand participated in at least five other conference calls regarding\\nCapitol Services, the Property’s “management issues,” and\\nVesta’s “interests.” Id. ¶¶ 17–18, J.A. 8.\\n\\n Things came to a head on May 2, 2014, when the District\\nexercised its default remedy under the Deed of Trust and\\nimmediately took over the Property. That same day, without\\nproviding any notice to Capitol Services, the District entered\\ninto an “emergency contract” with Vesta, authorizing it to take\\nover property management duties from Capitol Services the\\nvery next day.\\n\\n Capitol Services filed suit against the District in Superior\\nCourt in July 2014. It filed an amended complaint against the\\nDistrict in October 2014, asserting (as relevant here) claims of\\ntortious interference with contract and with business\\nopportunity. In July 2016, as part of discovery in that action,\\nCapitol Services deposed Milton Bailey. In Capitol Services’\\nview, Bailey’s testimony “revealed considerable evidence” of\\nVesta’s alleged tortious interference. J.A. 57. Yet with respect\\nto its claims against the District, Capitol Services came up\\nshort. In a May 2017 opinion, the Superior Court entered\\njudgment for the District, finding that the court lacked\\njurisdiction because the District was entitled to sovereign\\nimmunity. The court also ruled that Capitol Services had failed\\nto establish a prima facie case of tortious interference by the\\nDistrict.\\n\\f 4\\n B\\n\\n On August 28, 2017, Capitol Services filed suit against\\nVesta in the United States District Court for the District of\\nColumbia. Its complaint asserted claims for tortious\\ninterference with business relations and tortious interference\\nwith reasonable expectation of prospective economic\\nadvantage. Vesta moved to dismiss, asserting that (i) Capitol\\nServices’ claims were time-barred; (ii) Capitol Services was\\ncollaterally estopped from relitigating issues decided in the\\nSuperior Court action; and (iii) Capitol Services failed to state\\nclaims for which relief could be granted.\\n\\n Reaching only the first issue, the district court agreed that\\nthe statute of limitations barred the suit. The court ruled that\\nthe District’s three-year statute of limitations for tortious\\ninterference claims began to run on May 3, 2014, when the\\nDistrict terminated Capitol Services’ contract and substituted\\nVesta in its place. Because Capitol Services did not file suit\\nuntil August 28, 2017, the court dismissed the lawsuit as\\nuntimely.\\n\\n In so ruling, the district court rejected Capitol Services’\\nrejoinder that the discovery rule delayed the start of the\\nlimitations period until July 2016, when Capitol Services\\ndeposed Milton Bailey as part of the Superior Court action. In\\nthe district court’s view, Capitol Services was on inquiry notice\\nof its claims against Vesta as soon as the District ended the\\ncontract because, at that point, Capitol Services had “reason to\\nsuspect that [Vesta] did some wrong.” J.A. 111. The court\\nadded that Capitol Services surely was on inquiry notice far\\nearlier than Bailey’s deposition because Capitol Services’\\nSuperior Court amended complaint against the District alleged\\nVesta’s involvement in terminating its management agreement.\\n\\f 5\\n Capitol Services timely appealed.\\n\\n II\\n\\n The district court had diversity jurisdiction under 28\\nU.S.C. § 1332(a)(1). This court’s jurisdiction rests on 28\\nU.S.C. § 1291.\\n\\n We review de novo a Rule 12(b)(6) dismissal on statute-\\nof-limitations grounds, accepting plaintiff’s well-pleaded\\nfactual allegations as true and drawing all reasonable\\ninferences in plaintiff’s favor. Momenian v. Davison, 878 F.3d\\n381, 387 (D.C. Cir. 2017). At the motion to dismiss stage,\\ndismissal on statute-of-limitations grounds is proper “only if\\nthe complaint on its face is conclusively time-barred.”\\nCommonwealth Land Title Ins. Co. v. KCI Techs., Inc., 922\\nF.3d 459, 464 (D.C. Cir. 2019); accord Firestone v. Firestone,\\n76 F.3d 1205, 1209 (D.C. Cir. 1996).\\n\\n Capitol Services argues that the district court effectively\\nconverted the case to summary judgment by considering its\\npleadings in the Superior Court action, and so the summary-\\njudgment standard of review should govern. But no such\\nconversion occurred. District courts may, at the Rule 12(b)(6)\\nstage, take judicial notice of publicly filed pleadings in related\\nactions as evidence of what was alleged in the other actions,\\nalthough they may not treat as true the matter alleged. Hurd v.\\nDistrict of Columbia, 864 F.3d 671, 686 (D.C. Cir. 2017); see\\nalso FED. R. EVID. 201(b).\\n\\n That is all the district court did here. The district court\\nproperly took notice of Capitol Services’ allegations in the\\nSuperior Court action not for their truth, but for the fact that\\nCapitol Services believed those allegations to be true and\\n\\f 6\\nviable as legal claims at the time they were made. See Watkins\\nv. United States, 854 F.3d 947, 949 (7th Cir. 2017).\\n\\n So this remains an appeal of a motion to dismiss under\\nFederal Rule of Civil Procedure 12(b)(6).\\n\\n III\\n\\n On appeal, Capitol Services asserts that its complaint\\nagainst Vesta was timely because the three-year statute of\\nlimitations did not begin to run until it had inquiry notice of\\n“all of the essential elements of its cause[s] of action” against\\nVesta. Capitol Servs. Br. 32. In Capitol Services’ view, that\\ndid not happen until it deposed Milton Bailey in July 2016.\\nCapitol Servs. Br. 33–39. From that deposition, reports Capitol\\nServices, it learned that Vesta had the motive or purpose to\\n“intentional[ly] interfere[]” with its business dealings\\nmanaging the Property. Capitol Servs. Br. 34–39. Vesta, for\\nits part, insists that Capitol Services had sufficient notice to\\ntrigger the statute of limitations at the earliest when it was\\nousted from its contract in May 2014 and Vesta took over, and\\nat the latest in July 2014 when it filed suit against the District.\\nSee Vesta Br. 23–33.\\n\\n As it happens, neither party is correct. As explained\\nbelow, inquiry notice was triggered sometime after the filing of\\nthe original July 2014 complaint against the District, but before\\nthe filing of the October 2014 amended complaint, which\\nreferenced Vesta’s interference. When exactly during that\\nthree-month period Capitol Services was on inquiry notice of\\nVesta’s alleged role is an unresolved factual question on this\\nrecord, making dismissal under Rule 12(b)(6) erroneous.\\n\\f 7\\n A\\n\\n Before reaching the merits, we first address Vesta’s\\nargument that Capitol Services did not properly preserve its\\nstatute-of-limitations argument for appellate review. We find\\nno merit to this objection.\\n\\n As an appellate court, “we are a court of review, not of first\\nview[.]” Cutter v. Wilkinson, 544 U.S. 709, 718 n.7 (2005).\\nFor that reason, absent “exceptional circumstances,” this court\\ndoes not entertain issues first raised on appeal. Woodruff v.\\nPeters, 482 F.3d 521, 525 (D.C. Cir. 2007).\\n\\n Vesta objects that, before the district court, Capitol\\nServices argued that it had inquiry notice only upon learning of\\nthe “depth and breadth of Vesta’s involvement and\\ncommunications with the District.” Vesta Br. 20 (quoting J.A.\\n57). According to Vesta, Capitol Services has shifted gears on\\nappeal in arguing that it lacked any knowledge of Vesta’s\\n“motive” or “intent” to interfere. Vesta Br. 20.\\n\\n That is a distinction without a difference. Both before the\\ndistrict court and this court, Capitol Services has argued that it\\nwas not on inquiry notice of Vesta’s alleged tortious\\ninterference at the time its contract was terminated in May 2014\\nand, indeed, until Bailey’s July 2016 deposition revealed\\ndetails about Vesta’s role. Although its emphasis has changed,\\nCapitol Services has argued consistently that the facts it learned\\nat that deposition were necessary for it to have inquiry notice\\nof its causes of action. See Yee v. City of Escondido, 503 U.S.\\n519, 534 (1992) (“Once a * * * claim is properly presented, a\\nparty can make any argument in support of that claim; parties\\nare not limited to the precise arguments they made below.”); In\\nre Harman Int’l Indus., Inc. Sec. Litig., 791 F.3d 90, 100 (D.C.\\nCir. 2015) (“On appeal, a party may refine and clarify its\\n\\f 8\\nanalysis in light of the district court’s ruling, including citing\\nadditional support.”) (formatting modified); Koch v. Cox, 489\\nF.3d 384, 391 (D.C. Cir. 2007) (A party may cite “additional\\nsupport for his side of an issue upon which the district court did\\nrule, much like citing a case for the first time on appeal.”).\\n\\n On to the merits.\\n\\n B\\n\\n We take as given the parties’ agreement that District of\\nColumbia law governs this tort dispute. See Patton Boggs LLP\\nv. Chevron Corp., 683 F.3d 397, 403 (D.C. Cir. 2012)\\n(applying District of Columbia law “which both parties assume\\napplies”). Under District law, Capitol Services’ claims for\\ntortious interference with business relations and tortious\\ninterference with reasonable expectation of prospective\\neconomic advantage are subject to a three-year statute of\\nlimitations. D.C. CODE § 12-301(8); see Beard v. Edmondson\\n& Gallagher, 790 A.2d 541, 546 (D.C. 2002); Carr v. Brown,\\n395 A.2d 79, 83 (D.C. 1978). See generally Hartford Accident\\n& Indem. Co. v. Pro-Football, Inc., 127 F.3d 1111, 1118 (D.C.\\nCir. 1997) (“Since [plaintiff’s] claims arise under District law,\\nthe applicable statute of limitations is also that of the\\nDistrict.”).\\n\\n Ordinarily under District law, the statute of limitations for\\na tort claim starts to run when the plaintiff is injured. See\\nAmobi v. District of Columbia Dep’t of Corr., 755 F.3d 980,\\n994 (D.C. Cir. 2014). All agree that Capitol Services was\\ninjured on May 3, 2014, when the District terminated its\\nmanagement contract. If that were the end of the story, basic\\nmath would render Capitol Services’ August 28, 2017 filing of\\nthis lawsuit untimely.\\n\\f 9\\n But Capitol Services contends, and Vesta does not\\nmeaningfully dispute, that the District of Columbia’s discovery\\nrule delays the start of the statute of limitations until Capitol\\nServices had actual or inquiry notice of its potential causes of\\naction against Vesta. See Momenian, 878 F.3d at 388. 1\\n\\n Where Capitol Services and Vesta part ways is over (i)\\nwhat “inquiry notice” demands, and (ii) whether it is\\n“conclusive” on the face of the complaint that Capitol Services\\nwas on inquiry notice of its claims against Vesta more than\\nthree years before it filed suit.\\n\\n 1\\n\\n A claim accrues under the discovery rule when the plaintiff\\n“either has actual knowledge of a cause of action” or is\\notherwise “charged with knowledge of that cause of action.”\\nCevenini v. Archbishop of Wash., 707 A.2d 768, 771 (D.C.\\n1998). The latter is known as “inquiry notice.” Inquiry notice\\nis “that notice which a plaintiff would have possessed after due\\ninvestigation.” Diamond v. Davis, 680 A.2d 364, 372 (D.C.\\n1996). It does not refer to “the amount of information that\\ntriggers a duty to investigate,” but rather the “quantum of\\nknowledge required to” “charge[] [a plaintiff] with knowledge\\n\\n 1\\n Vesta dedicated a measly three parenthetical words in its brief\\nto the applicability of the discovery rule. See Vesta Br. 27 (“In the\\ncase at bar, the district court * * * concluded that even if the\\ndiscovery rule applied (it does not), Capitol Services’ claim\\nnonetheless accrued in March 2014.”) (emphasis added). That was\\nnot enough to put the issue in controversy. See Davis v. Pension\\nBenefit Guar. Corp., 734 F.3d 1161, 1166–1167 (D.C. Cir. 2013) (A\\nparty may not “mention a possible argument in the most skeletal way,\\nleaving the court to do counsel’s work, create the ossature for the\\nargument, and put flesh on its bones.”) (internal quotation marks\\nomitted).\\n\\f 10\\nof [his] cause of action[.]” Id.; see BDO Seidman v. Morgan,\\nLewis & Bockius LLP, 89 A.3d 492, 500 (D.C. 2014). Under\\nDistrict of Columbia law, the time at which a plaintiff acquires\\ninquiry notice “is a question of fact.” Diamond, 680 A.2d at\\n372.\\n\\n Following multiple decisions of the District of Columbia\\nCourt of Appeals, this Court has repeatedly applied the\\ndiscovery rule to postpone the running of a statute of\\nlimitations until the plaintiff “knows or by the exercise of\\nreasonable diligence should know: (1) of the injury; (2) the\\ninjury’s cause in fact; and (3) of some evidence of\\nwrongdoing.” Commonwealth Land, 922 F.3d at 464 (quoting\\nCapital Place I Assocs. L.P. v. George Hyman Constr. Co., 673\\nA.2d 194, 199 (D.C. 1996)). 2\\n\\n Under that test, Capitol Services’ claims accrued when,\\nafter due investigation, it would have learned of its injury, the\\ninjury’s cause in fact, and “some evidence” of Vesta’s\\nwrongdoing. Commonwealth Land, 922 F.3d at 464. Yet as\\nCapitol Services notes, the District’s Court of Appeals has\\nsometimes articulated its discovery rule differently, stating that\\nclaims do not accrue until the plaintiff has “discovered or\\nreasonably should have discovered all of the essential elements\\nof his possible cause of action.” Price v. Washington Metro.\\n\\n 2\\n Accord Hancock v. HomEq Servicing Corp., 526 F.3d 785,\\n786 (D.C. Cir. 2008) (quoting Diamond, 680 A.2d at 381); Bradley\\nv. National Ass’n of Sec. Dealers Dispute Resolution, Inc., 433 F.3d\\n846, 849 (D.C. Cir. 2005) (quoting Bussineau v. President & Dirs.\\nof Georgetown College, 518 A.2d 423, 425 (D.C. 1986)); Jung v.\\nMundy, Holt & Mance, P.C., 372 F.3d 429, 433 (D.C. Cir. 2004)\\n(quoting Knight v. Furlow, 553 A.2d 1232, 1234 (D.C. 1989));\\nGoldman v. Bequai, 19 F.3d 666, 671–672 (D.C. Cir. 1994) (quoting\\nKnight, 553 A.2d at 1234); Williams v. Mordofsky, 901 F.2d 158, 162\\n(D.C. Cir. 1990) (quoting Knight, 553 A.2d at 1234).\\n\\f 11\\nArea Transit Auth., 41 A.3d 526, 533 (D.C. 2012) (formatting\\nmodified); accord Owens-Corning Fiberglass Corp. v. Henkel,\\n689 A.2d 1224, 1231 (D.C. 1997); Arrington v. District of\\nColumbia, 673 A.2d 674, 678 (D.C. 1996); Colbert v.\\nGeorgetown Univ., 641 A.2d 469, 473 (D.C. 1994) (en banc).\\nAnd because a prima facie case of tortious interference with\\nbusiness relations or with reasonable expectation of\\nprospective economic advantage requires that a plaintiff plead\\n“intentional interference,” Jankovic v. International Crisis\\nGroup, 593 F.3d 22, 29 (D.C. Cir. 2010); NCRIC, Inc. v.\\nColumbia Hosp. for Women Med. Ctr., Inc., 957 A.2d 890, 900\\n(D.C. 2008), Capitol Services reasons that the statute of\\nlimitations did not begin running until it “received notice that\\nVesta intentionally interfered with [its] expectation that it\\nwould continue to perform property management services” at\\nthe Property, Capitol Servs. Br. 34.\\n\\n Predicting whether the District’s Court of Appeals would\\nfind light between the “some evidence of wrongdoing” and “all\\nelements” tests would be no easy task. See Earle v. District of\\nColumbia, 707 F.3d 299, 310 (D.C. Cir. 2012). There is\\ncaselaw on both sides of the ledger. Compare Hardi v.\\nMezzanotte, 818 A.2d 974, 979 (D.C. 2003), and East v.\\nGraphic Arts Indus. Joint Pension Trust, 718 A.2d 153, 157\\n(D.C. 1998), with Morton v. National Med. Enters., Inc., 725\\nA.2d 462, 469 (D.C. 1999), and Cevenini, 707 A.2d at 771.\\nUltimately, that is a job for another day. Even assuming the\\n“some evidence of wrongdoing” standard is stricter for\\nplaintiffs, Capitol Services’ lawsuit is timely.\\n\\n 2\\n\\n There is no dispute that, as of May 3, 2014, Capitol\\nServices had “actual notice” of its injury—the cancellation of\\nits property management contract. And Capitol Services knew\\n\\f 12\\nthen that the injury was caused by the District’s decision to\\naward an “emergency contract” to Vesta. See Vesta Br. 27–28;\\nCapitol Servs. Br. 33–34. There is also no meaningful dispute\\nthat Capitol Services lacked “actual notice” of Vesta’s\\nwrongdoing as of that date. See Vesta Br. 28–31; Capitol\\nServs. Br. 30, 36–37.\\n\\n The district court, instead, ruled that Capitol Services was\\non inquiry notice as of May 3rd. Specifically, the district court\\nreasoned that Capitol Services had “reason to suspect that\\nVesta did some wrong” because it knew that (i) the District had\\nissued an “emergency contract” to Vesta, (ii) the District\\n“inappropriately terminated” Capitol Services’ agreement, and\\n(iii) Capitol Services’ relationship with Park Southern “ended.”\\nJ.A. 110–111.\\n\\n While those facts certainly put Capitol Services on notice\\nof its claims against the District, there is nothing in the\\ndescription of the District’s actions that hinted at Vesta’s role\\nin the contract’s termination. “[P]laintiff’s knowledge of\\nwrongdoing on the part of” the District itself “did not cause\\naccrual of his action against another[.]” Diamond, 680 A.2d at\\n380; see also Cevenini, 707 A.2d at 773 (“[K]nowledge of\\nmisconduct on the part of one defendant will not automatically\\ncreate inquiry notice of claims against a potential co-\\ndefendant[.]”).\\n\\n Nor did the district court identify what evidence a “diligent\\ninvestigation” at that point would have uncovered about\\nVesta’s participation in the District’s decisionmaking.\\nDiamond, 680 A.2d at 372; see Commonwealth Land, 922 F.3d\\nat 464. Because Capitol Services had reported to Park\\nSouthern, not the District, and because the District had declined\\nto provide notice to Capitol Services of its ouster, nothing in\\nthe initial complaint suggests that Capitol Services had any\\n\\f 13\\nknowledge or even suspicions of background relationships or\\nconnections between Vesta and the District government. Even\\nthe written communication between the District and Vesta were\\nby “private e-mails, which effectively circumvented FOIA\\nscrutiny.” Complaint ¶ 17, J.A. 8. The complaint does not\\n“conclusively establish” any facts uncoverable by Capitol\\nServices on May 3, 2014, or even by August 28, 2014 (that is,\\nthree years before the complaint against Vesta was first filed)\\nthat would have implicated Vesta in its injury within the three-\\nyear limitations period.\\n\\n Vesta argues, in the alternative, that Capitol Services was\\non inquiry notice in July 2014, when it filed its complaint\\nagainst the District in Superior Court. This is so, Vesta\\nexplains, because the complaint repeatedly alleges Vesta’s role\\nin inducing Capitol Services’ contract termination. See, e.g.,\\nJ.A. 32 ¶ 18 (The District “terminated the contract and awarded\\nit immediately to Vesta because of political and personal\\nconnections on the pretense of an ‘emergency contract’ award.\\nUpon information and belief, certain [District] representatives\\njointly engaged in this decision. Said officials also\\ncollaborated with Vesta regarding the government’s actions in\\nadvance of [the District’s] May 3 * * * takeover of the\\nproperty.”); J.A. 32 ¶ 19 (The District’s “purported\\nqualification of the contract award to Vesta as an ‘Emergency’\\nwas pretextual. It was intended to cause the removal of\\n[Capitol Services] and the substitution of Vesta as the\\nproperty’s property manager.”); J.A. 34 ¶ 29 (“Upon\\ninformation and belief, * * * one or more [District] officials\\nconsulted improperly and illegally with yet known [sic] Vesta\\nofficials to achieve the termination of [Capitol Services’]\\nAgreement and the award of an ‘emergency’ contract to Vesta\\nin order to block [Capitol Services’] business opportunities.”);\\nJ.A. 38 ¶ 48 (The District “impeded [Capitol Services’]\\nbusiness relationship and prospective advantages/opportunities\\n\\f 14\\nin order to advantage the interests of [Vesta], declared a non-\\nexisting pretextual ‘emergency’ situation and awarded an\\nillegal ‘emergency [contract]’ to [Vesta].”); J.A. 39 (asserting\\na claim against the District for civil conspiracy); see also D.C.\\nR. CIV. P. 11(b)(3) (imposing an obligation on counsel to assert\\nonly factual contentions having evidentiary support).\\n\\n That would be a really good argument if Vesta were\\ndiscussing the right version of the complaint. But those\\nallegations appear nowhere in Capitol Services’ July 2014\\ncomplaint against the District. The allegations in the July 2014\\ncomplaint reference Vesta only in innocuous and fleeting\\nfactual recitations. See Complaint ¶ 17, Capitol Servs. Mgmt.\\nInc. v. District of Columbia, No. 2014 CA 004551 B (D.C.\\nSuper. Ct. July 24, 2014) (The District “entered into an\\n‘emergency contract’ with Vesta * * * on May 2, 2014, to\\nauthorize Vesta to take over the management of Park Southern\\nApartments on May 3, 2014.”); id. ¶ 18 (“In executing said\\ncontract with Vesta, [the District] immediately terminated the\\nmanagement agreement between [Park Southern] and [Capitol\\nServices].”); id. ¶ 31.\\n\\n The allegations that Vesta cites and on which Vesta’s\\nargument relies come from Capitol Services’ amended\\nSuperior Court complaint. That was filed in October 2014—\\nsquarely within the three-year period preceding the lawsuit\\nagainst Vesta.\\n\\n Finally, in light of the allegations in its amended\\ncomplaint, Capitol Services’ insistence that it had no notice\\nuntil its deposition of Milton Bailey in July 2016 pushes things\\ntoo far. Inquiry notice commences when, with due diligence, a\\nplaintiff would have been on notice that a particular\\ndefendant’s wrongdoing caused or contributed to a potential\\ninjury. See, e.g., Diamond, 680 A.2d at 372. It does not wait\\n\\f 15\\nuntil a plaintiff has actually assembled the evidence in\\ndeposition form.\\n\\n So all that the limited record shows at this Rule 12(b)(6)\\nstage is that Capitol Services was not on actual or inquiry\\nnotice in either May or July of 2014, but was on inquiry notice\\nat least by October 2014. The record is inconclusive as to\\nCapitol Services’ knowledge about Vesta’s role prior to August\\n28, 2014—three years before this suit against Vesta was filed.\\nWhen during that intervening period Capitol Services had\\ninquiry notice of Vesta’s potential role thus is an open factual\\nquestion that cannot be resolved at this early juncture.\\n\\n IV\\n\\n This court can affirm a district court’s judgment on any\\nbasis supported by the record. See United States v. Hicks, 911\\nF.3d 623, 626 (D.C. Cir. 2018). Vesta asks this court to do just\\nthat, arguing that even if the complaint is timely, Capitol\\nServices nonetheless is collaterally estopped from relitigating\\nthe question of alleged tortious interference with its contract.\\nVesta points to the prior Superior Court judgment dismissing\\nCapitol Services’ case against the District arising from\\ntermination of the Property contract. See J.A. 43–49.\\n\\n Vesta fares no better on this front. The Superior Court held\\nthat it lacked jurisdiction to adjudicate the suit against the\\nDistrict, and so a central element for collateral estoppel is\\nmissing.\\n\\n Collateral estoppel, also known as issue preclusion,\\ngenerally confines plaintiffs to one bite at the litigation apple.\\nSee B&B Hardware, Inc. v. Hargis Indus., Inc., 135 S. Ct.\\n1293, 1302–1303 (2015). Specifically, collateral estoppel bars\\nsuccessive litigation of an issue of fact or law when\\n\\f 16\\n\\n (1) the issue is actually litigated; (2) determined by a\\n valid, final judgment on the merits; (3) after a full and\\n fair opportunity for litigation by the parties or their\\n privies; and (4) under circumstances where the\\n determination was essential to the judgment, and not\\n merely dictum.\\n\\nWalker v. FedEx Office & Print Servs., Inc., 123 A.3d 160, 164\\n(D.C. 2015) (quoting Hogue v. Hopper, 728 A.2d 611, 614\\n(D.C. 1999)). See generally Bode & Grenier, LLP v. Knight,\\n808 F.3d 852, 857–858 (D.C. Cir. 2015) (applying preclusion\\nrules of the jurisdiction in which the judgment was rendered).\\n\\n Vesta directs our attention to language in the Superior\\nCourt’s 2017 decision stating that it found “no evidence” “that\\nthe District of Columbia actually interfered with [Capitol\\nServices’] contract or its business relationship with [Park\\nSouthern] because neither the District of Columbia nor [Park\\nSouthern] took any action to terminate it,” and instead “it\\nappear[ed] to the court that [Capitol Services] voluntarily\\ndecided not to enforce its claimed rights[.]” J.A. 48. Because\\nCapitol Services failed to show any interference in that case,\\nVesta argues, Capitol Services should be precluded from\\nrelitigating whether it suffered tortious interference at Vesta’s\\nhands.\\n\\n The Achilles heel of that argument is that, immediately\\nbefore remarking on contract interference, the Superior Court\\nruled that it could not adjudicate the lawsuit because the\\nDistrict enjoyed sovereign immunity from suit, and so was\\n“entitled to summary judgment” on that basis. J.A. 48. The\\nDistrict’s sovereign immunity is a jurisdictional bar to\\nlitigation—if it is not waived, the sovereign’s entitlement to\\nimmunity deprives the court of subject matter jurisdiction over\\n\\f 17\\nthe action. See Washington Metro. Area Transit Auth. v.\\nBarksdale-Showell, 965 A.2d 16, 21 n.4 (D.C. 2009). So once\\nthe Superior Court concluded that it was without jurisdiction,\\nit could not “reach beyond that determination to consider\\nwhether th[e plaintiff’s] claims would be meritorious if the\\ndefect in [jurisdiction] were cured[.]” UMC Dev., LLC v.\\nDistrict of Columbia, 120 A.3d 37, 49 (D.C. 2015).\\n\\n The absence of jurisdiction meant the Superior Court\\ncould not and did not actually resolve “on the merits” Capitol\\nServices’ tortious interference claim against the District, and so\\nthe court’s comments on the subject were “merely dictum[,]”\\nin no way “essential to the judgment[.]” Walker, 123 A.3d at\\n164. 3 As a result, the critical elements of collateral estoppel\\nhave not been established in this case.\\n\\n V\\n\\n Lastly, Vesta asks that we affirm on the ground that\\nCapitol Services failed to state a claim on the merits for either\\ntortious interference with business relations, or tortious\\ninterference with reasonable expectation of prospective\\neconomic advantage. Because the district court did not address\\nthose questions and the parties have dedicated scant appellate\\n\\n 3\\n See also United States ex rel. Bledsoe v. Community Health\\nSys., Inc., 501 F.3d 493, 507 (6th Cir. 2007) (holding that the\\nconclusion in a prior court’s opinion was “more properly\\ncharacterized as dicta than an alternative holding[]” because the\\ncourt had “concluded that it lacked subject matter jurisdiction”);\\nMoreland v. Federal Bureau of Prisons, 431 F.3d 180, 185 (5th Cir.\\n2005) (holding that the conclusion in a prior court’s opinion was\\n“dicta because the petition in [that opinion] was dismissed for lack\\nof subject-matter jurisdiction,” and was not an “alternative holding\\nbecause it could not support the actual judgment in that case, which\\nwas dismissal for lack of subject-matter jurisdiction”).\\n\\f 18\\nbriefing to the issues, it would be unwise for us to wade in at\\nthis time and on this record. The issue, of course, remains open\\nfor the district court to address on remand.\\n\\n VI\\n\\n For the foregoing reasons, we reverse the district court’s\\njudgment dismissing the case under Federal Rule of Civil\\nProcedure 12(b)(6) on statute-of-limitations grounds and\\nremand for further proceedings consistent with this opinion.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4176,"cells":{"id":{"kind":"number","value":7336847,"string":"7,336,847"},"name":{"kind":"string","value":"COMMITTEE ON the JUDICIARY, United States House of Representatives v. Donald F. MCGAHN II"},"name_abbreviation":{"kind":"string","value":"Committee on Judiciary v. McGahn"},"decision_date":{"kind":"string","value":"2019-08-14"},"docket_number":{"kind":"string","value":"Civil Action No. 19-2379 (BAH)"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Howell\", \"parties\": \"\", \"opinions\": [{\"author\": \"\", \"type\": \"020lead\", \"text\": \"\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/7254767/\", \"author_raw\": \"\"}]}"},"judge1":{"kind":"string","value":"HOWELL"},"judge2":{"kind":"string","value":"HOWELL"},"judge3":{"kind":"string","value":""},"exactly_three_caps":{"kind":"number","value":0,"string":"0"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/7336847/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":0,"string":"0"},"020lead_count":{"kind":"number","value":1,"string":"1"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: COMMITTEE ON the JUDICIARY, United States House of Representatives v. Donald F. MCGAHN II\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-14\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\n\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: COMMITTEE ON the JUDICIARY, United States House of Representatives v. Donald F. MCGAHN II\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-14\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\n\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4177,"cells":{"id":{"kind":"number","value":4648997,"string":"4,648,997"},"name":{"kind":"string","value":"TWIN RIVERS PAPER COMPANY LLC, Et Al., Petitioners v. SECURITIES AND EXCHANGE COMMISSION, Respondent"},"name_abbreviation":{"kind":"string","value":"Twin Rivers Paper Co. v. SEC. & Exch. Comm'n"},"decision_date":{"kind":"string","value":"2019-08-16"},"docket_number":{"kind":"string","value":"18-1213"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Henderson, Rogers, Katsas\", \"parties\": \"\", \"opinions\": [{\"author\": \"KATSAS, Circuit Judge\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued March 12, 2019 Decided August 16, 2019\\n\\n No. 18-1213\\n\\n TWIN RIVERS PAPER COMPANY LLC, ET AL.,\\n PETITIONERS\\n\\n v.\\n\\n SECURITIES AND EXCHANGE COMMISSION,\\n RESPONDENT\\n\\n\\n On Petition for Review of an Order of\\n the Securities & Exchange Commission\\n\\n\\n Jane C. Luxton argued the cause and filed the briefs for\\npetitioners.\\n\\n Alan W. Bakowski was on the brief for amici curiae Public\\nand Private Companies, Nonprofit Organizations, and Labor\\nUnion in support of petitioners.\\n\\n Tracey A. Hardin, Assistant General Counsel, Securities\\nand Exchange Commission, argued the cause for respondent.\\nWith her on the brief were Michael A. Conley, Solicitor, and\\nDaniel E. Matro, Senior Counsel.\\n\\n Eugene Scalia, Jacob T. Spencer, and Paul S. Stevens were\\non the brief for amici curiae Investment Company Institute and\\nIndependent Directors Council in support of respondent.\\n\\f 2\\n\\n Before: HENDERSON, ROGERS, and KATSAS, Circuit\\nJudges.\\n\\n KATSAS, Circuit Judge: In 2018, the Securities and\\nExchange Commission adopted a rule allowing investment\\ncompanies to post shareholder reports online and mail paper\\ncopies to shareholders upon request. The petitioners—a\\nconsumer-advocacy organization and representatives of the\\npaper industry—argue that the SEC did not adequately\\nconsider the interests of shareholders who prefer reports in\\npaper form. Because the consumer organization lacks\\nconstitutional standing and the paper-industry representatives\\nassert interests beyond those protected or regulated by the\\nsecurities laws, we deny the petition for review.\\n\\n I\\n\\n The Securities and Exchange Commission requires\\ninvestment companies, such as mutual funds, to transmit\\nperiodic shareholder reports to their investors. See, e.g., 17\\nC.F.R. § 270.30e-1. Previously, the SEC required the\\ncompanies to mail paper copies unless an investor affirmatively\\nselected electronic delivery. In 2018, the Commission adopted\\nRule 30e-3, which allows companies to change their default\\nmethod of transmission. See Optional Internet Availability of\\nInvestment Company Shareholder Reports, 83 Fed. Reg.\\n29,158 (June 22, 2018) (Shareholder Reports). The rule now\\npermits investment funds to post shareholder reports online and\\nnotify investors of their availability. See 17 C.F.R. § 270.30e-\\n3(b)–(d). Investment companies must mail shareholder reports\\nonly to investors who expressly request a paper copy. See id.\\n§ 270.30e-3(e).\\n\\f 3\\n The SEC gave two principal rationales for this change.\\nFirst, it projected that the rule would save investment funds\\nabout $140 million annually. Shareholder Reports, 83 Fed.\\nReg. at 29,187. It explained that because printing costs are paid\\nfrom fund assets, these savings ordinarily will be “passed along\\nto investors.” Id. at 29,183. Second, the Commission\\nconcluded that “many investors would prefer enhanced\\navailability of fund information on the internet.” Id. at 29,165.\\nIt explained that “an investor looking for a fund’s annual report\\nis most likely to seek it out on the fund’s website, rather than\\nrequest it by mail or phone.” Id. at 29,165 n.96. Moreover,\\ninternet usage “has continued to increase rapidly” over the last\\ndecade, including among “households owning mutual funds.”\\nId. at 29,165 n.97. In short, the SEC expected the rule to match\\nshareholder preferences and save them money.\\n\\n The Commission recognized that some investors may still\\nprefer paper delivery of shareholder reports. See, e.g.,\\nShareholder Reports, 83 Fed. Reg. at 29,165–66. Several\\nfeatures of Rule 30e-3 accommodate that preference. The rule\\nprescribes an extended transition period: investment\\ncompanies must continue to deliver paper copies of the reports\\nuntil at least January 1, 2021, and they must include with each\\nmailing a statement advising shareholders of the coming\\nchange in the default mode of transmission. See 17 C.F.R.\\n§ 270.30e-3(i). Then, after making the switch, a fund must\\nmail a paper notice to all shareholders for each report that it\\nposts online. See id. § 270.30e-3(c). The notice must include\\na toll-free telephone number that investors may call to request\\na paper copy of the individual report or all future reports. See\\nid. § 270.30e-3(c)(1)(i)–(v). The notice also may specify other\\nways to request paper reports, such as by e-mail or online. See\\nid. § 270.30e-3(c)(2). Finally, once an investor requests a\\npaper copy, the fund must mail the report within three business\\ndays at no cost to the investor. Id. § 270.30e-3(e).\\n\\f 4\\n This case presents two sets of petitioners. Consumer\\nAction is a non-profit membership organization that represents\\ncertain consumer interests. Twin Rivers Paper Company and\\nthree industry organizations—which we call the Industry\\nPetitioners—represent the interests of the American paper\\nindustry. Together, the petitioners argue that the SEC adopted\\nRule 30e-3 in violation of three securities laws and the\\nAdministrative Procedure Act. Among other grounds, they\\ncontend that the Commission did not adequately protect\\nshareholders who prefer paper delivery.\\n\\n II\\n\\n We begin with the question whether Consumer Action has\\nconstitutional standing to challenge Rule 30e-3. The\\nConstitution limits the “judicial Power of the United States” to\\n“Cases” or “Controversies,” U.S. Const. art. III, §§ 1–2, and\\nthe requirement of standing is “rooted in the traditional\\nunderstanding of a case or controversy,” Spokeo, Inc. v. Robins,\\n136 S. Ct. 1540, 1547 (2016). To establish standing under\\nArticle III, a party “must have (1) suffered an injury in fact,\\n(2) that is fairly traceable to the challenged conduct of the\\ndefendant, and (3) that is likely to be redressed by a favorable\\njudicial decision.” Id.\\n\\n Consumer Action claims standing on behalf of its\\nmembers. To establish representational standing, an\\norganization must prove, among other things, that its members\\nwould “have standing to sue in their own right.” Hunt v. Wash.\\nState Apple Advert. Comm’n, 432 U.S. 333, 343 (1977). In\\nturn, this requires proof that at least one member suffered an\\ninjury in fact—“an invasion of a legally protected interest\\nwhich is (a) concrete and particularized; and (b) actual or\\nimminent, not conjectural or hypothetical.” Lujan v. Defs. of\\nWildlife, 504 U.S. 555, 560 (1992) (cleaned up). “A ‘concrete’\\n\\f 5\\ninjury must be ‘de facto’; that is, it must actually exist.”\\nSpokeo, 136 S. Ct. at 1548. Moreover, a party cannot rest on\\n“abstract,” id., or “conclusory” assertions of injury, Block v.\\nMeese, 793 F.2d 1303, 1308 (D.C. Cir. 1986), but must point\\nto “specific, concrete facts demonstrating … harm,” Warth v.\\nSeldin, 422 U.S. 490, 508 (1975).\\n\\n In Sierra Club v. EPA, 292 F.3d 895 (D.C. Cir. 2002), we\\nexplained the procedures for proving standing when a\\npetitioner seeks review of agency action directly in a court of\\nappeals. “The party invoking federal jurisdiction bears the\\nburden of establishing” standing, which “must be supported in\\nthe same way as any other matter on which the [party] bears\\nthe burden of proof.” Defs. of Wildlife, 504 U.S. at 561. So a\\npetitioner “must either identify in [the administrative] record\\nevidence sufficient to support its standing to seek review or, if\\nthere is none because standing was not an issue before the\\nagency, submit additional evidence to the court of appeals.”\\nSierra Club, 292 F.3d at 899. Moreover, because “full\\ndevelopment of the arguments for and against standing requires\\nthe same tried and true adversarial procedure we use for the\\npresentation of arguments on the merits,” the petitioner must\\nmake this evidentiary presentation no later than when it files\\nthe opening brief. Id. at 900. We have reiterated these\\nprinciples many times. See, e.g., Del. Dep’t of Nat. Res. &\\nEnvt’l Control v. EPA, 785 F.3d 1, 8 (D.C. Cir. 2015); Texas v.\\nEPA, 726 F.3d 180, 198 (D.C. Cir. 2013); Int’l Bhd. of\\nTeamsters v. TSA, 429 F.3d 1130, 1134–36 (D.C. Cir. 2005).\\nAnd we have codified them in our Circuit Rule 28(a)(7), which\\nprovides that, “[i]n cases involving direct review in this court\\nof administrative actions, the brief of the appellant or petitioner\\nmust set forth the basis for the claim of standing,” and that,\\n“[w]hen the appellant’s or petitioner’s standing is not apparent\\nfrom the administrative record, the brief must include\\narguments and evidence establishing the claim of standing.”\\n\\f 6\\n Consumer Action’s initial submissions fail to show that its\\nmembers suffered or will suffer an injury in fact. To establish\\nsuch an injury, an organization must provide “individual\\naffidavits” from “members who have suffered the requisite\\nharm.” Summers v. Earth Island Inst., 555 U.S. 488, 499\\n(2009). It is “not enough to aver that unidentified members\\nhave been injured.” Chamber of Commerce v. EPA, 642 F.3d\\n192, 199 (D.C. Cir. 2011). Despite these settled rules,\\nConsumer Action failed to submit any member affidavits with\\nits opening brief, and its own affidavit fails to identify any\\nindividual members. Moreover, the affidavit barely describes\\neven the general contours of its membership. The affidavit\\nstates that “members, followers and supporters include retirees\\nand retirement savers,” as well as “representatives of a national\\nnetwork of nearly 7,000 community-based organizations that\\nserve seniors, minority Americans, disabled Americans, and\\nindividuals living in rural areas.” Pet. Add. at 433. Left\\nunclear is whether the members are retirees and retirement\\nsavers, members of other unnamed organizations, or perhaps\\nneither. To be sure, the opening brief states that Consumer\\nAction’s “members” include seniors, Americans who lack\\ninternet access, and others. Pet. Br. at 17. But the brief still\\nfails to identify individual members, as required by Earth\\nIsland Institute. And in any event, a petitioner must support its\\nstanding “by affidavit or other evidence,” Sierra Club, 292\\nF.3d at 899 (quotation marks omitted), and briefs “are not\\nevidence,” id. at 901. Finally, the affidavit states only that\\nunspecified members or others “prefer a choice to have paper\\ncommunications” and wish to avoid the “burdens imposed by\\nSEC’s Rule 30e-3.” Pet. Add. at 433. The former statement is\\npuzzling because Rule 30e-3 preserves investors’ ability to\\nchoose between electronic and paper communications. As for\\nthe latter statement, “general allegations of injury are\\ninsufficient,” Conservation Force, Inc. v. Jewell, 733 F.3d\\n1200, 1207 (D.C. Cir. 2013), so a reference to unspecified\\n\\f 7\\n“burdens” is not proof of “direct, real, and palpable” injury,\\nFood & Water Watch, Inc. v. Vilsack, 808 F.3d 905, 914 (D.C.\\nCir. 2015) (quotation marks omitted).\\n\\n Likewise, nothing in the administrative record shows a\\nconcrete injury to identified members of Consumer Action.\\nThe organization notes that, in various comments made during\\nthe rulemaking, it warned that Rule 30e-3 would decrease\\nreadership of shareholder reports, investor access to\\ninformation, and transparency. But none of the comments tied\\nthese harms to any identified members. Consumer Action\\nfurther notes that the SEC acknowledged its comments that the\\nrule would disadvantage seniors and minorities. See\\nShareholder Reports, 83 Fed. Reg. at 29,162 n.51. But again,\\nneither the comments themselves, nor the SEC’s response,\\naddressed whether the harms alleged would befall identified\\nmembers of Consumer Action. This should hardly be\\nsurprising: Article III standing requirements do not apply to\\nagency rulemaking, so Consumer Action would have had no\\noccasion to argue, and the SEC would have had no occasion to\\ndecide, whether the proposed rule would inflict an injury on\\nidentified members of the organization. And because\\nConsumer Action was not itself an object of the rulemaking, its\\nstanding would not likely have been apparent. See Sierra Club,\\n292 F.3d at 899–900.\\n\\n Consumer Action submitted affidavits from individual\\nmembers with its reply brief, but they came too late. We may\\nexcuse forfeitures of non-jurisdictional preservation\\nrequirements for “good cause.” Sierra Club, 292 F.3d at 900.\\nIn the context of Sierra Club and Circuit Rule 28(a)(7), we\\nhave found such good cause in two circumstances: where “the\\nparties reasonably, but mistakenly, believed that the initial\\nfilings before the court had sufficiently demonstrated\\nstanding,” Ctr. for Sustainable Econ. v. Jewell, 779 F.3d 588,\\n\\f 8\\n599 (D.C. Cir. 2015) (quoting Ams. for Safe Access v. DEA,\\n706 F.3d 438, 443 (D.C. Cir. 2013)); and where the parties\\n“reasonably assumed that [their] standing was self-evident”\\nfrom the administrative record, Am. Library Ass’n v. FCC, 401\\nF.3d 489, 494 (D.C. Cir. 2005); see Del. Dep’t of Nat. Res. &\\nEnvt’l Control, 785 F.3d at 8; Town of Barnstable v. FAA, 740\\nF.3d 681, 691 (D.C. Cir. 2014). In this case, Consumer Action\\ncould not reasonably have believed that its barebones affidavit,\\nvaguely describing the preferences and burdens of unnamed\\nmembers and others, sufficed to prove its representational\\nstanding. Nor could it reasonably have believed that its\\nstanding was self-evident from the rulemaking record.\\n\\n Consumer Action suggests that Communities Against\\nRunway Expansion, Inc. v. FAA, 355 F.3d 678 (D.C. Cir. 2004)\\n(CARE), allows a petitioner to prove standing for the first time\\nin a reply brief, so long as standing is obvious. We do not read\\nCARE that broadly. There, an organization challenged the\\napproval of a runway expansion at Logan Airport. With its\\nopening brief, the organization submitted affidavits from\\nmembers who lived nearby, but the affidavits lacked “facts\\nsufficient to support a finding that the declarants would be\\nexposed to a concrete and particularized ‘injury in fact’ as a\\nresult of the contested project.” Id. at 684. With its reply brief,\\nthe organization submitted additional affidavits from other\\nmembers who declared that they would experience “increased\\nnoise from aircraft operations at Logan.” Id. CARE thus\\ninvolved new factual material tendered to shore up deficient\\nindividual affidavits submitted with the opening brief. In that\\ncircumstance, we considered the reply affidavits—which, we\\nstressed, made the organization’s standing “patently obvious”\\nand “irrefutable.” Id. at 685.\\n\\n This case differs from CARE in three critical respects.\\nFirst, Consumer Action made a far less substantial showing of\\n\\f 9\\nstanding in its initial submissions. In CARE, the organization\\ntendered affidavits from members who lived near Logan\\nAirport and alleged injury in at least general terms. See 355\\nF.3d at 684. So, there was at least arguably good cause to\\nexcuse the forfeiture. See Ctr. for Sustainable Econ., 779 F.3d\\nat 599. Here, in contrast, Consumer Action tendered no\\nmember affidavits, identified no specific members, and failed\\nto describe its membership in even general terms. This cannot\\namount to good cause.\\n\\n Second, the reply affidavits in this case raise an entirely\\nnew theory of standing. In CARE, the injury pinned down in\\nthe reply affidavits—increased airport noise—was hardly\\nsurprising in light of initial submissions from individuals living\\nnear an airport and complaining about a runway expansion.\\nHere, in contrast, the opening submissions give no sense of the\\ntheory of injury introduced in reply. Consumer Action’s\\nopening affidavit speaks vaguely of “burdens,” and its opening\\nbrief elaborates that Rule 30e-3 will harm members “with\\nrespect to their ability not only to access but also to understand\\ncritical shareholder information.” Pet. Br. at 17. Yet, in the\\nreply submissions, some affiants stress ideological opposition\\nto Rule 30e-3, see Pet. Reply Add. at 50 (“I do not believe I\\nshould have to take these extra steps to continue receiving\\npaper copies of my information”); id. at 59 (“I do not like the\\nidea of fund managers relying on ‘implied consent’ to stop\\nsending paper copies”), while others do not wish to “spend\\n[their] valuable time” to make a toll-free telephone call to\\nreceive paper copies of shareholder reports, id. at 55–56, 58.\\nNo member alleges impeded access to or understanding of\\nshareholder reports. In short, the original suggestion of\\nimpaired access and understanding has morphed into an\\nobjection about having to make a free phone call.\\n\\f 10\\n We cannot read CARE to permit reply affidavits that\\npropose a new theory of injury—whether obvious or not.\\nBecause standing must be shown in the same way as other\\nissues, Defs. of Wildlife, 504 U.S. at 561, we have held that “the\\nordinary rules of forfeiture apply to standing,” Gov’t of\\nManitoba v. Bernhardt, 923 F.3d 173, 179 (D.C. Cir. 2019).\\nThose rules include the basic precept that arguments generally\\nare forfeited if raised for the first time in reply. See, e.g.,\\nUnited States v. Van Smith, 530 F.3d 967, 973 (D.C. Cir. 2008).\\nLikewise, an argument is forfeited if the petitioners “were\\nobscure on the issue in their opening brief and only warmed to\\nthe issue in their reply brief.” Novak v. Capital Mgmt. & Dev.\\nCorp., 570 F.3d 305, 316 n.5 (D.C. Cir. 2009) (cleaned up). At\\nbest, that is what Consumer Action has done here, for its\\nopening submissions did not fairly raise the theory of injury\\nthat it now seeks to press. We recognize that Americans for\\nSafe Access considered a theory of injury first raised in a\\nsupplemental brief ordered by this Court. See 706 F.3d at 443–\\n45. But we reasoned that the agency, by not arguing in its own\\nlater submission that the theory was raised too late, forfeited its\\nforfeiture argument. Id. at 444. Here, in contrast, the SEC had\\nno comparable opportunity to respond to the theory of injury\\nfirst raised by Consumer Action in its reply. And in any event,\\nit has stressed all along that “[a] petitioner is generally required\\nto meet its burden to establish standing in its opening brief.”\\nResp. Br. at 22.\\n\\n Third, the reply affidavits in this case hardly make\\nstanding patently obvious. The affiants stating ideological\\nopposition to Rule 30e-3 plainly lack standing, for an “interest\\nin the proper administration of the laws” is quintessentially\\n“nonconcrete.” Earth Island Inst., 555 U.S. at 497 (quotation\\nmarks omitted). The affiants complaining of lost time present\\na closer question. On the one hand, even a small financial\\ninjury confers Article III standing. See, e.g., Carpenters Indus.\\n\\f 11\\nCouncil v. Zinke, 854 F.3d 1, 5 (D.C. Cir. 2017). On the other\\nhand, at least where intangible injuries are at issue, either the\\ninjury must have “a close relationship to a harm that has\\ntraditionally been regarded as providing a basis for a lawsuit in\\nEnglish or American courts,” or a statute must make the injury\\n“legally cognizable.” Spokeo, 136 S. Ct. at 1549 (quotation\\nmarks omitted). Consumer Action cites an out-of-circuit\\ndecision predicating Article III standing on “the occupation of\\n[a] fax machine for … one minute.” Palm Beach Golf Ctr.–\\nBoca, Inc. v. John G. Sarris, D.D.S., P.A., 781 F.3d 1245, 1251\\n(11th Cir. 2015) (parentheses omitted). But Palm Beach Golf\\narose under the Telephone Consumer Protection Act, which\\ncreated statutory protection against unwanted phone or fax\\nsolicitations. See id. at 1252. Here, in contrast, we are aware\\nof no analogous statute that protects shareholders from having\\nto use the telephone. Nor does Consumer Action contend that\\nthe minimal time lost in making a free phone call bears any\\nclose relationship to injuries traditionally deemed adequate in\\nlaw. These may be debatable questions, but Consumer Action\\nfrustrated the adversarial process by teeing them up for the first\\ntime in reply. Under these circumstances, we cannot find\\nstanding based on its reply submissions.\\n\\n III\\n\\n To seek judicial review, the Industry Petitioners must\\nassert interests falling “arguably within the zone of interests to\\nbe protected or regulated” by the laws that they invoke. Clarke\\nv. Secs. Indus. Ass’n, 479 U.S. 388, 396 (1987) (quotation\\nmarks omitted). This rule was once described as one of\\n“prudential standing,” but then, in a case originating in district\\ncourt, was recast as one for “determining who may invoke the\\ncause of action” providing the basis for the lawsuit. Lexmark\\nInt’l, Inc. v. Static Control Components, Inc., 572 U.S. 118,\\n130 (2014). The zone-of-interests requirement also limits who\\n\\f 12\\nmay seek judicial review directly in a court of appeals, as we\\nhave recognized both before and after Lexmark. See, e.g.,\\nSierra Club v. EPA, 755 F.3d 968, 976 (D.C. Cir. 2014);\\nHazardous Waste Treatment Council v. Thomas, 885 F.2d 918,\\n921–22 (D.C. Cir. 1989) (HWTC). Protected interests are ones\\nasserted either by “intended beneficiaries” of the statute at\\nissue or by other “suitable challengers”—i.e., parties whose\\ninterests coincide “systemically, not fortuitously” with those of\\nintended beneficiaries. HWTC, 885 F.2d at 922–24. These\\nrules are designed to prevent litigation by parties “whose suits\\nare more likely to frustrate than to further statutory objectives.”\\nId. at 922 (quotation marks omitted).\\n\\n The Industry Petitioners contend that Rule 30e-3 violates\\nthe Securities Act of 1933, the Securities Exchange Act of\\n1934, and the Investment Company Act of 1940. Two of those\\nstatutes permit any person aggrieved by an SEC regulation to\\nseek judicial review in this Court. See 15 U.S.C. §§ 77i(a),\\n80a-42(a); N.Y. Republican State Comm. v. SEC, 799 F.3d\\n1126, 1130–34 (D.C. Cir. 2015).1 The Industry Petitioners also\\ninvoke the Administrative Procedure Act, but it neither creates\\njurisdiction, Califano v. Sanders, 430 U.S. 99, 104–07 (1977),\\nnor augments specific-review provisions like those in the\\nsecurities statutes, see 5 U.S.C. § 704; Bennett v. Spear, 520\\nU.S. 154, 161–62 (1997). As their basis for constitutional\\nstanding and statutory aggrievement, the Industry Petitioners\\nallege that Rule 30e-3 will harm paper companies by reducing\\nthe demand for their products. So the dispositive question is\\n\\n1\\n All three statutes permit any person aggrieved by an SEC “order”\\nto seek judicial review in this Court. 15 U.S.C. §§ 77i(a), 78y(b)(1),\\n80a-42(a). The word “order” encompasses SEC rules in the context\\nof the Securities Act and the Investment Company Act, see N.Y.\\nRepublican State Comm., 799 F.3d at 1130–34, but not in the context\\nof the Exchange Act, Am. Petrol. Inst. v. SEC, 714 F.3d 1329, 1333–\\n37 (D.C. Cir. 2013).\\n\\f 13\\nwhether that interest—in selling more paper—is one arguably\\nprotected by the securities laws.\\n\\n As sellers of paper, the Industry Petitioners are not\\nintended beneficiaries of the securities laws. The Securities\\nAct “regulates initial distributions of securities,” and the\\nExchange Act “regulates post-distribution trading.” Cent.\\nBank of Denver, N.A. v. First Interstate Bank of Denver, N.A.,\\n511 U.S. 164, 171 (1994). These statutes “embrace a\\nfundamental purpose to substitute a philosophy of full\\ndisclosure for the philosophy of caveat emptor.” Id. (cleaned\\nup). Likewise, the Investment Company Act regulates\\ninvestment companies to protect investors. See 15 U.S.C.\\n§ 80a-1(b). In sum, “shareholders [are] the direct and intended\\nbeneficiaries” of the securities laws, Piper v. Chris-Craft\\nIndus., Inc., 430 U.S. 1, 32 (1977)—and paper sellers are not.\\n\\n That leaves the question whether the interests asserted by\\nthe Industry Petitioners systematically coincide with those of\\nshareholders. In HWTC, we held that a group representing the\\ninterests of waste treatment plants could not challenge the\\nalleged laxity of regulations under the Resource Conservation\\nand Recovery Act. We reasoned that overly stringent\\nregulation might sometimes harm environmental interests, but\\nwaste treatment facilities—which stand to profit from it—\\nwould urge stricter regulation “whether the effect on health and\\nthe environment [was] good, bad, or indifferent.” 885 F.2d at\\n924–25. We have applied this holding repeatedly, see Sierra\\nClub, 755 F.3d at 976, and we have extended it to bar\\nchallenges by manufacturers of pollution-control equipment to\\nthe alleged laxity of regulations under the Clean Air Act, see\\nCement Kiln Recycling Coal. v. EPA, 255 F.3d 855, 870–71\\n(D.C. Cir. 2001).\\n\\f 14\\n Those cases control this one. As paper companies, the\\nIndustry Petitioners would prefer paper disclosure for all\\nshareholders. There is no reason to think that this unqualified\\npreference systematically aligns with the interests of\\nshareholders. In fact, there is good reason to believe the\\nopposite. As the SEC explained, “an investor looking for a\\nfund’s annual report is most likely to seek it out on the fund’s\\nwebsite, rather than request it by mail or phone.” Shareholder\\nReports, 83 Fed. Reg. at 29,165 n.96. So, the Industry\\nPetitioners already are opposed to the largest cross-section of\\nindividual shareholders. Moreover, there is a pronounced and\\ngrowing trend in favor of internet usage, especially among\\nhouseholds that own mutual funds. See id. at 29,165 n.97.\\nThus, the conflict between the interests of paper sellers and\\nthose of shareholders is likely to increase over time. This\\nsuggests a systematic misalignment with shareholder\\npreferences, which makes paper companies distinctly\\nunqualified to advance the interests of shareholders.\\n\\n The Industry Petitioners invoke First National Bank &\\nTrust Co. v. National Credit Union Administration, 988 F.2d\\n1272 (D.C. Cir. 1993), and Honeywell International, Inc. v.\\nEPA, 374 F.3d 1363 (D.C. Cir. 2004). These cases permitted\\nsuits to enforce specific entry restrictions imposed on\\ncompetitors of the plaintiffs or petitioners. In First National\\nBank, we allowed a bank to challenge an agency decision\\nallowing a rival credit union to expand its membership. 988\\nF.2d at 1275–79. The bank argued that the decision violated a\\nstatute limiting membership to “groups having a common bond\\nof occupation or design.” Id. at 1273 (quotation marks\\nomitted). In Honeywell, we permitted a company to challenge\\nan agency decision allowing a competitor to market certain\\nproducts. 374 F.3d at 1370–71. The company argued that the\\nagency impermissibly had rested on economic rather than\\nenvironmental considerations, in violation of section 612(c) of\\n\\f 15\\nthe Clean Air Act. See id. at 1365, 1371–72. In both cases, we\\ndistinguished HWTC on the ground that “the potentially\\nlimitless incentives of competitors were channeled by the terms\\nof the statute into suits of a limited nature brought to enforce\\nthe statutory demarcation.” First Nat’l Bank, 988 F.2d at 1278;\\nsee Honeywell, 374 F.3d at 1370–71. In other words, both\\ncases involved statutes that “constrain[ed] competitors to a\\nlimited role in guarding a congressionally drawn boundary.”\\nFirst Nat’l Bank, 988 F.2d at 1278.\\n\\n The challenge here is not so limited. The SEC adopted\\nRule 30e-3 under general grants of rulemaking authority to\\ncarry out the purposes of the Securities Act, the Exchange Act,\\nand the Investment Company Act. See 15 U.S.C. §§ 77s(a),\\n78w(a), 80a-37(a). In exercising these authorities, the\\nCommission must consider “the public interest,” “the\\nprotection of investors,” and “whether the action will promote\\nefficiency, competition, and capital formation.” 15 U.S.C.\\n§§ 77b(b), 78c(f), 80a-2(c). The Exchange Act further\\nprohibits rules that impose “a burden on competition not\\nnecessary or appropriate in furtherance of the purposes of this\\nchapter.” Id. § 78w(a)(2). The Industry Petitioners argue that\\nthe Commission violated these provisions, but none of them\\nforms a discrete “statutory demarcation” enforceable by “suits\\nof a limited nature.” First Nat’l Bank, 988 F.2d at 1278.\\nBecause the securities laws impose no meaningful constraint\\non the Industry Petitioners’ ability and incentive to push paper\\nregardless of the interests or preferences of shareholders, the\\ncontrolling precedent here is HWTC.\\n\\n IV\\n\\n Consumer Action lacks Article III standing, and the\\nIndustry Petitioners assert interests beyond those arguably\\n\\f 16\\nprotected or regulated by the securities laws. Accordingly, we\\ndeny the petition for review.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4426250/\", \"author_raw\": \"KATSAS, Circuit Judge\"}]}"},"judge1":{"kind":"string","value":"HENDERSON"},"judge2":{"kind":"string","value":"ROGERS"},"judge3":{"kind":"string","value":"KATSAS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4648997/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: TWIN RIVERS PAPER COMPANY LLC, Et Al., Petitioners v. SECURITIES AND EXCHANGE COMMISSION, Respondent\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-16\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued March 12, 2019 Decided August 16, 2019\n\n No. 18-1213\n\n TWIN RIVERS PAPER COMPANY LLC, ET AL.,\n PETITIONERS\n\n v.\n\n SECURITIES AND EXCHANGE COMMISSION,\n RESPONDENT\n\n\n On Petition for Review of an Order of\n the Securities & Exchange Commission\n\n\n Jane C. Luxton argued the cause and filed the briefs for\npetitioners.\n\n Alan W. Bakowski was on the brief for amici curiae Public\nand Private Companies, Nonprofit Organizations, and Labor\nUnion in support of petitioners.\n\n Tracey A. Hardin, Assistant General Counsel, Securities\nand Exchange Commission, argued the cause for respondent.\nWith her on the brief were Michael A. Conley, Solicitor, and\nDaniel E. Matro, Senior Counsel.\n\n Eugene Scalia, Jacob T. Spencer, and Paul S. Stevens were\non the brief for amici curiae Investment Company Institute and\nIndependent Directors Council in support of respondent.\n\f 2\n\n Before: HENDERSON, ROGERS, and KATSAS, Circuit\nJudges.\n\n KATSAS, Circuit Judge: In 2018, the Securities and\nExchange Commission adopted a rule allowing investment\ncompanies to post shareholder reports online and mail paper\ncopies to shareholders upon request. The petitioners—a\nconsumer-advocacy organization and representatives of the\npaper industry—argue that the SEC did not adequately\nconsider the interests of shareholders who prefer reports in\npaper form. Because the consumer organization lacks\nconstitutional standing and the paper-industry representatives\nassert interests beyond those protected or regulated by the\nsecurities laws, we deny the petition for review.\n\n I\n\n The Securities and Exchange Commission requires\ninvestment companies, such as mutual funds, to transmit\nperiodic shareholder reports to their investors. See, e.g., 17\nC.F.R. § 270.30e-1. Previously, the SEC required the\ncompanies to mail paper copies unless an investor affirmatively\nselected electronic delivery. In 2018, the Commission adopted\nRule 30e-3, which allows companies to change their default\nmethod of transmission. See Optional Internet Availability of\nInvestment Company Shareholder Reports, 83 Fed. Reg.\n29,158 (June 22, 2018) (Shareholder Reports). The rule now\npermits investment funds to post shareholder reports online and\nnotify investors of their availability. See 17 C.F.R. § 270.30e-\n3(b)–(d). Investment companies must mail shareholder reports\nonly to investors who expressly request a paper copy. See id.\n§ 270.30e-3(e).\n\f 3\n The SEC gave two principal rationales for this change.\nFirst, it projected that the rule would save investment funds\nabout $140 million annually. Shareholder Reports, 83 Fed.\nReg. at 29,187. It explained that because printing costs are paid\nfrom fund assets, these savings ordinarily will be “passed along\nto investors.” Id. at 29,183. Second, the Commission\nconcluded that “many investors would prefer enhanced\navailability of fund information on the internet.” Id. at 29,165.\nIt explained that “an investor looking for a fund’s annual report\nis most likely to seek it out on the fund’s website, rather than\nrequest it by mail or phone.” Id. at 29,165 n.96. Moreover,\ninternet usage “has continued to increase rapidly” over the last\ndecade, including among “households owning mutual funds.”\nId. at 29,165 n.97. In short, the SEC expected the rule to match\nshareholder preferences and save them money.\n\n The Commission recognized that some investors may still\nprefer paper delivery of shareholder reports. See, e.g.,\nShareholder Reports, 83 Fed. Reg. at 29,165–66. Several\nfeatures of Rule 30e-3 accommodate that preference. The rule\nprescribes an extended transition period: investment\ncompanies must continue to deliver paper copies of the reports\nuntil at least January 1, 2021, and they must include with each\nmailing a statement advising shareholders of the coming\nchange in the default mode of transmission. See 17 C.F.R.\n§ 270.30e-3(i). Then, after making the switch, a fund must\nmail a paper notice to all shareholders for each report that it\nposts online. See id. § 270.30e-3(c). The notice must include\na toll-free telephone number that investors may call to request\na paper copy of the individual report or all future reports. See\nid. § 270.30e-3(c)(1)(i)–(v). The notice also may specify other\nways to request paper reports, such as by e-mail or online. See\nid. § 270.30e-3(c)(2). Finally, once an investor requests a\npaper copy, the fund must mail the report within three business\ndays at no cost to the investor. Id. § 270.30e-3(e).\n\f 4\n This case presents two sets of petitioners. Consumer\nAction is a non-profit membership organization that represents\ncertain consumer interests. Twin Rivers Paper Company and\nthree industry organizations—which we call the Industry\nPetitioners—represent the interests of the American paper\nindustry. Together, the petitioners argue that the SEC adopted\nRule 30e-3 in violation of three securities laws and the\nAdministrative Procedure Act. Among other grounds, they\ncontend that the Commission did not adequately protect\nshareholders who prefer paper delivery.\n\n II\n\n We begin with the question whether Consumer Action has\nconstitutional standing to challenge Rule 30e-3. The\nConstitution limits the “judicial Power of the United States” to\n“Cases” or “Controversies,” U.S. Const. art. III, §§ 1–2, and\nthe requirement of standing is “rooted in the traditional\nunderstanding of a case or controversy,” Spokeo, Inc. v. Robins,\n136 S. Ct. 1540, 1547 (2016). To establish standing under\nArticle III, a party “must have (1) suffered an injury in fact,\n(2) that is fairly traceable to the challenged conduct of the\ndefendant, and (3) that is likely to be redressed by a favorable\njudicial decision.” Id.\n\n Consumer Action claims standing on behalf of its\nmembers. To establish representational standing, an\norganization must prove, among other things, that its members\nwould “have standing to sue in their own right.” Hunt v. Wash.\nState Apple Advert. Comm’n, 432 U.S. 333, 343 (1977). In\nturn, this requires proof that at least one member suffered an\ninjury in fact—“an invasion of a legally protected interest\nwhich is (a) concrete and particularized; and (b) actual or\nimminent, not conjectural or hypothetical.” Lujan v. Defs. of\nWildlife, 504 U.S. 555, 560 (1992) (cleaned up). “A ‘concrete’\n\f 5\ninjury must be ‘de facto’; that is, it must actually exist.”\nSpokeo, 136 S. Ct. at 1548. Moreover, a party cannot rest on\n“abstract,” id., or “conclusory” assertions of injury, Block v.\nMeese, 793 F.2d 1303, 1308 (D.C. Cir. 1986), but must point\nto “specific, concrete facts demonstrating … harm,” Warth v.\nSeldin, 422 U.S. 490, 508 (1975).\n\n In Sierra Club v. EPA, 292 F.3d 895 (D.C. Cir. 2002), we\nexplained the procedures for proving standing when a\npetitioner seeks review of agency action directly in a court of\nappeals. “The party invoking federal jurisdiction bears the\nburden of establishing” standing, which “must be supported in\nthe same way as any other matter on which the [party] bears\nthe burden of proof.” Defs. of Wildlife, 504 U.S. at 561. So a\npetitioner “must either identify in [the administrative] record\nevidence sufficient to support its standing to seek review or, if\nthere is none because standing was not an issue before the\nagency, submit additional evidence to the court of appeals.”\nSierra Club, 292 F.3d at 899. Moreover, because “full\ndevelopment of the arguments for and against standing requires\nthe same tried and true adversarial procedure we use for the\npresentation of arguments on the merits,” the petitioner must\nmake this evidentiary presentation no later than when it files\nthe opening brief. Id. at 900. We have reiterated these\nprinciples many times. See, e.g., Del. Dep’t of Nat. Res. &\nEnvt’l Control v. EPA, 785 F.3d 1, 8 (D.C. Cir. 2015); Texas v.\nEPA, 726 F.3d 180, 198 (D.C. Cir. 2013); Int’l Bhd. of\nTeamsters v. TSA, 429 F.3d 1130, 1134–36 (D.C. Cir. 2005).\nAnd we have codified them in our Circuit Rule 28(a)(7), which\nprovides that, “[i]n cases involving direct review in this court\nof administrative actions, the brief of the appellant or petitioner\nmust set forth the basis for the claim of standing,” and that,\n“[w]hen the appellant’s or petitioner’s standing is not apparent\nfrom the administrative record, the brief must include\narguments and evidence establishing the claim of standing.”\n\f 6\n Consumer Action’s initial submissions fail to show that its\nmembers suffered or will suffer an injury in fact. To establish\nsuch an injury, an organization must provide “individual\naffidavits” from “members who have suffered the requisite\nharm.” Summers v. Earth Island Inst., 555 U.S. 488, 499\n(2009). It is “not enough to aver that unidentified members\nhave been injured.” Chamber of Commerce v. EPA, 642 F.3d\n192, 199 (D.C. Cir. 2011). Despite these settled rules,\nConsumer Action failed to submit any member affidavits with\nits opening brief, and its own affidavit fails to identify any\nindividual members. Moreover, the affidavit barely describes\neven the general contours of its membership. The affidavit\nstates that “members, followers and supporters include retirees\nand retirement savers,” as well as “representatives of a national\nnetwork of nearly 7,000 community-based organizations that\nserve seniors, minority Americans, disabled Americans, and\nindividuals living in rural areas.” Pet. Add. at 433. Left\nunclear is whether the members are retirees and retirement\nsavers, members of other unnamed organizations, or perhaps\nneither. To be sure, the opening brief states that Consumer\nAction’s “members” include seniors, Americans who lack\ninternet access, and others. Pet. Br. at 17. But the brief still\nfails to identify individual members, as required by Earth\nIsland Institute. And in any event, a petitioner must support its\nstanding “by affidavit or other evidence,” Sierra Club, 292\nF.3d at 899 (quotation marks omitted), and briefs “are not\nevidence,” id. at 901. Finally, the affidavit states only that\nunspecified members or others “prefer a choice to have paper\ncommunications” and wish to avoid the “burdens imposed by\nSEC’s Rule 30e-3.” Pet. Add. at 433. The former statement is\npuzzling because Rule 30e-3 preserves investors’ ability to\nchoose between electronic and paper communications. As for\nthe latter statement, “general allegations of injury are\ninsufficient,” Conservation Force, Inc. v. Jewell, 733 F.3d\n1200, 1207 (D.C. Cir. 2013), so a reference to unspecified\n\f 7\n“burdens” is not proof of “direct, real, and palpable” injury,\nFood & Water Watch, Inc. v. Vilsack, 808 F.3d 905, 914 (D.C.\nCir. 2015) (quotation marks omitted).\n\n Likewise, nothing in the administrative record shows a\nconcrete injury to identified members of Consumer Action.\nThe organization notes that, in various comments made during\nthe rulemaking, it warned that Rule 30e-3 would decrease\nreadership of shareholder reports, investor access to\ninformation, and transparency. But none of the comments tied\nthese harms to any identified members. Consumer Action\nfurther notes that the SEC acknowledged its comments that the\nrule would disadvantage seniors and minorities. See\nShareholder Reports, 83 Fed. Reg. at 29,162 n.51. But again,\nneither the comments themselves, nor the SEC’s response,\naddressed whether the harms alleged would befall identified\nmembers of Consumer Action. This should hardly be\nsurprising: Article III standing requirements do not apply to\nagency rulemaking, so Consumer Action would have had no\noccasion to argue, and the SEC would have had no occasion to\ndecide, whether the proposed rule would inflict an injury on\nidentified members of the organization. And because\nConsumer Action was not itself an object of the rulemaking, its\nstanding would not likely have been apparent. See Sierra Club,\n292 F.3d at 899–900.\n\n Consumer Action submitted affidavits from individual\nmembers with its reply brief, but they came too late. We may\nexcuse forfeitures of non-jurisdictional preservation\nrequirements for “good cause.” Sierra Club, 292 F.3d at 900.\nIn the context of Sierra Club and Circuit Rule 28(a)(7), we\nhave found such good cause in two circumstances: where “the\nparties reasonably, but mistakenly, believed that the initial\nfilings before the court had sufficiently demonstrated\nstanding,” Ctr. for Sustainable Econ. v. Jewell, 779 F.3d 588,\n\f 8\n599 (D.C. Cir. 2015) (quoting Ams. for Safe Access v. DEA,\n706 F.3d 438, 443 (D.C. Cir. 2013)); and where the parties\n“reasonably assumed that [their] standing was self-evident”\nfrom the administrative record, Am. Library Ass’n v. FCC, 401\nF.3d 489, 494 (D.C. Cir. 2005); see Del. Dep’t of Nat. Res. &\nEnvt’l Control, 785 F.3d at 8; Town of Barnstable v. FAA, 740\nF.3d 681, 691 (D.C. Cir. 2014). In this case, Consumer Action\ncould not reasonably have believed that its barebones affidavit,\nvaguely describing the preferences and burdens of unnamed\nmembers and others, sufficed to prove its representational\nstanding. Nor could it reasonably have believed that its\nstanding was self-evident from the rulemaking record.\n\n Consumer Action suggests that Communities Against\nRunway Expansion, Inc. v. FAA, 355 F.3d 678 (D.C. Cir. 2004)\n(CARE), allows a petitioner to prove standing for the first time\nin a reply brief, so long as standing is obvious. We do not read\nCARE that broadly. There, an organization challenged the\napproval of a runway expansion at Logan Airport. With its\nopening brief, the organization submitted affidavits from\nmembers who lived nearby, but the affidavits lacked “facts\nsufficient to support a finding that the declarants would be\nexposed to a concrete and particularized ‘injury in fact’ as a\nresult of the contested project.” Id. at 684. With its reply brief,\nthe organization submitted additional affidavits from other\nmembers who declared that they would experience “increased\nnoise from aircraft operations at Logan.” Id. CARE thus\ninvolved new factual material tendered to shore up deficient\nindividual affidavits submitted with the opening brief. In that\ncircumstance, we considered the reply affidavits—which, we\nstressed, made the organization’s standing “patently obvious”\nand “irrefutable.” Id. at 685.\n\n This case differs from CARE in three critical respects.\nFirst, Consumer Action made a far less substantial showing of\n\f 9\nstanding in its initial submissions. In CARE, the organization\ntendered affidavits from members who lived near Logan\nAirport and alleged injury in at least general terms. See 355\nF.3d at 684. So, there was at least arguably good cause to\nexcuse the forfeiture. See Ctr. for Sustainable Econ., 779 F.3d\nat 599. Here, in contrast, Consumer Action tendered no\nmember affidavits, identified no specific members, and failed\nto describe its membership in even general terms. This cannot\namount to good cause.\n\n Second, the reply affidavits in this case raise an entirely\nnew theory of standing. In CARE, the injury pinned down in\nthe reply affidavits—increased airport noise—was hardly\nsurprising in light of initial submissions from individuals living\nnear an airport and complaining about a runway expansion.\nHere, in contrast, the opening submissions give no sense of the\ntheory of injury introduced in reply. Consumer Action’s\nopening affidavit speaks vaguely of “burdens,” and its opening\nbrief elaborates that Rule 30e-3 will harm members “with\nrespect to their ability not only to access but also to understand\ncritical shareholder information.” Pet. Br. at 17. Yet, in the\nreply submissions, some affiants stress ideological opposition\nto Rule 30e-3, see Pet. Reply Add. at 50 (“I do not believe I\nshould have to take these extra steps to continue receiving\npaper copies of my information”); id. at 59 (“I do not like the\nidea of fund managers relying on ‘implied consent’ to stop\nsending paper copies”), while others do not wish to “spend\n[their] valuable time” to make a toll-free telephone call to\nreceive paper copies of shareholder reports, id. at 55–56, 58.\nNo member alleges impeded access to or understanding of\nshareholder reports. In short, the original suggestion of\nimpaired access and understanding has morphed into an\nobjection about having to make a free phone call.\n\f 10\n We cannot read CARE to permit reply affidavits that\npropose a new theory of injury—whether obvious or not.\nBecause standing must be shown in the same way as other\nissues, Defs. of Wildlife, 504 U.S. at 561, we have held that “the\nordinary rules of forfeiture apply to standing,” Gov’t of\nManitoba v. Bernhardt, 923 F.3d 173, 179 (D.C. Cir. 2019).\nThose rules include the basic precept that arguments generally\nare forfeited if raised for the first time in reply. See, e.g.,\nUnited States v. Van Smith, 530 F.3d 967, 973 (D.C. Cir. 2008).\nLikewise, an argument is forfeited if the petitioners “were\nobscure on the issue in their opening brief and only warmed to\nthe issue in their reply brief.” Novak v. Capital Mgmt. & Dev.\nCorp., 570 F.3d 305, 316 n.5 (D.C. Cir. 2009) (cleaned up). At\nbest, that is what Consumer Action has done here, for its\nopening submissions did not fairly raise the theory of injury\nthat it now seeks to press. We recognize that Americans for\nSafe Access considered a theory of injury first raised in a\nsupplemental brief ordered by this Court. See 706 F.3d at 443–\n45. But we reasoned that the agency, by not arguing in its own\nlater submission that the theory was raised too late, forfeited its\nforfeiture argument. Id. at 444. Here, in contrast, the SEC had\nno comparable opportunity to respond to the theory of injury\nfirst raised by Consumer Action in its reply. And in any event,\nit has stressed all along that “[a] petitioner is generally required\nto meet its burden to establish standing in its opening brief.”\nResp. Br. at 22.\n\n Third, the reply affidavits in this case hardly make\nstanding patently obvious. The affiants stating ideological\nopposition to Rule 30e-3 plainly lack standing, for an “interest\nin the proper administration of the laws” is quintessentially\n“nonconcrete.” Earth Island Inst., 555 U.S. at 497 (quotation\nmarks omitted). The affiants complaining of lost time present\na closer question. On the one hand, even a small financial\ninjury confers Article III standing. See, e.g., Carpenters Indus.\n\f 11\nCouncil v. Zinke, 854 F.3d 1, 5 (D.C. Cir. 2017). On the other\nhand, at least where intangible injuries are at issue, either the\ninjury must have “a close relationship to a harm that has\ntraditionally been regarded as providing a basis for a lawsuit in\nEnglish or American courts,” or a statute must make the injury\n“legally cognizable.” Spokeo, 136 S. Ct. at 1549 (quotation\nmarks omitted). Consumer Action cites an out-of-circuit\ndecision predicating Article III standing on “the occupation of\n[a] fax machine for … one minute.” Palm Beach Golf Ctr.–\nBoca, Inc. v. John G. Sarris, D.D.S., P.A., 781 F.3d 1245, 1251\n(11th Cir. 2015) (parentheses omitted). But Palm Beach Golf\narose under the Telephone Consumer Protection Act, which\ncreated statutory protection against unwanted phone or fax\nsolicitations. See id. at 1252. Here, in contrast, we are aware\nof no analogous statute that protects shareholders from having\nto use the telephone. Nor does Consumer Action contend that\nthe minimal time lost in making a free phone call bears any\nclose relationship to injuries traditionally deemed adequate in\nlaw. These may be debatable questions, but Consumer Action\nfrustrated the adversarial process by teeing them up for the first\ntime in reply. Under these circumstances, we cannot find\nstanding based on its reply submissions.\n\n III\n\n To seek judicial review, the Industry Petitioners must\nassert interests falling “arguably within the zone of interests to\nbe protected or regulated” by the laws that they invoke. Clarke\nv. Secs. Indus. Ass’n, 479 U.S. 388, 396 (1987) (quotation\nmarks omitted). This rule was once described as one of\n“prudential standing,” but then, in a case originating in district\ncourt, was recast as one for “determining who may invoke the\ncause of action” providing the basis for the lawsuit. Lexmark\nInt’l, Inc. v. Static Control Components, Inc., 572 U.S. 118,\n130 (2014). The zone-of-interests requirement also limits who\n\f 12\nmay seek judicial review directly in a court of appeals, as we\nhave recognized both before and after Lexmark. See, e.g.,\nSierra Club v. EPA, 755 F.3d 968, 976 (D.C. Cir. 2014);\nHazardous Waste Treatment Council v. Thomas, 885 F.2d 918,\n921–22 (D.C. Cir. 1989) (HWTC). Protected interests are ones\nasserted either by “intended beneficiaries” of the statute at\nissue or by other “suitable challengers”—i.e., parties whose\ninterests coincide “systemically, not fortuitously” with those of\nintended beneficiaries. HWTC, 885 F.2d at 922–24. These\nrules are designed to prevent litigation by parties “whose suits\nare more likely to frustrate than to further statutory objectives.”\nId. at 922 (quotation marks omitted).\n\n The Industry Petitioners contend that Rule 30e-3 violates\nthe Securities Act of 1933, the Securities Exchange Act of\n1934, and the Investment Company Act of 1940. Two of those\nstatutes permit any person aggrieved by an SEC regulation to\nseek judicial review in this Court. See 15 U.S.C. §§ 77i(a),\n80a-42(a); N.Y. Republican State Comm. v. SEC, 799 F.3d\n1126, 1130–34 (D.C. Cir. 2015).1 The Industry Petitioners also\ninvoke the Administrative Procedure Act, but it neither creates\njurisdiction, Califano v. Sanders, 430 U.S. 99, 104–07 (1977),\nnor augments specific-review provisions like those in the\nsecurities statutes, see 5 U.S.C. § 704; Bennett v. Spear, 520\nU.S. 154, 161–62 (1997). As their basis for constitutional\nstanding and statutory aggrievement, the Industry Petitioners\nallege that Rule 30e-3 will harm paper companies by reducing\nthe demand for their products. So the dispositive question is\n\n1\n All three statutes permit any person aggrieved by an SEC “order”\nto seek judicial review in this Court. 15 U.S.C. §§ 77i(a), 78y(b)(1),\n80a-42(a). The word “order” encompasses SEC rules in the context\nof the Securities Act and the Investment Company Act, see N.Y.\nRepublican State Comm., 799 F.3d at 1130–34, but not in the context\nof the Exchange Act, Am. Petrol. Inst. v. SEC, 714 F.3d 1329, 1333–\n37 (D.C. Cir. 2013).\n\f 13\nwhether that interest—in selling more paper—is one arguably\nprotected by the securities laws.\n\n As sellers of paper, the Industry Petitioners are not\nintended beneficiaries of the securities laws. The Securities\nAct “regulates initial distributions of securities,” and the\nExchange Act “regulates post-distribution trading.” Cent.\nBank of Denver, N.A. v. First Interstate Bank of Denver, N.A.,\n511 U.S. 164, 171 (1994). These statutes “embrace a\nfundamental purpose to substitute a philosophy of full\ndisclosure for the philosophy of caveat emptor.” Id. (cleaned\nup). Likewise, the Investment Company Act regulates\ninvestment companies to protect investors. See 15 U.S.C.\n§ 80a-1(b). In sum, “shareholders [are] the direct and intended\nbeneficiaries” of the securities laws, Piper v. Chris-Craft\nIndus., Inc., 430 U.S. 1, 32 (1977)—and paper sellers are not.\n\n That leaves the question whether the interests asserted by\nthe Industry Petitioners systematically coincide with those of\nshareholders. In HWTC, we held that a group representing the\ninterests of waste treatment plants could not challenge the\nalleged laxity of regulations under the Resource Conservation\nand Recovery Act. We reasoned that overly stringent\nregulation might sometimes harm environmental interests, but\nwaste treatment facilities—which stand to profit from it—\nwould urge stricter regulation “whether the effect on health and\nthe environment [was] good, bad, or indifferent.” 885 F.2d at\n924–25. We have applied this holding repeatedly, see Sierra\nClub, 755 F.3d at 976, and we have extended it to bar\nchallenges by manufacturers of pollution-control equipment to\nthe alleged laxity of regulations under the Clean Air Act, see\nCement Kiln Recycling Coal. v. EPA, 255 F.3d 855, 870–71\n(D.C. Cir. 2001).\n\f 14\n Those cases control this one. As paper companies, the\nIndustry Petitioners would prefer paper disclosure for all\nshareholders. There is no reason to think that this unqualified\npreference systematically aligns with the interests of\nshareholders. In fact, there is good reason to believe the\nopposite. As the SEC explained, “an investor looking for a\nfund’s annual report is most likely to seek it out on the fund’s\nwebsite, rather than request it by mail or phone.” Shareholder\nReports, 83 Fed. Reg. at 29,165 n.96. So, the Industry\nPetitioners already are opposed to the largest cross-section of\nindividual shareholders. Moreover, there is a pronounced and\ngrowing trend in favor of internet usage, especially among\nhouseholds that own mutual funds. See id. at 29,165 n.97.\nThus, the conflict between the interests of paper sellers and\nthose of shareholders is likely to increase over time. This\nsuggests a systematic misalignment with shareholder\npreferences, which makes paper companies distinctly\nunqualified to advance the interests of shareholders.\n\n The Industry Petitioners invoke First National Bank &\nTrust Co. v. National Credit Union Administration, 988 F.2d\n1272 (D.C. Cir. 1993), and Honeywell International, Inc. v.\nEPA, 374 F.3d 1363 (D.C. Cir. 2004). These cases permitted\nsuits to enforce specific entry restrictions imposed on\ncompetitors of the plaintiffs or petitioners. In First National\nBank, we allowed a bank to challenge an agency decision\nallowing a rival credit union to expand its membership. 988\nF.2d at 1275–79. The bank argued that the decision violated a\nstatute limiting membership to “groups having a common bond\nof occupation or design.” Id. at 1273 (quotation marks\nomitted). In Honeywell, we permitted a company to challenge\nan agency decision allowing a competitor to market certain\nproducts. 374 F.3d at 1370–71. The company argued that the\nagency impermissibly had rested on economic rather than\nenvironmental considerations, in violation of section 612(c) of\n\f 15\nthe Clean Air Act. See id. at 1365, 1371–72. In both cases, we\ndistinguished HWTC on the ground that “the potentially\nlimitless incentives of competitors were channeled by the terms\nof the statute into suits of a limited nature brought to enforce\nthe statutory demarcation.” First Nat’l Bank, 988 F.2d at 1278;\nsee Honeywell, 374 F.3d at 1370–71. In other words, both\ncases involved statutes that “constrain[ed] competitors to a\nlimited role in guarding a congressionally drawn boundary.”\nFirst Nat’l Bank, 988 F.2d at 1278.\n\n The challenge here is not so limited. The SEC adopted\nRule 30e-3 under general grants of rulemaking authority to\ncarry out the purposes of the Securities Act, the Exchange Act,\nand the Investment Company Act. See 15 U.S.C. §§ 77s(a),\n78w(a), 80a-37(a). In exercising these authorities, the\nCommission must consider “the public interest,” “the\nprotection of investors,” and “whether the action will promote\nefficiency, competition, and capital formation.” 15 U.S.C.\n§§ 77b(b), 78c(f), 80a-2(c). The Exchange Act further\nprohibits rules that impose “a burden on competition not\nnecessary or appropriate in furtherance of the purposes of this\nchapter.” Id. § 78w(a)(2). The Industry Petitioners argue that\nthe Commission violated these provisions, but none of them\nforms a discrete “statutory demarcation” enforceable by “suits\nof a limited nature.” First Nat’l Bank, 988 F.2d at 1278.\nBecause the securities laws impose no meaningful constraint\non the Industry Petitioners’ ability and incentive to push paper\nregardless of the interests or preferences of shareholders, the\ncontrolling precedent here is HWTC.\n\n IV\n\n Consumer Action lacks Article III standing, and the\nIndustry Petitioners assert interests beyond those arguably\n\f 16\nprotected or regulated by the securities laws. Accordingly, we\ndeny the petition for review.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: TWIN RIVERS PAPER COMPANY LLC, Et Al., Petitioners v. SECURITIES AND EXCHANGE COMMISSION, Respondent\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-16\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued March 12, 2019 Decided August 16, 2019\\n\\n No. 18-1213\\n\\n TWIN RIVERS PAPER COMPANY LLC, ET AL.,\\n PETITIONERS\\n\\n v.\\n\\n SECURITIES AND EXCHANGE COMMISSION,\\n RESPONDENT\\n\\n\\n On Petition for Review of an Order of\\n the Securities & Exchange Commission\\n\\n\\n Jane C. Luxton argued the cause and filed the briefs for\\npetitioners.\\n\\n Alan W. Bakowski was on the brief for amici curiae Public\\nand Private Companies, Nonprofit Organizations, and Labor\\nUnion in support of petitioners.\\n\\n Tracey A. Hardin, Assistant General Counsel, Securities\\nand Exchange Commission, argued the cause for respondent.\\nWith her on the brief were Michael A. Conley, Solicitor, and\\nDaniel E. Matro, Senior Counsel.\\n\\n Eugene Scalia, Jacob T. Spencer, and Paul S. Stevens were\\non the brief for amici curiae Investment Company Institute and\\nIndependent Directors Council in support of respondent.\\n\\f 2\\n\\n Before: HENDERSON, ROGERS, and KATSAS, Circuit\\nJudges.\\n\\n KATSAS, Circuit Judge: In 2018, the Securities and\\nExchange Commission adopted a rule allowing investment\\ncompanies to post shareholder reports online and mail paper\\ncopies to shareholders upon request. The petitioners—a\\nconsumer-advocacy organization and representatives of the\\npaper industry—argue that the SEC did not adequately\\nconsider the interests of shareholders who prefer reports in\\npaper form. Because the consumer organization lacks\\nconstitutional standing and the paper-industry representatives\\nassert interests beyond those protected or regulated by the\\nsecurities laws, we deny the petition for review.\\n\\n I\\n\\n The Securities and Exchange Commission requires\\ninvestment companies, such as mutual funds, to transmit\\nperiodic shareholder reports to their investors. See, e.g., 17\\nC.F.R. § 270.30e-1. Previously, the SEC required the\\ncompanies to mail paper copies unless an investor affirmatively\\nselected electronic delivery. In 2018, the Commission adopted\\nRule 30e-3, which allows companies to change their default\\nmethod of transmission. See Optional Internet Availability of\\nInvestment Company Shareholder Reports, 83 Fed. Reg.\\n29,158 (June 22, 2018) (Shareholder Reports). The rule now\\npermits investment funds to post shareholder reports online and\\nnotify investors of their availability. See 17 C.F.R. § 270.30e-\\n3(b)–(d). Investment companies must mail shareholder reports\\nonly to investors who expressly request a paper copy. See id.\\n§ 270.30e-3(e).\\n\\f 3\\n The SEC gave two principal rationales for this change.\\nFirst, it projected that the rule would save investment funds\\nabout $140 million annually. Shareholder Reports, 83 Fed.\\nReg. at 29,187. It explained that because printing costs are paid\\nfrom fund assets, these savings ordinarily will be “passed along\\nto investors.” Id. at 29,183. Second, the Commission\\nconcluded that “many investors would prefer enhanced\\navailability of fund information on the internet.” Id. at 29,165.\\nIt explained that “an investor looking for a fund’s annual report\\nis most likely to seek it out on the fund’s website, rather than\\nrequest it by mail or phone.” Id. at 29,165 n.96. Moreover,\\ninternet usage “has continued to increase rapidly” over the last\\ndecade, including among “households owning mutual funds.”\\nId. at 29,165 n.97. In short, the SEC expected the rule to match\\nshareholder preferences and save them money.\\n\\n The Commission recognized that some investors may still\\nprefer paper delivery of shareholder reports. See, e.g.,\\nShareholder Reports, 83 Fed. Reg. at 29,165–66. Several\\nfeatures of Rule 30e-3 accommodate that preference. The rule\\nprescribes an extended transition period: investment\\ncompanies must continue to deliver paper copies of the reports\\nuntil at least January 1, 2021, and they must include with each\\nmailing a statement advising shareholders of the coming\\nchange in the default mode of transmission. See 17 C.F.R.\\n§ 270.30e-3(i). Then, after making the switch, a fund must\\nmail a paper notice to all shareholders for each report that it\\nposts online. See id. § 270.30e-3(c). The notice must include\\na toll-free telephone number that investors may call to request\\na paper copy of the individual report or all future reports. See\\nid. § 270.30e-3(c)(1)(i)–(v). The notice also may specify other\\nways to request paper reports, such as by e-mail or online. See\\nid. § 270.30e-3(c)(2). Finally, once an investor requests a\\npaper copy, the fund must mail the report within three business\\ndays at no cost to the investor. Id. § 270.30e-3(e).\\n\\f 4\\n This case presents two sets of petitioners. Consumer\\nAction is a non-profit membership organization that represents\\ncertain consumer interests. Twin Rivers Paper Company and\\nthree industry organizations—which we call the Industry\\nPetitioners—represent the interests of the American paper\\nindustry. Together, the petitioners argue that the SEC adopted\\nRule 30e-3 in violation of three securities laws and the\\nAdministrative Procedure Act. Among other grounds, they\\ncontend that the Commission did not adequately protect\\nshareholders who prefer paper delivery.\\n\\n II\\n\\n We begin with the question whether Consumer Action has\\nconstitutional standing to challenge Rule 30e-3. The\\nConstitution limits the “judicial Power of the United States” to\\n“Cases” or “Controversies,” U.S. Const. art. III, §§ 1–2, and\\nthe requirement of standing is “rooted in the traditional\\nunderstanding of a case or controversy,” Spokeo, Inc. v. Robins,\\n136 S. Ct. 1540, 1547 (2016). To establish standing under\\nArticle III, a party “must have (1) suffered an injury in fact,\\n(2) that is fairly traceable to the challenged conduct of the\\ndefendant, and (3) that is likely to be redressed by a favorable\\njudicial decision.” Id.\\n\\n Consumer Action claims standing on behalf of its\\nmembers. To establish representational standing, an\\norganization must prove, among other things, that its members\\nwould “have standing to sue in their own right.” Hunt v. Wash.\\nState Apple Advert. Comm’n, 432 U.S. 333, 343 (1977). In\\nturn, this requires proof that at least one member suffered an\\ninjury in fact—“an invasion of a legally protected interest\\nwhich is (a) concrete and particularized; and (b) actual or\\nimminent, not conjectural or hypothetical.” Lujan v. Defs. of\\nWildlife, 504 U.S. 555, 560 (1992) (cleaned up). “A ‘concrete’\\n\\f 5\\ninjury must be ‘de facto’; that is, it must actually exist.”\\nSpokeo, 136 S. Ct. at 1548. Moreover, a party cannot rest on\\n“abstract,” id., or “conclusory” assertions of injury, Block v.\\nMeese, 793 F.2d 1303, 1308 (D.C. Cir. 1986), but must point\\nto “specific, concrete facts demonstrating … harm,” Warth v.\\nSeldin, 422 U.S. 490, 508 (1975).\\n\\n In Sierra Club v. EPA, 292 F.3d 895 (D.C. Cir. 2002), we\\nexplained the procedures for proving standing when a\\npetitioner seeks review of agency action directly in a court of\\nappeals. “The party invoking federal jurisdiction bears the\\nburden of establishing” standing, which “must be supported in\\nthe same way as any other matter on which the [party] bears\\nthe burden of proof.” Defs. of Wildlife, 504 U.S. at 561. So a\\npetitioner “must either identify in [the administrative] record\\nevidence sufficient to support its standing to seek review or, if\\nthere is none because standing was not an issue before the\\nagency, submit additional evidence to the court of appeals.”\\nSierra Club, 292 F.3d at 899. Moreover, because “full\\ndevelopment of the arguments for and against standing requires\\nthe same tried and true adversarial procedure we use for the\\npresentation of arguments on the merits,” the petitioner must\\nmake this evidentiary presentation no later than when it files\\nthe opening brief. Id. at 900. We have reiterated these\\nprinciples many times. See, e.g., Del. Dep’t of Nat. Res. &\\nEnvt’l Control v. EPA, 785 F.3d 1, 8 (D.C. Cir. 2015); Texas v.\\nEPA, 726 F.3d 180, 198 (D.C. Cir. 2013); Int’l Bhd. of\\nTeamsters v. TSA, 429 F.3d 1130, 1134–36 (D.C. Cir. 2005).\\nAnd we have codified them in our Circuit Rule 28(a)(7), which\\nprovides that, “[i]n cases involving direct review in this court\\nof administrative actions, the brief of the appellant or petitioner\\nmust set forth the basis for the claim of standing,” and that,\\n“[w]hen the appellant’s or petitioner’s standing is not apparent\\nfrom the administrative record, the brief must include\\narguments and evidence establishing the claim of standing.”\\n\\f 6\\n Consumer Action’s initial submissions fail to show that its\\nmembers suffered or will suffer an injury in fact. To establish\\nsuch an injury, an organization must provide “individual\\naffidavits” from “members who have suffered the requisite\\nharm.” Summers v. Earth Island Inst., 555 U.S. 488, 499\\n(2009). It is “not enough to aver that unidentified members\\nhave been injured.” Chamber of Commerce v. EPA, 642 F.3d\\n192, 199 (D.C. Cir. 2011). Despite these settled rules,\\nConsumer Action failed to submit any member affidavits with\\nits opening brief, and its own affidavit fails to identify any\\nindividual members. Moreover, the affidavit barely describes\\neven the general contours of its membership. The affidavit\\nstates that “members, followers and supporters include retirees\\nand retirement savers,” as well as “representatives of a national\\nnetwork of nearly 7,000 community-based organizations that\\nserve seniors, minority Americans, disabled Americans, and\\nindividuals living in rural areas.” Pet. Add. at 433. Left\\nunclear is whether the members are retirees and retirement\\nsavers, members of other unnamed organizations, or perhaps\\nneither. To be sure, the opening brief states that Consumer\\nAction’s “members” include seniors, Americans who lack\\ninternet access, and others. Pet. Br. at 17. But the brief still\\nfails to identify individual members, as required by Earth\\nIsland Institute. And in any event, a petitioner must support its\\nstanding “by affidavit or other evidence,” Sierra Club, 292\\nF.3d at 899 (quotation marks omitted), and briefs “are not\\nevidence,” id. at 901. Finally, the affidavit states only that\\nunspecified members or others “prefer a choice to have paper\\ncommunications” and wish to avoid the “burdens imposed by\\nSEC’s Rule 30e-3.” Pet. Add. at 433. The former statement is\\npuzzling because Rule 30e-3 preserves investors’ ability to\\nchoose between electronic and paper communications. As for\\nthe latter statement, “general allegations of injury are\\ninsufficient,” Conservation Force, Inc. v. Jewell, 733 F.3d\\n1200, 1207 (D.C. Cir. 2013), so a reference to unspecified\\n\\f 7\\n“burdens” is not proof of “direct, real, and palpable” injury,\\nFood & Water Watch, Inc. v. Vilsack, 808 F.3d 905, 914 (D.C.\\nCir. 2015) (quotation marks omitted).\\n\\n Likewise, nothing in the administrative record shows a\\nconcrete injury to identified members of Consumer Action.\\nThe organization notes that, in various comments made during\\nthe rulemaking, it warned that Rule 30e-3 would decrease\\nreadership of shareholder reports, investor access to\\ninformation, and transparency. But none of the comments tied\\nthese harms to any identified members. Consumer Action\\nfurther notes that the SEC acknowledged its comments that the\\nrule would disadvantage seniors and minorities. See\\nShareholder Reports, 83 Fed. Reg. at 29,162 n.51. But again,\\nneither the comments themselves, nor the SEC’s response,\\naddressed whether the harms alleged would befall identified\\nmembers of Consumer Action. This should hardly be\\nsurprising: Article III standing requirements do not apply to\\nagency rulemaking, so Consumer Action would have had no\\noccasion to argue, and the SEC would have had no occasion to\\ndecide, whether the proposed rule would inflict an injury on\\nidentified members of the organization. And because\\nConsumer Action was not itself an object of the rulemaking, its\\nstanding would not likely have been apparent. See Sierra Club,\\n292 F.3d at 899–900.\\n\\n Consumer Action submitted affidavits from individual\\nmembers with its reply brief, but they came too late. We may\\nexcuse forfeitures of non-jurisdictional preservation\\nrequirements for “good cause.” Sierra Club, 292 F.3d at 900.\\nIn the context of Sierra Club and Circuit Rule 28(a)(7), we\\nhave found such good cause in two circumstances: where “the\\nparties reasonably, but mistakenly, believed that the initial\\nfilings before the court had sufficiently demonstrated\\nstanding,” Ctr. for Sustainable Econ. v. Jewell, 779 F.3d 588,\\n\\f 8\\n599 (D.C. Cir. 2015) (quoting Ams. for Safe Access v. DEA,\\n706 F.3d 438, 443 (D.C. Cir. 2013)); and where the parties\\n“reasonably assumed that [their] standing was self-evident”\\nfrom the administrative record, Am. Library Ass’n v. FCC, 401\\nF.3d 489, 494 (D.C. Cir. 2005); see Del. Dep’t of Nat. Res. &\\nEnvt’l Control, 785 F.3d at 8; Town of Barnstable v. FAA, 740\\nF.3d 681, 691 (D.C. Cir. 2014). In this case, Consumer Action\\ncould not reasonably have believed that its barebones affidavit,\\nvaguely describing the preferences and burdens of unnamed\\nmembers and others, sufficed to prove its representational\\nstanding. Nor could it reasonably have believed that its\\nstanding was self-evident from the rulemaking record.\\n\\n Consumer Action suggests that Communities Against\\nRunway Expansion, Inc. v. FAA, 355 F.3d 678 (D.C. Cir. 2004)\\n(CARE), allows a petitioner to prove standing for the first time\\nin a reply brief, so long as standing is obvious. We do not read\\nCARE that broadly. There, an organization challenged the\\napproval of a runway expansion at Logan Airport. With its\\nopening brief, the organization submitted affidavits from\\nmembers who lived nearby, but the affidavits lacked “facts\\nsufficient to support a finding that the declarants would be\\nexposed to a concrete and particularized ‘injury in fact’ as a\\nresult of the contested project.” Id. at 684. With its reply brief,\\nthe organization submitted additional affidavits from other\\nmembers who declared that they would experience “increased\\nnoise from aircraft operations at Logan.” Id. CARE thus\\ninvolved new factual material tendered to shore up deficient\\nindividual affidavits submitted with the opening brief. In that\\ncircumstance, we considered the reply affidavits—which, we\\nstressed, made the organization’s standing “patently obvious”\\nand “irrefutable.” Id. at 685.\\n\\n This case differs from CARE in three critical respects.\\nFirst, Consumer Action made a far less substantial showing of\\n\\f 9\\nstanding in its initial submissions. In CARE, the organization\\ntendered affidavits from members who lived near Logan\\nAirport and alleged injury in at least general terms. See 355\\nF.3d at 684. So, there was at least arguably good cause to\\nexcuse the forfeiture. See Ctr. for Sustainable Econ., 779 F.3d\\nat 599. Here, in contrast, Consumer Action tendered no\\nmember affidavits, identified no specific members, and failed\\nto describe its membership in even general terms. This cannot\\namount to good cause.\\n\\n Second, the reply affidavits in this case raise an entirely\\nnew theory of standing. In CARE, the injury pinned down in\\nthe reply affidavits—increased airport noise—was hardly\\nsurprising in light of initial submissions from individuals living\\nnear an airport and complaining about a runway expansion.\\nHere, in contrast, the opening submissions give no sense of the\\ntheory of injury introduced in reply. Consumer Action’s\\nopening affidavit speaks vaguely of “burdens,” and its opening\\nbrief elaborates that Rule 30e-3 will harm members “with\\nrespect to their ability not only to access but also to understand\\ncritical shareholder information.” Pet. Br. at 17. Yet, in the\\nreply submissions, some affiants stress ideological opposition\\nto Rule 30e-3, see Pet. Reply Add. at 50 (“I do not believe I\\nshould have to take these extra steps to continue receiving\\npaper copies of my information”); id. at 59 (“I do not like the\\nidea of fund managers relying on ‘implied consent’ to stop\\nsending paper copies”), while others do not wish to “spend\\n[their] valuable time” to make a toll-free telephone call to\\nreceive paper copies of shareholder reports, id. at 55–56, 58.\\nNo member alleges impeded access to or understanding of\\nshareholder reports. In short, the original suggestion of\\nimpaired access and understanding has morphed into an\\nobjection about having to make a free phone call.\\n\\f 10\\n We cannot read CARE to permit reply affidavits that\\npropose a new theory of injury—whether obvious or not.\\nBecause standing must be shown in the same way as other\\nissues, Defs. of Wildlife, 504 U.S. at 561, we have held that “the\\nordinary rules of forfeiture apply to standing,” Gov’t of\\nManitoba v. Bernhardt, 923 F.3d 173, 179 (D.C. Cir. 2019).\\nThose rules include the basic precept that arguments generally\\nare forfeited if raised for the first time in reply. See, e.g.,\\nUnited States v. Van Smith, 530 F.3d 967, 973 (D.C. Cir. 2008).\\nLikewise, an argument is forfeited if the petitioners “were\\nobscure on the issue in their opening brief and only warmed to\\nthe issue in their reply brief.” Novak v. Capital Mgmt. & Dev.\\nCorp., 570 F.3d 305, 316 n.5 (D.C. Cir. 2009) (cleaned up). At\\nbest, that is what Consumer Action has done here, for its\\nopening submissions did not fairly raise the theory of injury\\nthat it now seeks to press. We recognize that Americans for\\nSafe Access considered a theory of injury first raised in a\\nsupplemental brief ordered by this Court. See 706 F.3d at 443–\\n45. But we reasoned that the agency, by not arguing in its own\\nlater submission that the theory was raised too late, forfeited its\\nforfeiture argument. Id. at 444. Here, in contrast, the SEC had\\nno comparable opportunity to respond to the theory of injury\\nfirst raised by Consumer Action in its reply. And in any event,\\nit has stressed all along that “[a] petitioner is generally required\\nto meet its burden to establish standing in its opening brief.”\\nResp. Br. at 22.\\n\\n Third, the reply affidavits in this case hardly make\\nstanding patently obvious. The affiants stating ideological\\nopposition to Rule 30e-3 plainly lack standing, for an “interest\\nin the proper administration of the laws” is quintessentially\\n“nonconcrete.” Earth Island Inst., 555 U.S. at 497 (quotation\\nmarks omitted). The affiants complaining of lost time present\\na closer question. On the one hand, even a small financial\\ninjury confers Article III standing. See, e.g., Carpenters Indus.\\n\\f 11\\nCouncil v. Zinke, 854 F.3d 1, 5 (D.C. Cir. 2017). On the other\\nhand, at least where intangible injuries are at issue, either the\\ninjury must have “a close relationship to a harm that has\\ntraditionally been regarded as providing a basis for a lawsuit in\\nEnglish or American courts,” or a statute must make the injury\\n“legally cognizable.” Spokeo, 136 S. Ct. at 1549 (quotation\\nmarks omitted). Consumer Action cites an out-of-circuit\\ndecision predicating Article III standing on “the occupation of\\n[a] fax machine for … one minute.” Palm Beach Golf Ctr.–\\nBoca, Inc. v. John G. Sarris, D.D.S., P.A., 781 F.3d 1245, 1251\\n(11th Cir. 2015) (parentheses omitted). But Palm Beach Golf\\narose under the Telephone Consumer Protection Act, which\\ncreated statutory protection against unwanted phone or fax\\nsolicitations. See id. at 1252. Here, in contrast, we are aware\\nof no analogous statute that protects shareholders from having\\nto use the telephone. Nor does Consumer Action contend that\\nthe minimal time lost in making a free phone call bears any\\nclose relationship to injuries traditionally deemed adequate in\\nlaw. These may be debatable questions, but Consumer Action\\nfrustrated the adversarial process by teeing them up for the first\\ntime in reply. Under these circumstances, we cannot find\\nstanding based on its reply submissions.\\n\\n III\\n\\n To seek judicial review, the Industry Petitioners must\\nassert interests falling “arguably within the zone of interests to\\nbe protected or regulated” by the laws that they invoke. Clarke\\nv. Secs. Indus. Ass’n, 479 U.S. 388, 396 (1987) (quotation\\nmarks omitted). This rule was once described as one of\\n“prudential standing,” but then, in a case originating in district\\ncourt, was recast as one for “determining who may invoke the\\ncause of action” providing the basis for the lawsuit. Lexmark\\nInt’l, Inc. v. Static Control Components, Inc., 572 U.S. 118,\\n130 (2014). The zone-of-interests requirement also limits who\\n\\f 12\\nmay seek judicial review directly in a court of appeals, as we\\nhave recognized both before and after Lexmark. See, e.g.,\\nSierra Club v. EPA, 755 F.3d 968, 976 (D.C. Cir. 2014);\\nHazardous Waste Treatment Council v. Thomas, 885 F.2d 918,\\n921–22 (D.C. Cir. 1989) (HWTC). Protected interests are ones\\nasserted either by “intended beneficiaries” of the statute at\\nissue or by other “suitable challengers”—i.e., parties whose\\ninterests coincide “systemically, not fortuitously” with those of\\nintended beneficiaries. HWTC, 885 F.2d at 922–24. These\\nrules are designed to prevent litigation by parties “whose suits\\nare more likely to frustrate than to further statutory objectives.”\\nId. at 922 (quotation marks omitted).\\n\\n The Industry Petitioners contend that Rule 30e-3 violates\\nthe Securities Act of 1933, the Securities Exchange Act of\\n1934, and the Investment Company Act of 1940. Two of those\\nstatutes permit any person aggrieved by an SEC regulation to\\nseek judicial review in this Court. See 15 U.S.C. §§ 77i(a),\\n80a-42(a); N.Y. Republican State Comm. v. SEC, 799 F.3d\\n1126, 1130–34 (D.C. Cir. 2015).1 The Industry Petitioners also\\ninvoke the Administrative Procedure Act, but it neither creates\\njurisdiction, Califano v. Sanders, 430 U.S. 99, 104–07 (1977),\\nnor augments specific-review provisions like those in the\\nsecurities statutes, see 5 U.S.C. § 704; Bennett v. Spear, 520\\nU.S. 154, 161–62 (1997). As their basis for constitutional\\nstanding and statutory aggrievement, the Industry Petitioners\\nallege that Rule 30e-3 will harm paper companies by reducing\\nthe demand for their products. So the dispositive question is\\n\\n1\\n All three statutes permit any person aggrieved by an SEC “order”\\nto seek judicial review in this Court. 15 U.S.C. §§ 77i(a), 78y(b)(1),\\n80a-42(a). The word “order” encompasses SEC rules in the context\\nof the Securities Act and the Investment Company Act, see N.Y.\\nRepublican State Comm., 799 F.3d at 1130–34, but not in the context\\nof the Exchange Act, Am. Petrol. Inst. v. SEC, 714 F.3d 1329, 1333–\\n37 (D.C. Cir. 2013).\\n\\f 13\\nwhether that interest—in selling more paper—is one arguably\\nprotected by the securities laws.\\n\\n As sellers of paper, the Industry Petitioners are not\\nintended beneficiaries of the securities laws. The Securities\\nAct “regulates initial distributions of securities,” and the\\nExchange Act “regulates post-distribution trading.” Cent.\\nBank of Denver, N.A. v. First Interstate Bank of Denver, N.A.,\\n511 U.S. 164, 171 (1994). These statutes “embrace a\\nfundamental purpose to substitute a philosophy of full\\ndisclosure for the philosophy of caveat emptor.” Id. (cleaned\\nup). Likewise, the Investment Company Act regulates\\ninvestment companies to protect investors. See 15 U.S.C.\\n§ 80a-1(b). In sum, “shareholders [are] the direct and intended\\nbeneficiaries” of the securities laws, Piper v. Chris-Craft\\nIndus., Inc., 430 U.S. 1, 32 (1977)—and paper sellers are not.\\n\\n That leaves the question whether the interests asserted by\\nthe Industry Petitioners systematically coincide with those of\\nshareholders. In HWTC, we held that a group representing the\\ninterests of waste treatment plants could not challenge the\\nalleged laxity of regulations under the Resource Conservation\\nand Recovery Act. We reasoned that overly stringent\\nregulation might sometimes harm environmental interests, but\\nwaste treatment facilities—which stand to profit from it—\\nwould urge stricter regulation “whether the effect on health and\\nthe environment [was] good, bad, or indifferent.” 885 F.2d at\\n924–25. We have applied this holding repeatedly, see Sierra\\nClub, 755 F.3d at 976, and we have extended it to bar\\nchallenges by manufacturers of pollution-control equipment to\\nthe alleged laxity of regulations under the Clean Air Act, see\\nCement Kiln Recycling Coal. v. EPA, 255 F.3d 855, 870–71\\n(D.C. Cir. 2001).\\n\\f 14\\n Those cases control this one. As paper companies, the\\nIndustry Petitioners would prefer paper disclosure for all\\nshareholders. There is no reason to think that this unqualified\\npreference systematically aligns with the interests of\\nshareholders. In fact, there is good reason to believe the\\nopposite. As the SEC explained, “an investor looking for a\\nfund’s annual report is most likely to seek it out on the fund’s\\nwebsite, rather than request it by mail or phone.” Shareholder\\nReports, 83 Fed. Reg. at 29,165 n.96. So, the Industry\\nPetitioners already are opposed to the largest cross-section of\\nindividual shareholders. Moreover, there is a pronounced and\\ngrowing trend in favor of internet usage, especially among\\nhouseholds that own mutual funds. See id. at 29,165 n.97.\\nThus, the conflict between the interests of paper sellers and\\nthose of shareholders is likely to increase over time. This\\nsuggests a systematic misalignment with shareholder\\npreferences, which makes paper companies distinctly\\nunqualified to advance the interests of shareholders.\\n\\n The Industry Petitioners invoke First National Bank &\\nTrust Co. v. National Credit Union Administration, 988 F.2d\\n1272 (D.C. Cir. 1993), and Honeywell International, Inc. v.\\nEPA, 374 F.3d 1363 (D.C. Cir. 2004). These cases permitted\\nsuits to enforce specific entry restrictions imposed on\\ncompetitors of the plaintiffs or petitioners. In First National\\nBank, we allowed a bank to challenge an agency decision\\nallowing a rival credit union to expand its membership. 988\\nF.2d at 1275–79. The bank argued that the decision violated a\\nstatute limiting membership to “groups having a common bond\\nof occupation or design.” Id. at 1273 (quotation marks\\nomitted). In Honeywell, we permitted a company to challenge\\nan agency decision allowing a competitor to market certain\\nproducts. 374 F.3d at 1370–71. The company argued that the\\nagency impermissibly had rested on economic rather than\\nenvironmental considerations, in violation of section 612(c) of\\n\\f 15\\nthe Clean Air Act. See id. at 1365, 1371–72. In both cases, we\\ndistinguished HWTC on the ground that “the potentially\\nlimitless incentives of competitors were channeled by the terms\\nof the statute into suits of a limited nature brought to enforce\\nthe statutory demarcation.” First Nat’l Bank, 988 F.2d at 1278;\\nsee Honeywell, 374 F.3d at 1370–71. In other words, both\\ncases involved statutes that “constrain[ed] competitors to a\\nlimited role in guarding a congressionally drawn boundary.”\\nFirst Nat’l Bank, 988 F.2d at 1278.\\n\\n The challenge here is not so limited. The SEC adopted\\nRule 30e-3 under general grants of rulemaking authority to\\ncarry out the purposes of the Securities Act, the Exchange Act,\\nand the Investment Company Act. See 15 U.S.C. §§ 77s(a),\\n78w(a), 80a-37(a). In exercising these authorities, the\\nCommission must consider “the public interest,” “the\\nprotection of investors,” and “whether the action will promote\\nefficiency, competition, and capital formation.” 15 U.S.C.\\n§§ 77b(b), 78c(f), 80a-2(c). The Exchange Act further\\nprohibits rules that impose “a burden on competition not\\nnecessary or appropriate in furtherance of the purposes of this\\nchapter.” Id. § 78w(a)(2). The Industry Petitioners argue that\\nthe Commission violated these provisions, but none of them\\nforms a discrete “statutory demarcation” enforceable by “suits\\nof a limited nature.” First Nat’l Bank, 988 F.2d at 1278.\\nBecause the securities laws impose no meaningful constraint\\non the Industry Petitioners’ ability and incentive to push paper\\nregardless of the interests or preferences of shareholders, the\\ncontrolling precedent here is HWTC.\\n\\n IV\\n\\n Consumer Action lacks Article III standing, and the\\nIndustry Petitioners assert interests beyond those arguably\\n\\f 16\\nprotected or regulated by the securities laws. Accordingly, we\\ndeny the petition for review.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4178,"cells":{"id":{"kind":"number","value":4648998,"string":"4,648,998"},"name":{"kind":"string","value":"In RE: RAIL FREIGHT FUEL SURCHARGE ANTITRUST LITIGATION - MDL NO. 1869, Dakota Granite Company, on Behalf of Itself and All Others Similarly Situated, Et Al., Appellants v. BNSF Railway Company, Et Al., Appellees"},"name_abbreviation":{"kind":"string","value":"Rail Freight Fuel Surcharge Antitrust Litig. - MDL No. 1896 v. BNSF Ry. Co."},"decision_date":{"kind":"string","value":"2019-08-16"},"docket_number":{"kind":"string","value":"18-7010"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Garland, Rogers, Katsas\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge KATSAS\", \"type\": \"010combined\", \"text\": \"PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n AnUcb $ttt ffnurI of ppitl\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued September 28, 201$ Decided August 16, 2019\\n\\n No. 18-7010\\n\\n IN RE: RAIL FREIGHT FUEL SURCHARGE ANTITRUST\\n LITIGATION - MDL No. 1 $69,\\n\\n DAKOTA GRANITE COMPANY, ON BEHALF OF ITSELF AND ALL\\n OTHERS SIMILARLY SITUATED, ET AL.,\\n APPELLANTS\\n\\n\\n V.\\n\\n\\n BNSF RAILWAY COMPANY, ET AL.,\\n APPELLEES\\n\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1 :07-mc-00489)\\n\\n\\n Kathleen M Sullivan argued the cause for appellants.\\nWith her on the briefs were Stephen R. Neuwirth, Sami H\\nRashid, Michael D. Hausfeld, and Michael?. Lehmann.\\n\\n Carter G. Phillips argued the cause for appellees. With\\nhim on the brief were Joseph R. Guerra, Kathleen Moriarty\\nMueller, Saul?. Morgenstern, Thomas A. Isaacson, John li\\nNannes, Tara L. Reinhart, J Scott Ballenger, Veronica S.\\nLewis, Samuel M Sipe, Jr., Linda S. Stein, Andrew S.\\nTulumello, Lucas C. Townsend, and Kent A. Gardiner.\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 2\\n\\n\\n Anton Metlitsky and Warren D. Postman were on the brief\\nfor amicus curiae Chamber of Commerce of the United States\\nof America in support of defendants-appellees.\\n\\n Before: GARLAND, ChiefJudge, and ROGERS and KATSAS,\\nCircuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge KATSAS.\\n\\n\\n KATSAS, Circuit Judge:’ This case involves a putative\\nclass of over 16,000 shippers allegedly harmed by a price-\\nfixing conspiracy among the nation’s largest freight railroads.\\nThe district court denied class certification because the\\nplaintiffs’ regression analysis—their evidence for proving\\ncausation, injury, and damages on a class-wide basis—\\nmeasured negative damages for over 2,000 members of the\\nproposed class. Based on that consideration, we affirm.\\n\\n I\\n\\n This appeal arises out of eighteen antitrust actions\\nconsolidated by the Multidistrict Litigation Panel. The\\ndefendants are the four largest freight railroads in the United\\nStates: BNSF Railway Company; CSX Transportation, Inc.;\\nNorfolk Southern Railway Company; and Union Pacific\\nRailroad Company. The plaintiffs, who are their customers,\\nallege that the railroads conspired to fix rate-based fuel\\nsurcharges. Railroads impose fuel surcharges—additional\\ncharges above the base shipping price—when the price of fuel\\nrises above a certain trigger price. Rate-based surcharges are\\ncalculated as a percentage of the base shipping price.\\n\\n\\n NOTE: Portions of this opinion contain Sealed Information,\\nwhich has been redacted.\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n Following consolidation, the action was divided into one\\ncase involving direct purchasers and another involving indirect\\npurchasers. All plaintiffs alleged that the railroads violated\\nsection 1 of the Sherman Act, 15 U.S.C. § 1, by conspiring to\\nfix prices. The direct purchasers sought treble damages under\\nsection 4 of the clayton Act, 15 U.S.C. § 15, and the district\\ncourt held that they stated a claim, In re Rail Freight Surcharge\\nAntitrustLitig., 587 F. Supp. 2d27(D.D.c. 2008). The indirect\\npurchasers sought injunctive relief under section 16 of the\\nClayton Act, 15 U.S.C. § 26, and raised various state-law\\nclaims. The district court held that the state claims were\\npreempted by federal law, but it declined to dismiss the federal\\nclaims. In re Rail Freight Surcharge Antitrust Litig., 593 F.\\nSupp. 2d 29 (D.D.C. 2008), afT d, Fayus Enters. v. BNSF Ry.\\nCo., 602 F.3d 444 (D.C. Cir. 2010).\\n\\n The eight named plaintiffs in the direct-purchaser case—\\nCarter Distributing Company; Dakota Granite Company;\\nDonnelly Commodities, Inc.; Dust Pro, Inc.; Nyrstar Taylor\\nChemicals, Inc.; Olin Corporation; Strates Shows, Inc.; and US\\nMagnesium LLC—rnoved to certify a class under Federal Rule\\nof Civil Procedure 23(b)(3). The proposed class consisted of\\nall shippers who paid rate-based fuel surcharges for\\nunregulated services purchased from the defendants between\\nJuly 1,2003 and December 31, 2008. To show that causation,\\ninjury, and damages could be proved on a class-wide basis, the\\nplaintiffs invoked two regression models constructed by their\\neconomist, Dr. Gordon Rausser. The “common factor model”\\nidentified seven variables said to determine the price of the\\ndefendants’ services, including fuel surcharges. The “damages\\nmodel,” controlling for those variables, sought to isolate price\\nincreases attributable to the alleged conspiracy. The railroads\\ncriticized these models on various grounds, including that they\\nmeasured damages for shipments made under legacy contracts\\nfixed before any conspiracy allegedly began.\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 4\\n The district court initially certified the class. It noted that\\nif individualized proof were necessary to establish causation\\nand injury, then the plaintiffs could not satisfy the Rule\\n23(b)(3) requirement that common questions predominate. See\\nIn re Rail Freight Fuel Surcharge Antitrust Litig., 287 F.R.D.\\n1, 43 (D.D.C. 2012). But the court found Dr. Rausser’s\\nregression analysis to be “plausible” and “workable,” so it\\nconcluded that causation, injury, and damages were\\n“susceptible to proof at trial through evidence common to the\\nclass.” Id. at 67. The court rejected many different criticisms\\nof the regression models, but it did not specifically address the\\nquestion of false positives for legacy contracts.\\n\\n On interlocutory review, we vacated the certification order\\nand remanded for reconsideration in light of Comcast Corp. V.\\nBehrend, 569 U.S. 27 (2013). In re Rail Freight Surcharge\\nAntitrustLitig.—MDLNo. 1869, 725 F.3d 244 (D.C. Cir. 2013)\\n(Rail Freight 1). We explained that, for an antitrust class\\naction, common questions “cannot predominate where there\\nexists no reliable means of proving classwide injury in fact.”\\nId. at 253. We expressed concern with the district court’s\\nfailure to address “the damages model’s propensity toward\\nfalse positives,” which left us with no way of knowing whether\\n“the overcharges the damages model calculates for class\\nmembers [are] any more accurate than the obviously false\\nestimates it produces for legacy shippers.” Id. at 254. Finally,\\nwe stressed that Rule 23, as construed in Comcast, requires a\\n“hard look at the soundness of statistical models that purport to\\nshow predominance.” Id. at 255.\\n\\n On remand, after permitting supplemental discovery and\\nexpert reports, the district court denied class certification. In\\nre Rail Freight Surcharge Antitrust Litig., 292 F. Supp. 3d 14\\n(D.D.C. 2017) (Rail Freight If). The court concluded that Dr.\\nRausser’s expert opinions were reliable enough to be\\nadmissible at trial. Id. at 49—63. But in assessing\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 5\\npredominance, the court identified three shortcomings in his\\ndamages model: first, it measured highly inflated damages for\\nintermodal traffic (i.e., shipments traveling by rail and another\\nmode of transportation such as trucks or airplanes), Id. at 122—\\n26; second, as we had noted in the earlier appeal, the model\\nerroneously measured damages for shipments made under\\nlegacy contracts, Id. at 126—31; and third, the model measured\\nnegative damages—and hence no injury—for over 2,000\\nmembers of the proposed class, id. at 132—41. The court\\nconcluded that any one of these problems was enough to defeat\\nthe plaintiffs’ argument for predominance. Id. at 122.\\n\\n The plaintiffs filed a petition for permission to appeal the\\nclass-certification decision under Federal Rule of Civil\\nProcedure 23(f). A motions panel of this Court granted the\\npetition without prejudice to reconsideration at the merits\\nstage. In re Rail freight fuel Surcharge Antitrust Litig.—MDL\\nNo. 1869, No. 17-8005 (D.C. Cir. Dec. 20, 2017).\\n\\n II\\n\\n We begin with the question of our jurisdiction. Orders\\ndenying class certification are neither final decisions under 28\\nU.S.C. § 1291, Coopers & Lybrand v. Livesay, 437 U.S. 463\\n(1978), nor injunctions immediately appealable under 28\\nU.S.C. § 1292(a)(l), Gardner v. Westinghouse Broad Co., 437\\nU.S. 478 (1978). However, 28 U.S.C. § 1292(e) permits the\\nSupreme Court to promulgate rules creating new categories of\\ndecisions appealable before final judgment. Exercising that\\nauthority, the Court has provided that “[a] court of appeals may\\npermit an appeal from an order granting or denying class-action\\ncertification,” if a “petition for permission to appeal” is timely\\nfiled. Fed. R. Civ. P. 23(f).\\n\\n In this case, the plaintiffs filed a timely petition for\\npermission to appeal, which was enough under Rule 23(f) to\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 6\\nsecure our jurisdiction. That jurisdiction is “discretionary,”\\nRail freight I, 725 F.3d at 250, and the railroads contest\\nwhether we should exercise it. But their argument on this point\\nis perfunctory—less than one page of briefing, with no case\\ncitations—and it almost entirely duplicates their merits\\narguments. According to the railroads, we should conclude\\nthat the denial of class certification was correct. Then, we\\nshould dismiss the appeal as raising neither a questionable\\ndecision nor an unsettled issue—considerations that bear on\\nwhether to permit the appeal in the first place, see id. at 250—\\n54. Because that disposition would make little sense at this\\njuncture, we decline to revisit the motions panel’s decision\\naccepting the appeal.\\n\\n III\\n\\n Federal Rule of Civil Procedure 23 sets forth various\\nrequirements for the certification of class actions. Rule 23(a)\\nprovides four “prerequisites” for any class certification,\\nincluding that there must be “questions of law or fact common\\nto the class.” If these prerequisites are met, Rule 23(b)(3)\\npermits certification if, among other things, “questions of law\\nor fact common to class members predominate over any\\nquestions affecting only individual members.” For purposes of\\nthese rules, a “common” question is one that is “capable of\\nclasswide resolution—which means that determination of its\\ntruth or falsity will resolve an issue that is central to the validity\\nof each one of the claims in one stroke.” Wal-Mart Stores, Inc.\\nv. Dukes, 564 U.S. 338, 350 (2011). In contrast, an\\n“individual” question is one for which “members of a proposed\\nclass will need to present evidence that varies from member to\\nmember.” Tyson Foods, Inc. v. Bouaphakeo, 136 S. Ct. 1036,\\n 1045 (2016) (quotation marks omitted).\\n\\n The party seeking class certification “must affirmatively\\ndemonstrate” that the commonality and predominance\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 7\\nrequirements are satisfied. Wal-Mart, 564 U.S. at 350. This\\nrequires a “rigorous analysis” that often will “overlap with the\\nmerits.” Id. at 351 (quotation marks omitted). Three recent\\ncases address the contours of this analysis. In Wal-Mart, the\\nSupreme Court ordered decertification of a Title VII class\\nwhere the plaintiffs presented insufficient proof that the\\ndefendant had engaged in a general policy of sex\\ndiscrimination. Id. at 352—55. In Corncast, the Court ordered\\ndecertification where the regression analysis used to show\\ncommon injury did not track the underlying theory of liability.\\n569 U.S. at 36—38. The Court rejected a contention that, at the\\nclass-certification stage, “any method of measurement is\\nacceptable so long as it can be applied classwide, no matter\\nhow arbitrary the measurements may be.” Id. at 36. But in\\nTyson foods, the Court held that concerns about the validity of\\na statistical sample used to prove class-wide averages—which\\nthe Court described as presenting a common objection to the\\nclaims of each class member—should be addressed “as a\\nmatter of summary judgment, not class certification.” 136 5.\\nCt. at 1047 (quotation marks omitted).\\n\\n IV\\n\\n The direct-purchaser plaintiffs raise claims under section\\n4 of the Clayton Act, which provides treble damages to any\\nperson “injured in his business or property by reason of\\nanything forbidden in the antitrust laws.” 15 U.S.C. § 15(a).\\nTo establish liability under section 4, each plaintiff must prove\\nnot only an antitrust violation, but also an injury to its business\\nor property and a causal relation between the two. Without\\ncommon proof of injury and causation, section 4 plaintiffs\\ncannot establish predominance. See, e.g., Comcast, 569 U.S.\\nat 36—38; Rail freight I, 725 F.3d at 252—53.\\n\\n The parties dispute the extent to which a court, in\\nconducting the “hard look” required by Rule 23, should assess\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 8\\nthe reliability of common evidence. Specifically, they contest\\nwhether Rule 23 requires reliability above and beyond what is\\nnecessary to establish the admissibility of expert testimony\\nunder Federal Rule of Evidence 702 and Daubert v. Merrell\\nDow Pharmaceuticals, Inc., 509 U.S. 579 (1993). That\\nquestion arose because the district court found Dr. Rausser’s\\nregression models to be reliable for admissibility purposes,\\nRail freight II, 292 F. Supp. 3d at 54—63; held that “reliability\\nunder Rule 23 is a higher standard than reliability under\\nDaubert,” id. at 91; and found no predominance based on\\nconcerns that “undermine[d] the reliability of Dr. Rausser’s\\ndamages model,” id. at 122. The defendants embrace this\\nreasoning; they argue that Wal-Mart and Comcast rejected\\ncertification based on concerns that common evidence was\\nunreliable, even though its admissibility was uncontested. The\\nplaintiffs disagree; they contend that predominance ultimately\\nturns on whether the relevant evidence is common or\\nindividualized, and they read Tyson foods as holding that a\\ncourt may assess its reliability only in deciding admissibility or\\n summary judgment.\\n\\n We need not resolve this dispute because Dr. Rausser’s\\ndamages model, even if sufficiently reliable, does not prove\\nclasswide injury. As the district court explained, his model\\nindicates that the proposed class consists of 16,065 shippers.\\nRail freight II, 292 F. $upp. 3d at 136. The plaintiffs maintain\\nthat the alleged conspiracy injured every one of them. Yet the\\ndamages model also indicates that 2,037 members of the\\nproposed class—or 12.7 percent—suffered “only negative\\novercharges” and thus no injury from any conspiracy. Id. at\\n137. So even assuming the model can reliably show injury and\\ncausation for 87.3 percent of the class, that still leaves the\\nplaintiffs with no common proof of those essential elements of\\nliability for the remaining 12.7 percent. The district court held\\nthat the need for “individualized inquiries to determine which\\nof at least 2,037 (and possibly more) class members were\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 9\\nactually injured by the alleged conspiracy,” Id. at 140,\\nprecluded a finding of predominance. Id. at 122. We review a\\ndistrict court’s assessment of predominance only for abuse of\\ndiscretion. Garcia v. Johanns, 444 F.3d 625, 631 (D.C. Cir.\\n2006). Here, we find no such abuse.\\n\\n The plaintiffs argue that their model measures these\\nnegative damages only because of normal prediction error. The\\ndistrict court found that prediction error could not “account for\\nall—or even a substantial portion of—the 2,037 shippers that\\nthe model shows to be uninjured.” Rail Freight II, 292 F. Supp.\\n3d at 139. The plaintiffs take issue with that finding, but it is\\nnot clearly erroneous. In any event, the plaintiffs’ argument\\nabout prediction error at most suggests that their damages\\nmodel might falsely have measured no injury for as many as\\n2,037 shippers. This line of reasoning describes a possible\\nproblem with their own evidence; it does not point to\\naffirmative evidence—much less common affirmative\\nevidence—that a conspiracy did in fact injure these shippers.\\n\\n The plaintiffs further argue that predominance does not\\nrequire common evidence extending to all class members.\\nThat contention appears inconsistent with our statement in Rail\\nfreight I that the plaintiffs, to establish predominance, must\\n“show that they can prove, through common evidence, that all\\nclass members were in fact injured by the alleged conspiracy.”\\n725 F.3d at 252; see also id. (“we do expect the common\\nevidence to show all class members suffered some injury”).\\nDespite these statements, the district court held that our opinion\\ndid not require common evidence of injury to all class\\nmembers. See Rail Freight II, 292 F. Supp. 3d at 132—34.\\nInstead, it agreed with the plaintiffs that common proof\\ncovering “virtually all” members of the proposed class, and\\nleaving only a “de minimis” number of cases requiring\\nindividualized proof of injury and causation, would be enough\\nto show predominance. Id. at 135. For the sake of argument,\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 10\\nwe assume that the district court correctly recognized a de\\nminimis exception to the general rule that, for claims under\\nsection 4 of the Clayton Act, causation and injury must be\\n“capable of classwide resolution,” Wal-Mart, 564 U.S. at 350.\\nThe court reasonably concluded that such a de minimis\\nexception would not encompass this case.\\n\\n In assessing how many individual adjudications are too\\nmany, both the district court and the parties invoke cases\\naddressing the question of when, if ever, a class may include\\nconcededly uninjured members. Strictly speaking, this case\\ndoes not present that question, for the plaintiffs here insist that\\neach member of the proposed class was injured. Nonetheless,\\nthe cited cases bear some similarity to this one: Uninjured class\\nmembers cannot prevail on the merits, so their claims must be\\nwinnowed away as part of the liability determination. And that\\nprospect raises the same kind of question at issue here—when\\ndoes the need for individualized proof of injury and causation\\ndestroy predominance? See, e.g., In re Asacol Antitrust Litig.,\\n907 F.3d 42, 51—58 (1st Cir. 2018); In re Nexium Antitrust\\nLitig., 777 F.3d 9, 18—22 (1st Cir. 2015).\\n\\n The plaintiffs complain that the district court arbitrarily\\nimposed a six-percent upper limit on the percentage of\\nuninjured parties who may be included in a certified class. In\\nfact, the court’s analysis was more nuanced. As the court\\nexplained, the “few reported decisions” involving uninjured\\nclass members “suggest that 5% to 6% constitutes the outer\\nlimits of a de minimis number.” Rail Freight ll 292 F. Supp.\\n3d at 137. The 12.7 percent figure in this case is more than\\ntwice that approximate upper bound reflected in analogous\\ncaselaw. Moreover, the district court considered raw numbers\\nas well as percentages: six percent of a “class totaling only\\nfifty-five” members might be de minimis, but 12.7 percent of\\nthis class yields “2,037 uninjured class members” (according\\nto the common proof), all of whom would need individualized\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 11\\nadjudications of causation and injury. Id. at 137—38. Finally,\\nthe district court stressed that the plaintiffs have proposed no\\n“further way”—short of full-blown, individual trials—”to\\nreduce this number and segregate the uninjured from the truly\\ninjured.” Id. at 138. None of this was an abuse of discretion.\\n\\n The absence of any winnowing mechanism sharply\\ndistinguishes Nexiurn, the plaintiffs’ best case. There, the class\\nincluded purchasers of a drug allegedly shielded from\\ncompetition by the unlawful suppression of a generic\\nalternative. 777 F.3d at 13—14. The problem of uninjured class\\nmembers arose because a small percentage of the class, due to\\nbrand loyalty, would have purchased the drug even if a less\\nexpensive generic alternative had been available. Id. at 19—20.\\nThe First Circuit held that the uninjured class members could\\nmanageably be winnowed by having individual consumers file\\nminimal, likely unrebutted affidavit testimony indicating\\nwhether, if given the choice, they would have purchased the\\nbranded drug or a generic alternative. Id. at 20—21.\\n\\n Nexium does not support class certification here. For one\\nthing, the First Circuit sharply limited that decision in Asacol.\\nThere, the Court explained that any winnowing mechanism\\nmust be truncated enough to ensure that the common issues\\npredominate, yet robust enough to preserve the defendants’\\nSeventh Amendment and due process rights to contest every\\nelement of liability and to present every colorable defense. 907\\nF.3d at 5 1—54. Moreover, the Court held that Nexium’s\\naffidavit mechanism could not satisfy both conditions where\\nthe defendant seeks to contest the question whether individual\\nclass members would have shifted from the branded drug to a\\nless expensive generic alternative. See Id. That would require\\nindividual trials because genuinely contested affidavits do not\\nsupport summary judgment and are inadmissible. Id. Here, the\\ndefendants intend to contest whether any of the 2,037 shippers\\nsuffered injury as a result of any conspiracy. And the question\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 12\\npresented in these individual challenges—regarding impacts\\non shippers of different sizes, shipping different products in\\ndifferent geographic markets, with different transportation\\noptions and different degrees of leverage—would be far more\\ncomplex than the single unitary question (branded or generic?)\\nat issue in Nexium and Asacol.\\n\\n The plaintiffs also invoke the Supreme Court’s discussion\\nof uninjured class members in Tyson foods. The defendants\\nthere sought review of the question whether a certified class\\nmay contain any uninjured members. But the Supreme Court\\nreserved that question, 136 5. Ct. at 1049, and the district court\\nhere decided it in the plaintiffs’ favor, Rail Freight II, 292 F.\\nSupp. 3d at 133—35. So, this aspect of Tyson foods does not\\nadvance the plaintiffs’ appeal. The Supreme Court also\\naddressed a further ‘new argument” raised for the first time in\\nmerits briefing—that plaintiffs at the certification stage must\\nprove that all class members were injured or establish a\\nmanageable process for culling out uninjured class members.\\n 136 5. Ct. at 1049. The Court held that this argument was\\n“premature” because the parties disagreed about what culling\\nmechanisms might be available and the district court had not\\nyet addressed the question. See 136 5. Ct. at 1050. This\\nholding rested on the inappropriateness of raising new issues\\nfor the first time in Supreme Court merits briefing. See id. It\\ndoes not, as the plaintiffs here contend, perniit district courts\\nconsidering class certification to defer questions about the\\nnumber and nature of any individualized inquiries that might\\nbe necessary to establish liability. To the contrary, confronting\\nsuch questions is part-and-parcel of the “hard look” required\\nby Wal-Mart and Comcast, as recognized even by those courts\\npermitting a class to include some small number of concededly\\nuninjured individuals. See, e.g., In re Asacol, 907 F.3d at 51—\\n54; In re Nexiurn,777 F.3d at 19—2 1.\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 1—,\\n 1.)\\n\\n\\n Finally, the plaintiffs argue that the 2,037 class members\\nfor whom their damages model shows no injury are a de\\nminimis portion of the class because their shipments make up\\nless than one percent of the railroads’ overall revenue from the\\nalleged conspiracy. But revenue is irrelevant to predominance,\\nwhich looks to whether elements such as causation and injury\\nmay be proved through common evidence, not how much the\\ndefendants benefited from any wrongdoing.\\n\\n V\\n\\n Looking beyond Dr. Rausser’s regression analysis, the\\nplaintiffs point to other evidence that they say can prove injury\\nand causation on a class-wide basis. In Rail freight I, we\\nconcluded that Dr. Rausser’s analysis was “essential” to the\\nplaintiffs’ case for certification: “No damages model, no\\npredominance, no class certification.” 725 F.3d at 253. The\\ndistrict court reached the same conclusion on remand, after\\ncareful review of all the documentary and expert evidence.\\nThis was not an abuse of discretion.\\n\\n The plaintiffs invoke documentary evidence that the\\ndefendants enforced fuel surcharges “uniformly and with few\\nexceptions.” Rail Freight II, 292 F. $upp. 3d at 122; see id. at\\n103—07. But imposing fuel surcharges does not show injury\\ncaused by a conspiracy. The parties vigorously dispute\\nwhether higher overall prices during the class period were\\nattributable to causes besides any conspiracy—such as the\\nmarked increase in fuel prices that occurred around the\\nbeginning of the class period, see l.A. 65 92—93. As the district\\ncourt explained, Dr. Rausser designed his regression models\\nprecisely to control for these kinds of potential alternative\\ncauses, see Rail Freight II, 292 F. $upp. 3d at 101, yet his\\ndamages model showed that 2,037 shippers were uninjured\\ndespite paying fuel surcharges. The plaintiffs’ evidence of\\nwidespread fuel surcharges helps explain why the proposed\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 14\\nclass is a large one, but it neither proves that the 2,037 shippers\\nwere injured by the alleged conspiracy nor otherwise compels\\na finding of predominance.\\n\\n Next, the plaintiffs invoke the expert testimony of Dr.\\nJames McClave, who argued that the 2,037 shippers must have\\nbeen harmed by the conspiracy. Dr. McClave reasoned that\\nbecause these shippers made fewer purchases on average, and\\nthus had less bargaining power than the rest of the class, they\\nmust have been more susceptible to injury. Moreover, Dr.\\nMcClave’s own study concluded that the defendants’ smallest\\ncustomers—roughly shippers who made only one\\npurchase I I—collectively\\npaid higher prices than did larger shippers.\\n\\n This analysis is not common proof of injury to the 2,037\\nshippers. For one thing, of the shippers studied by Dr.\\nMcClave, only were in the class, and Dr. McClave did\\nnot indicate how many of them were included among the 2,037\\nclass members for whom Dr. Rausser’s model showed no\\ninjury. Moreover, evidence that a group of one-time shippers\\nmay have paid higher prices collectively does not prove that all\\n(or almost all) of them were injured individually. And, as the\\ndistrict court explained, Dr. McClave did not attempt to\\nidentify how many of the one-time shippers did in fact pay\\nmore, let alone which of them were among the 2,037. See Rail\\nFreight II, 292 F. $upp. 3d at 139.\\n\\n The plaintiffs respond that Dr. McClave’s analysis, even if\\ninconclusive, at least suggests that reduced bargaining power\\nmade the 2,037 shippers more vulnerable to any conspiracy.\\nThis line of reasoning parallels one rejected by the Supreme\\nCourt in Wal-Mart. There, an expert opined that Wal-Mart’s\\n“strong corporate culture” made it “vulnerable to gender bias.”\\n564 U.S. at 354 (quotation marks omitted). The Court found\\nthis evidence insufficient to prove that injury could be\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 15\\nestablished on a class-wide basis, because the expert could not\\nsay what percentage of adverse employment decisions were in\\nfact caused by bias. Id. The McClave study is similarly\\nincomplete because, as explained above, it does not attempt to\\nidentify which of the small shippers, or what percentage of\\nthem, were in fact harmed by the alleged conspiracy.\\n\\n VI\\n\\n In Asacol, the First Circuit noted the absence of even a\\nsingle case “allowing, under Rule 23, a trial in which thousands\\nof class members testify.” 907 F.3d at 57—58. That Court\\ndeclined to create “the first such case.” Id. So do we. Given\\nthe need in this case for at least 2,037 individual determinations\\nof injury and causation, the district court did not abuse its\\ndiscretion in denying class certification on the ground that\\ncommon issues do not predominate.\\n\\n Affirmed\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4426251/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge KATSAS\"}]}"},"judge1":{"kind":"string","value":"GARLAND"},"judge2":{"kind":"string","value":"ROGERS"},"judge3":{"kind":"string","value":"KATSAS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4648998/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: In RE: RAIL FREIGHT FUEL SURCHARGE ANTITRUST LITIGATION - MDL NO. 1869, Dakota Granite Company, on Behalf of Itself and All Others Similarly Situated, Et Al., Appellants v. BNSF Railway Company, Et Al., Appellees\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-16\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nPUBLIC COPY SEALED INFORMATION DELETED\n -\n\n\n\n\n AnUcb $ttt ffnurI of ppitl\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued September 28, 201$ Decided August 16, 2019\n\n No. 18-7010\n\n IN RE: RAIL FREIGHT FUEL SURCHARGE ANTITRUST\n LITIGATION - MDL No. 1 $69,\n\n DAKOTA GRANITE COMPANY, ON BEHALF OF ITSELF AND ALL\n OTHERS SIMILARLY SITUATED, ET AL.,\n APPELLANTS\n\n\n V.\n\n\n BNSF RAILWAY COMPANY, ET AL.,\n APPELLEES\n\n\n\n Appeal from the United States District Court\n for the District of Columbia\n (No. 1 :07-mc-00489)\n\n\n Kathleen M Sullivan argued the cause for appellants.\nWith her on the briefs were Stephen R. Neuwirth, Sami H\nRashid, Michael D. Hausfeld, and Michael?. Lehmann.\n\n Carter G. Phillips argued the cause for appellees. With\nhim on the brief were Joseph R. Guerra, Kathleen Moriarty\nMueller, Saul?. Morgenstern, Thomas A. Isaacson, John li\nNannes, Tara L. Reinhart, J Scott Ballenger, Veronica S.\nLewis, Samuel M Sipe, Jr., Linda S. Stein, Andrew S.\nTulumello, Lucas C. Townsend, and Kent A. Gardiner.\n\f PUBLIC COPY - SEALED INFORMATION DELETED\n\n 2\n\n\n Anton Metlitsky and Warren D. Postman were on the brief\nfor amicus curiae Chamber of Commerce of the United States\nof America in support of defendants-appellees.\n\n Before: GARLAND, ChiefJudge, and ROGERS and KATSAS,\nCircuit Judges.\n\n Opinion for the Court filed by Circuit Judge KATSAS.\n\n\n KATSAS, Circuit Judge:’ This case involves a putative\nclass of over 16,000 shippers allegedly harmed by a price-\nfixing conspiracy among the nation’s largest freight railroads.\nThe district court denied class certification because the\nplaintiffs’ regression analysis—their evidence for proving\ncausation, injury, and damages on a class-wide basis—\nmeasured negative damages for over 2,000 members of the\nproposed class. Based on that consideration, we affirm.\n\n I\n\n This appeal arises out of eighteen antitrust actions\nconsolidated by the Multidistrict Litigation Panel. The\ndefendants are the four largest freight railroads in the United\nStates: BNSF Railway Company; CSX Transportation, Inc.;\nNorfolk Southern Railway Company; and Union Pacific\nRailroad Company. The plaintiffs, who are their customers,\nallege that the railroads conspired to fix rate-based fuel\nsurcharges. Railroads impose fuel surcharges—additional\ncharges above the base shipping price—when the price of fuel\nrises above a certain trigger price. Rate-based surcharges are\ncalculated as a percentage of the base shipping price.\n\n\n NOTE: Portions of this opinion contain Sealed Information,\nwhich has been redacted.\n\f PUBLIC COPY SEALED INFORMATION DELETED\n -\n\n\n\n\n Following consolidation, the action was divided into one\ncase involving direct purchasers and another involving indirect\npurchasers. All plaintiffs alleged that the railroads violated\nsection 1 of the Sherman Act, 15 U.S.C. § 1, by conspiring to\nfix prices. The direct purchasers sought treble damages under\nsection 4 of the clayton Act, 15 U.S.C. § 15, and the district\ncourt held that they stated a claim, In re Rail Freight Surcharge\nAntitrustLitig., 587 F. Supp. 2d27(D.D.c. 2008). The indirect\npurchasers sought injunctive relief under section 16 of the\nClayton Act, 15 U.S.C. § 26, and raised various state-law\nclaims. The district court held that the state claims were\npreempted by federal law, but it declined to dismiss the federal\nclaims. In re Rail Freight Surcharge Antitrust Litig., 593 F.\nSupp. 2d 29 (D.D.C. 2008), afT d, Fayus Enters. v. BNSF Ry.\nCo., 602 F.3d 444 (D.C. Cir. 2010).\n\n The eight named plaintiffs in the direct-purchaser case—\nCarter Distributing Company; Dakota Granite Company;\nDonnelly Commodities, Inc.; Dust Pro, Inc.; Nyrstar Taylor\nChemicals, Inc.; Olin Corporation; Strates Shows, Inc.; and US\nMagnesium LLC—rnoved to certify a class under Federal Rule\nof Civil Procedure 23(b)(3). The proposed class consisted of\nall shippers who paid rate-based fuel surcharges for\nunregulated services purchased from the defendants between\nJuly 1,2003 and December 31, 2008. To show that causation,\ninjury, and damages could be proved on a class-wide basis, the\nplaintiffs invoked two regression models constructed by their\neconomist, Dr. Gordon Rausser. The “common factor model”\nidentified seven variables said to determine the price of the\ndefendants’ services, including fuel surcharges. The “damages\nmodel,” controlling for those variables, sought to isolate price\nincreases attributable to the alleged conspiracy. The railroads\ncriticized these models on various grounds, including that they\nmeasured damages for shipments made under legacy contracts\nfixed before any conspiracy allegedly began.\n\f PUBLIC COPY SEALED INFORMATION DELETED\n -\n\n\n\n\n 4\n The district court initially certified the class. It noted that\nif individualized proof were necessary to establish causation\nand injury, then the plaintiffs could not satisfy the Rule\n23(b)(3) requirement that common questions predominate. See\nIn re Rail Freight Fuel Surcharge Antitrust Litig., 287 F.R.D.\n1, 43 (D.D.C. 2012). But the court found Dr. Rausser’s\nregression analysis to be “plausible” and “workable,” so it\nconcluded that causation, injury, and damages were\n“susceptible to proof at trial through evidence common to the\nclass.” Id. at 67. The court rejected many different criticisms\nof the regression models, but it did not specifically address the\nquestion of false positives for legacy contracts.\n\n On interlocutory review, we vacated the certification order\nand remanded for reconsideration in light of Comcast Corp. V.\nBehrend, 569 U.S. 27 (2013). In re Rail Freight Surcharge\nAntitrustLitig.—MDLNo. 1869, 725 F.3d 244 (D.C. Cir. 2013)\n(Rail Freight 1). We explained that, for an antitrust class\naction, common questions “cannot predominate where there\nexists no reliable means of proving classwide injury in fact.”\nId. at 253. We expressed concern with the district court’s\nfailure to address “the damages model’s propensity toward\nfalse positives,” which left us with no way of knowing whether\n“the overcharges the damages model calculates for class\nmembers [are] any more accurate than the obviously false\nestimates it produces for legacy shippers.” Id. at 254. Finally,\nwe stressed that Rule 23, as construed in Comcast, requires a\n“hard look at the soundness of statistical models that purport to\nshow predominance.” Id. at 255.\n\n On remand, after permitting supplemental discovery and\nexpert reports, the district court denied class certification. In\nre Rail Freight Surcharge Antitrust Litig., 292 F. Supp. 3d 14\n(D.D.C. 2017) (Rail Freight If). The court concluded that Dr.\nRausser’s expert opinions were reliable enough to be\nadmissible at trial. Id. at 49—63. But in assessing\n\f PUBLIC COPY - SEALED INFORMATION DELETED\n\n 5\npredominance, the court identified three shortcomings in his\ndamages model: first, it measured highly inflated damages for\nintermodal traffic (i.e., shipments traveling by rail and another\nmode of transportation such as trucks or airplanes), Id. at 122—\n26; second, as we had noted in the earlier appeal, the model\nerroneously measured damages for shipments made under\nlegacy contracts, Id. at 126—31; and third, the model measured\nnegative damages—and hence no injury—for over 2,000\nmembers of the proposed class, id. at 132—41. The court\nconcluded that any one of these problems was enough to defeat\nthe plaintiffs’ argument for predominance. Id. at 122.\n\n The plaintiffs filed a petition for permission to appeal the\nclass-certification decision under Federal Rule of Civil\nProcedure 23(f). A motions panel of this Court granted the\npetition without prejudice to reconsideration at the merits\nstage. In re Rail freight fuel Surcharge Antitrust Litig.—MDL\nNo. 1869, No. 17-8005 (D.C. Cir. Dec. 20, 2017).\n\n II\n\n We begin with the question of our jurisdiction. Orders\ndenying class certification are neither final decisions under 28\nU.S.C. § 1291, Coopers & Lybrand v. Livesay, 437 U.S. 463\n(1978), nor injunctions immediately appealable under 28\nU.S.C. § 1292(a)(l), Gardner v. Westinghouse Broad Co., 437\nU.S. 478 (1978). However, 28 U.S.C. § 1292(e) permits the\nSupreme Court to promulgate rules creating new categories of\ndecisions appealable before final judgment. Exercising that\nauthority, the Court has provided that “[a] court of appeals may\npermit an appeal from an order granting or denying class-action\ncertification,” if a “petition for permission to appeal” is timely\nfiled. Fed. R. Civ. P. 23(f).\n\n In this case, the plaintiffs filed a timely petition for\npermission to appeal, which was enough under Rule 23(f) to\n\f PUBLIC COPY - SEALED INFORMATION DELETED\n\n 6\nsecure our jurisdiction. That jurisdiction is “discretionary,”\nRail freight I, 725 F.3d at 250, and the railroads contest\nwhether we should exercise it. But their argument on this point\nis perfunctory—less than one page of briefing, with no case\ncitations—and it almost entirely duplicates their merits\narguments. According to the railroads, we should conclude\nthat the denial of class certification was correct. Then, we\nshould dismiss the appeal as raising neither a questionable\ndecision nor an unsettled issue—considerations that bear on\nwhether to permit the appeal in the first place, see id. at 250—\n54. Because that disposition would make little sense at this\njuncture, we decline to revisit the motions panel’s decision\naccepting the appeal.\n\n III\n\n Federal Rule of Civil Procedure 23 sets forth various\nrequirements for the certification of class actions. Rule 23(a)\nprovides four “prerequisites” for any class certification,\nincluding that there must be “questions of law or fact common\nto the class.” If these prerequisites are met, Rule 23(b)(3)\npermits certification if, among other things, “questions of law\nor fact common to class members predominate over any\nquestions affecting only individual members.” For purposes of\nthese rules, a “common” question is one that is “capable of\nclasswide resolution—which means that determination of its\ntruth or falsity will resolve an issue that is central to the validity\nof each one of the claims in one stroke.” Wal-Mart Stores, Inc.\nv. Dukes, 564 U.S. 338, 350 (2011). In contrast, an\n“individual” question is one for which “members of a proposed\nclass will need to present evidence that varies from member to\nmember.” Tyson Foods, Inc. v. Bouaphakeo, 136 S. Ct. 1036,\n 1045 (2016) (quotation marks omitted).\n\n The party seeking class certification “must affirmatively\ndemonstrate” that the commonality and predominance\n\f PUBLIC COPY SEALED INFORMATION DELETED\n -\n\n\n\n\n 7\nrequirements are satisfied. Wal-Mart, 564 U.S. at 350. This\nrequires a “rigorous analysis” that often will “overlap with the\nmerits.” Id. at 351 (quotation marks omitted). Three recent\ncases address the contours of this analysis. In Wal-Mart, the\nSupreme Court ordered decertification of a Title VII class\nwhere the plaintiffs presented insufficient proof that the\ndefendant had engaged in a general policy of sex\ndiscrimination. Id. at 352—55. In Corncast, the Court ordered\ndecertification where the regression analysis used to show\ncommon injury did not track the underlying theory of liability.\n569 U.S. at 36—38. The Court rejected a contention that, at the\nclass-certification stage, “any method of measurement is\nacceptable so long as it can be applied classwide, no matter\nhow arbitrary the measurements may be.” Id. at 36. But in\nTyson foods, the Court held that concerns about the validity of\na statistical sample used to prove class-wide averages—which\nthe Court described as presenting a common objection to the\nclaims of each class member—should be addressed “as a\nmatter of summary judgment, not class certification.” 136 5.\nCt. at 1047 (quotation marks omitted).\n\n IV\n\n The direct-purchaser plaintiffs raise claims under section\n4 of the Clayton Act, which provides treble damages to any\nperson “injured in his business or property by reason of\nanything forbidden in the antitrust laws.” 15 U.S.C. § 15(a).\nTo establish liability under section 4, each plaintiff must prove\nnot only an antitrust violation, but also an injury to its business\nor property and a causal relation between the two. Without\ncommon proof of injury and causation, section 4 plaintiffs\ncannot establish predominance. See, e.g., Comcast, 569 U.S.\nat 36—38; Rail freight I, 725 F.3d at 252—53.\n\n The parties dispute the extent to which a court, in\nconducting the “hard look” required by Rule 23, should assess\n\f PUBLIC COPY - SEALED INFORMATION DELETED\n\n 8\nthe reliability of common evidence. Specifically, they contest\nwhether Rule 23 requires reliability above and beyond what is\nnecessary to establish the admissibility of expert testimony\nunder Federal Rule of Evidence 702 and Daubert v. Merrell\nDow Pharmaceuticals, Inc., 509 U.S. 579 (1993). That\nquestion arose because the district court found Dr. Rausser’s\nregression models to be reliable for admissibility purposes,\nRail freight II, 292 F. Supp. 3d at 54—63; held that “reliability\nunder Rule 23 is a higher standard than reliability under\nDaubert,” id. at 91; and found no predominance based on\nconcerns that “undermine[d] the reliability of Dr. Rausser’s\ndamages model,” id. at 122. The defendants embrace this\nreasoning; they argue that Wal-Mart and Comcast rejected\ncertification based on concerns that common evidence was\nunreliable, even though its admissibility was uncontested. The\nplaintiffs disagree; they contend that predominance ultimately\nturns on whether the relevant evidence is common or\nindividualized, and they read Tyson foods as holding that a\ncourt may assess its reliability only in deciding admissibility or\n summary judgment.\n\n We need not resolve this dispute because Dr. Rausser’s\ndamages model, even if sufficiently reliable, does not prove\nclasswide injury. As the district court explained, his model\nindicates that the proposed class consists of 16,065 shippers.\nRail freight II, 292 F. $upp. 3d at 136. The plaintiffs maintain\nthat the alleged conspiracy injured every one of them. Yet the\ndamages model also indicates that 2,037 members of the\nproposed class—or 12.7 percent—suffered “only negative\novercharges” and thus no injury from any conspiracy. Id. at\n137. So even assuming the model can reliably show injury and\ncausation for 87.3 percent of the class, that still leaves the\nplaintiffs with no common proof of those essential elements of\nliability for the remaining 12.7 percent. The district court held\nthat the need for “individualized inquiries to determine which\nof at least 2,037 (and possibly more) class members were\n\f PUBLIC COPY SEALED INFORMATION DELETED\n -\n\n\n\n\n 9\nactually injured by the alleged conspiracy,” Id. at 140,\nprecluded a finding of predominance. Id. at 122. We review a\ndistrict court’s assessment of predominance only for abuse of\ndiscretion. Garcia v. Johanns, 444 F.3d 625, 631 (D.C. Cir.\n2006). Here, we find no such abuse.\n\n The plaintiffs argue that their model measures these\nnegative damages only because of normal prediction error. The\ndistrict court found that prediction error could not “account for\nall—or even a substantial portion of—the 2,037 shippers that\nthe model shows to be uninjured.” Rail Freight II, 292 F. Supp.\n3d at 139. The plaintiffs take issue with that finding, but it is\nnot clearly erroneous. In any event, the plaintiffs’ argument\nabout prediction error at most suggests that their damages\nmodel might falsely have measured no injury for as many as\n2,037 shippers. This line of reasoning describes a possible\nproblem with their own evidence; it does not point to\naffirmative evidence—much less common affirmative\nevidence—that a conspiracy did in fact injure these shippers.\n\n The plaintiffs further argue that predominance does not\nrequire common evidence extending to all class members.\nThat contention appears inconsistent with our statement in Rail\nfreight I that the plaintiffs, to establish predominance, must\n“show that they can prove, through common evidence, that all\nclass members were in fact injured by the alleged conspiracy.”\n725 F.3d at 252; see also id. (“we do expect the common\nevidence to show all class members suffered some injury”).\nDespite these statements, the district court held that our opinion\ndid not require common evidence of injury to all class\nmembers. See Rail Freight II, 292 F. Supp. 3d at 132—34.\nInstead, it agreed with the plaintiffs that common proof\ncovering “virtually all” members of the proposed class, and\nleaving only a “de minimis” number of cases requiring\nindividualized proof of injury and causation, would be enough\nto show predominance. Id. at 135. For the sake of argument,\n\f PUBLIC COPY SEALED INFORMATION DELETED\n -\n\n\n\n\n 10\nwe assume that the district court correctly recognized a de\nminimis exception to the general rule that, for claims under\nsection 4 of the Clayton Act, causation and injury must be\n“capable of classwide resolution,” Wal-Mart, 564 U.S. at 350.\nThe court reasonably concluded that such a de minimis\nexception would not encompass this case.\n\n In assessing how many individual adjudications are too\nmany, both the district court and the parties invoke cases\naddressing the question of when, if ever, a class may include\nconcededly uninjured members. Strictly speaking, this case\ndoes not present that question, for the plaintiffs here insist that\neach member of the proposed class was injured. Nonetheless,\nthe cited cases bear some similarity to this one: Uninjured class\nmembers cannot prevail on the merits, so their claims must be\nwinnowed away as part of the liability determination. And that\nprospect raises the same kind of question at issue here—when\ndoes the need for individualized proof of injury and causation\ndestroy predominance? See, e.g., In re Asacol Antitrust Litig.,\n907 F.3d 42, 51—58 (1st Cir. 2018); In re Nexium Antitrust\nLitig., 777 F.3d 9, 18—22 (1st Cir. 2015).\n\n The plaintiffs complain that the district court arbitrarily\nimposed a six-percent upper limit on the percentage of\nuninjured parties who may be included in a certified class. In\nfact, the court’s analysis was more nuanced. As the court\nexplained, the “few reported decisions” involving uninjured\nclass members “suggest that 5% to 6% constitutes the outer\nlimits of a de minimis number.” Rail Freight ll 292 F. Supp.\n3d at 137. The 12.7 percent figure in this case is more than\ntwice that approximate upper bound reflected in analogous\ncaselaw. Moreover, the district court considered raw numbers\nas well as percentages: six percent of a “class totaling only\nfifty-five” members might be de minimis, but 12.7 percent of\nthis class yields “2,037 uninjured class members” (according\nto the common proof), all of whom would need individualized\n\f PUBLIC COPY - SEALED INFORMATION DELETED\n\n 11\nadjudications of causation and injury. Id. at 137—38. Finally,\nthe district court stressed that the plaintiffs have proposed no\n“further way”—short of full-blown, individual trials—”to\nreduce this number and segregate the uninjured from the truly\ninjured.” Id. at 138. None of this was an abuse of discretion.\n\n The absence of any winnowing mechanism sharply\ndistinguishes Nexiurn, the plaintiffs’ best case. There, the class\nincluded purchasers of a drug allegedly shielded from\ncompetition by the unlawful suppression of a generic\nalternative. 777 F.3d at 13—14. The problem of uninjured class\nmembers arose because a small percentage of the class, due to\nbrand loyalty, would have purchased the drug even if a less\nexpensive generic alternative had been available. Id. at 19—20.\nThe First Circuit held that the uninjured class members could\nmanageably be winnowed by having individual consumers file\nminimal, likely unrebutted affidavit testimony indicating\nwhether, if given the choice, they would have purchased the\nbranded drug or a generic alternative. Id. at 20—21.\n\n Nexium does not support class certification here. For one\nthing, the First Circuit sharply limited that decision in Asacol.\nThere, the Court explained that any winnowing mechanism\nmust be truncated enough to ensure that the common issues\npredominate, yet robust enough to preserve the defendants’\nSeventh Amendment and due process rights to contest every\nelement of liability and to present every colorable defense. 907\nF.3d at 5 1—54. Moreover, the Court held that Nexium’s\naffidavit mechanism could not satisfy both conditions where\nthe defendant seeks to contest the question whether individual\nclass members would have shifted from the branded drug to a\nless expensive generic alternative. See Id. That would require\nindividual trials because genuinely contested affidavits do not\nsupport summary judgment and are inadmissible. Id. Here, the\ndefendants intend to contest whether any of the 2,037 shippers\nsuffered injury as a result of any conspiracy. And the question\n\f PUBLIC COPY SEALED INFORMATION DELETED\n -\n\n\n\n\n 12\npresented in these individual challenges—regarding impacts\non shippers of different sizes, shipping different products in\ndifferent geographic markets, with different transportation\noptions and different degrees of leverage—would be far more\ncomplex than the single unitary question (branded or generic?)\nat issue in Nexium and Asacol.\n\n The plaintiffs also invoke the Supreme Court’s discussion\nof uninjured class members in Tyson foods. The defendants\nthere sought review of the question whether a certified class\nmay contain any uninjured members. But the Supreme Court\nreserved that question, 136 5. Ct. at 1049, and the district court\nhere decided it in the plaintiffs’ favor, Rail Freight II, 292 F.\nSupp. 3d at 133—35. So, this aspect of Tyson foods does not\nadvance the plaintiffs’ appeal. The Supreme Court also\naddressed a further ‘new argument” raised for the first time in\nmerits briefing—that plaintiffs at the certification stage must\nprove that all class members were injured or establish a\nmanageable process for culling out uninjured class members.\n 136 5. Ct. at 1049. The Court held that this argument was\n“premature” because the parties disagreed about what culling\nmechanisms might be available and the district court had not\nyet addressed the question. See 136 5. Ct. at 1050. This\nholding rested on the inappropriateness of raising new issues\nfor the first time in Supreme Court merits briefing. See id. It\ndoes not, as the plaintiffs here contend, perniit district courts\nconsidering class certification to defer questions about the\nnumber and nature of any individualized inquiries that might\nbe necessary to establish liability. To the contrary, confronting\nsuch questions is part-and-parcel of the “hard look” required\nby Wal-Mart and Comcast, as recognized even by those courts\npermitting a class to include some small number of concededly\nuninjured individuals. See, e.g., In re Asacol, 907 F.3d at 51—\n54; In re Nexiurn,777 F.3d at 19—2 1.\n\f PUBLIC COPY SEALED INFORMATION DELETED\n -\n\n\n\n\n 1—,\n 1.)\n\n\n Finally, the plaintiffs argue that the 2,037 class members\nfor whom their damages model shows no injury are a de\nminimis portion of the class because their shipments make up\nless than one percent of the railroads’ overall revenue from the\nalleged conspiracy. But revenue is irrelevant to predominance,\nwhich looks to whether elements such as causation and injury\nmay be proved through common evidence, not how much the\ndefendants benefited from any wrongdoing.\n\n V\n\n Looking beyond Dr. Rausser’s regression analysis, the\nplaintiffs point to other evidence that they say can prove injury\nand causation on a class-wide basis. In Rail freight I, we\nconcluded that Dr. Rausser’s analysis was “essential” to the\nplaintiffs’ case for certification: “No damages model, no\npredominance, no class certification.” 725 F.3d at 253. The\ndistrict court reached the same conclusion on remand, after\ncareful review of all the documentary and expert evidence.\nThis was not an abuse of discretion.\n\n The plaintiffs invoke documentary evidence that the\ndefendants enforced fuel surcharges “uniformly and with few\nexceptions.” Rail Freight II, 292 F. $upp. 3d at 122; see id. at\n103—07. But imposing fuel surcharges does not show injury\ncaused by a conspiracy. The parties vigorously dispute\nwhether higher overall prices during the class period were\nattributable to causes besides any conspiracy—such as the\nmarked increase in fuel prices that occurred around the\nbeginning of the class period, see l.A. 65 92—93. As the district\ncourt explained, Dr. Rausser designed his regression models\nprecisely to control for these kinds of potential alternative\ncauses, see Rail Freight II, 292 F. $upp. 3d at 101, yet his\ndamages model showed that 2,037 shippers were uninjured\ndespite paying fuel surcharges. The plaintiffs’ evidence of\nwidespread fuel surcharges helps explain why the proposed\n\f PUBLIC COPY - SEALED INFORMATION DELETED\n\n 14\nclass is a large one, but it neither proves that the 2,037 shippers\nwere injured by the alleged conspiracy nor otherwise compels\na finding of predominance.\n\n Next, the plaintiffs invoke the expert testimony of Dr.\nJames McClave, who argued that the 2,037 shippers must have\nbeen harmed by the conspiracy. Dr. McClave reasoned that\nbecause these shippers made fewer purchases on average, and\nthus had less bargaining power than the rest of the class, they\nmust have been more susceptible to injury. Moreover, Dr.\nMcClave’s own study concluded that the defendants’ smallest\ncustomers—roughly shippers who made only one\npurchase I I—collectively\npaid higher prices than did larger shippers.\n\n This analysis is not common proof of injury to the 2,037\nshippers. For one thing, of the shippers studied by Dr.\nMcClave, only were in the class, and Dr. McClave did\nnot indicate how many of them were included among the 2,037\nclass members for whom Dr. Rausser’s model showed no\ninjury. Moreover, evidence that a group of one-time shippers\nmay have paid higher prices collectively does not prove that all\n(or almost all) of them were injured individually. And, as the\ndistrict court explained, Dr. McClave did not attempt to\nidentify how many of the one-time shippers did in fact pay\nmore, let alone which of them were among the 2,037. See Rail\nFreight II, 292 F. $upp. 3d at 139.\n\n The plaintiffs respond that Dr. McClave’s analysis, even if\ninconclusive, at least suggests that reduced bargaining power\nmade the 2,037 shippers more vulnerable to any conspiracy.\nThis line of reasoning parallels one rejected by the Supreme\nCourt in Wal-Mart. There, an expert opined that Wal-Mart’s\n“strong corporate culture” made it “vulnerable to gender bias.”\n564 U.S. at 354 (quotation marks omitted). The Court found\nthis evidence insufficient to prove that injury could be\n\f PUBLIC COPY SEALED INFORMATION DELETED\n -\n\n\n\n\n 15\nestablished on a class-wide basis, because the expert could not\nsay what percentage of adverse employment decisions were in\nfact caused by bias. Id. The McClave study is similarly\nincomplete because, as explained above, it does not attempt to\nidentify which of the small shippers, or what percentage of\nthem, were in fact harmed by the alleged conspiracy.\n\n VI\n\n In Asacol, the First Circuit noted the absence of even a\nsingle case “allowing, under Rule 23, a trial in which thousands\nof class members testify.” 907 F.3d at 57—58. That Court\ndeclined to create “the first such case.” Id. So do we. Given\nthe need in this case for at least 2,037 individual determinations\nof injury and causation, the district court did not abuse its\ndiscretion in denying class certification on the ground that\ncommon issues do not predominate.\n\n Affirmed\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: In RE: RAIL FREIGHT FUEL SURCHARGE ANTITRUST LITIGATION - MDL NO. 1869, Dakota Granite Company, on Behalf of Itself and All Others Similarly Situated, Et Al., Appellants v. BNSF Railway Company, Et Al., Appellees\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-16\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nPUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n AnUcb $ttt ffnurI of ppitl\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued September 28, 201$ Decided August 16, 2019\\n\\n No. 18-7010\\n\\n IN RE: RAIL FREIGHT FUEL SURCHARGE ANTITRUST\\n LITIGATION - MDL No. 1 $69,\\n\\n DAKOTA GRANITE COMPANY, ON BEHALF OF ITSELF AND ALL\\n OTHERS SIMILARLY SITUATED, ET AL.,\\n APPELLANTS\\n\\n\\n V.\\n\\n\\n BNSF RAILWAY COMPANY, ET AL.,\\n APPELLEES\\n\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1 :07-mc-00489)\\n\\n\\n Kathleen M Sullivan argued the cause for appellants.\\nWith her on the briefs were Stephen R. Neuwirth, Sami H\\nRashid, Michael D. Hausfeld, and Michael?. Lehmann.\\n\\n Carter G. Phillips argued the cause for appellees. With\\nhim on the brief were Joseph R. Guerra, Kathleen Moriarty\\nMueller, Saul?. Morgenstern, Thomas A. Isaacson, John li\\nNannes, Tara L. Reinhart, J Scott Ballenger, Veronica S.\\nLewis, Samuel M Sipe, Jr., Linda S. Stein, Andrew S.\\nTulumello, Lucas C. Townsend, and Kent A. Gardiner.\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 2\\n\\n\\n Anton Metlitsky and Warren D. Postman were on the brief\\nfor amicus curiae Chamber of Commerce of the United States\\nof America in support of defendants-appellees.\\n\\n Before: GARLAND, ChiefJudge, and ROGERS and KATSAS,\\nCircuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge KATSAS.\\n\\n\\n KATSAS, Circuit Judge:’ This case involves a putative\\nclass of over 16,000 shippers allegedly harmed by a price-\\nfixing conspiracy among the nation’s largest freight railroads.\\nThe district court denied class certification because the\\nplaintiffs’ regression analysis—their evidence for proving\\ncausation, injury, and damages on a class-wide basis—\\nmeasured negative damages for over 2,000 members of the\\nproposed class. Based on that consideration, we affirm.\\n\\n I\\n\\n This appeal arises out of eighteen antitrust actions\\nconsolidated by the Multidistrict Litigation Panel. The\\ndefendants are the four largest freight railroads in the United\\nStates: BNSF Railway Company; CSX Transportation, Inc.;\\nNorfolk Southern Railway Company; and Union Pacific\\nRailroad Company. The plaintiffs, who are their customers,\\nallege that the railroads conspired to fix rate-based fuel\\nsurcharges. Railroads impose fuel surcharges—additional\\ncharges above the base shipping price—when the price of fuel\\nrises above a certain trigger price. Rate-based surcharges are\\ncalculated as a percentage of the base shipping price.\\n\\n\\n NOTE: Portions of this opinion contain Sealed Information,\\nwhich has been redacted.\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n Following consolidation, the action was divided into one\\ncase involving direct purchasers and another involving indirect\\npurchasers. All plaintiffs alleged that the railroads violated\\nsection 1 of the Sherman Act, 15 U.S.C. § 1, by conspiring to\\nfix prices. The direct purchasers sought treble damages under\\nsection 4 of the clayton Act, 15 U.S.C. § 15, and the district\\ncourt held that they stated a claim, In re Rail Freight Surcharge\\nAntitrustLitig., 587 F. Supp. 2d27(D.D.c. 2008). The indirect\\npurchasers sought injunctive relief under section 16 of the\\nClayton Act, 15 U.S.C. § 26, and raised various state-law\\nclaims. The district court held that the state claims were\\npreempted by federal law, but it declined to dismiss the federal\\nclaims. In re Rail Freight Surcharge Antitrust Litig., 593 F.\\nSupp. 2d 29 (D.D.C. 2008), afT d, Fayus Enters. v. BNSF Ry.\\nCo., 602 F.3d 444 (D.C. Cir. 2010).\\n\\n The eight named plaintiffs in the direct-purchaser case—\\nCarter Distributing Company; Dakota Granite Company;\\nDonnelly Commodities, Inc.; Dust Pro, Inc.; Nyrstar Taylor\\nChemicals, Inc.; Olin Corporation; Strates Shows, Inc.; and US\\nMagnesium LLC—rnoved to certify a class under Federal Rule\\nof Civil Procedure 23(b)(3). The proposed class consisted of\\nall shippers who paid rate-based fuel surcharges for\\nunregulated services purchased from the defendants between\\nJuly 1,2003 and December 31, 2008. To show that causation,\\ninjury, and damages could be proved on a class-wide basis, the\\nplaintiffs invoked two regression models constructed by their\\neconomist, Dr. Gordon Rausser. The “common factor model”\\nidentified seven variables said to determine the price of the\\ndefendants’ services, including fuel surcharges. The “damages\\nmodel,” controlling for those variables, sought to isolate price\\nincreases attributable to the alleged conspiracy. The railroads\\ncriticized these models on various grounds, including that they\\nmeasured damages for shipments made under legacy contracts\\nfixed before any conspiracy allegedly began.\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 4\\n The district court initially certified the class. It noted that\\nif individualized proof were necessary to establish causation\\nand injury, then the plaintiffs could not satisfy the Rule\\n23(b)(3) requirement that common questions predominate. See\\nIn re Rail Freight Fuel Surcharge Antitrust Litig., 287 F.R.D.\\n1, 43 (D.D.C. 2012). But the court found Dr. Rausser’s\\nregression analysis to be “plausible” and “workable,” so it\\nconcluded that causation, injury, and damages were\\n“susceptible to proof at trial through evidence common to the\\nclass.” Id. at 67. The court rejected many different criticisms\\nof the regression models, but it did not specifically address the\\nquestion of false positives for legacy contracts.\\n\\n On interlocutory review, we vacated the certification order\\nand remanded for reconsideration in light of Comcast Corp. V.\\nBehrend, 569 U.S. 27 (2013). In re Rail Freight Surcharge\\nAntitrustLitig.—MDLNo. 1869, 725 F.3d 244 (D.C. Cir. 2013)\\n(Rail Freight 1). We explained that, for an antitrust class\\naction, common questions “cannot predominate where there\\nexists no reliable means of proving classwide injury in fact.”\\nId. at 253. We expressed concern with the district court’s\\nfailure to address “the damages model’s propensity toward\\nfalse positives,” which left us with no way of knowing whether\\n“the overcharges the damages model calculates for class\\nmembers [are] any more accurate than the obviously false\\nestimates it produces for legacy shippers.” Id. at 254. Finally,\\nwe stressed that Rule 23, as construed in Comcast, requires a\\n“hard look at the soundness of statistical models that purport to\\nshow predominance.” Id. at 255.\\n\\n On remand, after permitting supplemental discovery and\\nexpert reports, the district court denied class certification. In\\nre Rail Freight Surcharge Antitrust Litig., 292 F. Supp. 3d 14\\n(D.D.C. 2017) (Rail Freight If). The court concluded that Dr.\\nRausser’s expert opinions were reliable enough to be\\nadmissible at trial. Id. at 49—63. But in assessing\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 5\\npredominance, the court identified three shortcomings in his\\ndamages model: first, it measured highly inflated damages for\\nintermodal traffic (i.e., shipments traveling by rail and another\\nmode of transportation such as trucks or airplanes), Id. at 122—\\n26; second, as we had noted in the earlier appeal, the model\\nerroneously measured damages for shipments made under\\nlegacy contracts, Id. at 126—31; and third, the model measured\\nnegative damages—and hence no injury—for over 2,000\\nmembers of the proposed class, id. at 132—41. The court\\nconcluded that any one of these problems was enough to defeat\\nthe plaintiffs’ argument for predominance. Id. at 122.\\n\\n The plaintiffs filed a petition for permission to appeal the\\nclass-certification decision under Federal Rule of Civil\\nProcedure 23(f). A motions panel of this Court granted the\\npetition without prejudice to reconsideration at the merits\\nstage. In re Rail freight fuel Surcharge Antitrust Litig.—MDL\\nNo. 1869, No. 17-8005 (D.C. Cir. Dec. 20, 2017).\\n\\n II\\n\\n We begin with the question of our jurisdiction. Orders\\ndenying class certification are neither final decisions under 28\\nU.S.C. § 1291, Coopers & Lybrand v. Livesay, 437 U.S. 463\\n(1978), nor injunctions immediately appealable under 28\\nU.S.C. § 1292(a)(l), Gardner v. Westinghouse Broad Co., 437\\nU.S. 478 (1978). However, 28 U.S.C. § 1292(e) permits the\\nSupreme Court to promulgate rules creating new categories of\\ndecisions appealable before final judgment. Exercising that\\nauthority, the Court has provided that “[a] court of appeals may\\npermit an appeal from an order granting or denying class-action\\ncertification,” if a “petition for permission to appeal” is timely\\nfiled. Fed. R. Civ. P. 23(f).\\n\\n In this case, the plaintiffs filed a timely petition for\\npermission to appeal, which was enough under Rule 23(f) to\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 6\\nsecure our jurisdiction. That jurisdiction is “discretionary,”\\nRail freight I, 725 F.3d at 250, and the railroads contest\\nwhether we should exercise it. But their argument on this point\\nis perfunctory—less than one page of briefing, with no case\\ncitations—and it almost entirely duplicates their merits\\narguments. According to the railroads, we should conclude\\nthat the denial of class certification was correct. Then, we\\nshould dismiss the appeal as raising neither a questionable\\ndecision nor an unsettled issue—considerations that bear on\\nwhether to permit the appeal in the first place, see id. at 250—\\n54. Because that disposition would make little sense at this\\njuncture, we decline to revisit the motions panel’s decision\\naccepting the appeal.\\n\\n III\\n\\n Federal Rule of Civil Procedure 23 sets forth various\\nrequirements for the certification of class actions. Rule 23(a)\\nprovides four “prerequisites” for any class certification,\\nincluding that there must be “questions of law or fact common\\nto the class.” If these prerequisites are met, Rule 23(b)(3)\\npermits certification if, among other things, “questions of law\\nor fact common to class members predominate over any\\nquestions affecting only individual members.” For purposes of\\nthese rules, a “common” question is one that is “capable of\\nclasswide resolution—which means that determination of its\\ntruth or falsity will resolve an issue that is central to the validity\\nof each one of the claims in one stroke.” Wal-Mart Stores, Inc.\\nv. Dukes, 564 U.S. 338, 350 (2011). In contrast, an\\n“individual” question is one for which “members of a proposed\\nclass will need to present evidence that varies from member to\\nmember.” Tyson Foods, Inc. v. Bouaphakeo, 136 S. Ct. 1036,\\n 1045 (2016) (quotation marks omitted).\\n\\n The party seeking class certification “must affirmatively\\ndemonstrate” that the commonality and predominance\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 7\\nrequirements are satisfied. Wal-Mart, 564 U.S. at 350. This\\nrequires a “rigorous analysis” that often will “overlap with the\\nmerits.” Id. at 351 (quotation marks omitted). Three recent\\ncases address the contours of this analysis. In Wal-Mart, the\\nSupreme Court ordered decertification of a Title VII class\\nwhere the plaintiffs presented insufficient proof that the\\ndefendant had engaged in a general policy of sex\\ndiscrimination. Id. at 352—55. In Corncast, the Court ordered\\ndecertification where the regression analysis used to show\\ncommon injury did not track the underlying theory of liability.\\n569 U.S. at 36—38. The Court rejected a contention that, at the\\nclass-certification stage, “any method of measurement is\\nacceptable so long as it can be applied classwide, no matter\\nhow arbitrary the measurements may be.” Id. at 36. But in\\nTyson foods, the Court held that concerns about the validity of\\na statistical sample used to prove class-wide averages—which\\nthe Court described as presenting a common objection to the\\nclaims of each class member—should be addressed “as a\\nmatter of summary judgment, not class certification.” 136 5.\\nCt. at 1047 (quotation marks omitted).\\n\\n IV\\n\\n The direct-purchaser plaintiffs raise claims under section\\n4 of the Clayton Act, which provides treble damages to any\\nperson “injured in his business or property by reason of\\nanything forbidden in the antitrust laws.” 15 U.S.C. § 15(a).\\nTo establish liability under section 4, each plaintiff must prove\\nnot only an antitrust violation, but also an injury to its business\\nor property and a causal relation between the two. Without\\ncommon proof of injury and causation, section 4 plaintiffs\\ncannot establish predominance. See, e.g., Comcast, 569 U.S.\\nat 36—38; Rail freight I, 725 F.3d at 252—53.\\n\\n The parties dispute the extent to which a court, in\\nconducting the “hard look” required by Rule 23, should assess\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 8\\nthe reliability of common evidence. Specifically, they contest\\nwhether Rule 23 requires reliability above and beyond what is\\nnecessary to establish the admissibility of expert testimony\\nunder Federal Rule of Evidence 702 and Daubert v. Merrell\\nDow Pharmaceuticals, Inc., 509 U.S. 579 (1993). That\\nquestion arose because the district court found Dr. Rausser’s\\nregression models to be reliable for admissibility purposes,\\nRail freight II, 292 F. Supp. 3d at 54—63; held that “reliability\\nunder Rule 23 is a higher standard than reliability under\\nDaubert,” id. at 91; and found no predominance based on\\nconcerns that “undermine[d] the reliability of Dr. Rausser’s\\ndamages model,” id. at 122. The defendants embrace this\\nreasoning; they argue that Wal-Mart and Comcast rejected\\ncertification based on concerns that common evidence was\\nunreliable, even though its admissibility was uncontested. The\\nplaintiffs disagree; they contend that predominance ultimately\\nturns on whether the relevant evidence is common or\\nindividualized, and they read Tyson foods as holding that a\\ncourt may assess its reliability only in deciding admissibility or\\n summary judgment.\\n\\n We need not resolve this dispute because Dr. Rausser’s\\ndamages model, even if sufficiently reliable, does not prove\\nclasswide injury. As the district court explained, his model\\nindicates that the proposed class consists of 16,065 shippers.\\nRail freight II, 292 F. $upp. 3d at 136. The plaintiffs maintain\\nthat the alleged conspiracy injured every one of them. Yet the\\ndamages model also indicates that 2,037 members of the\\nproposed class—or 12.7 percent—suffered “only negative\\novercharges” and thus no injury from any conspiracy. Id. at\\n137. So even assuming the model can reliably show injury and\\ncausation for 87.3 percent of the class, that still leaves the\\nplaintiffs with no common proof of those essential elements of\\nliability for the remaining 12.7 percent. The district court held\\nthat the need for “individualized inquiries to determine which\\nof at least 2,037 (and possibly more) class members were\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 9\\nactually injured by the alleged conspiracy,” Id. at 140,\\nprecluded a finding of predominance. Id. at 122. We review a\\ndistrict court’s assessment of predominance only for abuse of\\ndiscretion. Garcia v. Johanns, 444 F.3d 625, 631 (D.C. Cir.\\n2006). Here, we find no such abuse.\\n\\n The plaintiffs argue that their model measures these\\nnegative damages only because of normal prediction error. The\\ndistrict court found that prediction error could not “account for\\nall—or even a substantial portion of—the 2,037 shippers that\\nthe model shows to be uninjured.” Rail Freight II, 292 F. Supp.\\n3d at 139. The plaintiffs take issue with that finding, but it is\\nnot clearly erroneous. In any event, the plaintiffs’ argument\\nabout prediction error at most suggests that their damages\\nmodel might falsely have measured no injury for as many as\\n2,037 shippers. This line of reasoning describes a possible\\nproblem with their own evidence; it does not point to\\naffirmative evidence—much less common affirmative\\nevidence—that a conspiracy did in fact injure these shippers.\\n\\n The plaintiffs further argue that predominance does not\\nrequire common evidence extending to all class members.\\nThat contention appears inconsistent with our statement in Rail\\nfreight I that the plaintiffs, to establish predominance, must\\n“show that they can prove, through common evidence, that all\\nclass members were in fact injured by the alleged conspiracy.”\\n725 F.3d at 252; see also id. (“we do expect the common\\nevidence to show all class members suffered some injury”).\\nDespite these statements, the district court held that our opinion\\ndid not require common evidence of injury to all class\\nmembers. See Rail Freight II, 292 F. Supp. 3d at 132—34.\\nInstead, it agreed with the plaintiffs that common proof\\ncovering “virtually all” members of the proposed class, and\\nleaving only a “de minimis” number of cases requiring\\nindividualized proof of injury and causation, would be enough\\nto show predominance. Id. at 135. For the sake of argument,\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 10\\nwe assume that the district court correctly recognized a de\\nminimis exception to the general rule that, for claims under\\nsection 4 of the Clayton Act, causation and injury must be\\n“capable of classwide resolution,” Wal-Mart, 564 U.S. at 350.\\nThe court reasonably concluded that such a de minimis\\nexception would not encompass this case.\\n\\n In assessing how many individual adjudications are too\\nmany, both the district court and the parties invoke cases\\naddressing the question of when, if ever, a class may include\\nconcededly uninjured members. Strictly speaking, this case\\ndoes not present that question, for the plaintiffs here insist that\\neach member of the proposed class was injured. Nonetheless,\\nthe cited cases bear some similarity to this one: Uninjured class\\nmembers cannot prevail on the merits, so their claims must be\\nwinnowed away as part of the liability determination. And that\\nprospect raises the same kind of question at issue here—when\\ndoes the need for individualized proof of injury and causation\\ndestroy predominance? See, e.g., In re Asacol Antitrust Litig.,\\n907 F.3d 42, 51—58 (1st Cir. 2018); In re Nexium Antitrust\\nLitig., 777 F.3d 9, 18—22 (1st Cir. 2015).\\n\\n The plaintiffs complain that the district court arbitrarily\\nimposed a six-percent upper limit on the percentage of\\nuninjured parties who may be included in a certified class. In\\nfact, the court’s analysis was more nuanced. As the court\\nexplained, the “few reported decisions” involving uninjured\\nclass members “suggest that 5% to 6% constitutes the outer\\nlimits of a de minimis number.” Rail Freight ll 292 F. Supp.\\n3d at 137. The 12.7 percent figure in this case is more than\\ntwice that approximate upper bound reflected in analogous\\ncaselaw. Moreover, the district court considered raw numbers\\nas well as percentages: six percent of a “class totaling only\\nfifty-five” members might be de minimis, but 12.7 percent of\\nthis class yields “2,037 uninjured class members” (according\\nto the common proof), all of whom would need individualized\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 11\\nadjudications of causation and injury. Id. at 137—38. Finally,\\nthe district court stressed that the plaintiffs have proposed no\\n“further way”—short of full-blown, individual trials—”to\\nreduce this number and segregate the uninjured from the truly\\ninjured.” Id. at 138. None of this was an abuse of discretion.\\n\\n The absence of any winnowing mechanism sharply\\ndistinguishes Nexiurn, the plaintiffs’ best case. There, the class\\nincluded purchasers of a drug allegedly shielded from\\ncompetition by the unlawful suppression of a generic\\nalternative. 777 F.3d at 13—14. The problem of uninjured class\\nmembers arose because a small percentage of the class, due to\\nbrand loyalty, would have purchased the drug even if a less\\nexpensive generic alternative had been available. Id. at 19—20.\\nThe First Circuit held that the uninjured class members could\\nmanageably be winnowed by having individual consumers file\\nminimal, likely unrebutted affidavit testimony indicating\\nwhether, if given the choice, they would have purchased the\\nbranded drug or a generic alternative. Id. at 20—21.\\n\\n Nexium does not support class certification here. For one\\nthing, the First Circuit sharply limited that decision in Asacol.\\nThere, the Court explained that any winnowing mechanism\\nmust be truncated enough to ensure that the common issues\\npredominate, yet robust enough to preserve the defendants’\\nSeventh Amendment and due process rights to contest every\\nelement of liability and to present every colorable defense. 907\\nF.3d at 5 1—54. Moreover, the Court held that Nexium’s\\naffidavit mechanism could not satisfy both conditions where\\nthe defendant seeks to contest the question whether individual\\nclass members would have shifted from the branded drug to a\\nless expensive generic alternative. See Id. That would require\\nindividual trials because genuinely contested affidavits do not\\nsupport summary judgment and are inadmissible. Id. Here, the\\ndefendants intend to contest whether any of the 2,037 shippers\\nsuffered injury as a result of any conspiracy. And the question\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 12\\npresented in these individual challenges—regarding impacts\\non shippers of different sizes, shipping different products in\\ndifferent geographic markets, with different transportation\\noptions and different degrees of leverage—would be far more\\ncomplex than the single unitary question (branded or generic?)\\nat issue in Nexium and Asacol.\\n\\n The plaintiffs also invoke the Supreme Court’s discussion\\nof uninjured class members in Tyson foods. The defendants\\nthere sought review of the question whether a certified class\\nmay contain any uninjured members. But the Supreme Court\\nreserved that question, 136 5. Ct. at 1049, and the district court\\nhere decided it in the plaintiffs’ favor, Rail Freight II, 292 F.\\nSupp. 3d at 133—35. So, this aspect of Tyson foods does not\\nadvance the plaintiffs’ appeal. The Supreme Court also\\naddressed a further ‘new argument” raised for the first time in\\nmerits briefing—that plaintiffs at the certification stage must\\nprove that all class members were injured or establish a\\nmanageable process for culling out uninjured class members.\\n 136 5. Ct. at 1049. The Court held that this argument was\\n“premature” because the parties disagreed about what culling\\nmechanisms might be available and the district court had not\\nyet addressed the question. See 136 5. Ct. at 1050. This\\nholding rested on the inappropriateness of raising new issues\\nfor the first time in Supreme Court merits briefing. See id. It\\ndoes not, as the plaintiffs here contend, perniit district courts\\nconsidering class certification to defer questions about the\\nnumber and nature of any individualized inquiries that might\\nbe necessary to establish liability. To the contrary, confronting\\nsuch questions is part-and-parcel of the “hard look” required\\nby Wal-Mart and Comcast, as recognized even by those courts\\npermitting a class to include some small number of concededly\\nuninjured individuals. See, e.g., In re Asacol, 907 F.3d at 51—\\n54; In re Nexiurn,777 F.3d at 19—2 1.\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 1—,\\n 1.)\\n\\n\\n Finally, the plaintiffs argue that the 2,037 class members\\nfor whom their damages model shows no injury are a de\\nminimis portion of the class because their shipments make up\\nless than one percent of the railroads’ overall revenue from the\\nalleged conspiracy. But revenue is irrelevant to predominance,\\nwhich looks to whether elements such as causation and injury\\nmay be proved through common evidence, not how much the\\ndefendants benefited from any wrongdoing.\\n\\n V\\n\\n Looking beyond Dr. Rausser’s regression analysis, the\\nplaintiffs point to other evidence that they say can prove injury\\nand causation on a class-wide basis. In Rail freight I, we\\nconcluded that Dr. Rausser’s analysis was “essential” to the\\nplaintiffs’ case for certification: “No damages model, no\\npredominance, no class certification.” 725 F.3d at 253. The\\ndistrict court reached the same conclusion on remand, after\\ncareful review of all the documentary and expert evidence.\\nThis was not an abuse of discretion.\\n\\n The plaintiffs invoke documentary evidence that the\\ndefendants enforced fuel surcharges “uniformly and with few\\nexceptions.” Rail Freight II, 292 F. $upp. 3d at 122; see id. at\\n103—07. But imposing fuel surcharges does not show injury\\ncaused by a conspiracy. The parties vigorously dispute\\nwhether higher overall prices during the class period were\\nattributable to causes besides any conspiracy—such as the\\nmarked increase in fuel prices that occurred around the\\nbeginning of the class period, see l.A. 65 92—93. As the district\\ncourt explained, Dr. Rausser designed his regression models\\nprecisely to control for these kinds of potential alternative\\ncauses, see Rail Freight II, 292 F. $upp. 3d at 101, yet his\\ndamages model showed that 2,037 shippers were uninjured\\ndespite paying fuel surcharges. The plaintiffs’ evidence of\\nwidespread fuel surcharges helps explain why the proposed\\n\\f PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n 14\\nclass is a large one, but it neither proves that the 2,037 shippers\\nwere injured by the alleged conspiracy nor otherwise compels\\na finding of predominance.\\n\\n Next, the plaintiffs invoke the expert testimony of Dr.\\nJames McClave, who argued that the 2,037 shippers must have\\nbeen harmed by the conspiracy. Dr. McClave reasoned that\\nbecause these shippers made fewer purchases on average, and\\nthus had less bargaining power than the rest of the class, they\\nmust have been more susceptible to injury. Moreover, Dr.\\nMcClave’s own study concluded that the defendants’ smallest\\ncustomers—roughly shippers who made only one\\npurchase I I—collectively\\npaid higher prices than did larger shippers.\\n\\n This analysis is not common proof of injury to the 2,037\\nshippers. For one thing, of the shippers studied by Dr.\\nMcClave, only were in the class, and Dr. McClave did\\nnot indicate how many of them were included among the 2,037\\nclass members for whom Dr. Rausser’s model showed no\\ninjury. Moreover, evidence that a group of one-time shippers\\nmay have paid higher prices collectively does not prove that all\\n(or almost all) of them were injured individually. And, as the\\ndistrict court explained, Dr. McClave did not attempt to\\nidentify how many of the one-time shippers did in fact pay\\nmore, let alone which of them were among the 2,037. See Rail\\nFreight II, 292 F. $upp. 3d at 139.\\n\\n The plaintiffs respond that Dr. McClave’s analysis, even if\\ninconclusive, at least suggests that reduced bargaining power\\nmade the 2,037 shippers more vulnerable to any conspiracy.\\nThis line of reasoning parallels one rejected by the Supreme\\nCourt in Wal-Mart. There, an expert opined that Wal-Mart’s\\n“strong corporate culture” made it “vulnerable to gender bias.”\\n564 U.S. at 354 (quotation marks omitted). The Court found\\nthis evidence insufficient to prove that injury could be\\n\\f PUBLIC COPY SEALED INFORMATION DELETED\\n -\\n\\n\\n\\n\\n 15\\nestablished on a class-wide basis, because the expert could not\\nsay what percentage of adverse employment decisions were in\\nfact caused by bias. Id. The McClave study is similarly\\nincomplete because, as explained above, it does not attempt to\\nidentify which of the small shippers, or what percentage of\\nthem, were in fact harmed by the alleged conspiracy.\\n\\n VI\\n\\n In Asacol, the First Circuit noted the absence of even a\\nsingle case “allowing, under Rule 23, a trial in which thousands\\nof class members testify.” 907 F.3d at 57—58. That Court\\ndeclined to create “the first such case.” Id. So do we. Given\\nthe need in this case for at least 2,037 individual determinations\\nof injury and causation, the district court did not abuse its\\ndiscretion in denying class certification on the ground that\\ncommon issues do not predominate.\\n\\n Affirmed\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4179,"cells":{"id":{"kind":"number","value":4649378,"string":"4,649,378"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"United States v. James Duckett (AMENDED OPINION)"},"decision_date":{"kind":"string","value":"2019-08-19"},"docket_number":{"kind":"string","value":"18-3026"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before: GRIFFITH and SRINIVASAN, Circuit Judges, and RANDOLPH, Senior Circuit Judge.\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Senior Circuit Judge RANDOLPH.\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 4, 2019 Decided May 24, 2019\\n Reissued August 19, 2019\\n\\n No. 18-3026\\n\\n UNITED STATES OF AMERICA,\\n APPELLEE\\n\\n v.\\n\\n JAMES DUCKETT,\\n APPELLANT\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:15-cr-00120-1)\\n\\n\\n Rosanna M. Taormina, Assistant Federal Public Defender,\\nargued the cause for appellant. With her on the briefs was A.J.\\nKramer, Federal Public Defender. Tony Axam Jr., Assistant\\nFederal Public Defender, entered an appearance.\\n\\n Elizabeth Gabriel, Assistant U.S. Attorney, argued the\\ncause for appellee. With her on the briefs were Jessie K. Liu,\\nU.S. Attorney, and Elizabeth Trosman and Stephen J. Gripkey,\\nAssistant U.S. Attorneys.\\n\\n Before: GRIFFITH and SRINIVASAN, Circuit Judges, and\\nRANDOLPH, Senior Circuit Judge.\\n\\f 2\\n\\n\\n\\n\\n Opinion for the Court filed by Senior Circuit Judge\\nRANDOLPH.\\n\\n RANDOLPH, Senior Circuit Judge: James T. Duckett\\nappeals his latest sentence of 24 months’ imprisonment. He\\nclaims that his sentence should have been lower and that the\\ndistrict court did not sufficiently explain why it was imposing a\\nterm of imprisonment greater than the Sentencing Guidelines\\nrange.\\n\\n We have framed the issue in terms of Duckett’s “latest\\nsentence” because Duckett, now in his mid-50s, has been in\\nnearly continuous custody for one offense or another since the\\n1980s. In his teens he began committing petty crimes and the\\nolder he grew the more violent he became. He moved on to\\ndistributing illegal drugs, stabbing, robbing, attempting murder,\\nand murdering.\\n\\n In the District of Columbia area during a nine-year period\\nbeginning when Duckett was 18 years old, he served time for\\nsimple assault, attempted petty larceny, attempted robbery,\\narmed robbery, theft, and attempted distribution of cocaine.\\nAfter his release on parole, he violated the terms of his release.\\nImprisoned again, this time in Virginia, Duckett stabbed another\\ninmate repeatedly – in the chest and in the back – with a\\n“shank,” and continued his onslaught even as a prison guard\\nintervened to protect the victim. A jury found Duckett guilty of\\nassault with intent to murder, assault causing serious bodily\\nharm, and possession of contraband (the shank). The court\\nsentenced him to 20 years’ imprisonment.\\n\\f 3\\n\\n In 1994, while serving this sentence at the Lewisburg,\\nPennsylvania, federal penitentiary, Duckett stabbed another\\ninmate to death. The Bureau of Prisons transferred him to\\n“ADX” (Administrative Maximum) Florence, Colorado, by\\nsome accounts the nation’s most secure federal prison.\\n\\n At ADX Florence, Duckett assisted another inmate –\\nDominic Stewart – in attacking prisoner Gregory Joiner and\\nbeating him to death. A federal grand jury in Colorado indicted\\nDuckett and Stewart for second degree murder and assault\\nresulting in serious bodily injury. Stewart faced an additional\\nfirst degree murder charge for the crime.\\n\\n Duckett negotiated a cooperation agreement with the\\nprosecution, pursuant to which he pled guilty to the assault\\ncharge and testified against Stewart. In return, the prosecutor\\nfiled a motion for a downward sentencing departure. The\\nfederal district court in Colorado concurred and sentenced\\nDuckett to 42 months of imprisonment, followed by three years\\nof supervised release. Duckett started serving this period of\\nsupervised release in October 2014.\\n\\n This brings us to the events leading to Duckett’s latest\\nsentence of 24 months’ imprisonment. Duckett moved back to\\nWashington, D.C. in the fall of 2014. Chief Judge Howell of the\\nfederal district court here assumed jurisdiction over his\\ncompliance with the terms of his supervised release from the\\nsentence imposed in Colorado.\\n\\n In December 2014, shortly after his arrival here, he was\\narrested for assault, aggressive panhandling, and carrying a\\nprohibited weapon. In April 2015, while detained, he assaulted\\na prison guard. In May 2015, the D.C. Superior Court sentenced\\nhim to 6 months’ probation for each of these offenses.\\n\\f 4\\n\\n In July 2015, Duckett assaulted a special police officer\\nwhile in possession of a prohibited weapon. On his plea of\\nguilty, the D.C. Superior Court sentenced him to 180 days’\\nimprisonment followed by three years of supervised probation.\\n\\n In April 2016, Chief Judge Howell revoked Duckett’s\\nsupervised release on his Colorado sentence because of these\\ntwo crimes. The Chief Judge sentenced him to 6 months’\\nimprisonment followed by 24 months of supervised release.\\n\\n Duckett finished his prison term in August 2016 and\\nreentered supervised release. Once again he violated a condition\\nof his release by leaving the halfway house to which he had been\\nassigned. Again Chief Judge Howell revoked Duckett’s release\\nand sentenced him to another 6 months’ imprisonment followed\\nby 12 months of supervised release.\\n\\n After serving this 6 month prison term, Duckett returned to\\nsociety and again began violating the conditions of his release.\\nHe tested positive for cocaine in August 2017. About the same\\ntime, he refused to provide a urine sample for a drug test at the\\nhalfway house. A staff member informed him that if he did not\\ncomply he would be sent back to prison. In response, Duckett\\ntold the staff member that he would “stretch him out” – that is,\\npunch him in the head to render him unconscious – before he\\nwould be sent back to prison. In response to Duckett’s threat,\\nthe halfway house evicted him. A few months later Duckett\\nattempted to steal a television from a Walmart store in the\\nDistrict. On his plea of guilty to second-degree theft, the\\nSuperior Court sentenced him to 30 days’ incarceration,\\nconsecutive to any other sentence. The Superior Court also\\nrevoked his probation, which triggered one additional year of\\nimprisonment for him.\\n\\f 5\\n\\n There matters stood in April 2018 when Duckett appeared\\nin federal court for a revocation hearing, his third such hearing\\nin this district in two years. Duckett conceded that he had again\\nviolated the terms of his release. Each of his current violations,\\nin combination with his criminal history category of VI, carried\\nan advisory Sentencing Guidelines range of 8 to 14 months, and\\na statutory maximum of 24 months under 18 U.S.C.\\n§ 3583(e)(3). The probation office and the prosecution\\nrecommended a sentence of 24 months’ imprisonment\\nconsecutive to Duckett’s Superior Court sentence.\\n\\n Chief Judge Howell agreed with the recommendations and\\norally announced that Duckett was to serve the statutory\\nmaximum sentence of 24 months’ imprisonment running\\nconsecutively to the 13 months ordered by the Superior Court.\\nAfter so stating, the Chief Judge asked Duckett’s counsel\\nwhether there was any objection. In response, the attorney from\\nthe Public Defender’s office said nothing about the adequacy of\\nthe Chief Judge’s explanation for imposing an above-Guidelines\\nsentence.\\n\\n Even so, now on appeal the Public Defender presents the\\nquestion whether the Chief Judge gave a sufficient reason for\\nDuckett’s sentence. Under 18 U.S.C. § 3553(c)(2), the\\nsentencing judge must provide “the specific reason” for a\\nsentence outside the Guidelines range. Given the absence of an\\nobjection here, Duckett has the burden of showing “plain error.”\\nSee Fed. R. Crim. P. 52(b).\\n\\n A “plain error” is a glaring or obvious judicial mistake of\\nfact or law that adversely “affects substantial rights” but was\\n“not brought to the court’s attention.” Id. In most cases even an\\nobvious error will not constitute a “plain error” unless it\\nprejudiced the defendant – that is, unless it “affected the\\noutcome of the district court proceedings.” United States v.\\n\\f 6\\n\\nOlano, 507 U.S. 725, 734 (1993). And even though the\\ndefendant carries his burden of showing that the district court\\ncommitted a “plain error” as so defined, that will not be enough.\\n“Rule 52(b) is permissive, not mandatory. If the forfeited error\\nis ‘plain’ and ‘affect[s] substantial rights,’ the court of appeals\\nhas authority to order correction, but is not required to do so.”\\nId. at 735 (alteration in original); see also Puckett v. United\\nStates, 556 U.S. 129, 142 (2009) (noting “that a ‘per se approach\\nto plain-error review is flawed’”) (quoting United States v.\\nYoung, 470 U.S. 1, 17, n.14 (1985)).\\n\\n Here it is apparent why defense counsel did not object to the\\nChief Judge’s explanation of Duckett’s 24-month sentence.\\nDuckett “and anyone else present at his sentencing hearing[]\\nmust have understood why the district court imposed an above-\\nGuidelines sentence of” 24 months. United States v. Jackson,\\n848 F.3d 460, 462 (D.C. Cir. 2017).\\n\\n Needless to say, Chief Judge Howell was quite familiar with\\nDuckett. This was the third time in two years he stood before\\nher in a revocation hearing. In the Colorado sentencing and in\\nthe two previous revocation hearings, Duckett received the\\nbenefit of below-Guidelines sentences. He and his counsel\\ntherefore should have anticipated that an above-Guidelines\\nsentence could be coming. The Sentencing Guidelines indicate\\nthat courts may impose lengthier sentences when prior sentences\\nwere lenient. See U.S.S.G. § 7B1.4, application note 4 (advising\\nthat when “the original sentence was the result of a downward\\ndeparture (e.g., as a reward for substantial assistance), . . . an\\nupward departure may be warranted”). At the hearing Duckett’s\\ncounsel acknowledged that this time around the “Court is going\\nto impose a significant sanction.” Hr’g Tr. 4/20/18 at 22.\\n\\n As the prosecutor argued, and as the Chief Judge concluded,\\nDuckett’s three violations triggering his latest revocation\\n\\f 7\\n\\nproceeding were only part of his story. The prosecutor gave this\\nsummary, which was not all-inclusive: Duckett “shank[ed]\\nsomeone in prison”; he participated in the deadly beating of an\\ninmate; he was arrested on nine separate occasions while under\\nsupervision; he was arrested and convicted of assaulting a police\\nofficer while under supervision in this case; and so forth. Hr’g\\nTr. 4/20/2018 at 12. It hardly needs explaining to understand\\nwhy the prosecutor and the probation office urged a 24-month\\nabove-Guidelines sentence to protect the community.\\n\\n Chief Judge Howell reviewed and explained Duckett’s\\n“very extensive” and “horrible” criminal record, as she put it,\\nand his repeated violations of the terms she and other judges had\\nimposed on him during his brief periods out of prison. Hr’g Tr.\\n4/20/2018 at 23. She told Duckett that she had given him\\n“opportunities in the past”; that he did not take advantage of\\nthose opportunities; and that now the court’s responsibility was\\n“to protect the public from future crimes by you.” Hr’g Tr.\\n4/20/2018 at 29–30. All of this sufficiently explained the Chief\\nJudge’s decision to send Duckett back to prison for 24 months.\\nSee Rita v. United States, 551 U.S. 338, 356–59 (2007); see also\\nUnited States v. Ransom, 756 F.3d 770, 774, 776 (D.C. Cir.\\n2014). At least for those 24 months Duckett would be off the\\nstreets, unable to “stretch out” any member of the public, or to\\nsteal television sets, or to assault police officers, or to commit\\nother forms of mayhem.\\n\\n Duckett also claims that the Chief Judge relied on mistakes\\nof fact. The supposed mistakes dealt with the extent of\\nDuckett’s prior educational and vocational training, substance\\nabuse and mental health treatment, and anger management\\ncounseling. There is a mistake here but it is on Duckett’s part.\\nThe Chief Judge did not justify the sentence with reference to\\nDuckett’s treatment or lack thereof in the past. Treatment came\\n\\f 8\\n\\nup only once, when the Chief Judge asked whether Duckett’s\\ndrug treatment had been successful.\\n\\n Duckett’s final claim of “plain error” is that the calculation\\nof his criminal history was in error. The federal district court in\\nColorado, in 2012, determined that Duckett was in Category VI\\nbecause he was a career offender. The court identified two\\npredicate offenses for career-offender status under U.S.S.G.\\n§ 4B1.1: a 1989 conviction for attempted distribution of cocaine\\nand a 1993 conviction for assault with intent to commit murder.\\n\\n Duckett argues the calculation was in error in light of this\\ncourt’s intervening decision in United States v. Winstead, 890\\nF.3d 1082, 1090–92 (D.C. Cir. 2018), holding that attempted\\ndrug distribution does not qualify as a predicate offense for\\ncareer-offender status. We are skeptical of this argument for\\nseveral reasons. One problem is that Winstead is the law of this\\ncircuit, not the law of the Tenth Circuit where Duckett was\\noriginally sentenced. Another problem is that the Guidelines\\ninstruct courts in revocation proceedings to use the criminal\\nhistory category “determined at the time the defendant originally\\nwas sentenced to the term of supervision,” U.S.S.G. § 7B1.4,\\napplication note 1, which is what Chief Judge Howell did\\nwithout objection from defense counsel. See United States v.\\nJones, 690 F. App’x 336, 337 (6th Cir. 2017) (unpublished\\nopinion); United States v. Johnson, 570 F. App’x 611, 612 (7th\\nCir. 2014) (unpublished order). The third problem with\\nDuckett’s argument is that if his attempted drug distribution is\\ndisregarded, his armed robbery conviction in 1984 would be\\nsubstituted, thus placing him in the same criminal history\\ncategory. We need not resolve these issues, however. Even if\\nthere was an error and even if the error was obvious, it did not\\namount to a “plain error” because there is no “reasonable\\nprobability that, but for the error, the outcome of the proceeding\\nwould have been different.” United States v. Mack, 841 F.3d\\n\\f 9\\n\\n514, 522 (D.C. Cir. 2016) (quoting Molina–Martinez v. United\\nStates, 136 S. Ct. 1338, 1343 (2016)). Dropping Duckett’s\\ncriminal history category from VI to V would cause only a\\ndecrease in the revocation Guidelines range from 8-14 months\\nto 7-13 months. Given Chief Judge Howell’s reasons for\\nimposing the maximum 24 month sentence, that one-month\\ndifference would not have mattered. See Molina-Martinez, 136\\nS. Ct. at 1346–47.\\n\\n We have considered and rejected Duckett’s remaining\\narguments.\\n\\n Affirmed.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4426631/\", \"author_raw\": \"Opinion for the Court filed by Senior Circuit Judge RANDOLPH.\"}]}"},"judge1":{"kind":"string","value":"GRIFFITH"},"judge2":{"kind":"string","value":"SRINIVASAN"},"judge3":{"kind":"string","value":"RANDOLPH"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"GRIFFITH\": \", Circuit\", \"SRINIVASAN\": \", Circuit\", \"RANDOLPH\": \", Senior Circuit\"}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4649378/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: United States v. James Duckett (AMENDED OPINION)\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-19\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued April 4, 2019 Decided May 24, 2019\n Reissued August 19, 2019\n\n No. 18-3026\n\n UNITED STATES OF AMERICA,\n APPELLEE\n\n v.\n\n JAMES DUCKETT,\n APPELLANT\n\n\n Appeal from the United States District Court\n for the District of Columbia\n (No. 1:15-cr-00120-1)\n\n\n Rosanna M. Taormina, Assistant Federal Public Defender,\nargued the cause for appellant. With her on the briefs was A.J.\nKramer, Federal Public Defender. Tony Axam Jr., Assistant\nFederal Public Defender, entered an appearance.\n\n Elizabeth Gabriel, Assistant U.S. Attorney, argued the\ncause for appellee. With her on the briefs were Jessie K. Liu,\nU.S. Attorney, and Elizabeth Trosman and Stephen J. Gripkey,\nAssistant U.S. Attorneys.\n\n Before: GRIFFITH and SRINIVASAN, Circuit Judges, and\nRANDOLPH, Senior Circuit Judge.\n\f 2\n\n\n\n\n Opinion for the Court filed by Senior Circuit Judge\nRANDOLPH.\n\n RANDOLPH, Senior Circuit Judge: James T. Duckett\nappeals his latest sentence of 24 months’ imprisonment. He\nclaims that his sentence should have been lower and that the\ndistrict court did not sufficiently explain why it was imposing a\nterm of imprisonment greater than the Sentencing Guidelines\nrange.\n\n We have framed the issue in terms of Duckett’s “latest\nsentence” because Duckett, now in his mid-50s, has been in\nnearly continuous custody for one offense or another since the\n1980s. In his teens he began committing petty crimes and the\nolder he grew the more violent he became. He moved on to\ndistributing illegal drugs, stabbing, robbing, attempting murder,\nand murdering.\n\n In the District of Columbia area during a nine-year period\nbeginning when Duckett was 18 years old, he served time for\nsimple assault, attempted petty larceny, attempted robbery,\narmed robbery, theft, and attempted distribution of cocaine.\nAfter his release on parole, he violated the terms of his release.\nImprisoned again, this time in Virginia, Duckett stabbed another\ninmate repeatedly – in the chest and in the back – with a\n“shank,” and continued his onslaught even as a prison guard\nintervened to protect the victim. A jury found Duckett guilty of\nassault with intent to murder, assault causing serious bodily\nharm, and possession of contraband (the shank). The court\nsentenced him to 20 years’ imprisonment.\n\f 3\n\n In 1994, while serving this sentence at the Lewisburg,\nPennsylvania, federal penitentiary, Duckett stabbed another\ninmate to death. The Bureau of Prisons transferred him to\n“ADX” (Administrative Maximum) Florence, Colorado, by\nsome accounts the nation’s most secure federal prison.\n\n At ADX Florence, Duckett assisted another inmate –\nDominic Stewart – in attacking prisoner Gregory Joiner and\nbeating him to death. A federal grand jury in Colorado indicted\nDuckett and Stewart for second degree murder and assault\nresulting in serious bodily injury. Stewart faced an additional\nfirst degree murder charge for the crime.\n\n Duckett negotiated a cooperation agreement with the\nprosecution, pursuant to which he pled guilty to the assault\ncharge and testified against Stewart. In return, the prosecutor\nfiled a motion for a downward sentencing departure. The\nfederal district court in Colorado concurred and sentenced\nDuckett to 42 months of imprisonment, followed by three years\nof supervised release. Duckett started serving this period of\nsupervised release in October 2014.\n\n This brings us to the events leading to Duckett’s latest\nsentence of 24 months’ imprisonment. Duckett moved back to\nWashington, D.C. in the fall of 2014. Chief Judge Howell of the\nfederal district court here assumed jurisdiction over his\ncompliance with the terms of his supervised release from the\nsentence imposed in Colorado.\n\n In December 2014, shortly after his arrival here, he was\narrested for assault, aggressive panhandling, and carrying a\nprohibited weapon. In April 2015, while detained, he assaulted\na prison guard. In May 2015, the D.C. Superior Court sentenced\nhim to 6 months’ probation for each of these offenses.\n\f 4\n\n In July 2015, Duckett assaulted a special police officer\nwhile in possession of a prohibited weapon. On his plea of\nguilty, the D.C. Superior Court sentenced him to 180 days’\nimprisonment followed by three years of supervised probation.\n\n In April 2016, Chief Judge Howell revoked Duckett’s\nsupervised release on his Colorado sentence because of these\ntwo crimes. The Chief Judge sentenced him to 6 months’\nimprisonment followed by 24 months of supervised release.\n\n Duckett finished his prison term in August 2016 and\nreentered supervised release. Once again he violated a condition\nof his release by leaving the halfway house to which he had been\nassigned. Again Chief Judge Howell revoked Duckett’s release\nand sentenced him to another 6 months’ imprisonment followed\nby 12 months of supervised release.\n\n After serving this 6 month prison term, Duckett returned to\nsociety and again began violating the conditions of his release.\nHe tested positive for cocaine in August 2017. About the same\ntime, he refused to provide a urine sample for a drug test at the\nhalfway house. A staff member informed him that if he did not\ncomply he would be sent back to prison. In response, Duckett\ntold the staff member that he would “stretch him out” – that is,\npunch him in the head to render him unconscious – before he\nwould be sent back to prison. In response to Duckett’s threat,\nthe halfway house evicted him. A few months later Duckett\nattempted to steal a television from a Walmart store in the\nDistrict. On his plea of guilty to second-degree theft, the\nSuperior Court sentenced him to 30 days’ incarceration,\nconsecutive to any other sentence. The Superior Court also\nrevoked his probation, which triggered one additional year of\nimprisonment for him.\n\f 5\n\n There matters stood in April 2018 when Duckett appeared\nin federal court for a revocation hearing, his third such hearing\nin this district in two years. Duckett conceded that he had again\nviolated the terms of his release. Each of his current violations,\nin combination with his criminal history category of VI, carried\nan advisory Sentencing Guidelines range of 8 to 14 months, and\na statutory maximum of 24 months under 18 U.S.C.\n§ 3583(e)(3). The probation office and the prosecution\nrecommended a sentence of 24 months’ imprisonment\nconsecutive to Duckett’s Superior Court sentence.\n\n Chief Judge Howell agreed with the recommendations and\norally announced that Duckett was to serve the statutory\nmaximum sentence of 24 months’ imprisonment running\nconsecutively to the 13 months ordered by the Superior Court.\nAfter so stating, the Chief Judge asked Duckett’s counsel\nwhether there was any objection. In response, the attorney from\nthe Public Defender’s office said nothing about the adequacy of\nthe Chief Judge’s explanation for imposing an above-Guidelines\nsentence.\n\n Even so, now on appeal the Public Defender presents the\nquestion whether the Chief Judge gave a sufficient reason for\nDuckett’s sentence. Under 18 U.S.C. § 3553(c)(2), the\nsentencing judge must provide “the specific reason” for a\nsentence outside the Guidelines range. Given the absence of an\nobjection here, Duckett has the burden of showing “plain error.”\nSee Fed. R. Crim. P. 52(b).\n\n A “plain error” is a glaring or obvious judicial mistake of\nfact or law that adversely “affects substantial rights” but was\n“not brought to the court’s attention.” Id. In most cases even an\nobvious error will not constitute a “plain error” unless it\nprejudiced the defendant – that is, unless it “affected the\noutcome of the district court proceedings.” United States v.\n\f 6\n\nOlano, 507 U.S. 725, 734 (1993). And even though the\ndefendant carries his burden of showing that the district court\ncommitted a “plain error” as so defined, that will not be enough.\n“Rule 52(b) is permissive, not mandatory. If the forfeited error\nis ‘plain’ and ‘affect[s] substantial rights,’ the court of appeals\nhas authority to order correction, but is not required to do so.”\nId. at 735 (alteration in original); see also Puckett v. United\nStates, 556 U.S. 129, 142 (2009) (noting “that a ‘per se approach\nto plain-error review is flawed’”) (quoting United States v.\nYoung, 470 U.S. 1, 17, n.14 (1985)).\n\n Here it is apparent why defense counsel did not object to the\nChief Judge’s explanation of Duckett’s 24-month sentence.\nDuckett “and anyone else present at his sentencing hearing[]\nmust have understood why the district court imposed an above-\nGuidelines sentence of” 24 months. United States v. Jackson,\n848 F.3d 460, 462 (D.C. Cir. 2017).\n\n Needless to say, Chief Judge Howell was quite familiar with\nDuckett. This was the third time in two years he stood before\nher in a revocation hearing. In the Colorado sentencing and in\nthe two previous revocation hearings, Duckett received the\nbenefit of below-Guidelines sentences. He and his counsel\ntherefore should have anticipated that an above-Guidelines\nsentence could be coming. The Sentencing Guidelines indicate\nthat courts may impose lengthier sentences when prior sentences\nwere lenient. See U.S.S.G. § 7B1.4, application note 4 (advising\nthat when “the original sentence was the result of a downward\ndeparture (e.g., as a reward for substantial assistance), . . . an\nupward departure may be warranted”). At the hearing Duckett’s\ncounsel acknowledged that this time around the “Court is going\nto impose a significant sanction.” Hr’g Tr. 4/20/18 at 22.\n\n As the prosecutor argued, and as the Chief Judge concluded,\nDuckett’s three violations triggering his latest revocation\n\f 7\n\nproceeding were only part of his story. The prosecutor gave this\nsummary, which was not all-inclusive: Duckett “shank[ed]\nsomeone in prison”; he participated in the deadly beating of an\ninmate; he was arrested on nine separate occasions while under\nsupervision; he was arrested and convicted of assaulting a police\nofficer while under supervision in this case; and so forth. Hr’g\nTr. 4/20/2018 at 12. It hardly needs explaining to understand\nwhy the prosecutor and the probation office urged a 24-month\nabove-Guidelines sentence to protect the community.\n\n Chief Judge Howell reviewed and explained Duckett’s\n“very extensive” and “horrible” criminal record, as she put it,\nand his repeated violations of the terms she and other judges had\nimposed on him during his brief periods out of prison. Hr’g Tr.\n4/20/2018 at 23. She told Duckett that she had given him\n“opportunities in the past”; that he did not take advantage of\nthose opportunities; and that now the court’s responsibility was\n“to protect the public from future crimes by you.” Hr’g Tr.\n4/20/2018 at 29–30. All of this sufficiently explained the Chief\nJudge’s decision to send Duckett back to prison for 24 months.\nSee Rita v. United States, 551 U.S. 338, 356–59 (2007); see also\nUnited States v. Ransom, 756 F.3d 770, 774, 776 (D.C. Cir.\n2014). At least for those 24 months Duckett would be off the\nstreets, unable to “stretch out” any member of the public, or to\nsteal television sets, or to assault police officers, or to commit\nother forms of mayhem.\n\n Duckett also claims that the Chief Judge relied on mistakes\nof fact. The supposed mistakes dealt with the extent of\nDuckett’s prior educational and vocational training, substance\nabuse and mental health treatment, and anger management\ncounseling. There is a mistake here but it is on Duckett’s part.\nThe Chief Judge did not justify the sentence with reference to\nDuckett’s treatment or lack thereof in the past. Treatment came\n\f 8\n\nup only once, when the Chief Judge asked whether Duckett’s\ndrug treatment had been successful.\n\n Duckett’s final claim of “plain error” is that the calculation\nof his criminal history was in error. The federal district court in\nColorado, in 2012, determined that Duckett was in Category VI\nbecause he was a career offender. The court identified two\npredicate offenses for career-offender status under U.S.S.G.\n§ 4B1.1: a 1989 conviction for attempted distribution of cocaine\nand a 1993 conviction for assault with intent to commit murder.\n\n Duckett argues the calculation was in error in light of this\ncourt’s intervening decision in United States v. Winstead, 890\nF.3d 1082, 1090–92 (D.C. Cir. 2018), holding that attempted\ndrug distribution does not qualify as a predicate offense for\ncareer-offender status. We are skeptical of this argument for\nseveral reasons. One problem is that Winstead is the law of this\ncircuit, not the law of the Tenth Circuit where Duckett was\noriginally sentenced. Another problem is that the Guidelines\ninstruct courts in revocation proceedings to use the criminal\nhistory category “determined at the time the defendant originally\nwas sentenced to the term of supervision,” U.S.S.G. § 7B1.4,\napplication note 1, which is what Chief Judge Howell did\nwithout objection from defense counsel. See United States v.\nJones, 690 F. App’x 336, 337 (6th Cir. 2017) (unpublished\nopinion); United States v. Johnson, 570 F. App’x 611, 612 (7th\nCir. 2014) (unpublished order). The third problem with\nDuckett’s argument is that if his attempted drug distribution is\ndisregarded, his armed robbery conviction in 1984 would be\nsubstituted, thus placing him in the same criminal history\ncategory. We need not resolve these issues, however. Even if\nthere was an error and even if the error was obvious, it did not\namount to a “plain error” because there is no “reasonable\nprobability that, but for the error, the outcome of the proceeding\nwould have been different.” United States v. Mack, 841 F.3d\n\f 9\n\n514, 522 (D.C. Cir. 2016) (quoting Molina–Martinez v. United\nStates, 136 S. Ct. 1338, 1343 (2016)). Dropping Duckett’s\ncriminal history category from VI to V would cause only a\ndecrease in the revocation Guidelines range from 8-14 months\nto 7-13 months. Given Chief Judge Howell’s reasons for\nimposing the maximum 24 month sentence, that one-month\ndifference would not have mattered. See Molina-Martinez, 136\nS. Ct. at 1346–47.\n\n We have considered and rejected Duckett’s remaining\narguments.\n\n Affirmed.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: United States v. James Duckett (AMENDED OPINION)\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-19\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 4, 2019 Decided May 24, 2019\\n Reissued August 19, 2019\\n\\n No. 18-3026\\n\\n UNITED STATES OF AMERICA,\\n APPELLEE\\n\\n v.\\n\\n JAMES DUCKETT,\\n APPELLANT\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:15-cr-00120-1)\\n\\n\\n Rosanna M. Taormina, Assistant Federal Public Defender,\\nargued the cause for appellant. With her on the briefs was A.J.\\nKramer, Federal Public Defender. Tony Axam Jr., Assistant\\nFederal Public Defender, entered an appearance.\\n\\n Elizabeth Gabriel, Assistant U.S. Attorney, argued the\\ncause for appellee. With her on the briefs were Jessie K. Liu,\\nU.S. Attorney, and Elizabeth Trosman and Stephen J. Gripkey,\\nAssistant U.S. Attorneys.\\n\\n Before: GRIFFITH and SRINIVASAN, Circuit Judges, and\\nRANDOLPH, Senior Circuit Judge.\\n\\f 2\\n\\n\\n\\n\\n Opinion for the Court filed by Senior Circuit Judge\\nRANDOLPH.\\n\\n RANDOLPH, Senior Circuit Judge: James T. Duckett\\nappeals his latest sentence of 24 months’ imprisonment. He\\nclaims that his sentence should have been lower and that the\\ndistrict court did not sufficiently explain why it was imposing a\\nterm of imprisonment greater than the Sentencing Guidelines\\nrange.\\n\\n We have framed the issue in terms of Duckett’s “latest\\nsentence” because Duckett, now in his mid-50s, has been in\\nnearly continuous custody for one offense or another since the\\n1980s. In his teens he began committing petty crimes and the\\nolder he grew the more violent he became. He moved on to\\ndistributing illegal drugs, stabbing, robbing, attempting murder,\\nand murdering.\\n\\n In the District of Columbia area during a nine-year period\\nbeginning when Duckett was 18 years old, he served time for\\nsimple assault, attempted petty larceny, attempted robbery,\\narmed robbery, theft, and attempted distribution of cocaine.\\nAfter his release on parole, he violated the terms of his release.\\nImprisoned again, this time in Virginia, Duckett stabbed another\\ninmate repeatedly – in the chest and in the back – with a\\n“shank,” and continued his onslaught even as a prison guard\\nintervened to protect the victim. A jury found Duckett guilty of\\nassault with intent to murder, assault causing serious bodily\\nharm, and possession of contraband (the shank). The court\\nsentenced him to 20 years’ imprisonment.\\n\\f 3\\n\\n In 1994, while serving this sentence at the Lewisburg,\\nPennsylvania, federal penitentiary, Duckett stabbed another\\ninmate to death. The Bureau of Prisons transferred him to\\n“ADX” (Administrative Maximum) Florence, Colorado, by\\nsome accounts the nation’s most secure federal prison.\\n\\n At ADX Florence, Duckett assisted another inmate –\\nDominic Stewart – in attacking prisoner Gregory Joiner and\\nbeating him to death. A federal grand jury in Colorado indicted\\nDuckett and Stewart for second degree murder and assault\\nresulting in serious bodily injury. Stewart faced an additional\\nfirst degree murder charge for the crime.\\n\\n Duckett negotiated a cooperation agreement with the\\nprosecution, pursuant to which he pled guilty to the assault\\ncharge and testified against Stewart. In return, the prosecutor\\nfiled a motion for a downward sentencing departure. The\\nfederal district court in Colorado concurred and sentenced\\nDuckett to 42 months of imprisonment, followed by three years\\nof supervised release. Duckett started serving this period of\\nsupervised release in October 2014.\\n\\n This brings us to the events leading to Duckett’s latest\\nsentence of 24 months’ imprisonment. Duckett moved back to\\nWashington, D.C. in the fall of 2014. Chief Judge Howell of the\\nfederal district court here assumed jurisdiction over his\\ncompliance with the terms of his supervised release from the\\nsentence imposed in Colorado.\\n\\n In December 2014, shortly after his arrival here, he was\\narrested for assault, aggressive panhandling, and carrying a\\nprohibited weapon. In April 2015, while detained, he assaulted\\na prison guard. In May 2015, the D.C. Superior Court sentenced\\nhim to 6 months’ probation for each of these offenses.\\n\\f 4\\n\\n In July 2015, Duckett assaulted a special police officer\\nwhile in possession of a prohibited weapon. On his plea of\\nguilty, the D.C. Superior Court sentenced him to 180 days’\\nimprisonment followed by three years of supervised probation.\\n\\n In April 2016, Chief Judge Howell revoked Duckett’s\\nsupervised release on his Colorado sentence because of these\\ntwo crimes. The Chief Judge sentenced him to 6 months’\\nimprisonment followed by 24 months of supervised release.\\n\\n Duckett finished his prison term in August 2016 and\\nreentered supervised release. Once again he violated a condition\\nof his release by leaving the halfway house to which he had been\\nassigned. Again Chief Judge Howell revoked Duckett’s release\\nand sentenced him to another 6 months’ imprisonment followed\\nby 12 months of supervised release.\\n\\n After serving this 6 month prison term, Duckett returned to\\nsociety and again began violating the conditions of his release.\\nHe tested positive for cocaine in August 2017. About the same\\ntime, he refused to provide a urine sample for a drug test at the\\nhalfway house. A staff member informed him that if he did not\\ncomply he would be sent back to prison. In response, Duckett\\ntold the staff member that he would “stretch him out” – that is,\\npunch him in the head to render him unconscious – before he\\nwould be sent back to prison. In response to Duckett’s threat,\\nthe halfway house evicted him. A few months later Duckett\\nattempted to steal a television from a Walmart store in the\\nDistrict. On his plea of guilty to second-degree theft, the\\nSuperior Court sentenced him to 30 days’ incarceration,\\nconsecutive to any other sentence. The Superior Court also\\nrevoked his probation, which triggered one additional year of\\nimprisonment for him.\\n\\f 5\\n\\n There matters stood in April 2018 when Duckett appeared\\nin federal court for a revocation hearing, his third such hearing\\nin this district in two years. Duckett conceded that he had again\\nviolated the terms of his release. Each of his current violations,\\nin combination with his criminal history category of VI, carried\\nan advisory Sentencing Guidelines range of 8 to 14 months, and\\na statutory maximum of 24 months under 18 U.S.C.\\n§ 3583(e)(3). The probation office and the prosecution\\nrecommended a sentence of 24 months’ imprisonment\\nconsecutive to Duckett’s Superior Court sentence.\\n\\n Chief Judge Howell agreed with the recommendations and\\norally announced that Duckett was to serve the statutory\\nmaximum sentence of 24 months’ imprisonment running\\nconsecutively to the 13 months ordered by the Superior Court.\\nAfter so stating, the Chief Judge asked Duckett’s counsel\\nwhether there was any objection. In response, the attorney from\\nthe Public Defender’s office said nothing about the adequacy of\\nthe Chief Judge’s explanation for imposing an above-Guidelines\\nsentence.\\n\\n Even so, now on appeal the Public Defender presents the\\nquestion whether the Chief Judge gave a sufficient reason for\\nDuckett’s sentence. Under 18 U.S.C. § 3553(c)(2), the\\nsentencing judge must provide “the specific reason” for a\\nsentence outside the Guidelines range. Given the absence of an\\nobjection here, Duckett has the burden of showing “plain error.”\\nSee Fed. R. Crim. P. 52(b).\\n\\n A “plain error” is a glaring or obvious judicial mistake of\\nfact or law that adversely “affects substantial rights” but was\\n“not brought to the court’s attention.” Id. In most cases even an\\nobvious error will not constitute a “plain error” unless it\\nprejudiced the defendant – that is, unless it “affected the\\noutcome of the district court proceedings.” United States v.\\n\\f 6\\n\\nOlano, 507 U.S. 725, 734 (1993). And even though the\\ndefendant carries his burden of showing that the district court\\ncommitted a “plain error” as so defined, that will not be enough.\\n“Rule 52(b) is permissive, not mandatory. If the forfeited error\\nis ‘plain’ and ‘affect[s] substantial rights,’ the court of appeals\\nhas authority to order correction, but is not required to do so.”\\nId. at 735 (alteration in original); see also Puckett v. United\\nStates, 556 U.S. 129, 142 (2009) (noting “that a ‘per se approach\\nto plain-error review is flawed’”) (quoting United States v.\\nYoung, 470 U.S. 1, 17, n.14 (1985)).\\n\\n Here it is apparent why defense counsel did not object to the\\nChief Judge’s explanation of Duckett’s 24-month sentence.\\nDuckett “and anyone else present at his sentencing hearing[]\\nmust have understood why the district court imposed an above-\\nGuidelines sentence of” 24 months. United States v. Jackson,\\n848 F.3d 460, 462 (D.C. Cir. 2017).\\n\\n Needless to say, Chief Judge Howell was quite familiar with\\nDuckett. This was the third time in two years he stood before\\nher in a revocation hearing. In the Colorado sentencing and in\\nthe two previous revocation hearings, Duckett received the\\nbenefit of below-Guidelines sentences. He and his counsel\\ntherefore should have anticipated that an above-Guidelines\\nsentence could be coming. The Sentencing Guidelines indicate\\nthat courts may impose lengthier sentences when prior sentences\\nwere lenient. See U.S.S.G. § 7B1.4, application note 4 (advising\\nthat when “the original sentence was the result of a downward\\ndeparture (e.g., as a reward for substantial assistance), . . . an\\nupward departure may be warranted”). At the hearing Duckett’s\\ncounsel acknowledged that this time around the “Court is going\\nto impose a significant sanction.” Hr’g Tr. 4/20/18 at 22.\\n\\n As the prosecutor argued, and as the Chief Judge concluded,\\nDuckett’s three violations triggering his latest revocation\\n\\f 7\\n\\nproceeding were only part of his story. The prosecutor gave this\\nsummary, which was not all-inclusive: Duckett “shank[ed]\\nsomeone in prison”; he participated in the deadly beating of an\\ninmate; he was arrested on nine separate occasions while under\\nsupervision; he was arrested and convicted of assaulting a police\\nofficer while under supervision in this case; and so forth. Hr’g\\nTr. 4/20/2018 at 12. It hardly needs explaining to understand\\nwhy the prosecutor and the probation office urged a 24-month\\nabove-Guidelines sentence to protect the community.\\n\\n Chief Judge Howell reviewed and explained Duckett’s\\n“very extensive” and “horrible” criminal record, as she put it,\\nand his repeated violations of the terms she and other judges had\\nimposed on him during his brief periods out of prison. Hr’g Tr.\\n4/20/2018 at 23. She told Duckett that she had given him\\n“opportunities in the past”; that he did not take advantage of\\nthose opportunities; and that now the court’s responsibility was\\n“to protect the public from future crimes by you.” Hr’g Tr.\\n4/20/2018 at 29–30. All of this sufficiently explained the Chief\\nJudge’s decision to send Duckett back to prison for 24 months.\\nSee Rita v. United States, 551 U.S. 338, 356–59 (2007); see also\\nUnited States v. Ransom, 756 F.3d 770, 774, 776 (D.C. Cir.\\n2014). At least for those 24 months Duckett would be off the\\nstreets, unable to “stretch out” any member of the public, or to\\nsteal television sets, or to assault police officers, or to commit\\nother forms of mayhem.\\n\\n Duckett also claims that the Chief Judge relied on mistakes\\nof fact. The supposed mistakes dealt with the extent of\\nDuckett’s prior educational and vocational training, substance\\nabuse and mental health treatment, and anger management\\ncounseling. There is a mistake here but it is on Duckett’s part.\\nThe Chief Judge did not justify the sentence with reference to\\nDuckett’s treatment or lack thereof in the past. Treatment came\\n\\f 8\\n\\nup only once, when the Chief Judge asked whether Duckett’s\\ndrug treatment had been successful.\\n\\n Duckett’s final claim of “plain error” is that the calculation\\nof his criminal history was in error. The federal district court in\\nColorado, in 2012, determined that Duckett was in Category VI\\nbecause he was a career offender. The court identified two\\npredicate offenses for career-offender status under U.S.S.G.\\n§ 4B1.1: a 1989 conviction for attempted distribution of cocaine\\nand a 1993 conviction for assault with intent to commit murder.\\n\\n Duckett argues the calculation was in error in light of this\\ncourt’s intervening decision in United States v. Winstead, 890\\nF.3d 1082, 1090–92 (D.C. Cir. 2018), holding that attempted\\ndrug distribution does not qualify as a predicate offense for\\ncareer-offender status. We are skeptical of this argument for\\nseveral reasons. One problem is that Winstead is the law of this\\ncircuit, not the law of the Tenth Circuit where Duckett was\\noriginally sentenced. Another problem is that the Guidelines\\ninstruct courts in revocation proceedings to use the criminal\\nhistory category “determined at the time the defendant originally\\nwas sentenced to the term of supervision,” U.S.S.G. § 7B1.4,\\napplication note 1, which is what Chief Judge Howell did\\nwithout objection from defense counsel. See United States v.\\nJones, 690 F. App’x 336, 337 (6th Cir. 2017) (unpublished\\nopinion); United States v. Johnson, 570 F. App’x 611, 612 (7th\\nCir. 2014) (unpublished order). The third problem with\\nDuckett’s argument is that if his attempted drug distribution is\\ndisregarded, his armed robbery conviction in 1984 would be\\nsubstituted, thus placing him in the same criminal history\\ncategory. We need not resolve these issues, however. Even if\\nthere was an error and even if the error was obvious, it did not\\namount to a “plain error” because there is no “reasonable\\nprobability that, but for the error, the outcome of the proceeding\\nwould have been different.” United States v. Mack, 841 F.3d\\n\\f 9\\n\\n514, 522 (D.C. Cir. 2016) (quoting Molina–Martinez v. United\\nStates, 136 S. Ct. 1338, 1343 (2016)). Dropping Duckett’s\\ncriminal history category from VI to V would cause only a\\ndecrease in the revocation Guidelines range from 8-14 months\\nto 7-13 months. Given Chief Judge Howell’s reasons for\\nimposing the maximum 24 month sentence, that one-month\\ndifference would not have mattered. See Molina-Martinez, 136\\nS. Ct. at 1346–47.\\n\\n We have considered and rejected Duckett’s remaining\\narguments.\\n\\n Affirmed.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4180,"cells":{"id":{"kind":"number","value":4649683,"string":"4,649,683"},"name":{"kind":"string","value":"UNITED STATES of America, Appellee v. Pheerayuth BURDEN, Appellant"},"name_abbreviation":{"kind":"string","value":"United States v. Pheerayuth Burden"},"decision_date":{"kind":"string","value":"2019-08-20"},"docket_number":{"kind":"string","value":"17-3018; C/w 17-3019"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Henderson, Pillard, Rogers\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge PILLARD\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued January 16, 2019 Decided August 20, 2019\\n\\n No. 17-3018\\n\\n UNITED STATES OF AMERICA,\\n APPELLEE\\n\\n v.\\n\\n PHEERAYUTH BURDEN,\\n APPELLANT\\n\\n\\n Consolidated with 17-3019\\n\\n\\n Appeals from the United States District Court\\n for the District of Columbia\\n (No. 1:14-cr-00069-2)\\n (No. 1:14-cr-00069-3)\\n\\n\\n Lindsay C. Harrison, pro bono, argued the cause and filed\\nthe briefs for appellant. James Dawson entered an appearance.\\n\\n Daniel J. Lenerz, Assistant U.S. Attorney, argued the\\ncause for appellee. With him on the brief were Jessie K. Liu,\\nU.S. Attorney, and Elizabeth Trosman and Elizabeth H.\\nDanello, Assistant U.S. Attorneys.\\n\\n Before: HENDERSON, ROGERS and PILLARD, Circuit\\nJudges.\\n\\n 2\\n Opinion for the Court filed by Circuit Judge PILLARD.\\n\\n Opinion concurring in part and concurring in the judgment\\nfiled by Circuit Judge ROGERS.\\n\\n PILLARD, Circuit Judge: The Arms Export Control Act\\n(AECA) criminalizes exporting defense articles without a\\nlicense. 22 U.S.C. § 2778(b)(2), (c). Pheerayuth Burden, a\\nU.S.-resident Thai national who ran a business exporting goods\\nfrom the United States to Thailand, and his export business,\\nWing-On LLC (collectively, the defendants), exported five\\nassault-rifle magazines and a grenade-launcher mount.\\nFollowing a three-week trial, a jury convicted the defendants\\nof conspiracy to violate the AECA, unlawful export in violation\\nof the AECA, and conspiracy to launder money.\\n\\n The defendants contend that three of the district court’s\\nrulings are reversible error. First, they argue that the court\\nerred in admitting video deposition testimony by a key witness\\nover a Confrontation Clause objection where the government\\nitself rendered the witness “unavailable” at trial by deporting\\nhim shortly before trial without first making reasonable efforts\\nto arrange his return. Second, they challenge a jury instruction\\ndefining the “willfulness” element of unlawful exportation of\\ndefense articles as requiring only proof that the defendants\\n“acted with knowledge that the conduct was unlawful.” That\\ninstruction was inadequate, they contend, because it failed to\\ntie the willfulness finding to the pertinent conduct and law,\\ncreating an impermissible risk that the jury relied on evidence\\nthat Burden thought he was violating Thai import law. Third,\\ndefendants claim that the district court erred in admitting\\nBurden’s non-Mirandized statements because it failed to\\naccount for his limited English abilities in determining that he\\nwas not in custody when agents interrogated him.\\n\\n 3\\n We hold that the district court erred in admitting the\\ndeposition testimony because the government failed to make\\nreasonable efforts before it deported the witness to procure his\\npresence at trial. We conclude that the jury instruction was\\ncorrect as far as it went in instructing the jury to find that “the\\ndefendant knew that his conduct was unlawful,” and that\\n“willfully” violating the law does not require proof “that a\\ndefendant had read, was aware of, or had consulted the\\nlicensing provisions of the Arms Export Control Act” as such.\\nAppellants’ Appendix (App.) 66. But we suggest clarification\\nof the willfulness instruction to more squarely require a finding\\nthat defendants were aware of and knowingly violated their\\nlegal obligation not to commit the charged actus reus. A case\\nsuch as this one—that includes evidence of consciousness of\\nguilt relating to distinct actus reus arguably violating different,\\nuncharged legal obligations—creates some risk of the jury\\nrelying on evidence of consciousness of guilt unrelated to the\\ncharged crime. We affirm the district court’s determination\\nthat Burden was not in custody because, even assuming\\nlanguage proficiency is relevant to the custody inquiry, a\\nreasonable officer would not have thought Burden’s imperfect\\nEnglish meant a reasonable person in his position would have\\nbelieved himself detained during the interview.\\n\\n Because the error we identify was not harmless, we vacate\\nthe judgments and remand for proceedings consistent with this\\nopinion.\\n\\n BACKGROUND\\n\\n A. Legal Background\\n\\n The AECA establishes executive-branch control over the\\nexport and import of “defense articles,” meaning arms or other\\nmilitary items. See 22 U.S.C. § 2278. It authorizes the\\n\\n 4\\nPresident, “[i]n furtherance of world peace and the security and\\nforeign policy of the United States,” to control the export of\\ndefense articles and services, designate which items count as\\ndefense articles and services, and promulgate regulations for\\nthose purposes. Id. § 2778(a)(1). The designated defense\\narticles make up the United States Munitions List (the\\nMunitions List or the List). Id. With certain enumerated\\nexceptions, “no defense articles or defense services designated\\nby the President” as part of the Munitions List “may be\\nexported or imported without a license for such export or\\nimport, issued in accordance with” the AECA and its\\nassociated regulations. Id. § 2778(b)(2). The State Department\\nis responsible for issuing licenses. See id.; 22 C.F.R. §§ 120.1,\\n120.20. The decision whether to issue an export license\\nimplicates sensitive issues of national security and foreign\\npolicy. It must “take into account whether the export of an\\narticle would contribute to an arms race, aid in the development\\nof weapons of mass destruction, support international\\nterrorism, increase the possibility of outbreak or escalation of\\nconflict, or prejudice the development of . . . arms control or\\nnonproliferation agreements or other arrangements.” 22\\nU.S.C. § 2778(a)(2). The statute criminalizes “willfully\\nviolat[ing] any provision of this section . . . or any rule or\\nregulation issued under this section.” Id. § 2278(c). It thus\\ncriminalizes willfully exporting defense articles without a\\nlicense.\\n\\n The President delegated to the Secretary of State the\\nauthority to designate defense articles and promulgate\\nregulations under the AECA, see Exec. Order No. 13637, 78\\nFed. Reg. 16,129 (2013); 22 C.F.R. § 120.1(a), and the\\nSecretary accordingly promulgated the International Traffic in\\nArms Regulations (ITAR), 22 C.F.R. §§ 120.1-130.17. The\\nITAR prohibits exporting defense articles and services without\\n“obtaining the required license or other written approval” from\\n\\n 5\\nthe appropriate office of the State Department. 22 C.F.R.\\n§ 127.1(a). The ITAR also includes the Munitions List, which\\nruns to over forty pages in the Code of Federal Regulations.\\nSee id. § 121.1. The covered defense articles are described with\\nvarying levels of specificity, such as “[r]iflescopes\\nmanufactured to military specifications,” id. (Category I(f)),\\n“[g]uns over caliber .50,” id. (Category II(a)), “[i]ron powder\\n. . . with particle size of 3 micrometers or less produced by\\nreduction of iron oxide with hydrogen,” id. (Category\\nV(c)(4)(i)(B)), and “[h]elmets . . . providing a protection level\\nequal to or greater than NIJ Type IV,” id. (Category X(a)(6)).\\nThe convictions in this case relate to items in Category I(h) of\\nthe Munitions List: “Components, parts, accessories and\\nattachments” for the firearms listed in Category I(a)-(g). See\\nApp. 89-90.\\n\\n B. Factual and Procedural Background\\n\\n Burden started Wing-On LLC (Wing-On), a freight-\\nforwarding business that shipped American goods to Thailand,\\naround 2008. In 2010, Kitibordee Yindeear-Rom became one\\nof Burden’s customers. A Thai national living in Thailand,\\nYindeear-Rom had a business importing many different types\\nof goods from the United States to Thailand. As part of that\\nbusiness, he helped his customers get gun parts and accessories\\nfrom the U.S. that they could not purchase directly because\\nU.S. companies would neither accept Thai credit cards nor ship\\nthe parts to Thailand. According to Yindeear-Rom, Burden\\ninitially ordered gun parts for him from U.S. vendors, received\\nthem in the United States, then shipped them to Thailand.\\nSupplemental Appendix (S.A.) 291A-91B. Yindeear-Rom\\nlater began placing the orders himself using a debit card\\nattached to a U.S. bank account Burden opened. S.A. 294-96,\\n479. Yindeear-Rom testified in his deposition that he\\nreimbursed Burden for the purchases he made on Burden’s\\n\\n 6\\ndebit card by transferring money to Thai bank accounts\\nbelonging to Burden and Burden’s associate. S.A. 298-300.\\nNeither Burden nor Wing-On had a license to export defense\\narticles on the Munitions List.\\n\\n In October 2013, Yindeear-Rom took a vacation to the\\nUnited States, where he was stopped and interviewed by\\nDepartment of Homeland Security (DHS) agents. He was\\narrested two days later for conspiracy to violate American\\nexport laws. He later pleaded guilty and was sentenced to\\nthirty-six months in prison. At Yindeear-Rom’s initial court\\nappearance, the DHS agents saw in the courtroom two people\\nthey believed to be Burden’s wife and roommate, respectively.\\nConcerned that Burden might have been alerted to the\\ninvestigation, the agents went immediately to Wing-On’s\\nwarehouse. Burden was not there, but the agents met one of\\nhis employees, who helped the agents call him. They called\\nhim again later that day and arranged an interview for that\\nevening at the warehouse.\\n\\n The DHS agents interviewed Burden in English without an\\ninterpreter. They did not advise Burden of his rights with the\\nfamiliar warnings officials must give suspects in custodial\\ninterrogation under Miranda v. Arizona, 384 U.S. 436, 444\\n(1966). At the beginning of the interview, one of the agents\\ngave the following preamble (recounted here with Burden’s\\naffirmative interjections omitted):\\n\\n We are federal agents for the U.S. Government so\\n I have to let you know that you have to be honest\\n with us[,] okay? If you don’t want to answer\\n something, you don’t have to answer but you\\n cannot lie to us. All right? And you can’t withhold\\n relevant information. If you do, that is a crime.\\n\\n 7\\n Okay? Punishable by up to five years in prison so\\n just please be honest.\\n\\nApp. 174. The agent then asked, “Is your English good?” Id.\\nBurden replied, “A little bit.” App. 175. “If there’s anything\\nthat I say that you don’t understand, ask me,” the agent said.\\nId. “Okay,” Burden replied. Id. During the interview, Burden\\nadmitted he had shipped gun parts to Yindeear-Rom and\\nfalsified customs declarations. App. 266-71. The agents did\\nnot arrest Burden at the end of the interview. S.A. 420-21. The\\ncourt’s eventual admission of Burden’s statement at his trial\\nover defense objections is the subject of his Miranda claim on\\nappeal. App. 429-33.\\n\\n Burden was arrested six months later, in May 2014. Trial\\nwas initially scheduled for November 2015, but was continued\\ntwice, first to April 2016 and then to September 2016. The\\ndistrict court granted the second continuance because many\\ndocuments remained to be translated into English. The court\\nnoted that the defense “can’t actually do this without translated\\ndocuments” and that “it’s not that [the defense has] been less\\nthan diligent about it.” App. 381. That continuance introduced\\na wrinkle into the trial, however: Yindeear-Rom was\\nscheduled to be released from prison in June 2016, three\\nmonths before the new trial date, and was to be deported after\\nhis release. The government had a clear path to remove\\nYindeear-Rom upon his release because he had stipulated when\\nhe pleaded guilty to an order of removal that would “render[]\\nhim permanently inadmissible to the United States,” which\\nassured that the government, “promptly upon his release from\\nconfinement . . . may execute the order of removal according\\nto the applicable laws and regulations.” App. 47 (alteration in\\noriginal). The government in February 2016 moved to take\\nYindeear-Rom’s deposition under Federal Rule of Criminal\\nProcedure 15, which governs depositions taken to preserve a\\n\\n 8\\npotential witness’s testimony for trial. S.A. 1; Fed. R. Crim. P.\\n15(a)(1). The district court granted the motion over the\\ndefendants’ objections. The defense objected to the deposition\\nfor some of the same reasons it asked for the second trial\\npostponement: the government had not produced sufficient\\ndiscovery to allow them to prepare for Yindeear-Rom’s\\ntestimony. The court nonetheless allowed the deposition to be\\ntaken to preserve evidence in the event that Yindeear-Rom\\nwould be unavailable to testify. Yindeear-Rom’s videotaped,\\nin-court deposition took place over four days in March and\\nApril. The court granted the government’s motion to reduce\\nYindeear-Rom’s sentence in exchange for his testimony.\\n\\n The United States deported Yindeear-Rom to Thailand in\\nApril 2016. Even though the government had substantial\\nbargaining leverage before it moved for his sentence reduction,\\nthere is no record that it made any efforts before deporting him\\nto secure Yindeear-Rom’s presence at trial. It was only once\\nYindeear-Rom was back in Thailand that the government\\nbegan to make such efforts.\\n\\n In seeking to bring Yindeear-Rom back a few months\\nlater, the government contacted Yindeear-Rom’s counsel by\\nphone and mail. United States v. Burden, No. 14-cr-0069\\n(RMC), 2016 WL 5108010, at *3 (D.D.C. Sept. 20, 2016). Its\\nletter to counsel included a subpoena for Yindeear-Rom’s\\ntestimony at trial and a promise to help him obtain a visa and\\nto pay his travel expenses, “including but not limited to round-\\ntrip airfare, transportation, room, board, and per diem witness\\nfee.” Id. (quoting letter, S.A. 51). Yindeear-Rom’s lawyer\\n“forwarded the letter and subpoena to Mr. Yindeear-Rom in\\nThailand, but was unable to confirm receipt or make any\\nrepresentations about [his] willingness to testify.” Id. at *3.\\nThe government also sent the letter and subpoena directly to\\nYindeear-Rom’s last known email and physical address in\\n\\n 9\\nThailand. Id. at *4. The government received no response to\\neither, but confirmed that the letter was signed for by “K. Yen,”\\nwhich it believed to be Yindeear-Rom. Id. DHS personnel in\\nThailand eventually reached Yindeear-Rom by phone and\\nlearned that he had received the email and letter, but that he\\n“had no desire to travel to the United States to cooperate in any\\nway.” Id. (quoting DHS Report of Investigation, S.A. 75).\\n\\n At trial, the court granted the government’s motion to\\nadmit the Rule 15 deposition over the defendants’ objections.\\nDefendants argued that the government should have sought to\\nkeep Yindeear-Rom in the country between his release from\\nprison and the trial. The court concluded that the witness was\\nunavailable, and that “the use of a videotaped deposition taken\\nin court, before the trial judge and including the presence of\\nMr. Burden and cross examination by both defense lawyers,”\\nwas “a very good substitute” for Yindeear-Rom’s live trial\\ntestimony “that would allow the jury to observe his demeanor\\nand preserve the Defendants’ rights to confront witnesses\\nagainst them.” Id. at *3, *8.\\n\\n The government’s trial evidence included Burden’s\\nstatement to DHS agents at the Wing-On warehouse that he\\nmislabeled customs declarations for shipments containing gun\\nparts, and packed gun parts hidden among other items for\\nshipping. App. 269-72. In his defense, Burden highlighted his\\nstatements that he took those steps to evade Thai customs. He\\ntold the agents that he concealed gun parts among other items\\nbecause “[y]ou want to hide from the custom in Thailand,”\\nApp. 271, and that he falsified customs documents “[b]ecause\\nof tax in Thailand,” App. 281.\\n\\n The government, for its part, pointed to circumstantial\\nevidence tending to show that Burden had reason to know that\\nhe was violating U.S. arms-export law. For instance, Burden\\n\\n 10\\nacknowledged in his statement to DHS agents that people\\n“[n]eed a license” to ship certain things, like gun parts. App.\\n203-04. Yindeear-Rom received a notice that a seized\\nshipment of gun parts violated the ITAR, see S.A. 212-13, and\\neven though he had ordered that shipment through a different\\nshipper (not Wing-On), Yindeear-Rom forwarded the notice to\\nBurden asking what he should do, see S.A. 282-83. Yindeear-\\nRom then testified that the notice he forwarded informed both\\nof them of the requirements of the ITAR. S.A. 369. Burden\\nhad also received a notice directly from U.S. Customs and\\nBorder Protection that it had seized a rifle scope (controlled\\nunder a distinct set of regulations analogous to the ITAR but\\ncovering different items) because the scope could not be\\nexported without a license. See App. 171.\\n\\n Further evidence, however, suggested that Burden either\\ndid not realize he was shipping real gun parts, or thought it was\\nlegal under U.S. export law to ship those parts if they were to\\nbe used with toy guns. As part of their business, the defendants\\nshipped BB guns (air guns that shoot small metal balls) and\\nAirsoft toys (which are similar to BB guns and shoot plastic\\npellets). See App. 311-12, 574. Airsoft and BB guns\\nthemselves may lawfully be exported without a State\\nDepartment license. The defense’s expert witness on firearms\\nand Airsoft identification testified that an Airsoft toy “looks\\nlike a gun in every way, shape or form from the outside, same\\nlength, weight, contour, field markings, but it won’t kill\\nanybody.” S.A. 672-73. They have all the same parts as the\\nreal guns they mimic; in fact, real gun parts can be used with\\nAirsoft toys. App. 508-10; S.A. 676, 678. In his statement to\\nDHS agents, Burden said that his Thai customers were not\\nusing gun parts “for the gun,” but “for the BB gun . . . for the\\npaintball [gun].” App. 259. A Wing-On manager testified that\\nemployees were instructed not to ship parts for real guns but\\nthat they could ship parts to “be used for toys for Airsoft items.”\\n\\n 11\\nApp. 554-56, 574. Burden affirmed to one of his customers\\nthat if a part was for a BB gun “then there’s no problem”\\nshipping it. App. 361. That evidence tended to support\\nBurden’s defense that he concealed the contents of shipments\\nto evade Thai customs law.\\n\\n Evidence also showed that Burden tried not to ship real\\ngun parts after realizing it was illegal. Burden sent an email to\\nYindeear-Rom saying, “I have warned you many times that I\\ndo not accept gun parts. . . . Stop sending them to me\\nabsolutely!” App. 358. The Wing-On warehouse had a “no-\\ngo” shelf for gun parts, where employees would segregate\\nitems that they could not lawfully ship. See App. 350, 554-55.\\nThere was ambiguous evidence suggesting that Wing-On may\\nhave ultimately shipped some gun parts on the no-go shelf.\\n\\n The conflicting evidence regarding Burden’s intent\\noccasioned a dispute over the jury instruction defining\\n“willfully” under the AECA. The defendants proposed using\\nthe Fifth Circuit pattern jury instruction, which requires the\\njury to find that a defendant exported articles on the Munitions\\nList without obtaining a license from the Department of State;\\nand “[t]hat the defendant acted ‘willfully,’ that is, that the\\ndefendant knew such license . . . was required for the export of\\nthese articles and intended to violate the law by exporting them\\nwithout such license.” App. 99.\\n\\n Instead, the district court adopted the government’s\\nproposed instruction, which described the requisite state of\\nmind as follows:\\n\\n [A]n act is done willfully if it is committed with the\\n knowledge that it was prohibited by law or was\\n done in disregard of a known legal obligation. The\\n government must prove that a defendant acted with\\n knowledge that the conduct was unlawful. While\\n\\n 12\\n the government must show that a defendant knew\\n that the conduct was unlawful, it is not necessary\\n for the Government to show that a defendant was\\n aware of the specific law, rule, or regulation that\\n the conduct may have violated.\\n\\n In other words, the government need not prove that\\n a defendant had read, was aware of, or had\\n consulted the licensing provisions of the Arms\\n Export Control Act or the International Traffic in\\n Arms Regulations, or the Munitions List. The\\n government, however, must prove beyond a\\n reasonable doubt, by reference to facts and\\n circumstances surrounding the case, that a\\n defendant knew that the conduct was unlawful.\\n\\nApp. 66; see App. 435-37. The jury convicted the\\ndefendants on all three counts—conspiracy to violate the\\nAECA, unlawful export in violation of the AECA, and\\nconspiracy to launder money. The defendants timely appealed\\nthe district court’s admission of Yindeear-Rom’s deposition\\nand Burden’s statement to DHS, as well as the district court’s\\njury instruction on the definition of “willfully.”\\n\\n ANALYSIS\\n\\n A. Yindeear-Rom Was Not “Unavailable” for\\n Purposes of the Confrontation Clause.\\n\\n We review legal conclusions regarding the Confrontation\\nClause de novo, and reverse any error unless it was harmless\\nbeyond a reasonable doubt. United States v. Moore, 651 F.3d\\n30, 69 (D.C. Cir. 2011) (citing constitutional harmless-error\\nstandard in Chapman v. California, 386 U.S. 18, 24 (1967)).\\nBecause the government concedes that, if it was error,\\nadmitting Yindeear-Rom’s video deposition testimony instead\\n\\n 13\\nof producing him at trial was not harmless, we need only\\nconsider the claimed error itself. See Appellee’s Br. 55 n.5.\\n\\n The Confrontation Clause of the Sixth Amendment\\nguarantees that, “[i]n all criminal prosecutions, the accused\\nshall enjoy the right . . . to be confronted with the witnesses\\nagainst him.” U.S. Const. amend. VI. Courts have long\\nrecognized the critical importance of a criminal defendant’s\\n“opportunity to cross-examine and impeach a witness at trial\\nbefore the jury that will decide his innocence or guilt.” United\\nStates v. Lynch, 499 F.2d 1011, 1022 (D.C. Cir. 1974). Live\\nwitness testimony gives the jury the opportunity to assess the\\nwitness’s demeanor and allows counsel to adjust examination\\nto other evidence and to the jury’s apparent reactions as the\\nwitness testifies. “William Blackstone long ago recognized\\nthis virtue of the right to confrontation, stressing that through\\nlive testimony, ‘and this [procedure] only, the persons who are\\nto decide upon the evidence have an opportunity of observing\\nthe quality, age, education, understanding, behavior, and\\ninclinations of the witness.’” United States v. Yida, 498 F.3d\\n945, 950 (9th Cir. 2007) (alteration in original) (quoting 3\\nWilliam Blackstone, Commentaries on the Laws of England\\n373-74 (1768)).\\n\\n A testimonial statement by a person who does not appear\\nat trial may be admitted “only where the declarant is\\nunavailable, and only where the defendant has had a prior\\nopportunity to cross-examine.” Crawford v. Washington, 541\\nU.S. 36, 59 (2004); see also Fed. R. Evid. 804(b)(1).\\nUnavailability and prior opportunity to cross-examine are\\nindependent criteria: Even when a defendant had a prior\\nopportunity to cross-examine the witness, if the government\\ndoes not establish that the witness is unavailable, the testimony\\nmust be excluded. See id. at 57. Here, the parties agree that\\ndefendants had an opportunity to cross-examine Yindeear-\\n\\n 14\\nRom during his videotaped deposition. The sole question is\\nwhether he was “unavailable” for purposes of the\\nConfrontation Clause.\\n\\n Defendants argue that Yindeear-Rom was not truly\\nunavailable because the government procured Yindeear-Rom’s\\nunavailability by deporting him, and, in any case, it did not\\nmake reasonable efforts before it deported him to ensure his\\npresence at trial. For its part, the government contends that\\nthere is no per se presumption that the government fails the\\nConfrontation Clause’s test of good-faith and reasonable\\nefforts when it deports a witness, and that its reasonable, good-\\nfaith efforts after it deported Yindeear-Rom sufficed.\\n\\n As a general matter, a witness is considered unavailable\\nonly if the prosecution cannot procure her with good-faith,\\nreasonable efforts. See Ohio v. Roberts, 448 U.S. 56, 74\\n(1980), abrogated in part on other grounds by Crawford, 541\\nU.S. 36. It is the prosecution’s burden to establish that its\\nactions meet that test. Id. at 74-75. “The law does not require\\nthe doing of a futile act” such as producing a witness who has\\ndied, but “if there is a possibility, albeit remote, that affirmative\\nmeasures might produce the declarant, the obligation of good\\nfaith may demand their effectuation.” Id. at 74. “The lengths\\nto which the prosecution must go to produce a witness . . . is a\\nquestion of reasonableness.” Id. (alteration in original)\\n(quoting California v. Green, 399 U.S. 149, 189 n.22 (1970)\\n(Harlan, J., concurring)). In Lynch, we held good-faith but\\nnon-exhaustive efforts to find a witness were inadequate to\\nrender her unavailable and justify admitting pretrial hearing\\ntestimony. 499 F.2d at 1024. We treated the witness as\\navailable because she was within the court’s jurisdiction and\\nthe government had not “inquired at the local hospitals, area\\npolice departments, the morgue, or of [the witness’s]\\nemployer.” Id. In contrast, a witness whose “whereabouts\\n\\n 15\\nwere entirely unknown,” who had been sent numerous\\nsubpoenas at her parents’ house, was last heard from outside\\nthe jurisdiction, and was unreachable by her family even in\\ncase of an emergency was held in Ohio v. Roberts to be\\nunavailable even though, as Justice Brennan pointed out in\\ndissent, the government had not followed every possible lead\\nto find her. Compare 448 U.S. at 60-61, 75, with id. at 79-82\\n(Brennan, J., dissenting).\\n\\n When the government seeks to rely on prior recorded\\nstatements of a witness on the ground that the witness is\\nunavailable, it bears the burden of establishing that its\\nunsuccessful efforts to procure the witness’s appearance at trial\\nwere “as vigorous as that which the government would\\nundertake to [secure] a critical witness if it has no [prior]\\ntestimony to rely upon in the event of ‘unavailability.’” Lynch,\\n499 F.2d at 1023; see also United States v. Mann, 590 F.2d 361,\\n367 (1st Cir. 1978). Where the government itself bears some\\nresponsibility for the difficulty of procuring the witness, such\\nas by deporting the witness, the government will have to make\\ngreater exertions to satisfy the standard of good-faith and\\nreasonable efforts than it would have if it had not played any\\nrole. Failing to factor the government’s own contribution to\\nthe witness’s absence into the Confrontation Clause analysis\\nwould warp the government’s incentives. “This relatively high\\ngood faith standard cannot be satisfied by perfunctory efforts,\\nif the rule is not to sanction the government’s procuring\\ndepositions of witnesses, especially shaky witnesses, but then\\ndiscourage attempts to bring the witness to trial so long as the\\ngovernment is satisfied with” the witness’s recorded testimony.\\nMann, 590 F.2d at 367. Rather, the analysis of good-faith,\\nreasonable efforts should “include an assessment of the\\ngovernment’s affirmative conduct” that allowed the witness to\\nbecome unavailable “in the first instance.” Yida, 498 F.3d at\\n955-56. That analysis should account for the good faith and\\n\\n 16\\nreasonableness of the government’s deportation of the witness\\nand of its attempts to secure witnesses it does deport.\\n\\n In a case such as this one, in which the government knew\\nor should have known of the potential need for the witness’s\\ntestimony before he was deported, the government’s duty to\\nmake good-faith, reasonable efforts to ensure the witness’s\\npresence arises before the witness leaves the United States.\\nOther courts that have addressed this question in the context of\\nwitness deportation agree. In United States v. Tirado-Tirado,\\nthe Fifth Circuit refused to deem unavailable a witness whom\\nthe government had deported where the government “failed to\\nmake any concrete arrangements with [the witness] prior to his\\ndeportation,” and only shortly before trial made what the court\\nacknowledged were “fairly exhaustive” efforts to bring him\\nback from Mexico. 563 F.3d 117, 123, 125 (5th Cir. 2009).\\nThe government had orally told the witness before deporting\\nhim that his appearance would be required if the case went to\\ntrial—a step not taken here—but the court faulted it for not\\ntaking several other pre-deportation steps that might have\\nencouraged the witness to appear, including serving him with\\na subpoena or other written notice. Id. at 123-24. In United\\nStates v. Foster, the court followed Tirado-Tirado to exclude\\nvideotaped depositions notwithstanding that the government\\nhad advised the witnesses before deporting them that it might\\nneed them to return, promised to allow their reentry and pay for\\ntheir travel, and collected contact information for them in\\nMexico. 753 F. App’x 307, 309 (5th Cir. 2018). Emphasizing\\nthe constitutional importance of taking “reasonable measures,\\nunder the circumstances, that are likely to ensure that the\\nwitness will return for trial,” the court declined to find the\\nunreachable witnesses “unavailable” because the government\\nhad not verified their contact information from the outset,\\nsought alternative contact information, or remained in contact\\n\\n 17\\nwith them over the months between their deportation and the\\ntrial. Id. at 312.\\n\\n The Ninth Circuit in Yida similarly recognized “the\\ngovernment’s obligation to use ‘reasonable means’ to ‘procure\\nthe declarant’s testimony’ in the context of the government’s\\naffirmative role in a witness’s deportation,” 498 F.3d at 953,\\nand refused to admit testimony from a prior mistrial “when the\\ngovernment itself share[d] some of the responsibility for its\\ninability to produce the witness at [the second] trial,” id. at 956.\\nThe government’s efforts were insufficient where it had chosen\\nto detain the admittedly untrustworthy and skittish witness\\nprior to the first trial but, once it had that transcript in hand as\\na substitute, deported the witness without considering the many\\nalternatives to confinement in federal prison that might have\\nprevented the witness’s absence. Id. at 959-60.\\n\\n The First Circuit, too, in Mann, excluded deposition\\ntestimony of a crucial prosecution witness in lieu of live\\ntestimony where, after the deposition, the government had\\nreturned the witness’s passport and plane ticket that it had\\nseized upon her arrest. 590 F.2d at 366, 368. The court held\\ninsufficient the government’s later offer to pay to bring the\\nwitness back, similar to the government’s offer here. Id. at\\n363. “Implicit . . . in the duty to use reasonable means to\\nprocure the presence of an absent witness is the duty to use\\nreasonable means to prevent a present witness from becoming\\nabsent.” Id. at 368. Where the government fails to take\\nreasonable steps to prevent a witness from becoming absent,\\nthe defendant should not suffer from the government’s choice.\\nSee id.\\n\\n The cases on which the government relies are not to the\\ncontrary. Foster distinguishes United States v. Allie, 978 F.2d\\n1401, 1403-08 (5th Cir. 1992), and United States v. Calderon-\\n\\n 18\\nLopez, 268 F. App’x 279, 289 (5th Cir. 2008), on the basis of\\nextensive efforts the government made before it deported the\\nwitnesses in those cases, none of which were made here.\\nFoster, 753 F. App’x at 311. And in United States v. Eufracio-\\nTorres, it was the defendant who initially opposed the\\ngovernment’s motion to detain seven material witnesses\\npending trial. 890 F.2d 266, 268 (10th Cir. 1989). The court\\nthought it significant that none of those witnesses had been\\ncharged with any crime, yet they were detained while\\ndefendant Eufracio-Torres was free on bond. Id. Eufracio-\\nTorres changed his position to request their detention pending\\ntrial only after their depositions were taken—a request the\\ncourt denied. Id. When the witnesses could not be brought\\nback for trial and the government sought to use the depositions,\\nthe court held that the government’s good-faith, reasonable\\nefforts supported its request: Before they left the country, the\\ngovernment had given the witnesses subpoenas, instructions\\nfor reentry, travel reimbursement, appearance fees, and\\nobtained their promises to return to testify. Id. No such steps\\nwere taken here. The government’s reliance on United States\\nv. Rivera is similarly misplaced. 859 F.2d 1204, 1207-09 (4th\\nCir. 1988) (witnesses but not defendant were being detained,\\nand court found no prejudice claimed or suffered from use of\\nwitness depositions taken at the witnesses’ own request before\\nthey were released and voluntarily left the United States).\\n\\n Under the applicable standard, the government failed to\\nshow that Yindeear-Rom was “unavailable” for purposes of the\\nConfrontation Clause. The government’s efforts to secure his\\npresence at trial did not begin until after he was deported. See\\nBurden, 2016 WL 5108010, at *3-4. Before his deportation,\\nthe government did not give Yindeear-Rom a subpoena, offer\\nto permit and pay for him either to remain in the U.S. or to\\nreturn here from Thailand, obtain his commitment to appear,\\nconfirm his contact information, or take any other measures.\\n\\n 19\\nSee id. at *9. Its only efforts began once he was out of custody,\\nout of the jurisdiction, and no longer dependent on the\\ngovernment’s good graces for lenient treatment. Yindeear-\\nRom’s eagerness to return to Thailand helped to persuade the\\ndistrict court that further efforts to persuade him to testify at\\ntrial would have been futile. See id. But in these circumstances\\nthat eagerness cuts the other way. Given the government’s\\nduty to make good-faith, reasonable efforts before Yindeear-\\nRom’s deportation, “a witness’s known reluctance to testify\\nadds to the government’s burden to show that it made\\n‘reasonable, good faith efforts’ to secure her appearance\\nbecause it makes her failure to appear voluntarily all the more\\nforeseeable.” Brooks v. United States, 39 A.3d 873, 886 (D.C.\\n2012). This is a case where the “possibility, albeit remote, that\\naffirmative measures might produce the declarant . . .\\ndemand[ed] their effectuation.” Roberts, 448 U.S. at 74. Any\\nchance the government had of securing Yindeear-Rom’s\\nappearance at trial would have been far greater had it addressed\\nthe problem as soon as it knew it would rely on his testimony.\\nInstead, its own approach appears to have ensured the futility\\nof the post-deportation efforts.\\n\\n We recognize that it may not always be reasonable to\\nexpect the government to postpone removal until trial—\\nparticularly if the government would have to detain the witness\\nin order to keep her in the country. See Aguilar-Ayala v. Ruiz,\\n973 F.2d 411, 419 (5th Cir. 1992) (enjoining Department of\\nJustice policy of automatically detaining undocumented\\nimmigrant material witnesses because “undocumented aliens\\nhave an overriding liberty interest in not being detained as\\nmaterial witnesses”). And we assume that “in some cases the\\nneed for the criminal defendant to confront the witness at trial\\n(rather than at deposition) might outweigh the material witness’\\nliberty interest in being released immediately,” whereas in\\nother cases it will not. Id. (emphasis omitted). In order to\\n\\n 20\\nidentify the existence and extent of any such conflict, however,\\nthe unavailability inquiry must account for whether the\\ngovernment has addressed potentially reasonable means of\\nsecuring removable witnesses’ live testimony short of\\ndetention, such as placing them in “lesser custody,” “supplying\\nmaintenance, and retaining [their] passport[s] and ticket[s],”\\nand “plac[ing] [them] under subpoena.” Mann, 590 F.2d at\\n366; see also Allie, 978 F.2d at 1407 (holding that the\\ngovernment’s efforts were reasonable where it offered\\nwitnesses work permits to keep them in the United States).\\n\\n Because the government’s omissions place its efforts\\nbelow the standard the Confrontation Clause demands, we need\\nnot decide precisely how the government should have sought\\nto prevent the witness from becoming unavailable. While the\\ngovernment’s deportation of a witness may sometimes fail the\\nstandard of good-faith and reasonable efforts, we reject any per\\nse rule that no witness the government deports can be\\nconsidered unavailable under the Confrontation Clause.\\nConsistent with the fact-intensive nature of the standard, the\\ngovernment decries any per se rule, see Appellee’s Br. at 45-\\n46, the defendants do not advocate one, see Appellants’ Reply\\nBr. 24 n.12, and no circuit has adopted any such categorical\\napproach.\\n\\n We hold that the duty to use reasonable means to procure\\na witness’s presence at trial includes the duty to use reasonable\\nefforts to prevent a witness from becoming absent in the first\\nplace. The government does not dispute that it made no efforts\\nbefore deporting Yindeear-Rom to secure his presence at trial.\\nThe witness thus was not “unavailable” such that prior\\ntestimony could be admitted consistent with the Confrontation\\nClause. Because admitting his deposition was not harmless\\nbeyond a reasonable doubt, we vacate the convictions and\\nremand for a new trial.\\n\\n 21\\n B. Guidance Regarding the Jury Instruction’s\\n Definition of “Willfully”\\n\\n Although our resolution of the Confrontation Clause issue\\nis sufficient to dispose of this appeal, we provide some\\nguidance regarding the jury instruction’s definition of the term\\n“willfully,” an issue that was fully briefed and argued.\\n\\n The parties agree that defendants “willfully” violate the\\nAECA only where they act with knowledge that their conduct\\nis unlawful. The Supreme Court has explained that “willfully”\\nis “a word of many meanings whose construction is often\\ndependent on the context in which it appears.” Bryan v. United\\nStates, 524 U.S. 184, 191 (1998) (internal quotation marks\\nomitted). “As a general matter, when used in the criminal\\ncontext, a ‘willful’ act is one undertaken with a bad purpose.\\nIn other words, in order to establish a ‘willful’ violation of a\\nstatute, the Government must prove that the defendant acted\\nwith knowledge that his conduct was unlawful.” Id. at 191-92\\n(internal quotation marks omitted). The parties disagree about\\nhow specific that knowledge must be under the AECA’s\\nwillfulness standard.\\n\\n Defendants argue that the AECA’s prohibition on\\n“willfully violat[ing] any provision of this section,” 22 U.S.C.\\n§ 2778(c), means that they should not have been convicted\\nunless they “were aware of the specific law, rule, or regulation\\nthat the[ir] conduct may have violated.” Appellants’ Br. 17\\n(alteration in original). Before us, they frame this standard as\\n“requir[ing] the government to prove an individual was aware\\nthe items he exported were on the [Munitions List],” id. at 18-\\n19, though their requested jury instruction would instead have\\ndemanded proof that they knew that there was a license\\nrequirement for their exports, App. 99. The heart of\\ndefendants’ case at trial was that they “had not willfully\\n\\n 22\\nviolated [the] AECA because they believed the parts were\\nintended for use in Airsoft BB guns, not real guns, and they\\nbelieved such parts could be exported without a license.”\\nAppellants’ Br. 12. Under the court’s jury instructions, they\\ncontend, “intent to violate Thai customs duties or Thai gun\\ncontrol laws could support the jury’s finding of willfulness.”\\nId. at 31. The government defends the district court’s\\ninstruction on the ground that “the AECA is not one of the\\nextremely rare, highly technical statutes that the Supreme\\nCourt has found to require a heightened willfulness showing,”\\nso the government did not have to prove defendants had\\nknowledge of the license requirement or the Munitions List as\\nsuch. Appellee Br. 16.\\n\\n As discussed below, because it was ambiguous as to what\\n“conduct” defendants had to know was “unlawful,” the district\\ncourt’s jury instruction arguably fell short of the baseline\\nrequirement that the mens rea relate to the charged actus reus.\\nBut we believe that the district court correctly instructed that,\\nif defendants knew exporting the charged items without a\\nlicense was unlawful, they did not need specific knowledge of\\nthe Munitions List. Thus, the district court was right that “the\\ngovernment need not prove that a defendant had read, was\\naware of, or had consulted the licensing provisions of the Arms\\nExport Control Act or the International Traffic in Arms\\nRegulations, or the Munitions List.” App. 66. For purposes of\\nthe AECA, a requirement of proof that defendants knew the\\nproscribed conduct was unlawful adequately protects against\\n“the danger of ensnaring individuals engaged in apparently\\ninnocent conduct.” Bryan, 524 U.S. at 194.\\n\\n Most criminal prohibitions require only proof that the\\ncrime was committed “knowingly,” meaning that the defendant\\nknew of the facts that made his act illegal, even if he did not\\nknow the act was illegal. When Congress wants to ensure that\\n\\n 23\\ndefendants will be convicted only if they have a more culpable\\nstate of mind, it limits the crime to conduct that a defendant\\nengages in “willfully.” See id. at 191-92. The Court has\\ndeveloped an approach to “willfulness” that is calibrated to the\\nstatutes in which it appears, consistently reading it to require\\nproof that the defendant had a sufficiently culpable state of\\nmind, but not to require proof so specific as to stymie the\\nstatute’s enforcement. It has required proof that a defendant\\nknow which law he was breaking in only two contexts:\\ncriminal tax evasion, and currency structuring.\\n\\n In Cheek v. United States, a tax case, the Court explained\\nCongress’s inclusion of a willfulness requirement for felony\\ntax-evasion as resting partly on the recognition that “[t]he\\nproliferation of statutes and regulations has sometimes made it\\ndifficult for the average citizen to know and comprehend the\\nextent of the duties and obligations imposed by the tax laws.”\\n498 U.S. 192, 199-200 (1991). The Court held that the\\nwillfulness requirement could be defeated by the defendant’s\\ngood-faith belief that he “was not a person required to file a\\nreturn or to pay income taxes and that wages are not taxable\\nincome,” even if that belief was unreasonable. Id. at 203.\\n\\n In Ratzlaf v. United States, the Court interpreted\\n“willfully” in the context of a statute criminalizing “currency\\nstructuring.” 510 U.S. 135, 136, 144 (1994). The anti-\\nstructuring law prohibited a customer from breaking up a large\\nfinancial transaction into multiple smaller ones to avoid\\ntriggering a bank’s legal duty to report transactions over\\n$10,000. Id. at 145. The Court held that “willfulness” meant\\nthat the defendant had to be aware not only of reporting\\nobligations applicable to banks, but of his own distinct\\nobligation not to structure his transactions for the purpose of\\navoiding triggering the bank’s obligation to report. Id. at 137-\\n38. In other words, “[t]o convict Ratzlaf of the crime with\\n\\n 24\\nwhich he was charged . . . the jury had to find he knew the\\nstructuring in which he was engaged was unlawful.” Id. at 149.\\n\\n In Bryan, which concerned unlicensed firearms-dealing,\\nthe Court concluded that “requiring only knowledge that the\\nconduct is unlawful [was] fully consistent” with Congress’s\\npurpose in adding a willfulness requirement to the firearms\\nstatute “to protect law-abiding citizens who might\\ninadvertently violate the law.” 524 U.S. at 195 n.23. Bryan\\ndistinguished the firearms law before it from the “highly\\ntechnical” statutes at issue in Cheek and Ratzlaf, which had\\njustified “‘carv[ing] out an exception to the traditional rule’\\nthat ignorance of the law is no excuse and requir[ing] that the\\ndefendant have knowledge of the law”—that is, the specific\\nlaw he violated. Bryan, 524 U.S. at 194-95 (quoting Cheek,\\n498 U.S. at 200).\\n\\n Following Bryan, most courts of appeals to consider the\\nissue have interpreted the AECA’s requirement that a violation\\nbe “willful” not to require specific awareness of the Munitions\\nList as such. What they have required—as we do here—is\\nproof that defendants knew it was illegal to export the items\\nthey shipped without a license. See United States v. Henry, 888\\nF.3d 589, 598-600 (2d Cir. 2018) (“[W]illfulness requires only\\nthat the defendant know that what he was doing was illegal,\\nand not that he know that his conduct was prohibited under a\\nspecific AECA provision or related regulation.”); United States\\nv. Bishop, 740 F.3d 927, 932-33 (4th Cir. 2014) (government\\nneed not show defendant knew “that the ammunition was\\nspecifically covered by the AECA,” but only that he had\\n“general knowledge of an export’s illegality”); United States v.\\nChi Mak, 683 F.3d 1126, 1138 (9th Cir. 2012) (affirming\\ninstruction that conviction did not require proof that “the\\ndefendant had read, was aware of, or had consulted the specific\\nregulations governing his activities”); United States v. Roth,\\n\\n 25\\n628 F.3d 827, 835 (6th Cir. 2011) (the AECA “does not require\\na defendant to know that the items being exported are on the\\nMunitions List,” but only “knowledge that the underlying\\naction is unlawful”); United States v. Tsai, 954 F.2d 155, 160-\\n62 & n.3 (3d Cir. 1992) (jury need not find that defendant\\n“knew all of the specifics of the law,” but “what it had to find\\nwas that that the defendant knew that he could not export that\\nparticular item”); United States v. Murphy, 852 F.2d 1, 6-7 (1st\\nCir. 1988) (requiring proof that “defendant must know that his\\nconduct in exporting from the United States articles proscribed\\nby the statute is violative of the law,” but rejecting requirement\\nthat he must know specifically that the article was on the\\nMunitions List or that there was a licensing requirement).\\n\\n The Eleventh Circuit in United States v. Wenxia Man\\ncharacterized the willfulness requirement of the AECA as more\\ndemanding than that of the statute in Bryan, but its bottom line\\nis not materially different from that of the other circuits. See\\n891 F.3d 1253, 1268 (11th Cir. 2018). That court hewed to its\\npre-Bryan precedent that “aware[ness] of the generally\\nunlawful nature of [one’s] actions is insufficient,” but was\\nsatisfied with proof “that a conspirator actually knew that it was\\nunlawful to export the [items] and intentionally violated [the]\\nknown legal duty not to export [them].” Id. (internal quotation\\nmarks omitted) (alterations in original) (quoting United States\\nv. Adames, 878 F.2d 1374, 1377 (11th Cir. 1989)). The Fifth\\nCircuit held in United States v. Hernandez that the AECA\\ndemands a finding that the defendant “knew he was unlawfully\\nexporting weapons on the Munitions List,” but that case\\npredated Bryan, Cheek, and Ratzlaf. 662 F.2d 289, 292 (5th\\nCir. 1981). The Seventh Circuit did so more recently, but it,\\ntoo, took no account of those Supreme Court cases. See United\\nStates v. Dobek, 789 F.3d 698, 700 (7th Cir. 2015).\\n\\n 26\\n We hold that the district court’s jury instruction was\\ncorrect insofar as it rejected defendants’ position that the\\nwillfulness requirement in the AECA is akin to those at issue\\nin Cheek and Ratzlaf, and clarified that the government need\\nnot prove defendants knew the specific law their conduct\\nviolated.\\n\\n That does not quite resolve the issue, however, because of\\nambiguity in the meaning of “conduct” in the instruction that a\\nwillful violation of the AECA requires that the defendant\\n“acted with knowledge that his conduct was unlawful.” App.\\n66. More is required where, as here, there is evidence of\\nwillfully unlawful conduct apart from the charged offenses,\\ncreating a risk that the jury may consider any and all evidence\\nof the defendant’s guilty mind—whatever its object—as\\nsupporting willful commission of the charged offense. The\\nwillfulness instruction arguably fell short in not specifying that\\nthe unlawful “conduct” the jury must find the defendants to\\nhave willfully done was the actus reus that violated the AECA:\\nunlawfully exporting the magazines and mount without a\\nlicense. To be sure, the rest of the instruction made clear that\\nthe actus reus of the charged offense was exporting the items\\nwithout a license, and we review jury instructions as a whole.\\nUnited States v. McGill, 815 F.3d 846, 888 (D.C. Cir. 2016).\\nBut we nonetheless detect a potential problem specific to the\\ninstruction’s discussion of willfulness—an issue that is both of\\ncentral importance and easily muddied. It is natural enough\\nthat jurors who are told they may convict upon finding the\\ndefendants did the requisite act, so long as the defendant had a\\nguilty mind, may not parse the object of the guilty mind. To\\nprompt that important precision, the instruction’s definition of\\n“willfully”—the word at issue here—should specify that the\\nactus reus was the only “conduct” to which defendants’ guilty\\nmental state could apply for a finding of willfulness.\\n\\n 27\\n The requirement that the mens rea relate to the charged\\nactus reus is the baseline for any criminal mental standard. See\\nDixon v. United States, 548 U.S. 1, 6 (2006) (“[C]riminal\\nliability is normally based upon the concurrence of two factors,\\nan evil-meaning mind [and] an evil-doing hand.” (second\\nalteration in original) (internal quotation marks omitted)). If a\\nstatute requires that the defendant knowingly perform the act\\nthat violates the law—even where he need not also know that\\nthe act is illegal—then he must knowingly perform the charged\\nactus reus, not some other, uncharged act. See Bryan, 524 U.S.\\nat 193 (“[T]he term ‘knowingly’ merely requires proof of\\nknowledge of the facts that constitute the offense.” (emphasis\\nadded)). So, too, with willfulness. The district court and the\\nparties agree that willfulness requires that the defendant “acted\\nwith knowledge that his conduct was unlawful.” App. 66; see\\nBryan, 524 U.S. at 193. The conduct that he must know was\\nunlawful is the actus reus of the crime with which he is\\ncharged.\\n\\n Because the willfulness instruction required only that the\\ndefendants acted with knowledge that “the conduct” was\\nunlawful, App. 66, there is some chance that the jury convicted\\nbased in part on defendants’ evasiveness in importing to\\nThailand. On retrial, the instruction should make clear that an\\nAECA conviction requires that defendants knew of the\\nunlawfulness of the charged unlicensed export of the items\\nfrom the United States, and that a willfulness finding cannot\\ndraw on evidence that they knew the related, but legally and\\nfactually distinct, import of those items into Thailand was\\nillegal.\\n\\n 28\\n C. Burden’s Limited English Did Not Render His\\n Interrogation Custodial for Miranda Purposes.\\n\\n We review de novo the determination whether Burden was\\nin custody and thus entitled to Miranda warnings before any\\ninterrogation, and we review the underlying factual findings for\\nclear error. United States v. Hallford, 756 F. App’x 1, 5 (D.C.\\nCir. 2018). If the court erroneously admits a non-Mirandized\\nstatement, we must reverse unless the admission was harmless\\nbeyond a reasonable doubt. United States v. Brinson-Scott, 714\\nF.3d 616, 622 (D.C. Cir. 2013).\\n\\n In Miranda, the Court held that “the prosecution may not\\nuse statements, whether exculpatory or inculpatory, stemming\\nfrom custodial interrogation of the defendant unless it\\ndemonstrates the use of procedural safeguards effective to\\nsecure the privilege against self-incrimination.” 384 U.S. at\\n444. Defendants argue that Burden’s limited English abilities\\nrendered the interview “custodial.” We disagree. We see no\\nerror in the district court’s admission of Burden’s non-\\nMirandized statement because Burden was not in custody when\\nthe agents questioned him at his warehouse. His English\\nproficiency was not an “objective circumstance surrounding\\nthe interrogation” that a reasonable officer would have\\nassumed bore on whether Burden felt free to leave the\\ninterview.\\n\\n Miranda warnings are required as a bulwark against the\\ncoercive power of being taken into police custody and\\ninterrogated. “An individual swept from familiar surroundings\\ninto police custody, surrounded by antagonistic forces, and\\nsubjected to . . . techniques of persuasion . . . cannot be\\notherwise than under compulsion to speak.” Miranda, 384\\nU.S. at 461. Police custody is sometimes self-evident, as in\\nMiranda’s own case: The police arrested him and took him to\\n\\n 29\\na police station interrogation room where they questioned him.\\nId. at 491. But a person is also in custody when he is\\n“otherwise deprived of his freedom of action in any significant\\nway.” Id. at 444. In “Miranda case law, ‘custody’ is a term of\\nart that specifies circumstances that are thought generally to\\npresent a serious danger of coercion.” Howes v. Fields, 565\\nU.S. 499, 508 (2012).\\n\\n The Supreme Court has laid out guidelines for the custody\\nanalysis. “In determining whether a person is in custody,”\\ntriggering the duty to give Miranda warnings,\\n\\n the initial step is to ascertain whether, in light of\\n the objective circumstances of the interrogation, a\\n reasonable person [would] have felt he or she was\\n not at liberty to terminate the interrogation and\\n leave. And in order to determine how a suspect\\n would have gauge[d] his freedom of movement,\\n courts must examine all of the circumstances\\n surrounding the interrogation. Relevant factors\\n include the location of the questioning, its duration,\\n statements made during the interview, the presence\\n or absence of physical restraints during the\\n questioning, and the release of the interviewee at\\n the end of the questioning.\\n\\nId. at 509 (alterations in original) (internal quotation marks and\\ncitations omitted).\\n\\n “[W]hether a suspect is ‘in custody’ is an objective\\ninquiry.” J.D.B. v. North Carolina, 564 U.S. 261, 270 (2011).\\nThat means that, while the analysis accounts for “any\\ncircumstance that would have affected how a reasonable person\\nin the suspect’s position would perceive his or her freedom to\\nleave,” it “involves no consideration of the actual mindset of\\nthe particular suspect subjected to police questioning.” Id. at\\n\\n 30\\n271 (internal quotation marks omitted). The benefit of the\\nobjective test is practical: It “avoids burdening police with the\\ntask of anticipating the idiosyncrasies of every individual\\nsuspect and divining how those particular traits affect each\\nperson’s subjective state of mind.” Id.\\n\\n The test, while objective, is also contextual. Because all\\nof the “objective circumstances of the interrogation” must be\\nconsidered, Howes, 565 U.S. at 508, individual characteristics\\nthat have an “objectively discernible relationship to a\\nreasonable person’s understanding of his freedom of action,”\\nsuch as a child suspect’s age, J.D.B., 564 U.S. at 275, must be\\ntaken into account. Under circumstances in which “a\\nreasonable child subjected to police questioning will . . . feel\\npressured to submit when a reasonable adult would feel free to\\ngo[,] . . . courts can account for that reality without doing any\\ndamage to the objective nature of the custody analysis.” Id. at\\n272. That is because age “generates commonsense conclusions\\nabout behavior and perception” that “apply broadly to children\\nas a class.” Id. A reasonable eight-year-old will not necessarily\\nfeel free to leave when a reasonable adult would.\\n\\n Because the test is designed to guide police, a person’s\\nyouth—or analogous circumstances bearing on a reasonable\\nperson’s perception of her freedom to leave—only factors into\\nthe custody analysis where it “was known to the officer at the\\ntime of the interview, or would have been objectively apparent\\nto any reasonable officer.” Id. at 274. The test thus includes a\\ndouble inquiry: whether it would have been apparent to a\\nreasonable officer that a reasonable person in the suspect’s\\nposition would not have felt free to leave. In holding that the\\nsuspect’s youth was an objective circumstance relevant to the\\ncustody analysis, the Supreme Court suggested that other\\npersonal characteristics, like blindness, could be similarly\\nrelevant. Id. at 278. The Court underscored: “Not once have\\n\\n 31\\nwe excluded from the custody analysis a circumstance that we\\ndetermined was relevant and objective, simply to make the\\nfault line between custodial and noncustodial ‘brighter.’” Id.\\nat 280.\\n\\n Under some circumstances, English language capabilities\\nmight have an “objectively discernible relationship to a\\nreasonable person’s understanding of his freedom of action”\\nthat would bear on the custody analysis for purposes of\\nMiranda. Id. at 275. Some courts have accordingly factored\\nlimited English abilities into the custody inquiry. The Eighth\\nCircuit explained that “the ultimate issue is whether a\\nreasonable police officer conducting [an] otherwise non-\\ncustodial interview would have given Miranda warnings\\nbecause he realized that the questioning would be perceived by\\n[the defendant] as custodial due to his limited English language\\nskills.” Thatsaphone v. Weber, 137 F.3d 1041, 1045 (8th Cir.\\n1998). In United States v. Kim, the Ninth Circuit weighed\\nlimited English proficiency (among other factors the court\\ndeemed important) because, under the circumstances, it bore\\non the defendant’s ability to understand whether she was a\\ncriminal suspect. 292 F.3d 969, 977 (9th Cir. 2002). Kim had\\narrived at the store her family owned to find many police cars\\noutside and the door locked; the police let her, but not her\\nhusband, enter, locked the door, did not let her son (who was\\nalready inside) speak to her, and forced her to speak only\\nEnglish, though she and her son informed the officers that she\\ndid not speak English well. Id. at 971-72. The court held that\\nKim’s interrogation was custodial partly because the police\\n“temporarily took over complete control of Kim’s store,\\ncreating ‘a police-dominated atmosphere,’ in which the police\\nkept Kim physically isolated from two family members who\\ncould have provided both moral support and, given her limited\\nEnglish, a more complete understanding of the overall\\nsituation.” Id. at 977. Both Thatsaphone and Kim preceded the\\n\\n 32\\nCourt’s application of the custody analysis to a child in J.D.B.,\\nbut both decisions comport with J.D.B. insofar as they consider\\nwhether a reasonable officer would have been able to discern\\nthat the language limitations of a reasonable person in the\\nsuspect’s position would have contributed to that person’s not\\nfeeling free to leave. Since J.D.B., at least one other court has\\nheld a suspect’s limited English proficiency relevant to\\nJ.D.B.’s objective inquiry. See United States v. Han, 199 F.\\nSupp. 3d 38, 52-54 (D.D.C. 2016).\\n\\n The district court in this case correctly held that a\\nreasonable officer would not have thought that Burden’s\\nlanguage abilities prevented him from feeling free to leave, and\\nthus properly admitted Burden’s non-Mirandized statement.\\nEven though Burden sometimes had trouble formulating\\nresponses and appeared to lack perfect comprehension of all\\nthe questions, the evidence does not suggest that it would have\\nbeen apparent to a reasonable officer that Burden was not\\nunderstanding what was being said.\\n\\n The defendants identify only one moment, when the DHS\\nofficer was explaining the purpose and terms of the interview,\\nwhen they believe that a reasonable officer should have\\nrecognized that Burden’s English skills would affect the\\nperception of a reasonable person in his position as to whether\\nhe was free to leave:\\n\\n UNIDENTIFIED AGENT NO. 1: We are federal\\n agents for the U.S. Government so I have to let you\\n know that you have to be honest with us —\\n\\n PHEERAYUTH BURDEN: Uh-huh.\\n\\n UNIDENTIFIED AGENT NO. 1: — okay? If you\\n don’t want to answer something, you don’t have to\\n answer but you cannot lie to us.\\n\\n 33\\n PHEERAYUTH BURDEN: Okay.\\n\\n UNIDENTIFIED AGENT NO. 1: All right? And\\n you can’t withhold relevant information.\\n\\n PHEERAYUTH BURDEN: Uh-huh.\\n\\n UNIDENTIFIED AGENT NO. 1: If you do, that is\\n a crime.\\n\\n PHEERAYUTH BURDEN: Okay\\n\\n UNIDENTIFIED AGENT NO. 1: Okay?\\n Punishable by up to five years in prison so just\\n please be honest.\\n\\n PHEERAYUTH BURDEN: Yeah.\\n\\nApp. 174. While the agent’s statement was somewhat\\nconfusing, it is not clear from this exchange that Burden was\\nnot comprehending what the agent was saying to him or\\nsomehow believed he could not leave an interview at his own\\nwarehouse that he had agreed to by phone and shown up for of\\nhis own accord.\\n\\n The district court did not refuse to consider Burden’s\\nproficiency entirely, as defendants assert. It properly applied\\nthe custody test by evaluating how a reasonable officer would\\nhave perceived Burden’s comprehension. It noted “that for\\neach of the points that the Special Agent was communicating\\nto Mr. Burden, Mr. Burden said okay or yes or [uh-huh],\\nreflect[ing] responses like that which would give one\\nreasonably the understanding that Mr. Burden understood what\\nwas being said.” App. 430-31.\\n\\n 34\\n The custody question was not otherwise close. The district\\ncourt found that Burden “arrived with his wife at his own work\\nplace, at the time that he had set and there was no evident effort\\nto overcome his will. There was no effort to put him in\\nhandcuffs, no threats, other threats during the course of the\\ninterview.” App. 431. Burden himself “chose where he was\\ngoing to sit, he chose the room in which they were going to\\ntalk,” and he “showed that he knew he could get up, go out,\\nopen the door and talk to somebody outside throughout the\\ncourse of this.” App. 432. One of the two agents who\\nquestioned Burden testified that Burden had chosen where and\\nwhen the interview would take place, S.A. 400, and that the\\nagents wore plainclothes and did not display their badges or\\nweapons, S.A. 407. The interview lasted no more than three\\nhours and Burden left when it was over. In sum, “the location\\nof the questioning, its duration, statements made during the\\ninterview, the presence or absence of physical restraints during\\nthe questioning, . . . the release of the interviewee at the end of\\nthe questioning,” and other factors all support the conclusion\\nthat Burden was not in custody. Howes, 565 U.S. at 509\\n(internal quotation marks and citations omitted). Because\\nBurden’s imperfect English would not have given a reasonable\\nofficer the impression that a reasonable person in Burden’s\\nposition would have believed himself detained, it does not\\nchange that determination here.\\n\\n * * *\\n\\n For the foregoing reasons, we affirm the district court’s\\nadmission of Burden’s non-Mirandized statement to DHS\\nagents. We nonetheless vacate the convictions in view of the\\nerror in the admission of Yindeear-Rom’s deposition\\ntestimony.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4426936/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge PILLARD\"}, {\"author\": \"ROGERS, Circuit Judge, concurring\", \"type\": \"concurrence\", \"text\": \"ROGERS, Circuit Judge, concurring in part, and concurring\\nin the judgments: I join the court in reversing the judgments of\\nconviction and remanding the case for a new trial because the\\ndistrict court erred in admitting at trial the deposition of a key\\ngovernment witness taken pursuant to Federal Rule of Criminal\\nProcedure 15. See Op. 12–21. As the proponent of the prior\\ntestimony, the government had the burden of establishing\\nYindeear-Rom’s unavailability by making reasonable efforts to\\nprocure his presence at trial, see United States v. Lynch, 499\\nF.2d 1011, 1022 (D.C. Cir. 1974); FED. R. EVID. 804(a)(5)(A),\\nand failed to offer evidence that it did so. See Op. 15–17. The\\ndistrict court’s error violated defendants’ Confrontation rights\\nunder the Sixth Amendment to the Constitution. See Ohio v.\\nRoberts, 448 U.S. 56, 74 (1980), abrogated on other grounds\\nby Crawford v. Washington, 541 U.S. 36 (2004). The\\ngovernment concedes that if there was error, it was not\\nharmless beyond a reasonable doubt. Appellee’s Br. 55 n.5.\\n\\n In view of our remand for a new trial, see United States v.\\nHite, 769 F.3d 1154, 1167 (D.C. Cir. 2014), I also join two of\\nthe court’s other conclusions. First, I agree the district court\\ndid not err in denying defendant Pheerayuth Burden’s motion\\nto suppress statements that he had made during an interview\\nwith Department of Homeland Security agents at Burden’s\\nsuggestion at his warehouse, while his wife was present. Op.\\n28–34. The district court could reasonably conclude that,\\nunder the circumstances, it would not have been “objectively\\napparent to any reasonable officer,” J.D.B. v. North Carolina,\\n564 U.S. 261, 274 (2011), that Burden’s lack of fluency in the\\nEnglish language would have prevented him from\\nunderstanding that he was not detained at the time.\\n\\n Second, I agree the district court did not err in rejecting the\\ndefendants’ requested jury instruction on the meaning of\\n“willfully” under the Arms Export Control Act (“AECA”). See\\nOp. 21–26. This court now joins the majority of circuit courts\\nof appeal to have considered the issue that the AECA does not\\n\\n 2\\nrequire the government to prove that defendants know the\\nspecific items they exported were on the Munitions List.\\n\\n But I cannot agree that the instruction on “Willful\\nViolation of Arms Export Control Act” (“AECA instruction”),\\nApp. 66, failed adequately to define “the conduct” that the jury\\nhad to find was “willfully” committed by the defendants. Op.\\n26–27. When reviewing an instruction for legal error, the\\n“pertinent question is whether, taken as a whole, the\\ninstructions accurately state the governing law and provide the\\njury with sufficient understanding of those issues and\\napplicable standards.” United States v. Vega, 826 F.3d 514,\\n524 (D.C. Cir. 2016) (quoting United States v. Wilson, 605 F.3d\\n985, 1018 (D.C. Cir. 2010)). Courts must “consider not just\\nthe challenged phrases, but the instruction as a whole.” United\\nStates v. Washington, 106 F.3d 983, 997 (D.C. Cir. 1997)\\n(quoting United States v. Merlos, 984 F.2d 1239, 1242 (D.C.\\nCir. 1993)).\\n\\n The district court instructed the jury that “[t]hrough the\\nArms Export Control Act and the International Traffic in Arms\\nRegulations, Congress and the President have made it a\\ncriminal offense for anyone willfully to export or attempt to\\nexport from the United States any defense article without first\\nobtaining a license or written approval from the U.S.\\nDepartment of State.” App. 66 (emphasis added). In the next\\nsentence, the court instructed that “[t]he defendants are charged\\n. . . with specific instances of willfully exporting United States\\nMunitions List items . . . in violation of the export control\\nlaws.” Id. (emphasis added). Adding specific details about the\\nalleged unlawful conduct, the district court instructed the jury\\nthat Count 2 of the superseding indictment charged the\\ndefendants with exporting, in violation of the AECA, two types\\nof items on the U.S. Munitions List: “(1) five AR Style, NATO\\n5.56, 30 round magazines, and (2) a KAC-Knight Armament\\n\\n 3\\nM203 Qd Mount.” Id. Regarding the word “willfully,” the\\ndistrict court instructed the jury that the requisite unlawful\\nintent concerns U.S. export law: “[T]he government need not\\nprove that a defendant had read, was aware of, or had consulted\\nthe licensing provisions of the Arms Export Control Act or the\\nInternational Traffic in Arms Regulations, or the Munitions\\nList . . . [but it] must prove . . . that a defendant knew that the\\nconduct was unlawful.” Id. The defendants did not otherwise\\nrequest any instruction on the conduct at issue.\\n\\n Consistent with the presumption that juries follow\\ninstructions, see Richardson v. Marsh, 481 U.S. 200, 211\\n(1987), “a conscientious and attentive juror viewing the\\ninstructions as a whole” would not have convicted defendants\\non the basis of finding a willful mental state with respect to any\\nconduct other than exporting in violation of U.S. law. United\\nStates v. Lemire, 720 F.2d 1327, 1341 (D.C. Cir. 1983). The\\nAECA instruction made clear that the alleged criminal act was\\nexporting from the United States. Because Thai customs law\\ngoverns importing into Thailand, see Appellants’ Br. 30–31, if\\nthe jury understood Burden’s only unlawful intent was to evade\\nThai customs, then it would not have found Burden understood\\nexporting from the United States to be unlawful. Viewed as a\\nwhole, the AECA instruction “fairly present[s] the applicable\\nlegal principles and standards,” Joy v. Bell Helicopter Textron,\\nInc., 999 F.2d 549, 556 (D.C. Cir. 1993) (internal quotation\\nmarks omitted), and passes the “critical inquiry [of] whether\\nthe instructions, viewed in the aggregate, properly guided the\\njurors in their deliberations,” 9C CHARLES ALAN WRIGHT &\\nARTHUR R. MILLER, FEDERAL PRACTICE AND PROCEDURE\\n§ 2558 (3d ed.).\\n\\n The court nevertheless perceives an instructional\\nambiguity about the relevant unlawful conduct. See Op. 26.\\nSpeculating about juror conduct, the court concludes that\\n\\n 4\\n“there is some chance that the jury convicted based in part on\\ndefendants’ evasiveness in importing to Thailand.” Op. 27.\\nInstead of analyzing the whole of the district court’s AECA\\ninstruction, the court focuses on the instruction about\\nwillfulness, and imposes a sua sponte obligation on the district\\ncourt upon retrial to instruct the jury that it must find that the\\n“defendants knew of the unlawfulness of the charged\\nunlicensed export of the items from the United States, and that\\na willfulness finding cannot draw on evidence that they knew\\nthe related, but legally and factually distinct, import of those\\nitems into Thailand was illegal.” Op. 27.\\n\\n Of course, instructional clarity is desirable. But “the\\ndefense had ample opportunity to make clarifying suggestions”\\nas this court now requires, and it did not, Lemire, 720 F.2d at\\n1343 — likely because defense counsel told the jury in closing\\nargument that evidence of Burden’s interest in evading Thai\\ncustoms law explained his behavior without providing\\nevidence of the requisite guilty mind with respect to the\\nviolations of U.S. law with which he was charged, see Supp.\\nApp. 809. Defense silence may further indicate the instructions\\nmade clear the precise nature of the willful conduct required to\\nfind guilt of the AECA charges. In any event, as noted, there\\nis no basis to conclude there is a “reasonable likelihood” the\\njury misapplied the challenged instruction, Boyde v.\\nCalifornia, 494 U.S. 370, 380, 383 (1990), and “[t]he choice of\\nthe [words] to be used in a particular instruction . . . is reviewed\\nonly for abuse of discretion,” Joy, 999 F.2d at 556, for it is well\\nsettled that “[t]he district judge need not use any particular\\nform of words or sequence of ideas so long as the charge as a\\nwhole conveys to the jury a clear and correct understanding of\\nthe applicable substantive law without confusing or misleading\\nthem.” WRIGHT & MILLER § 2556. “Jurors do not sit in solitary\\nisolation booths parsing instructions for subtle shades of\\nmeaning in the same way that lawyers might”; instead,\\n\\n 5\\n“commonsense understanding of the instructions in the light of\\nall that has taken place at the trial [is] likely to prevail over\\ntechnical hairsplitting.” Boyde, 494 U.S. at 380–81.\\n\\n For these reasons I concur in part and concur in the\\njudgments.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4426936/\", \"author_raw\": \"ROGERS, Circuit Judge, concurring\"}]}"},"judge1":{"kind":"string","value":"HENDERSON"},"judge2":{"kind":"string","value":"PILLARD"},"judge3":{"kind":"string","value":"ROGERS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":1,"string":"1"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4649683/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: UNITED STATES of America, Appellee v. Pheerayuth BURDEN, Appellant\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-20\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: Yes\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued January 16, 2019 Decided August 20, 2019\n\n No. 17-3018\n\n UNITED STATES OF AMERICA,\n APPELLEE\n\n v.\n\n PHEERAYUTH BURDEN,\n APPELLANT\n\n\n Consolidated with 17-3019\n\n\n Appeals from the United States District Court\n for the District of Columbia\n (No. 1:14-cr-00069-2)\n (No. 1:14-cr-00069-3)\n\n\n Lindsay C. Harrison, pro bono, argued the cause and filed\nthe briefs for appellant. James Dawson entered an appearance.\n\n Daniel J. Lenerz, Assistant U.S. Attorney, argued the\ncause for appellee. With him on the brief were Jessie K. Liu,\nU.S. Attorney, and Elizabeth Trosman and Elizabeth H.\nDanello, Assistant U.S. Attorneys.\n\n Before: HENDERSON, ROGERS and PILLARD, Circuit\nJudges.\n\n 2\n Opinion for the Court filed by Circuit Judge PILLARD.\n\n Opinion concurring in part and concurring in the judgment\nfiled by Circuit Judge ROGERS.\n\n PILLARD, Circuit Judge: The Arms Export Control Act\n(AECA) criminalizes exporting defense articles without a\nlicense. 22 U.S.C. § 2778(b)(2), (c). Pheerayuth Burden, a\nU.S.-resident Thai national who ran a business exporting goods\nfrom the United States to Thailand, and his export business,\nWing-On LLC (collectively, the defendants), exported five\nassault-rifle magazines and a grenade-launcher mount.\nFollowing a three-week trial, a jury convicted the defendants\nof conspiracy to violate the AECA, unlawful export in violation\nof the AECA, and conspiracy to launder money.\n\n The defendants contend that three of the district court’s\nrulings are reversible error. First, they argue that the court\nerred in admitting video deposition testimony by a key witness\nover a Confrontation Clause objection where the government\nitself rendered the witness “unavailable” at trial by deporting\nhim shortly before trial without first making reasonable efforts\nto arrange his return. Second, they challenge a jury instruction\ndefining the “willfulness” element of unlawful exportation of\ndefense articles as requiring only proof that the defendants\n“acted with knowledge that the conduct was unlawful.” That\ninstruction was inadequate, they contend, because it failed to\ntie the willfulness finding to the pertinent conduct and law,\ncreating an impermissible risk that the jury relied on evidence\nthat Burden thought he was violating Thai import law. Third,\ndefendants claim that the district court erred in admitting\nBurden’s non-Mirandized statements because it failed to\naccount for his limited English abilities in determining that he\nwas not in custody when agents interrogated him.\n\n 3\n We hold that the district court erred in admitting the\ndeposition testimony because the government failed to make\nreasonable efforts before it deported the witness to procure his\npresence at trial. We conclude that the jury instruction was\ncorrect as far as it went in instructing the jury to find that “the\ndefendant knew that his conduct was unlawful,” and that\n“willfully” violating the law does not require proof “that a\ndefendant had read, was aware of, or had consulted the\nlicensing provisions of the Arms Export Control Act” as such.\nAppellants’ Appendix (App.) 66. But we suggest clarification\nof the willfulness instruction to more squarely require a finding\nthat defendants were aware of and knowingly violated their\nlegal obligation not to commit the charged actus reus. A case\nsuch as this one—that includes evidence of consciousness of\nguilt relating to distinct actus reus arguably violating different,\nuncharged legal obligations—creates some risk of the jury\nrelying on evidence of consciousness of guilt unrelated to the\ncharged crime. We affirm the district court’s determination\nthat Burden was not in custody because, even assuming\nlanguage proficiency is relevant to the custody inquiry, a\nreasonable officer would not have thought Burden’s imperfect\nEnglish meant a reasonable person in his position would have\nbelieved himself detained during the interview.\n\n Because the error we identify was not harmless, we vacate\nthe judgments and remand for proceedings consistent with this\nopinion.\n\n BACKGROUND\n\n A. Legal Background\n\n The AECA establishes executive-branch control over the\nexport and import of “defense articles,” meaning arms or other\nmilitary items. See 22 U.S.C. § 2278. It authorizes the\n\n 4\nPresident, “[i]n furtherance of world peace and the security and\nforeign policy of the United States,” to control the export of\ndefense articles and services, designate which items count as\ndefense articles and services, and promulgate regulations for\nthose purposes. Id. § 2778(a)(1). The designated defense\narticles make up the United States Munitions List (the\nMunitions List or the List). Id. With certain enumerated\nexceptions, “no defense articles or defense services designated\nby the President” as part of the Munitions List “may be\nexported or imported without a license for such export or\nimport, issued in accordance with” the AECA and its\nassociated regulations. Id. § 2778(b)(2). The State Department\nis responsible for issuing licenses. See id.; 22 C.F.R. §§ 120.1,\n120.20. The decision whether to issue an export license\nimplicates sensitive issues of national security and foreign\npolicy. It must “take into account whether the export of an\narticle would contribute to an arms race, aid in the development\nof weapons of mass destruction, support international\nterrorism, increase the possibility of outbreak or escalation of\nconflict, or prejudice the development of . . . arms control or\nnonproliferation agreements or other arrangements.” 22\nU.S.C. § 2778(a)(2). The statute criminalizes “willfully\nviolat[ing] any provision of this section . . . or any rule or\nregulation issued under this section.” Id. § 2278(c). It thus\ncriminalizes willfully exporting defense articles without a\nlicense.\n\n The President delegated to the Secretary of State the\nauthority to designate defense articles and promulgate\nregulations under the AECA, see Exec. Order No. 13637, 78\nFed. Reg. 16,129 (2013); 22 C.F.R. § 120.1(a), and the\nSecretary accordingly promulgated the International Traffic in\nArms Regulations (ITAR), 22 C.F.R. §§ 120.1-130.17. The\nITAR prohibits exporting defense articles and services without\n“obtaining the required license or other written approval” from\n\n 5\nthe appropriate office of the State Department. 22 C.F.R.\n§ 127.1(a). The ITAR also includes the Munitions List, which\nruns to over forty pages in the Code of Federal Regulations.\nSee id. § 121.1. The covered defense articles are described with\nvarying levels of specificity, such as “[r]iflescopes\nmanufactured to military specifications,” id. (Category I(f)),\n“[g]uns over caliber .50,” id. (Category II(a)), “[i]ron powder\n. . . with particle size of 3 micrometers or less produced by\nreduction of iron oxide with hydrogen,” id. (Category\nV(c)(4)(i)(B)), and “[h]elmets . . . providing a protection level\nequal to or greater than NIJ Type IV,” id. (Category X(a)(6)).\nThe convictions in this case relate to items in Category I(h) of\nthe Munitions List: “Components, parts, accessories and\nattachments” for the firearms listed in Category I(a)-(g). See\nApp. 89-90.\n\n B. Factual and Procedural Background\n\n Burden started Wing-On LLC (Wing-On), a freight-\nforwarding business that shipped American goods to Thailand,\naround 2008. In 2010, Kitibordee Yindeear-Rom became one\nof Burden’s customers. A Thai national living in Thailand,\nYindeear-Rom had a business importing many different types\nof goods from the United States to Thailand. As part of that\nbusiness, he helped his customers get gun parts and accessories\nfrom the U.S. that they could not purchase directly because\nU.S. companies would neither accept Thai credit cards nor ship\nthe parts to Thailand. According to Yindeear-Rom, Burden\ninitially ordered gun parts for him from U.S. vendors, received\nthem in the United States, then shipped them to Thailand.\nSupplemental Appendix (S.A.) 291A-91B. Yindeear-Rom\nlater began placing the orders himself using a debit card\nattached to a U.S. bank account Burden opened. S.A. 294-96,\n479. Yindeear-Rom testified in his deposition that he\nreimbursed Burden for the purchases he made on Burden’s\n\n 6\ndebit card by transferring money to Thai bank accounts\nbelonging to Burden and Burden’s associate. S.A. 298-300.\nNeither Burden nor Wing-On had a license to export defense\narticles on the Munitions List.\n\n In October 2013, Yindeear-Rom took a vacation to the\nUnited States, where he was stopped and interviewed by\nDepartment of Homeland Security (DHS) agents. He was\narrested two days later for conspiracy to violate American\nexport laws. He later pleaded guilty and was sentenced to\nthirty-six months in prison. At Yindeear-Rom’s initial court\nappearance, the DHS agents saw in the courtroom two people\nthey believed to be Burden’s wife and roommate, respectively.\nConcerned that Burden might have been alerted to the\ninvestigation, the agents went immediately to Wing-On’s\nwarehouse. Burden was not there, but the agents met one of\nhis employees, who helped the agents call him. They called\nhim again later that day and arranged an interview for that\nevening at the warehouse.\n\n The DHS agents interviewed Burden in English without an\ninterpreter. They did not advise Burden of his rights with the\nfamiliar warnings officials must give suspects in custodial\ninterrogation under Miranda v. Arizona, 384 U.S. 436, 444\n(1966). At the beginning of the interview, one of the agents\ngave the following preamble (recounted here with Burden’s\naffirmative interjections omitted):\n\n We are federal agents for the U.S. Government so\n I have to let you know that you have to be honest\n with us[,] okay? If you don’t want to answer\n something, you don’t have to answer but you\n cannot lie to us. All right? And you can’t withhold\n relevant information. If you do, that is a crime.\n\n 7\n Okay? Punishable by up to five years in prison so\n just please be honest.\n\nApp. 174. The agent then asked, “Is your English good?” Id.\nBurden replied, “A little bit.” App. 175. “If there’s anything\nthat I say that you don’t understand, ask me,” the agent said.\nId. “Okay,” Burden replied. Id. During the interview, Burden\nadmitted he had shipped gun parts to Yindeear-Rom and\nfalsified customs declarations. App. 266-71. The agents did\nnot arrest Burden at the end of the interview. S.A. 420-21. The\ncourt’s eventual admission of Burden’s statement at his trial\nover defense objections is the subject of his Miranda claim on\nappeal. App. 429-33.\n\n Burden was arrested six months later, in May 2014. Trial\nwas initially scheduled for November 2015, but was continued\ntwice, first to April 2016 and then to September 2016. The\ndistrict court granted the second continuance because many\ndocuments remained to be translated into English. The court\nnoted that the defense “can’t actually do this without translated\ndocuments” and that “it’s not that [the defense has] been less\nthan diligent about it.” App. 381. That continuance introduced\na wrinkle into the trial, however: Yindeear-Rom was\nscheduled to be released from prison in June 2016, three\nmonths before the new trial date, and was to be deported after\nhis release. The government had a clear path to remove\nYindeear-Rom upon his release because he had stipulated when\nhe pleaded guilty to an order of removal that would “render[]\nhim permanently inadmissible to the United States,” which\nassured that the government, “promptly upon his release from\nconfinement . . . may execute the order of removal according\nto the applicable laws and regulations.” App. 47 (alteration in\noriginal). The government in February 2016 moved to take\nYindeear-Rom’s deposition under Federal Rule of Criminal\nProcedure 15, which governs depositions taken to preserve a\n\n 8\npotential witness’s testimony for trial. S.A. 1; Fed. R. Crim. P.\n15(a)(1). The district court granted the motion over the\ndefendants’ objections. The defense objected to the deposition\nfor some of the same reasons it asked for the second trial\npostponement: the government had not produced sufficient\ndiscovery to allow them to prepare for Yindeear-Rom’s\ntestimony. The court nonetheless allowed the deposition to be\ntaken to preserve evidence in the event that Yindeear-Rom\nwould be unavailable to testify. Yindeear-Rom’s videotaped,\nin-court deposition took place over four days in March and\nApril. The court granted the government’s motion to reduce\nYindeear-Rom’s sentence in exchange for his testimony.\n\n The United States deported Yindeear-Rom to Thailand in\nApril 2016. Even though the government had substantial\nbargaining leverage before it moved for his sentence reduction,\nthere is no record that it made any efforts before deporting him\nto secure Yindeear-Rom’s presence at trial. It was only once\nYindeear-Rom was back in Thailand that the government\nbegan to make such efforts.\n\n In seeking to bring Yindeear-Rom back a few months\nlater, the government contacted Yindeear-Rom’s counsel by\nphone and mail. United States v. Burden, No. 14-cr-0069\n(RMC), 2016 WL 5108010, at *3 (D.D.C. Sept. 20, 2016). Its\nletter to counsel included a subpoena for Yindeear-Rom’s\ntestimony at trial and a promise to help him obtain a visa and\nto pay his travel expenses, “including but not limited to round-\ntrip airfare, transportation, room, board, and per diem witness\nfee.” Id. (quoting letter, S.A. 51). Yindeear-Rom’s lawyer\n“forwarded the letter and subpoena to Mr. Yindeear-Rom in\nThailand, but was unable to confirm receipt or make any\nrepresentations about [his] willingness to testify.” Id. at *3.\nThe government also sent the letter and subpoena directly to\nYindeear-Rom’s last known email and physical address in\n\n 9\nThailand. Id. at *4. The government received no response to\neither, but confirmed that the letter was signed for by “K. Yen,”\nwhich it believed to be Yindeear-Rom. Id. DHS personnel in\nThailand eventually reached Yindeear-Rom by phone and\nlearned that he had received the email and letter, but that he\n“had no desire to travel to the United States to cooperate in any\nway.” Id. (quoting DHS Report of Investigation, S.A. 75).\n\n At trial, the court granted the government’s motion to\nadmit the Rule 15 deposition over the defendants’ objections.\nDefendants argued that the government should have sought to\nkeep Yindeear-Rom in the country between his release from\nprison and the trial. The court concluded that the witness was\nunavailable, and that “the use of a videotaped deposition taken\nin court, before the trial judge and including the presence of\nMr. Burden and cross examination by both defense lawyers,”\nwas “a very good substitute” for Yindeear-Rom’s live trial\ntestimony “that would allow the jury to observe his demeanor\nand preserve the Defendants’ rights to confront witnesses\nagainst them.” Id. at *3, *8.\n\n The government’s trial evidence included Burden’s\nstatement to DHS agents at the Wing-On warehouse that he\nmislabeled customs declarations for shipments containing gun\nparts, and packed gun parts hidden among other items for\nshipping. App. 269-72. In his defense, Burden highlighted his\nstatements that he took those steps to evade Thai customs. He\ntold the agents that he concealed gun parts among other items\nbecause “[y]ou want to hide from the custom in Thailand,”\nApp. 271, and that he falsified customs documents “[b]ecause\nof tax in Thailand,” App. 281.\n\n The government, for its part, pointed to circumstantial\nevidence tending to show that Burden had reason to know that\nhe was violating U.S. arms-export law. For instance, Burden\n\n 10\nacknowledged in his statement to DHS agents that people\n“[n]eed a license” to ship certain things, like gun parts. App.\n203-04. Yindeear-Rom received a notice that a seized\nshipment of gun parts violated the ITAR, see S.A. 212-13, and\neven though he had ordered that shipment through a different\nshipper (not Wing-On), Yindeear-Rom forwarded the notice to\nBurden asking what he should do, see S.A. 282-83. Yindeear-\nRom then testified that the notice he forwarded informed both\nof them of the requirements of the ITAR. S.A. 369. Burden\nhad also received a notice directly from U.S. Customs and\nBorder Protection that it had seized a rifle scope (controlled\nunder a distinct set of regulations analogous to the ITAR but\ncovering different items) because the scope could not be\nexported without a license. See App. 171.\n\n Further evidence, however, suggested that Burden either\ndid not realize he was shipping real gun parts, or thought it was\nlegal under U.S. export law to ship those parts if they were to\nbe used with toy guns. As part of their business, the defendants\nshipped BB guns (air guns that shoot small metal balls) and\nAirsoft toys (which are similar to BB guns and shoot plastic\npellets). See App. 311-12, 574. Airsoft and BB guns\nthemselves may lawfully be exported without a State\nDepartment license. The defense’s expert witness on firearms\nand Airsoft identification testified that an Airsoft toy “looks\nlike a gun in every way, shape or form from the outside, same\nlength, weight, contour, field markings, but it won’t kill\nanybody.” S.A. 672-73. They have all the same parts as the\nreal guns they mimic; in fact, real gun parts can be used with\nAirsoft toys. App. 508-10; S.A. 676, 678. In his statement to\nDHS agents, Burden said that his Thai customers were not\nusing gun parts “for the gun,” but “for the BB gun . . . for the\npaintball [gun].” App. 259. A Wing-On manager testified that\nemployees were instructed not to ship parts for real guns but\nthat they could ship parts to “be used for toys for Airsoft items.”\n\n 11\nApp. 554-56, 574. Burden affirmed to one of his customers\nthat if a part was for a BB gun “then there’s no problem”\nshipping it. App. 361. That evidence tended to support\nBurden’s defense that he concealed the contents of shipments\nto evade Thai customs law.\n\n Evidence also showed that Burden tried not to ship real\ngun parts after realizing it was illegal. Burden sent an email to\nYindeear-Rom saying, “I have warned you many times that I\ndo not accept gun parts. . . . Stop sending them to me\nabsolutely!” App. 358. The Wing-On warehouse had a “no-\ngo” shelf for gun parts, where employees would segregate\nitems that they could not lawfully ship. See App. 350, 554-55.\nThere was ambiguous evidence suggesting that Wing-On may\nhave ultimately shipped some gun parts on the no-go shelf.\n\n The conflicting evidence regarding Burden’s intent\noccasioned a dispute over the jury instruction defining\n“willfully” under the AECA. The defendants proposed using\nthe Fifth Circuit pattern jury instruction, which requires the\njury to find that a defendant exported articles on the Munitions\nList without obtaining a license from the Department of State;\nand “[t]hat the defendant acted ‘willfully,’ that is, that the\ndefendant knew such license . . . was required for the export of\nthese articles and intended to violate the law by exporting them\nwithout such license.” App. 99.\n\n Instead, the district court adopted the government’s\nproposed instruction, which described the requisite state of\nmind as follows:\n\n [A]n act is done willfully if it is committed with the\n knowledge that it was prohibited by law or was\n done in disregard of a known legal obligation. The\n government must prove that a defendant acted with\n knowledge that the conduct was unlawful. While\n\n 12\n the government must show that a defendant knew\n that the conduct was unlawful, it is not necessary\n for the Government to show that a defendant was\n aware of the specific law, rule, or regulation that\n the conduct may have violated.\n\n In other words, the government need not prove that\n a defendant had read, was aware of, or had\n consulted the licensing provisions of the Arms\n Export Control Act or the International Traffic in\n Arms Regulations, or the Munitions List. The\n government, however, must prove beyond a\n reasonable doubt, by reference to facts and\n circumstances surrounding the case, that a\n defendant knew that the conduct was unlawful.\n\nApp. 66; see App. 435-37. The jury convicted the\ndefendants on all three counts—conspiracy to violate the\nAECA, unlawful export in violation of the AECA, and\nconspiracy to launder money. The defendants timely appealed\nthe district court’s admission of Yindeear-Rom’s deposition\nand Burden’s statement to DHS, as well as the district court’s\njury instruction on the definition of “willfully.”\n\n ANALYSIS\n\n A. Yindeear-Rom Was Not “Unavailable” for\n Purposes of the Confrontation Clause.\n\n We review legal conclusions regarding the Confrontation\nClause de novo, and reverse any error unless it was harmless\nbeyond a reasonable doubt. United States v. Moore, 651 F.3d\n30, 69 (D.C. Cir. 2011) (citing constitutional harmless-error\nstandard in Chapman v. California, 386 U.S. 18, 24 (1967)).\nBecause the government concedes that, if it was error,\nadmitting Yindeear-Rom’s video deposition testimony instead\n\n 13\nof producing him at trial was not harmless, we need only\nconsider the claimed error itself. See Appellee’s Br. 55 n.5.\n\n The Confrontation Clause of the Sixth Amendment\nguarantees that, “[i]n all criminal prosecutions, the accused\nshall enjoy the right . . . to be confronted with the witnesses\nagainst him.” U.S. Const. amend. VI. Courts have long\nrecognized the critical importance of a criminal defendant’s\n“opportunity to cross-examine and impeach a witness at trial\nbefore the jury that will decide his innocence or guilt.” United\nStates v. Lynch, 499 F.2d 1011, 1022 (D.C. Cir. 1974). Live\nwitness testimony gives the jury the opportunity to assess the\nwitness’s demeanor and allows counsel to adjust examination\nto other evidence and to the jury’s apparent reactions as the\nwitness testifies. “William Blackstone long ago recognized\nthis virtue of the right to confrontation, stressing that through\nlive testimony, ‘and this [procedure] only, the persons who are\nto decide upon the evidence have an opportunity of observing\nthe quality, age, education, understanding, behavior, and\ninclinations of the witness.’” United States v. Yida, 498 F.3d\n945, 950 (9th Cir. 2007) (alteration in original) (quoting 3\nWilliam Blackstone, Commentaries on the Laws of England\n373-74 (1768)).\n\n A testimonial statement by a person who does not appear\nat trial may be admitted “only where the declarant is\nunavailable, and only where the defendant has had a prior\nopportunity to cross-examine.” Crawford v. Washington, 541\nU.S. 36, 59 (2004); see also Fed. R. Evid. 804(b)(1).\nUnavailability and prior opportunity to cross-examine are\nindependent criteria: Even when a defendant had a prior\nopportunity to cross-examine the witness, if the government\ndoes not establish that the witness is unavailable, the testimony\nmust be excluded. See id. at 57. Here, the parties agree that\ndefendants had an opportunity to cross-examine Yindeear-\n\n 14\nRom during his videotaped deposition. The sole question is\nwhether he was “unavailable” for purposes of the\nConfrontation Clause.\n\n Defendants argue that Yindeear-Rom was not truly\nunavailable because the government procured Yindeear-Rom’s\nunavailability by deporting him, and, in any case, it did not\nmake reasonable efforts before it deported him to ensure his\npresence at trial. For its part, the government contends that\nthere is no per se presumption that the government fails the\nConfrontation Clause’s test of good-faith and reasonable\nefforts when it deports a witness, and that its reasonable, good-\nfaith efforts after it deported Yindeear-Rom sufficed.\n\n As a general matter, a witness is considered unavailable\nonly if the prosecution cannot procure her with good-faith,\nreasonable efforts. See Ohio v. Roberts, 448 U.S. 56, 74\n(1980), abrogated in part on other grounds by Crawford, 541\nU.S. 36. It is the prosecution’s burden to establish that its\nactions meet that test. Id. at 74-75. “The law does not require\nthe doing of a futile act” such as producing a witness who has\ndied, but “if there is a possibility, albeit remote, that affirmative\nmeasures might produce the declarant, the obligation of good\nfaith may demand their effectuation.” Id. at 74. “The lengths\nto which the prosecution must go to produce a witness . . . is a\nquestion of reasonableness.” Id. (alteration in original)\n(quoting California v. Green, 399 U.S. 149, 189 n.22 (1970)\n(Harlan, J., concurring)). In Lynch, we held good-faith but\nnon-exhaustive efforts to find a witness were inadequate to\nrender her unavailable and justify admitting pretrial hearing\ntestimony. 499 F.2d at 1024. We treated the witness as\navailable because she was within the court’s jurisdiction and\nthe government had not “inquired at the local hospitals, area\npolice departments, the morgue, or of [the witness’s]\nemployer.” Id. In contrast, a witness whose “whereabouts\n\n 15\nwere entirely unknown,” who had been sent numerous\nsubpoenas at her parents’ house, was last heard from outside\nthe jurisdiction, and was unreachable by her family even in\ncase of an emergency was held in Ohio v. Roberts to be\nunavailable even though, as Justice Brennan pointed out in\ndissent, the government had not followed every possible lead\nto find her. Compare 448 U.S. at 60-61, 75, with id. at 79-82\n(Brennan, J., dissenting).\n\n When the government seeks to rely on prior recorded\nstatements of a witness on the ground that the witness is\nunavailable, it bears the burden of establishing that its\nunsuccessful efforts to procure the witness’s appearance at trial\nwere “as vigorous as that which the government would\nundertake to [secure] a critical witness if it has no [prior]\ntestimony to rely upon in the event of ‘unavailability.’” Lynch,\n499 F.2d at 1023; see also United States v. Mann, 590 F.2d 361,\n367 (1st Cir. 1978). Where the government itself bears some\nresponsibility for the difficulty of procuring the witness, such\nas by deporting the witness, the government will have to make\ngreater exertions to satisfy the standard of good-faith and\nreasonable efforts than it would have if it had not played any\nrole. Failing to factor the government’s own contribution to\nthe witness’s absence into the Confrontation Clause analysis\nwould warp the government’s incentives. “This relatively high\ngood faith standard cannot be satisfied by perfunctory efforts,\nif the rule is not to sanction the government’s procuring\ndepositions of witnesses, especially shaky witnesses, but then\ndiscourage attempts to bring the witness to trial so long as the\ngovernment is satisfied with” the witness’s recorded testimony.\nMann, 590 F.2d at 367. Rather, the analysis of good-faith,\nreasonable efforts should “include an assessment of the\ngovernment’s affirmative conduct” that allowed the witness to\nbecome unavailable “in the first instance.” Yida, 498 F.3d at\n955-56. That analysis should account for the good faith and\n\n 16\nreasonableness of the government’s deportation of the witness\nand of its attempts to secure witnesses it does deport.\n\n In a case such as this one, in which the government knew\nor should have known of the potential need for the witness’s\ntestimony before he was deported, the government’s duty to\nmake good-faith, reasonable efforts to ensure the witness’s\npresence arises before the witness leaves the United States.\nOther courts that have addressed this question in the context of\nwitness deportation agree. In United States v. Tirado-Tirado,\nthe Fifth Circuit refused to deem unavailable a witness whom\nthe government had deported where the government “failed to\nmake any concrete arrangements with [the witness] prior to his\ndeportation,” and only shortly before trial made what the court\nacknowledged were “fairly exhaustive” efforts to bring him\nback from Mexico. 563 F.3d 117, 123, 125 (5th Cir. 2009).\nThe government had orally told the witness before deporting\nhim that his appearance would be required if the case went to\ntrial—a step not taken here—but the court faulted it for not\ntaking several other pre-deportation steps that might have\nencouraged the witness to appear, including serving him with\na subpoena or other written notice. Id. at 123-24. In United\nStates v. Foster, the court followed Tirado-Tirado to exclude\nvideotaped depositions notwithstanding that the government\nhad advised the witnesses before deporting them that it might\nneed them to return, promised to allow their reentry and pay for\ntheir travel, and collected contact information for them in\nMexico. 753 F. App’x 307, 309 (5th Cir. 2018). Emphasizing\nthe constitutional importance of taking “reasonable measures,\nunder the circumstances, that are likely to ensure that the\nwitness will return for trial,” the court declined to find the\nunreachable witnesses “unavailable” because the government\nhad not verified their contact information from the outset,\nsought alternative contact information, or remained in contact\n\n 17\nwith them over the months between their deportation and the\ntrial. Id. at 312.\n\n The Ninth Circuit in Yida similarly recognized “the\ngovernment’s obligation to use ‘reasonable means’ to ‘procure\nthe declarant’s testimony’ in the context of the government’s\naffirmative role in a witness’s deportation,” 498 F.3d at 953,\nand refused to admit testimony from a prior mistrial “when the\ngovernment itself share[d] some of the responsibility for its\ninability to produce the witness at [the second] trial,” id. at 956.\nThe government’s efforts were insufficient where it had chosen\nto detain the admittedly untrustworthy and skittish witness\nprior to the first trial but, once it had that transcript in hand as\na substitute, deported the witness without considering the many\nalternatives to confinement in federal prison that might have\nprevented the witness’s absence. Id. at 959-60.\n\n The First Circuit, too, in Mann, excluded deposition\ntestimony of a crucial prosecution witness in lieu of live\ntestimony where, after the deposition, the government had\nreturned the witness’s passport and plane ticket that it had\nseized upon her arrest. 590 F.2d at 366, 368. The court held\ninsufficient the government’s later offer to pay to bring the\nwitness back, similar to the government’s offer here. Id. at\n363. “Implicit . . . in the duty to use reasonable means to\nprocure the presence of an absent witness is the duty to use\nreasonable means to prevent a present witness from becoming\nabsent.” Id. at 368. Where the government fails to take\nreasonable steps to prevent a witness from becoming absent,\nthe defendant should not suffer from the government’s choice.\nSee id.\n\n The cases on which the government relies are not to the\ncontrary. Foster distinguishes United States v. Allie, 978 F.2d\n1401, 1403-08 (5th Cir. 1992), and United States v. Calderon-\n\n 18\nLopez, 268 F. App’x 279, 289 (5th Cir. 2008), on the basis of\nextensive efforts the government made before it deported the\nwitnesses in those cases, none of which were made here.\nFoster, 753 F. App’x at 311. And in United States v. Eufracio-\nTorres, it was the defendant who initially opposed the\ngovernment’s motion to detain seven material witnesses\npending trial. 890 F.2d 266, 268 (10th Cir. 1989). The court\nthought it significant that none of those witnesses had been\ncharged with any crime, yet they were detained while\ndefendant Eufracio-Torres was free on bond. Id. Eufracio-\nTorres changed his position to request their detention pending\ntrial only after their depositions were taken—a request the\ncourt denied. Id. When the witnesses could not be brought\nback for trial and the government sought to use the depositions,\nthe court held that the government’s good-faith, reasonable\nefforts supported its request: Before they left the country, the\ngovernment had given the witnesses subpoenas, instructions\nfor reentry, travel reimbursement, appearance fees, and\nobtained their promises to return to testify. Id. No such steps\nwere taken here. The government’s reliance on United States\nv. Rivera is similarly misplaced. 859 F.2d 1204, 1207-09 (4th\nCir. 1988) (witnesses but not defendant were being detained,\nand court found no prejudice claimed or suffered from use of\nwitness depositions taken at the witnesses’ own request before\nthey were released and voluntarily left the United States).\n\n Under the applicable standard, the government failed to\nshow that Yindeear-Rom was “unavailable” for purposes of the\nConfrontation Clause. The government’s efforts to secure his\npresence at trial did not begin until after he was deported. See\nBurden, 2016 WL 5108010, at *3-4. Before his deportation,\nthe government did not give Yindeear-Rom a subpoena, offer\nto permit and pay for him either to remain in the U.S. or to\nreturn here from Thailand, obtain his commitment to appear,\nconfirm his contact information, or take any other measures.\n\n 19\nSee id. at *9. Its only efforts began once he was out of custody,\nout of the jurisdiction, and no longer dependent on the\ngovernment’s good graces for lenient treatment. Yindeear-\nRom’s eagerness to return to Thailand helped to persuade the\ndistrict court that further efforts to persuade him to testify at\ntrial would have been futile. See id. But in these circumstances\nthat eagerness cuts the other way. Given the government’s\nduty to make good-faith, reasonable efforts before Yindeear-\nRom’s deportation, “a witness’s known reluctance to testify\nadds to the government’s burden to show that it made\n‘reasonable, good faith efforts’ to secure her appearance\nbecause it makes her failure to appear voluntarily all the more\nforeseeable.” Brooks v. United States, 39 A.3d 873, 886 (D.C.\n2012). This is a case where the “possibility, albeit remote, that\naffirmative measures might produce the declarant . . .\ndemand[ed] their effectuation.” Roberts, 448 U.S. at 74. Any\nchance the government had of securing Yindeear-Rom’s\nappearance at trial would have been far greater had it addressed\nthe problem as soon as it knew it would rely on his testimony.\nInstead, its own approach appears to have ensured the futility\nof the post-deportation efforts.\n\n We recognize that it may not always be reasonable to\nexpect the government to postpone removal until trial—\nparticularly if the government would have to detain the witness\nin order to keep her in the country. See Aguilar-Ayala v. Ruiz,\n973 F.2d 411, 419 (5th Cir. 1992) (enjoining Department of\nJustice policy of automatically detaining undocumented\nimmigrant material witnesses because “undocumented aliens\nhave an overriding liberty interest in not being detained as\nmaterial witnesses”). And we assume that “in some cases the\nneed for the criminal defendant to confront the witness at trial\n(rather than at deposition) might outweigh the material witness’\nliberty interest in being released immediately,” whereas in\nother cases it will not. Id. (emphasis omitted). In order to\n\n 20\nidentify the existence and extent of any such conflict, however,\nthe unavailability inquiry must account for whether the\ngovernment has addressed potentially reasonable means of\nsecuring removable witnesses’ live testimony short of\ndetention, such as placing them in “lesser custody,” “supplying\nmaintenance, and retaining [their] passport[s] and ticket[s],”\nand “plac[ing] [them] under subpoena.” Mann, 590 F.2d at\n366; see also Allie, 978 F.2d at 1407 (holding that the\ngovernment’s efforts were reasonable where it offered\nwitnesses work permits to keep them in the United States).\n\n Because the government’s omissions place its efforts\nbelow the standard the Confrontation Clause demands, we need\nnot decide precisely how the government should have sought\nto prevent the witness from becoming unavailable. While the\ngovernment’s deportation of a witness may sometimes fail the\nstandard of good-faith and reasonable efforts, we reject any per\nse rule that no witness the government deports can be\nconsidered unavailable under the Confrontation Clause.\nConsistent with the fact-intensive nature of the standard, the\ngovernment decries any per se rule, see Appellee’s Br. at 45-\n46, the defendants do not advocate one, see Appellants’ Reply\nBr. 24 n.12, and no circuit has adopted any such categorical\napproach.\n\n We hold that the duty to use reasonable means to procure\na witness’s presence at trial includes the duty to use reasonable\nefforts to prevent a witness from becoming absent in the first\nplace. The government does not dispute that it made no efforts\nbefore deporting Yindeear-Rom to secure his presence at trial.\nThe witness thus was not “unavailable” such that prior\ntestimony could be admitted consistent with the Confrontation\nClause. Because admitting his deposition was not harmless\nbeyond a reasonable doubt, we vacate the convictions and\nremand for a new trial.\n\n 21\n B. Guidance Regarding the Jury Instruction’s\n Definition of “Willfully”\n\n Although our resolution of the Confrontation Clause issue\nis sufficient to dispose of this appeal, we provide some\nguidance regarding the jury instruction’s definition of the term\n“willfully,” an issue that was fully briefed and argued.\n\n The parties agree that defendants “willfully” violate the\nAECA only where they act with knowledge that their conduct\nis unlawful. The Supreme Court has explained that “willfully”\nis “a word of many meanings whose construction is often\ndependent on the context in which it appears.” Bryan v. United\nStates, 524 U.S. 184, 191 (1998) (internal quotation marks\nomitted). “As a general matter, when used in the criminal\ncontext, a ‘willful’ act is one undertaken with a bad purpose.\nIn other words, in order to establish a ‘willful’ violation of a\nstatute, the Government must prove that the defendant acted\nwith knowledge that his conduct was unlawful.” Id. at 191-92\n(internal quotation marks omitted). The parties disagree about\nhow specific that knowledge must be under the AECA’s\nwillfulness standard.\n\n Defendants argue that the AECA’s prohibition on\n“willfully violat[ing] any provision of this section,” 22 U.S.C.\n§ 2778(c), means that they should not have been convicted\nunless they “were aware of the specific law, rule, or regulation\nthat the[ir] conduct may have violated.” Appellants’ Br. 17\n(alteration in original). Before us, they frame this standard as\n“requir[ing] the government to prove an individual was aware\nthe items he exported were on the [Munitions List],” id. at 18-\n19, though their requested jury instruction would instead have\ndemanded proof that they knew that there was a license\nrequirement for their exports, App. 99. The heart of\ndefendants’ case at trial was that they “had not willfully\n\n 22\nviolated [the] AECA because they believed the parts were\nintended for use in Airsoft BB guns, not real guns, and they\nbelieved such parts could be exported without a license.”\nAppellants’ Br. 12. Under the court’s jury instructions, they\ncontend, “intent to violate Thai customs duties or Thai gun\ncontrol laws could support the jury’s finding of willfulness.”\nId. at 31. The government defends the district court’s\ninstruction on the ground that “the AECA is not one of the\nextremely rare, highly technical statutes that the Supreme\nCourt has found to require a heightened willfulness showing,”\nso the government did not have to prove defendants had\nknowledge of the license requirement or the Munitions List as\nsuch. Appellee Br. 16.\n\n As discussed below, because it was ambiguous as to what\n“conduct” defendants had to know was “unlawful,” the district\ncourt’s jury instruction arguably fell short of the baseline\nrequirement that the mens rea relate to the charged actus reus.\nBut we believe that the district court correctly instructed that,\nif defendants knew exporting the charged items without a\nlicense was unlawful, they did not need specific knowledge of\nthe Munitions List. Thus, the district court was right that “the\ngovernment need not prove that a defendant had read, was\naware of, or had consulted the licensing provisions of the Arms\nExport Control Act or the International Traffic in Arms\nRegulations, or the Munitions List.” App. 66. For purposes of\nthe AECA, a requirement of proof that defendants knew the\nproscribed conduct was unlawful adequately protects against\n“the danger of ensnaring individuals engaged in apparently\ninnocent conduct.” Bryan, 524 U.S. at 194.\n\n Most criminal prohibitions require only proof that the\ncrime was committed “knowingly,” meaning that the defendant\nknew of the facts that made his act illegal, even if he did not\nknow the act was illegal. When Congress wants to ensure that\n\n 23\ndefendants will be convicted only if they have a more culpable\nstate of mind, it limits the crime to conduct that a defendant\nengages in “willfully.” See id. at 191-92. The Court has\ndeveloped an approach to “willfulness” that is calibrated to the\nstatutes in which it appears, consistently reading it to require\nproof that the defendant had a sufficiently culpable state of\nmind, but not to require proof so specific as to stymie the\nstatute’s enforcement. It has required proof that a defendant\nknow which law he was breaking in only two contexts:\ncriminal tax evasion, and currency structuring.\n\n In Cheek v. United States, a tax case, the Court explained\nCongress’s inclusion of a willfulness requirement for felony\ntax-evasion as resting partly on the recognition that “[t]he\nproliferation of statutes and regulations has sometimes made it\ndifficult for the average citizen to know and comprehend the\nextent of the duties and obligations imposed by the tax laws.”\n498 U.S. 192, 199-200 (1991). The Court held that the\nwillfulness requirement could be defeated by the defendant’s\ngood-faith belief that he “was not a person required to file a\nreturn or to pay income taxes and that wages are not taxable\nincome,” even if that belief was unreasonable. Id. at 203.\n\n In Ratzlaf v. United States, the Court interpreted\n“willfully” in the context of a statute criminalizing “currency\nstructuring.” 510 U.S. 135, 136, 144 (1994). The anti-\nstructuring law prohibited a customer from breaking up a large\nfinancial transaction into multiple smaller ones to avoid\ntriggering a bank’s legal duty to report transactions over\n$10,000. Id. at 145. The Court held that “willfulness” meant\nthat the defendant had to be aware not only of reporting\nobligations applicable to banks, but of his own distinct\nobligation not to structure his transactions for the purpose of\navoiding triggering the bank’s obligation to report. Id. at 137-\n38. In other words, “[t]o convict Ratzlaf of the crime with\n\n 24\nwhich he was charged . . . the jury had to find he knew the\nstructuring in which he was engaged was unlawful.” Id. at 149.\n\n In Bryan, which concerned unlicensed firearms-dealing,\nthe Court concluded that “requiring only knowledge that the\nconduct is unlawful [was] fully consistent” with Congress’s\npurpose in adding a willfulness requirement to the firearms\nstatute “to protect law-abiding citizens who might\ninadvertently violate the law.” 524 U.S. at 195 n.23. Bryan\ndistinguished the firearms law before it from the “highly\ntechnical” statutes at issue in Cheek and Ratzlaf, which had\njustified “‘carv[ing] out an exception to the traditional rule’\nthat ignorance of the law is no excuse and requir[ing] that the\ndefendant have knowledge of the law”—that is, the specific\nlaw he violated. Bryan, 524 U.S. at 194-95 (quoting Cheek,\n498 U.S. at 200).\n\n Following Bryan, most courts of appeals to consider the\nissue have interpreted the AECA’s requirement that a violation\nbe “willful” not to require specific awareness of the Munitions\nList as such. What they have required—as we do here—is\nproof that defendants knew it was illegal to export the items\nthey shipped without a license. See United States v. Henry, 888\nF.3d 589, 598-600 (2d Cir. 2018) (“[W]illfulness requires only\nthat the defendant know that what he was doing was illegal,\nand not that he know that his conduct was prohibited under a\nspecific AECA provision or related regulation.”); United States\nv. Bishop, 740 F.3d 927, 932-33 (4th Cir. 2014) (government\nneed not show defendant knew “that the ammunition was\nspecifically covered by the AECA,” but only that he had\n“general knowledge of an export’s illegality”); United States v.\nChi Mak, 683 F.3d 1126, 1138 (9th Cir. 2012) (affirming\ninstruction that conviction did not require proof that “the\ndefendant had read, was aware of, or had consulted the specific\nregulations governing his activities”); United States v. Roth,\n\n 25\n628 F.3d 827, 835 (6th Cir. 2011) (the AECA “does not require\na defendant to know that the items being exported are on the\nMunitions List,” but only “knowledge that the underlying\naction is unlawful”); United States v. Tsai, 954 F.2d 155, 160-\n62 & n.3 (3d Cir. 1992) (jury need not find that defendant\n“knew all of the specifics of the law,” but “what it had to find\nwas that that the defendant knew that he could not export that\nparticular item”); United States v. Murphy, 852 F.2d 1, 6-7 (1st\nCir. 1988) (requiring proof that “defendant must know that his\nconduct in exporting from the United States articles proscribed\nby the statute is violative of the law,” but rejecting requirement\nthat he must know specifically that the article was on the\nMunitions List or that there was a licensing requirement).\n\n The Eleventh Circuit in United States v. Wenxia Man\ncharacterized the willfulness requirement of the AECA as more\ndemanding than that of the statute in Bryan, but its bottom line\nis not materially different from that of the other circuits. See\n891 F.3d 1253, 1268 (11th Cir. 2018). That court hewed to its\npre-Bryan precedent that “aware[ness] of the generally\nunlawful nature of [one’s] actions is insufficient,” but was\nsatisfied with proof “that a conspirator actually knew that it was\nunlawful to export the [items] and intentionally violated [the]\nknown legal duty not to export [them].” Id. (internal quotation\nmarks omitted) (alterations in original) (quoting United States\nv. Adames, 878 F.2d 1374, 1377 (11th Cir. 1989)). The Fifth\nCircuit held in United States v. Hernandez that the AECA\ndemands a finding that the defendant “knew he was unlawfully\nexporting weapons on the Munitions List,” but that case\npredated Bryan, Cheek, and Ratzlaf. 662 F.2d 289, 292 (5th\nCir. 1981). The Seventh Circuit did so more recently, but it,\ntoo, took no account of those Supreme Court cases. See United\nStates v. Dobek, 789 F.3d 698, 700 (7th Cir. 2015).\n\n 26\n We hold that the district court’s jury instruction was\ncorrect insofar as it rejected defendants’ position that the\nwillfulness requirement in the AECA is akin to those at issue\nin Cheek and Ratzlaf, and clarified that the government need\nnot prove defendants knew the specific law their conduct\nviolated.\n\n That does not quite resolve the issue, however, because of\nambiguity in the meaning of “conduct” in the instruction that a\nwillful violation of the AECA requires that the defendant\n“acted with knowledge that his conduct was unlawful.” App.\n66. More is required where, as here, there is evidence of\nwillfully unlawful conduct apart from the charged offenses,\ncreating a risk that the jury may consider any and all evidence\nof the defendant’s guilty mind—whatever its object—as\nsupporting willful commission of the charged offense. The\nwillfulness instruction arguably fell short in not specifying that\nthe unlawful “conduct” the jury must find the defendants to\nhave willfully done was the actus reus that violated the AECA:\nunlawfully exporting the magazines and mount without a\nlicense. To be sure, the rest of the instruction made clear that\nthe actus reus of the charged offense was exporting the items\nwithout a license, and we review jury instructions as a whole.\nUnited States v. McGill, 815 F.3d 846, 888 (D.C. Cir. 2016).\nBut we nonetheless detect a potential problem specific to the\ninstruction’s discussion of willfulness—an issue that is both of\ncentral importance and easily muddied. It is natural enough\nthat jurors who are told they may convict upon finding the\ndefendants did the requisite act, so long as the defendant had a\nguilty mind, may not parse the object of the guilty mind. To\nprompt that important precision, the instruction’s definition of\n“willfully”—the word at issue here—should specify that the\nactus reus was the only “conduct” to which defendants’ guilty\nmental state could apply for a finding of willfulness.\n\n 27\n The requirement that the mens rea relate to the charged\nactus reus is the baseline for any criminal mental standard. See\nDixon v. United States, 548 U.S. 1, 6 (2006) (“[C]riminal\nliability is normally based upon the concurrence of two factors,\nan evil-meaning mind [and] an evil-doing hand.” (second\nalteration in original) (internal quotation marks omitted)). If a\nstatute requires that the defendant knowingly perform the act\nthat violates the law—even where he need not also know that\nthe act is illegal—then he must knowingly perform the charged\nactus reus, not some other, uncharged act. See Bryan, 524 U.S.\nat 193 (“[T]he term ‘knowingly’ merely requires proof of\nknowledge of the facts that constitute the offense.” (emphasis\nadded)). So, too, with willfulness. The district court and the\nparties agree that willfulness requires that the defendant “acted\nwith knowledge that his conduct was unlawful.” App. 66; see\nBryan, 524 U.S. at 193. The conduct that he must know was\nunlawful is the actus reus of the crime with which he is\ncharged.\n\n Because the willfulness instruction required only that the\ndefendants acted with knowledge that “the conduct” was\nunlawful, App. 66, there is some chance that the jury convicted\nbased in part on defendants’ evasiveness in importing to\nThailand. On retrial, the instruction should make clear that an\nAECA conviction requires that defendants knew of the\nunlawfulness of the charged unlicensed export of the items\nfrom the United States, and that a willfulness finding cannot\ndraw on evidence that they knew the related, but legally and\nfactually distinct, import of those items into Thailand was\nillegal.\n\n 28\n C. Burden’s Limited English Did Not Render His\n Interrogation Custodial for Miranda Purposes.\n\n We review de novo the determination whether Burden was\nin custody and thus entitled to Miranda warnings before any\ninterrogation, and we review the underlying factual findings for\nclear error. United States v. Hallford, 756 F. App’x 1, 5 (D.C.\nCir. 2018). If the court erroneously admits a non-Mirandized\nstatement, we must reverse unless the admission was harmless\nbeyond a reasonable doubt. United States v. Brinson-Scott, 714\nF.3d 616, 622 (D.C. Cir. 2013).\n\n In Miranda, the Court held that “the prosecution may not\nuse statements, whether exculpatory or inculpatory, stemming\nfrom custodial interrogation of the defendant unless it\ndemonstrates the use of procedural safeguards effective to\nsecure the privilege against self-incrimination.” 384 U.S. at\n444. Defendants argue that Burden’s limited English abilities\nrendered the interview “custodial.” We disagree. We see no\nerror in the district court’s admission of Burden’s non-\nMirandized statement because Burden was not in custody when\nthe agents questioned him at his warehouse. His English\nproficiency was not an “objective circumstance surrounding\nthe interrogation” that a reasonable officer would have\nassumed bore on whether Burden felt free to leave the\ninterview.\n\n Miranda warnings are required as a bulwark against the\ncoercive power of being taken into police custody and\ninterrogated. “An individual swept from familiar surroundings\ninto police custody, surrounded by antagonistic forces, and\nsubjected to . . . techniques of persuasion . . . cannot be\notherwise than under compulsion to speak.” Miranda, 384\nU.S. at 461. Police custody is sometimes self-evident, as in\nMiranda’s own case: The police arrested him and took him to\n\n 29\na police station interrogation room where they questioned him.\nId. at 491. But a person is also in custody when he is\n“otherwise deprived of his freedom of action in any significant\nway.” Id. at 444. In “Miranda case law, ‘custody’ is a term of\nart that specifies circumstances that are thought generally to\npresent a serious danger of coercion.” Howes v. Fields, 565\nU.S. 499, 508 (2012).\n\n The Supreme Court has laid out guidelines for the custody\nanalysis. “In determining whether a person is in custody,”\ntriggering the duty to give Miranda warnings,\n\n the initial step is to ascertain whether, in light of\n the objective circumstances of the interrogation, a\n reasonable person [would] have felt he or she was\n not at liberty to terminate the interrogation and\n leave. And in order to determine how a suspect\n would have gauge[d] his freedom of movement,\n courts must examine all of the circumstances\n surrounding the interrogation. Relevant factors\n include the location of the questioning, its duration,\n statements made during the interview, the presence\n or absence of physical restraints during the\n questioning, and the release of the interviewee at\n the end of the questioning.\n\nId. at 509 (alterations in original) (internal quotation marks and\ncitations omitted).\n\n “[W]hether a suspect is ‘in custody’ is an objective\ninquiry.” J.D.B. v. North Carolina, 564 U.S. 261, 270 (2011).\nThat means that, while the analysis accounts for “any\ncircumstance that would have affected how a reasonable person\nin the suspect’s position would perceive his or her freedom to\nleave,” it “involves no consideration of the actual mindset of\nthe particular suspect subjected to police questioning.” Id. at\n\n 30\n271 (internal quotation marks omitted). The benefit of the\nobjective test is practical: It “avoids burdening police with the\ntask of anticipating the idiosyncrasies of every individual\nsuspect and divining how those particular traits affect each\nperson’s subjective state of mind.” Id.\n\n The test, while objective, is also contextual. Because all\nof the “objective circumstances of the interrogation” must be\nconsidered, Howes, 565 U.S. at 508, individual characteristics\nthat have an “objectively discernible relationship to a\nreasonable person’s understanding of his freedom of action,”\nsuch as a child suspect’s age, J.D.B., 564 U.S. at 275, must be\ntaken into account. Under circumstances in which “a\nreasonable child subjected to police questioning will . . . feel\npressured to submit when a reasonable adult would feel free to\ngo[,] . . . courts can account for that reality without doing any\ndamage to the objective nature of the custody analysis.” Id. at\n272. That is because age “generates commonsense conclusions\nabout behavior and perception” that “apply broadly to children\nas a class.” Id. A reasonable eight-year-old will not necessarily\nfeel free to leave when a reasonable adult would.\n\n Because the test is designed to guide police, a person’s\nyouth—or analogous circumstances bearing on a reasonable\nperson’s perception of her freedom to leave—only factors into\nthe custody analysis where it “was known to the officer at the\ntime of the interview, or would have been objectively apparent\nto any reasonable officer.” Id. at 274. The test thus includes a\ndouble inquiry: whether it would have been apparent to a\nreasonable officer that a reasonable person in the suspect’s\nposition would not have felt free to leave. In holding that the\nsuspect’s youth was an objective circumstance relevant to the\ncustody analysis, the Supreme Court suggested that other\npersonal characteristics, like blindness, could be similarly\nrelevant. Id. at 278. The Court underscored: “Not once have\n\n 31\nwe excluded from the custody analysis a circumstance that we\ndetermined was relevant and objective, simply to make the\nfault line between custodial and noncustodial ‘brighter.’” Id.\nat 280.\n\n Under some circumstances, English language capabilities\nmight have an “objectively discernible relationship to a\nreasonable person’s understanding of his freedom of action”\nthat would bear on the custody analysis for purposes of\nMiranda. Id. at 275. Some courts have accordingly factored\nlimited English abilities into the custody inquiry. The Eighth\nCircuit explained that “the ultimate issue is whether a\nreasonable police officer conducting [an] otherwise non-\ncustodial interview would have given Miranda warnings\nbecause he realized that the questioning would be perceived by\n[the defendant] as custodial due to his limited English language\nskills.” Thatsaphone v. Weber, 137 F.3d 1041, 1045 (8th Cir.\n1998). In United States v. Kim, the Ninth Circuit weighed\nlimited English proficiency (among other factors the court\ndeemed important) because, under the circumstances, it bore\non the defendant’s ability to understand whether she was a\ncriminal suspect. 292 F.3d 969, 977 (9th Cir. 2002). Kim had\narrived at the store her family owned to find many police cars\noutside and the door locked; the police let her, but not her\nhusband, enter, locked the door, did not let her son (who was\nalready inside) speak to her, and forced her to speak only\nEnglish, though she and her son informed the officers that she\ndid not speak English well. Id. at 971-72. The court held that\nKim’s interrogation was custodial partly because the police\n“temporarily took over complete control of Kim’s store,\ncreating ‘a police-dominated atmosphere,’ in which the police\nkept Kim physically isolated from two family members who\ncould have provided both moral support and, given her limited\nEnglish, a more complete understanding of the overall\nsituation.” Id. at 977. Both Thatsaphone and Kim preceded the\n\n 32\nCourt’s application of the custody analysis to a child in J.D.B.,\nbut both decisions comport with J.D.B. insofar as they consider\nwhether a reasonable officer would have been able to discern\nthat the language limitations of a reasonable person in the\nsuspect’s position would have contributed to that person’s not\nfeeling free to leave. Since J.D.B., at least one other court has\nheld a suspect’s limited English proficiency relevant to\nJ.D.B.’s objective inquiry. See United States v. Han, 199 F.\nSupp. 3d 38, 52-54 (D.D.C. 2016).\n\n The district court in this case correctly held that a\nreasonable officer would not have thought that Burden’s\nlanguage abilities prevented him from feeling free to leave, and\nthus properly admitted Burden’s non-Mirandized statement.\nEven though Burden sometimes had trouble formulating\nresponses and appeared to lack perfect comprehension of all\nthe questions, the evidence does not suggest that it would have\nbeen apparent to a reasonable officer that Burden was not\nunderstanding what was being said.\n\n The defendants identify only one moment, when the DHS\nofficer was explaining the purpose and terms of the interview,\nwhen they believe that a reasonable officer should have\nrecognized that Burden’s English skills would affect the\nperception of a reasonable person in his position as to whether\nhe was free to leave:\n\n UNIDENTIFIED AGENT NO. 1: We are federal\n agents for the U.S. Government so I have to let you\n know that you have to be honest with us —\n\n PHEERAYUTH BURDEN: Uh-huh.\n\n UNIDENTIFIED AGENT NO. 1: — okay? If you\n don’t want to answer something, you don’t have to\n answer but you cannot lie to us.\n\n 33\n PHEERAYUTH BURDEN: Okay.\n\n UNIDENTIFIED AGENT NO. 1: All right? And\n you can’t withhold relevant information.\n\n PHEERAYUTH BURDEN: Uh-huh.\n\n UNIDENTIFIED AGENT NO. 1: If you do, that is\n a crime.\n\n PHEERAYUTH BURDEN: Okay\n\n UNIDENTIFIED AGENT NO. 1: Okay?\n Punishable by up to five years in prison so just\n please be honest.\n\n PHEERAYUTH BURDEN: Yeah.\n\nApp. 174. While the agent’s statement was somewhat\nconfusing, it is not clear from this exchange that Burden was\nnot comprehending what the agent was saying to him or\nsomehow believed he could not leave an interview at his own\nwarehouse that he had agreed to by phone and shown up for of\nhis own accord.\n\n The district court did not refuse to consider Burden’s\nproficiency entirely, as defendants assert. It properly applied\nthe custody test by evaluating how a reasonable officer would\nhave perceived Burden’s comprehension. It noted “that for\neach of the points that the Special Agent was communicating\nto Mr. Burden, Mr. Burden said okay or yes or [uh-huh],\nreflect[ing] responses like that which would give one\nreasonably the understanding that Mr. Burden understood what\nwas being said.” App. 430-31.\n\n 34\n The custody question was not otherwise close. The district\ncourt found that Burden “arrived with his wife at his own work\nplace, at the time that he had set and there was no evident effort\nto overcome his will. There was no effort to put him in\nhandcuffs, no threats, other threats during the course of the\ninterview.” App. 431. Burden himself “chose where he was\ngoing to sit, he chose the room in which they were going to\ntalk,” and he “showed that he knew he could get up, go out,\nopen the door and talk to somebody outside throughout the\ncourse of this.” App. 432. One of the two agents who\nquestioned Burden testified that Burden had chosen where and\nwhen the interview would take place, S.A. 400, and that the\nagents wore plainclothes and did not display their badges or\nweapons, S.A. 407. The interview lasted no more than three\nhours and Burden left when it was over. In sum, “the location\nof the questioning, its duration, statements made during the\ninterview, the presence or absence of physical restraints during\nthe questioning, . . . the release of the interviewee at the end of\nthe questioning,” and other factors all support the conclusion\nthat Burden was not in custody. Howes, 565 U.S. at 509\n(internal quotation marks and citations omitted). Because\nBurden’s imperfect English would not have given a reasonable\nofficer the impression that a reasonable person in Burden’s\nposition would have believed himself detained, it does not\nchange that determination here.\n\n * * *\n\n For the foregoing reasons, we affirm the district court’s\nadmission of Burden’s non-Mirandized statement to DHS\nagents. We nonetheless vacate the convictions in view of the\nerror in the admission of Yindeear-Rom’s deposition\ntestimony.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: UNITED STATES of America, Appellee v. Pheerayuth BURDEN, Appellant\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-20\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: Yes\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued January 16, 2019 Decided August 20, 2019\\n\\n No. 17-3018\\n\\n UNITED STATES OF AMERICA,\\n APPELLEE\\n\\n v.\\n\\n PHEERAYUTH BURDEN,\\n APPELLANT\\n\\n\\n Consolidated with 17-3019\\n\\n\\n Appeals from the United States District Court\\n for the District of Columbia\\n (No. 1:14-cr-00069-2)\\n (No. 1:14-cr-00069-3)\\n\\n\\n Lindsay C. Harrison, pro bono, argued the cause and filed\\nthe briefs for appellant. James Dawson entered an appearance.\\n\\n Daniel J. Lenerz, Assistant U.S. Attorney, argued the\\ncause for appellee. With him on the brief were Jessie K. Liu,\\nU.S. Attorney, and Elizabeth Trosman and Elizabeth H.\\nDanello, Assistant U.S. Attorneys.\\n\\n Before: HENDERSON, ROGERS and PILLARD, Circuit\\nJudges.\\n\\n 2\\n Opinion for the Court filed by Circuit Judge PILLARD.\\n\\n Opinion concurring in part and concurring in the judgment\\nfiled by Circuit Judge ROGERS.\\n\\n PILLARD, Circuit Judge: The Arms Export Control Act\\n(AECA) criminalizes exporting defense articles without a\\nlicense. 22 U.S.C. § 2778(b)(2), (c). Pheerayuth Burden, a\\nU.S.-resident Thai national who ran a business exporting goods\\nfrom the United States to Thailand, and his export business,\\nWing-On LLC (collectively, the defendants), exported five\\nassault-rifle magazines and a grenade-launcher mount.\\nFollowing a three-week trial, a jury convicted the defendants\\nof conspiracy to violate the AECA, unlawful export in violation\\nof the AECA, and conspiracy to launder money.\\n\\n The defendants contend that three of the district court’s\\nrulings are reversible error. First, they argue that the court\\nerred in admitting video deposition testimony by a key witness\\nover a Confrontation Clause objection where the government\\nitself rendered the witness “unavailable” at trial by deporting\\nhim shortly before trial without first making reasonable efforts\\nto arrange his return. Second, they challenge a jury instruction\\ndefining the “willfulness” element of unlawful exportation of\\ndefense articles as requiring only proof that the defendants\\n“acted with knowledge that the conduct was unlawful.” That\\ninstruction was inadequate, they contend, because it failed to\\ntie the willfulness finding to the pertinent conduct and law,\\ncreating an impermissible risk that the jury relied on evidence\\nthat Burden thought he was violating Thai import law. Third,\\ndefendants claim that the district court erred in admitting\\nBurden’s non-Mirandized statements because it failed to\\naccount for his limited English abilities in determining that he\\nwas not in custody when agents interrogated him.\\n\\n 3\\n We hold that the district court erred in admitting the\\ndeposition testimony because the government failed to make\\nreasonable efforts before it deported the witness to procure his\\npresence at trial. We conclude that the jury instruction was\\ncorrect as far as it went in instructing the jury to find that “the\\ndefendant knew that his conduct was unlawful,” and that\\n“willfully” violating the law does not require proof “that a\\ndefendant had read, was aware of, or had consulted the\\nlicensing provisions of the Arms Export Control Act” as such.\\nAppellants’ Appendix (App.) 66. But we suggest clarification\\nof the willfulness instruction to more squarely require a finding\\nthat defendants were aware of and knowingly violated their\\nlegal obligation not to commit the charged actus reus. A case\\nsuch as this one—that includes evidence of consciousness of\\nguilt relating to distinct actus reus arguably violating different,\\nuncharged legal obligations—creates some risk of the jury\\nrelying on evidence of consciousness of guilt unrelated to the\\ncharged crime. We affirm the district court’s determination\\nthat Burden was not in custody because, even assuming\\nlanguage proficiency is relevant to the custody inquiry, a\\nreasonable officer would not have thought Burden’s imperfect\\nEnglish meant a reasonable person in his position would have\\nbelieved himself detained during the interview.\\n\\n Because the error we identify was not harmless, we vacate\\nthe judgments and remand for proceedings consistent with this\\nopinion.\\n\\n BACKGROUND\\n\\n A. Legal Background\\n\\n The AECA establishes executive-branch control over the\\nexport and import of “defense articles,” meaning arms or other\\nmilitary items. See 22 U.S.C. § 2278. It authorizes the\\n\\n 4\\nPresident, “[i]n furtherance of world peace and the security and\\nforeign policy of the United States,” to control the export of\\ndefense articles and services, designate which items count as\\ndefense articles and services, and promulgate regulations for\\nthose purposes. Id. § 2778(a)(1). The designated defense\\narticles make up the United States Munitions List (the\\nMunitions List or the List). Id. With certain enumerated\\nexceptions, “no defense articles or defense services designated\\nby the President” as part of the Munitions List “may be\\nexported or imported without a license for such export or\\nimport, issued in accordance with” the AECA and its\\nassociated regulations. Id. § 2778(b)(2). The State Department\\nis responsible for issuing licenses. See id.; 22 C.F.R. §§ 120.1,\\n120.20. The decision whether to issue an export license\\nimplicates sensitive issues of national security and foreign\\npolicy. It must “take into account whether the export of an\\narticle would contribute to an arms race, aid in the development\\nof weapons of mass destruction, support international\\nterrorism, increase the possibility of outbreak or escalation of\\nconflict, or prejudice the development of . . . arms control or\\nnonproliferation agreements or other arrangements.” 22\\nU.S.C. § 2778(a)(2). The statute criminalizes “willfully\\nviolat[ing] any provision of this section . . . or any rule or\\nregulation issued under this section.” Id. § 2278(c). It thus\\ncriminalizes willfully exporting defense articles without a\\nlicense.\\n\\n The President delegated to the Secretary of State the\\nauthority to designate defense articles and promulgate\\nregulations under the AECA, see Exec. Order No. 13637, 78\\nFed. Reg. 16,129 (2013); 22 C.F.R. § 120.1(a), and the\\nSecretary accordingly promulgated the International Traffic in\\nArms Regulations (ITAR), 22 C.F.R. §§ 120.1-130.17. The\\nITAR prohibits exporting defense articles and services without\\n“obtaining the required license or other written approval” from\\n\\n 5\\nthe appropriate office of the State Department. 22 C.F.R.\\n§ 127.1(a). The ITAR also includes the Munitions List, which\\nruns to over forty pages in the Code of Federal Regulations.\\nSee id. § 121.1. The covered defense articles are described with\\nvarying levels of specificity, such as “[r]iflescopes\\nmanufactured to military specifications,” id. (Category I(f)),\\n“[g]uns over caliber .50,” id. (Category II(a)), “[i]ron powder\\n. . . with particle size of 3 micrometers or less produced by\\nreduction of iron oxide with hydrogen,” id. (Category\\nV(c)(4)(i)(B)), and “[h]elmets . . . providing a protection level\\nequal to or greater than NIJ Type IV,” id. (Category X(a)(6)).\\nThe convictions in this case relate to items in Category I(h) of\\nthe Munitions List: “Components, parts, accessories and\\nattachments” for the firearms listed in Category I(a)-(g). See\\nApp. 89-90.\\n\\n B. Factual and Procedural Background\\n\\n Burden started Wing-On LLC (Wing-On), a freight-\\nforwarding business that shipped American goods to Thailand,\\naround 2008. In 2010, Kitibordee Yindeear-Rom became one\\nof Burden’s customers. A Thai national living in Thailand,\\nYindeear-Rom had a business importing many different types\\nof goods from the United States to Thailand. As part of that\\nbusiness, he helped his customers get gun parts and accessories\\nfrom the U.S. that they could not purchase directly because\\nU.S. companies would neither accept Thai credit cards nor ship\\nthe parts to Thailand. According to Yindeear-Rom, Burden\\ninitially ordered gun parts for him from U.S. vendors, received\\nthem in the United States, then shipped them to Thailand.\\nSupplemental Appendix (S.A.) 291A-91B. Yindeear-Rom\\nlater began placing the orders himself using a debit card\\nattached to a U.S. bank account Burden opened. S.A. 294-96,\\n479. Yindeear-Rom testified in his deposition that he\\nreimbursed Burden for the purchases he made on Burden’s\\n\\n 6\\ndebit card by transferring money to Thai bank accounts\\nbelonging to Burden and Burden’s associate. S.A. 298-300.\\nNeither Burden nor Wing-On had a license to export defense\\narticles on the Munitions List.\\n\\n In October 2013, Yindeear-Rom took a vacation to the\\nUnited States, where he was stopped and interviewed by\\nDepartment of Homeland Security (DHS) agents. He was\\narrested two days later for conspiracy to violate American\\nexport laws. He later pleaded guilty and was sentenced to\\nthirty-six months in prison. At Yindeear-Rom’s initial court\\nappearance, the DHS agents saw in the courtroom two people\\nthey believed to be Burden’s wife and roommate, respectively.\\nConcerned that Burden might have been alerted to the\\ninvestigation, the agents went immediately to Wing-On’s\\nwarehouse. Burden was not there, but the agents met one of\\nhis employees, who helped the agents call him. They called\\nhim again later that day and arranged an interview for that\\nevening at the warehouse.\\n\\n The DHS agents interviewed Burden in English without an\\ninterpreter. They did not advise Burden of his rights with the\\nfamiliar warnings officials must give suspects in custodial\\ninterrogation under Miranda v. Arizona, 384 U.S. 436, 444\\n(1966). At the beginning of the interview, one of the agents\\ngave the following preamble (recounted here with Burden’s\\naffirmative interjections omitted):\\n\\n We are federal agents for the U.S. Government so\\n I have to let you know that you have to be honest\\n with us[,] okay? If you don’t want to answer\\n something, you don’t have to answer but you\\n cannot lie to us. All right? And you can’t withhold\\n relevant information. If you do, that is a crime.\\n\\n 7\\n Okay? Punishable by up to five years in prison so\\n just please be honest.\\n\\nApp. 174. The agent then asked, “Is your English good?” Id.\\nBurden replied, “A little bit.” App. 175. “If there’s anything\\nthat I say that you don’t understand, ask me,” the agent said.\\nId. “Okay,” Burden replied. Id. During the interview, Burden\\nadmitted he had shipped gun parts to Yindeear-Rom and\\nfalsified customs declarations. App. 266-71. The agents did\\nnot arrest Burden at the end of the interview. S.A. 420-21. The\\ncourt’s eventual admission of Burden’s statement at his trial\\nover defense objections is the subject of his Miranda claim on\\nappeal. App. 429-33.\\n\\n Burden was arrested six months later, in May 2014. Trial\\nwas initially scheduled for November 2015, but was continued\\ntwice, first to April 2016 and then to September 2016. The\\ndistrict court granted the second continuance because many\\ndocuments remained to be translated into English. The court\\nnoted that the defense “can’t actually do this without translated\\ndocuments” and that “it’s not that [the defense has] been less\\nthan diligent about it.” App. 381. That continuance introduced\\na wrinkle into the trial, however: Yindeear-Rom was\\nscheduled to be released from prison in June 2016, three\\nmonths before the new trial date, and was to be deported after\\nhis release. The government had a clear path to remove\\nYindeear-Rom upon his release because he had stipulated when\\nhe pleaded guilty to an order of removal that would “render[]\\nhim permanently inadmissible to the United States,” which\\nassured that the government, “promptly upon his release from\\nconfinement . . . may execute the order of removal according\\nto the applicable laws and regulations.” App. 47 (alteration in\\noriginal). The government in February 2016 moved to take\\nYindeear-Rom’s deposition under Federal Rule of Criminal\\nProcedure 15, which governs depositions taken to preserve a\\n\\n 8\\npotential witness’s testimony for trial. S.A. 1; Fed. R. Crim. P.\\n15(a)(1). The district court granted the motion over the\\ndefendants’ objections. The defense objected to the deposition\\nfor some of the same reasons it asked for the second trial\\npostponement: the government had not produced sufficient\\ndiscovery to allow them to prepare for Yindeear-Rom’s\\ntestimony. The court nonetheless allowed the deposition to be\\ntaken to preserve evidence in the event that Yindeear-Rom\\nwould be unavailable to testify. Yindeear-Rom’s videotaped,\\nin-court deposition took place over four days in March and\\nApril. The court granted the government’s motion to reduce\\nYindeear-Rom’s sentence in exchange for his testimony.\\n\\n The United States deported Yindeear-Rom to Thailand in\\nApril 2016. Even though the government had substantial\\nbargaining leverage before it moved for his sentence reduction,\\nthere is no record that it made any efforts before deporting him\\nto secure Yindeear-Rom’s presence at trial. It was only once\\nYindeear-Rom was back in Thailand that the government\\nbegan to make such efforts.\\n\\n In seeking to bring Yindeear-Rom back a few months\\nlater, the government contacted Yindeear-Rom’s counsel by\\nphone and mail. United States v. Burden, No. 14-cr-0069\\n(RMC), 2016 WL 5108010, at *3 (D.D.C. Sept. 20, 2016). Its\\nletter to counsel included a subpoena for Yindeear-Rom’s\\ntestimony at trial and a promise to help him obtain a visa and\\nto pay his travel expenses, “including but not limited to round-\\ntrip airfare, transportation, room, board, and per diem witness\\nfee.” Id. (quoting letter, S.A. 51). Yindeear-Rom’s lawyer\\n“forwarded the letter and subpoena to Mr. Yindeear-Rom in\\nThailand, but was unable to confirm receipt or make any\\nrepresentations about [his] willingness to testify.” Id. at *3.\\nThe government also sent the letter and subpoena directly to\\nYindeear-Rom’s last known email and physical address in\\n\\n 9\\nThailand. Id. at *4. The government received no response to\\neither, but confirmed that the letter was signed for by “K. Yen,”\\nwhich it believed to be Yindeear-Rom. Id. DHS personnel in\\nThailand eventually reached Yindeear-Rom by phone and\\nlearned that he had received the email and letter, but that he\\n“had no desire to travel to the United States to cooperate in any\\nway.” Id. (quoting DHS Report of Investigation, S.A. 75).\\n\\n At trial, the court granted the government’s motion to\\nadmit the Rule 15 deposition over the defendants’ objections.\\nDefendants argued that the government should have sought to\\nkeep Yindeear-Rom in the country between his release from\\nprison and the trial. The court concluded that the witness was\\nunavailable, and that “the use of a videotaped deposition taken\\nin court, before the trial judge and including the presence of\\nMr. Burden and cross examination by both defense lawyers,”\\nwas “a very good substitute” for Yindeear-Rom’s live trial\\ntestimony “that would allow the jury to observe his demeanor\\nand preserve the Defendants’ rights to confront witnesses\\nagainst them.” Id. at *3, *8.\\n\\n The government’s trial evidence included Burden’s\\nstatement to DHS agents at the Wing-On warehouse that he\\nmislabeled customs declarations for shipments containing gun\\nparts, and packed gun parts hidden among other items for\\nshipping. App. 269-72. In his defense, Burden highlighted his\\nstatements that he took those steps to evade Thai customs. He\\ntold the agents that he concealed gun parts among other items\\nbecause “[y]ou want to hide from the custom in Thailand,”\\nApp. 271, and that he falsified customs documents “[b]ecause\\nof tax in Thailand,” App. 281.\\n\\n The government, for its part, pointed to circumstantial\\nevidence tending to show that Burden had reason to know that\\nhe was violating U.S. arms-export law. For instance, Burden\\n\\n 10\\nacknowledged in his statement to DHS agents that people\\n“[n]eed a license” to ship certain things, like gun parts. App.\\n203-04. Yindeear-Rom received a notice that a seized\\nshipment of gun parts violated the ITAR, see S.A. 212-13, and\\neven though he had ordered that shipment through a different\\nshipper (not Wing-On), Yindeear-Rom forwarded the notice to\\nBurden asking what he should do, see S.A. 282-83. Yindeear-\\nRom then testified that the notice he forwarded informed both\\nof them of the requirements of the ITAR. S.A. 369. Burden\\nhad also received a notice directly from U.S. Customs and\\nBorder Protection that it had seized a rifle scope (controlled\\nunder a distinct set of regulations analogous to the ITAR but\\ncovering different items) because the scope could not be\\nexported without a license. See App. 171.\\n\\n Further evidence, however, suggested that Burden either\\ndid not realize he was shipping real gun parts, or thought it was\\nlegal under U.S. export law to ship those parts if they were to\\nbe used with toy guns. As part of their business, the defendants\\nshipped BB guns (air guns that shoot small metal balls) and\\nAirsoft toys (which are similar to BB guns and shoot plastic\\npellets). See App. 311-12, 574. Airsoft and BB guns\\nthemselves may lawfully be exported without a State\\nDepartment license. The defense’s expert witness on firearms\\nand Airsoft identification testified that an Airsoft toy “looks\\nlike a gun in every way, shape or form from the outside, same\\nlength, weight, contour, field markings, but it won’t kill\\nanybody.” S.A. 672-73. They have all the same parts as the\\nreal guns they mimic; in fact, real gun parts can be used with\\nAirsoft toys. App. 508-10; S.A. 676, 678. In his statement to\\nDHS agents, Burden said that his Thai customers were not\\nusing gun parts “for the gun,” but “for the BB gun . . . for the\\npaintball [gun].” App. 259. A Wing-On manager testified that\\nemployees were instructed not to ship parts for real guns but\\nthat they could ship parts to “be used for toys for Airsoft items.”\\n\\n 11\\nApp. 554-56, 574. Burden affirmed to one of his customers\\nthat if a part was for a BB gun “then there’s no problem”\\nshipping it. App. 361. That evidence tended to support\\nBurden’s defense that he concealed the contents of shipments\\nto evade Thai customs law.\\n\\n Evidence also showed that Burden tried not to ship real\\ngun parts after realizing it was illegal. Burden sent an email to\\nYindeear-Rom saying, “I have warned you many times that I\\ndo not accept gun parts. . . . Stop sending them to me\\nabsolutely!” App. 358. The Wing-On warehouse had a “no-\\ngo” shelf for gun parts, where employees would segregate\\nitems that they could not lawfully ship. See App. 350, 554-55.\\nThere was ambiguous evidence suggesting that Wing-On may\\nhave ultimately shipped some gun parts on the no-go shelf.\\n\\n The conflicting evidence regarding Burden’s intent\\noccasioned a dispute over the jury instruction defining\\n“willfully” under the AECA. The defendants proposed using\\nthe Fifth Circuit pattern jury instruction, which requires the\\njury to find that a defendant exported articles on the Munitions\\nList without obtaining a license from the Department of State;\\nand “[t]hat the defendant acted ‘willfully,’ that is, that the\\ndefendant knew such license . . . was required for the export of\\nthese articles and intended to violate the law by exporting them\\nwithout such license.” App. 99.\\n\\n Instead, the district court adopted the government’s\\nproposed instruction, which described the requisite state of\\nmind as follows:\\n\\n [A]n act is done willfully if it is committed with the\\n knowledge that it was prohibited by law or was\\n done in disregard of a known legal obligation. The\\n government must prove that a defendant acted with\\n knowledge that the conduct was unlawful. While\\n\\n 12\\n the government must show that a defendant knew\\n that the conduct was unlawful, it is not necessary\\n for the Government to show that a defendant was\\n aware of the specific law, rule, or regulation that\\n the conduct may have violated.\\n\\n In other words, the government need not prove that\\n a defendant had read, was aware of, or had\\n consulted the licensing provisions of the Arms\\n Export Control Act or the International Traffic in\\n Arms Regulations, or the Munitions List. The\\n government, however, must prove beyond a\\n reasonable doubt, by reference to facts and\\n circumstances surrounding the case, that a\\n defendant knew that the conduct was unlawful.\\n\\nApp. 66; see App. 435-37. The jury convicted the\\ndefendants on all three counts—conspiracy to violate the\\nAECA, unlawful export in violation of the AECA, and\\nconspiracy to launder money. The defendants timely appealed\\nthe district court’s admission of Yindeear-Rom’s deposition\\nand Burden’s statement to DHS, as well as the district court’s\\njury instruction on the definition of “willfully.”\\n\\n ANALYSIS\\n\\n A. Yindeear-Rom Was Not “Unavailable” for\\n Purposes of the Confrontation Clause.\\n\\n We review legal conclusions regarding the Confrontation\\nClause de novo, and reverse any error unless it was harmless\\nbeyond a reasonable doubt. United States v. Moore, 651 F.3d\\n30, 69 (D.C. Cir. 2011) (citing constitutional harmless-error\\nstandard in Chapman v. California, 386 U.S. 18, 24 (1967)).\\nBecause the government concedes that, if it was error,\\nadmitting Yindeear-Rom’s video deposition testimony instead\\n\\n 13\\nof producing him at trial was not harmless, we need only\\nconsider the claimed error itself. See Appellee’s Br. 55 n.5.\\n\\n The Confrontation Clause of the Sixth Amendment\\nguarantees that, “[i]n all criminal prosecutions, the accused\\nshall enjoy the right . . . to be confronted with the witnesses\\nagainst him.” U.S. Const. amend. VI. Courts have long\\nrecognized the critical importance of a criminal defendant’s\\n“opportunity to cross-examine and impeach a witness at trial\\nbefore the jury that will decide his innocence or guilt.” United\\nStates v. Lynch, 499 F.2d 1011, 1022 (D.C. Cir. 1974). Live\\nwitness testimony gives the jury the opportunity to assess the\\nwitness’s demeanor and allows counsel to adjust examination\\nto other evidence and to the jury’s apparent reactions as the\\nwitness testifies. “William Blackstone long ago recognized\\nthis virtue of the right to confrontation, stressing that through\\nlive testimony, ‘and this [procedure] only, the persons who are\\nto decide upon the evidence have an opportunity of observing\\nthe quality, age, education, understanding, behavior, and\\ninclinations of the witness.’” United States v. Yida, 498 F.3d\\n945, 950 (9th Cir. 2007) (alteration in original) (quoting 3\\nWilliam Blackstone, Commentaries on the Laws of England\\n373-74 (1768)).\\n\\n A testimonial statement by a person who does not appear\\nat trial may be admitted “only where the declarant is\\nunavailable, and only where the defendant has had a prior\\nopportunity to cross-examine.” Crawford v. Washington, 541\\nU.S. 36, 59 (2004); see also Fed. R. Evid. 804(b)(1).\\nUnavailability and prior opportunity to cross-examine are\\nindependent criteria: Even when a defendant had a prior\\nopportunity to cross-examine the witness, if the government\\ndoes not establish that the witness is unavailable, the testimony\\nmust be excluded. See id. at 57. Here, the parties agree that\\ndefendants had an opportunity to cross-examine Yindeear-\\n\\n 14\\nRom during his videotaped deposition. The sole question is\\nwhether he was “unavailable” for purposes of the\\nConfrontation Clause.\\n\\n Defendants argue that Yindeear-Rom was not truly\\nunavailable because the government procured Yindeear-Rom’s\\nunavailability by deporting him, and, in any case, it did not\\nmake reasonable efforts before it deported him to ensure his\\npresence at trial. For its part, the government contends that\\nthere is no per se presumption that the government fails the\\nConfrontation Clause’s test of good-faith and reasonable\\nefforts when it deports a witness, and that its reasonable, good-\\nfaith efforts after it deported Yindeear-Rom sufficed.\\n\\n As a general matter, a witness is considered unavailable\\nonly if the prosecution cannot procure her with good-faith,\\nreasonable efforts. See Ohio v. Roberts, 448 U.S. 56, 74\\n(1980), abrogated in part on other grounds by Crawford, 541\\nU.S. 36. It is the prosecution’s burden to establish that its\\nactions meet that test. Id. at 74-75. “The law does not require\\nthe doing of a futile act” such as producing a witness who has\\ndied, but “if there is a possibility, albeit remote, that affirmative\\nmeasures might produce the declarant, the obligation of good\\nfaith may demand their effectuation.” Id. at 74. “The lengths\\nto which the prosecution must go to produce a witness . . . is a\\nquestion of reasonableness.” Id. (alteration in original)\\n(quoting California v. Green, 399 U.S. 149, 189 n.22 (1970)\\n(Harlan, J., concurring)). In Lynch, we held good-faith but\\nnon-exhaustive efforts to find a witness were inadequate to\\nrender her unavailable and justify admitting pretrial hearing\\ntestimony. 499 F.2d at 1024. We treated the witness as\\navailable because she was within the court’s jurisdiction and\\nthe government had not “inquired at the local hospitals, area\\npolice departments, the morgue, or of [the witness’s]\\nemployer.” Id. In contrast, a witness whose “whereabouts\\n\\n 15\\nwere entirely unknown,” who had been sent numerous\\nsubpoenas at her parents’ house, was last heard from outside\\nthe jurisdiction, and was unreachable by her family even in\\ncase of an emergency was held in Ohio v. Roberts to be\\nunavailable even though, as Justice Brennan pointed out in\\ndissent, the government had not followed every possible lead\\nto find her. Compare 448 U.S. at 60-61, 75, with id. at 79-82\\n(Brennan, J., dissenting).\\n\\n When the government seeks to rely on prior recorded\\nstatements of a witness on the ground that the witness is\\nunavailable, it bears the burden of establishing that its\\nunsuccessful efforts to procure the witness’s appearance at trial\\nwere “as vigorous as that which the government would\\nundertake to [secure] a critical witness if it has no [prior]\\ntestimony to rely upon in the event of ‘unavailability.’” Lynch,\\n499 F.2d at 1023; see also United States v. Mann, 590 F.2d 361,\\n367 (1st Cir. 1978). Where the government itself bears some\\nresponsibility for the difficulty of procuring the witness, such\\nas by deporting the witness, the government will have to make\\ngreater exertions to satisfy the standard of good-faith and\\nreasonable efforts than it would have if it had not played any\\nrole. Failing to factor the government’s own contribution to\\nthe witness’s absence into the Confrontation Clause analysis\\nwould warp the government’s incentives. “This relatively high\\ngood faith standard cannot be satisfied by perfunctory efforts,\\nif the rule is not to sanction the government’s procuring\\ndepositions of witnesses, especially shaky witnesses, but then\\ndiscourage attempts to bring the witness to trial so long as the\\ngovernment is satisfied with” the witness’s recorded testimony.\\nMann, 590 F.2d at 367. Rather, the analysis of good-faith,\\nreasonable efforts should “include an assessment of the\\ngovernment’s affirmative conduct” that allowed the witness to\\nbecome unavailable “in the first instance.” Yida, 498 F.3d at\\n955-56. That analysis should account for the good faith and\\n\\n 16\\nreasonableness of the government’s deportation of the witness\\nand of its attempts to secure witnesses it does deport.\\n\\n In a case such as this one, in which the government knew\\nor should have known of the potential need for the witness’s\\ntestimony before he was deported, the government’s duty to\\nmake good-faith, reasonable efforts to ensure the witness’s\\npresence arises before the witness leaves the United States.\\nOther courts that have addressed this question in the context of\\nwitness deportation agree. In United States v. Tirado-Tirado,\\nthe Fifth Circuit refused to deem unavailable a witness whom\\nthe government had deported where the government “failed to\\nmake any concrete arrangements with [the witness] prior to his\\ndeportation,” and only shortly before trial made what the court\\nacknowledged were “fairly exhaustive” efforts to bring him\\nback from Mexico. 563 F.3d 117, 123, 125 (5th Cir. 2009).\\nThe government had orally told the witness before deporting\\nhim that his appearance would be required if the case went to\\ntrial—a step not taken here—but the court faulted it for not\\ntaking several other pre-deportation steps that might have\\nencouraged the witness to appear, including serving him with\\na subpoena or other written notice. Id. at 123-24. In United\\nStates v. Foster, the court followed Tirado-Tirado to exclude\\nvideotaped depositions notwithstanding that the government\\nhad advised the witnesses before deporting them that it might\\nneed them to return, promised to allow their reentry and pay for\\ntheir travel, and collected contact information for them in\\nMexico. 753 F. App’x 307, 309 (5th Cir. 2018). Emphasizing\\nthe constitutional importance of taking “reasonable measures,\\nunder the circumstances, that are likely to ensure that the\\nwitness will return for trial,” the court declined to find the\\nunreachable witnesses “unavailable” because the government\\nhad not verified their contact information from the outset,\\nsought alternative contact information, or remained in contact\\n\\n 17\\nwith them over the months between their deportation and the\\ntrial. Id. at 312.\\n\\n The Ninth Circuit in Yida similarly recognized “the\\ngovernment’s obligation to use ‘reasonable means’ to ‘procure\\nthe declarant’s testimony’ in the context of the government’s\\naffirmative role in a witness’s deportation,” 498 F.3d at 953,\\nand refused to admit testimony from a prior mistrial “when the\\ngovernment itself share[d] some of the responsibility for its\\ninability to produce the witness at [the second] trial,” id. at 956.\\nThe government’s efforts were insufficient where it had chosen\\nto detain the admittedly untrustworthy and skittish witness\\nprior to the first trial but, once it had that transcript in hand as\\na substitute, deported the witness without considering the many\\nalternatives to confinement in federal prison that might have\\nprevented the witness’s absence. Id. at 959-60.\\n\\n The First Circuit, too, in Mann, excluded deposition\\ntestimony of a crucial prosecution witness in lieu of live\\ntestimony where, after the deposition, the government had\\nreturned the witness’s passport and plane ticket that it had\\nseized upon her arrest. 590 F.2d at 366, 368. The court held\\ninsufficient the government’s later offer to pay to bring the\\nwitness back, similar to the government’s offer here. Id. at\\n363. “Implicit . . . in the duty to use reasonable means to\\nprocure the presence of an absent witness is the duty to use\\nreasonable means to prevent a present witness from becoming\\nabsent.” Id. at 368. Where the government fails to take\\nreasonable steps to prevent a witness from becoming absent,\\nthe defendant should not suffer from the government’s choice.\\nSee id.\\n\\n The cases on which the government relies are not to the\\ncontrary. Foster distinguishes United States v. Allie, 978 F.2d\\n1401, 1403-08 (5th Cir. 1992), and United States v. Calderon-\\n\\n 18\\nLopez, 268 F. App’x 279, 289 (5th Cir. 2008), on the basis of\\nextensive efforts the government made before it deported the\\nwitnesses in those cases, none of which were made here.\\nFoster, 753 F. App’x at 311. And in United States v. Eufracio-\\nTorres, it was the defendant who initially opposed the\\ngovernment’s motion to detain seven material witnesses\\npending trial. 890 F.2d 266, 268 (10th Cir. 1989). The court\\nthought it significant that none of those witnesses had been\\ncharged with any crime, yet they were detained while\\ndefendant Eufracio-Torres was free on bond. Id. Eufracio-\\nTorres changed his position to request their detention pending\\ntrial only after their depositions were taken—a request the\\ncourt denied. Id. When the witnesses could not be brought\\nback for trial and the government sought to use the depositions,\\nthe court held that the government’s good-faith, reasonable\\nefforts supported its request: Before they left the country, the\\ngovernment had given the witnesses subpoenas, instructions\\nfor reentry, travel reimbursement, appearance fees, and\\nobtained their promises to return to testify. Id. No such steps\\nwere taken here. The government’s reliance on United States\\nv. Rivera is similarly misplaced. 859 F.2d 1204, 1207-09 (4th\\nCir. 1988) (witnesses but not defendant were being detained,\\nand court found no prejudice claimed or suffered from use of\\nwitness depositions taken at the witnesses’ own request before\\nthey were released and voluntarily left the United States).\\n\\n Under the applicable standard, the government failed to\\nshow that Yindeear-Rom was “unavailable” for purposes of the\\nConfrontation Clause. The government’s efforts to secure his\\npresence at trial did not begin until after he was deported. See\\nBurden, 2016 WL 5108010, at *3-4. Before his deportation,\\nthe government did not give Yindeear-Rom a subpoena, offer\\nto permit and pay for him either to remain in the U.S. or to\\nreturn here from Thailand, obtain his commitment to appear,\\nconfirm his contact information, or take any other measures.\\n\\n 19\\nSee id. at *9. Its only efforts began once he was out of custody,\\nout of the jurisdiction, and no longer dependent on the\\ngovernment’s good graces for lenient treatment. Yindeear-\\nRom’s eagerness to return to Thailand helped to persuade the\\ndistrict court that further efforts to persuade him to testify at\\ntrial would have been futile. See id. But in these circumstances\\nthat eagerness cuts the other way. Given the government’s\\nduty to make good-faith, reasonable efforts before Yindeear-\\nRom’s deportation, “a witness’s known reluctance to testify\\nadds to the government’s burden to show that it made\\n‘reasonable, good faith efforts’ to secure her appearance\\nbecause it makes her failure to appear voluntarily all the more\\nforeseeable.” Brooks v. United States, 39 A.3d 873, 886 (D.C.\\n2012). This is a case where the “possibility, albeit remote, that\\naffirmative measures might produce the declarant . . .\\ndemand[ed] their effectuation.” Roberts, 448 U.S. at 74. Any\\nchance the government had of securing Yindeear-Rom’s\\nappearance at trial would have been far greater had it addressed\\nthe problem as soon as it knew it would rely on his testimony.\\nInstead, its own approach appears to have ensured the futility\\nof the post-deportation efforts.\\n\\n We recognize that it may not always be reasonable to\\nexpect the government to postpone removal until trial—\\nparticularly if the government would have to detain the witness\\nin order to keep her in the country. See Aguilar-Ayala v. Ruiz,\\n973 F.2d 411, 419 (5th Cir. 1992) (enjoining Department of\\nJustice policy of automatically detaining undocumented\\nimmigrant material witnesses because “undocumented aliens\\nhave an overriding liberty interest in not being detained as\\nmaterial witnesses”). And we assume that “in some cases the\\nneed for the criminal defendant to confront the witness at trial\\n(rather than at deposition) might outweigh the material witness’\\nliberty interest in being released immediately,” whereas in\\nother cases it will not. Id. (emphasis omitted). In order to\\n\\n 20\\nidentify the existence and extent of any such conflict, however,\\nthe unavailability inquiry must account for whether the\\ngovernment has addressed potentially reasonable means of\\nsecuring removable witnesses’ live testimony short of\\ndetention, such as placing them in “lesser custody,” “supplying\\nmaintenance, and retaining [their] passport[s] and ticket[s],”\\nand “plac[ing] [them] under subpoena.” Mann, 590 F.2d at\\n366; see also Allie, 978 F.2d at 1407 (holding that the\\ngovernment’s efforts were reasonable where it offered\\nwitnesses work permits to keep them in the United States).\\n\\n Because the government’s omissions place its efforts\\nbelow the standard the Confrontation Clause demands, we need\\nnot decide precisely how the government should have sought\\nto prevent the witness from becoming unavailable. While the\\ngovernment’s deportation of a witness may sometimes fail the\\nstandard of good-faith and reasonable efforts, we reject any per\\nse rule that no witness the government deports can be\\nconsidered unavailable under the Confrontation Clause.\\nConsistent with the fact-intensive nature of the standard, the\\ngovernment decries any per se rule, see Appellee’s Br. at 45-\\n46, the defendants do not advocate one, see Appellants’ Reply\\nBr. 24 n.12, and no circuit has adopted any such categorical\\napproach.\\n\\n We hold that the duty to use reasonable means to procure\\na witness’s presence at trial includes the duty to use reasonable\\nefforts to prevent a witness from becoming absent in the first\\nplace. The government does not dispute that it made no efforts\\nbefore deporting Yindeear-Rom to secure his presence at trial.\\nThe witness thus was not “unavailable” such that prior\\ntestimony could be admitted consistent with the Confrontation\\nClause. Because admitting his deposition was not harmless\\nbeyond a reasonable doubt, we vacate the convictions and\\nremand for a new trial.\\n\\n 21\\n B. Guidance Regarding the Jury Instruction’s\\n Definition of “Willfully”\\n\\n Although our resolution of the Confrontation Clause issue\\nis sufficient to dispose of this appeal, we provide some\\nguidance regarding the jury instruction’s definition of the term\\n“willfully,” an issue that was fully briefed and argued.\\n\\n The parties agree that defendants “willfully” violate the\\nAECA only where they act with knowledge that their conduct\\nis unlawful. The Supreme Court has explained that “willfully”\\nis “a word of many meanings whose construction is often\\ndependent on the context in which it appears.” Bryan v. United\\nStates, 524 U.S. 184, 191 (1998) (internal quotation marks\\nomitted). “As a general matter, when used in the criminal\\ncontext, a ‘willful’ act is one undertaken with a bad purpose.\\nIn other words, in order to establish a ‘willful’ violation of a\\nstatute, the Government must prove that the defendant acted\\nwith knowledge that his conduct was unlawful.” Id. at 191-92\\n(internal quotation marks omitted). The parties disagree about\\nhow specific that knowledge must be under the AECA’s\\nwillfulness standard.\\n\\n Defendants argue that the AECA’s prohibition on\\n“willfully violat[ing] any provision of this section,” 22 U.S.C.\\n§ 2778(c), means that they should not have been convicted\\nunless they “were aware of the specific law, rule, or regulation\\nthat the[ir] conduct may have violated.” Appellants’ Br. 17\\n(alteration in original). Before us, they frame this standard as\\n“requir[ing] the government to prove an individual was aware\\nthe items he exported were on the [Munitions List],” id. at 18-\\n19, though their requested jury instruction would instead have\\ndemanded proof that they knew that there was a license\\nrequirement for their exports, App. 99. The heart of\\ndefendants’ case at trial was that they “had not willfully\\n\\n 22\\nviolated [the] AECA because they believed the parts were\\nintended for use in Airsoft BB guns, not real guns, and they\\nbelieved such parts could be exported without a license.”\\nAppellants’ Br. 12. Under the court’s jury instructions, they\\ncontend, “intent to violate Thai customs duties or Thai gun\\ncontrol laws could support the jury’s finding of willfulness.”\\nId. at 31. The government defends the district court’s\\ninstruction on the ground that “the AECA is not one of the\\nextremely rare, highly technical statutes that the Supreme\\nCourt has found to require a heightened willfulness showing,”\\nso the government did not have to prove defendants had\\nknowledge of the license requirement or the Munitions List as\\nsuch. Appellee Br. 16.\\n\\n As discussed below, because it was ambiguous as to what\\n“conduct” defendants had to know was “unlawful,” the district\\ncourt’s jury instruction arguably fell short of the baseline\\nrequirement that the mens rea relate to the charged actus reus.\\nBut we believe that the district court correctly instructed that,\\nif defendants knew exporting the charged items without a\\nlicense was unlawful, they did not need specific knowledge of\\nthe Munitions List. Thus, the district court was right that “the\\ngovernment need not prove that a defendant had read, was\\naware of, or had consulted the licensing provisions of the Arms\\nExport Control Act or the International Traffic in Arms\\nRegulations, or the Munitions List.” App. 66. For purposes of\\nthe AECA, a requirement of proof that defendants knew the\\nproscribed conduct was unlawful adequately protects against\\n“the danger of ensnaring individuals engaged in apparently\\ninnocent conduct.” Bryan, 524 U.S. at 194.\\n\\n Most criminal prohibitions require only proof that the\\ncrime was committed “knowingly,” meaning that the defendant\\nknew of the facts that made his act illegal, even if he did not\\nknow the act was illegal. When Congress wants to ensure that\\n\\n 23\\ndefendants will be convicted only if they have a more culpable\\nstate of mind, it limits the crime to conduct that a defendant\\nengages in “willfully.” See id. at 191-92. The Court has\\ndeveloped an approach to “willfulness” that is calibrated to the\\nstatutes in which it appears, consistently reading it to require\\nproof that the defendant had a sufficiently culpable state of\\nmind, but not to require proof so specific as to stymie the\\nstatute’s enforcement. It has required proof that a defendant\\nknow which law he was breaking in only two contexts:\\ncriminal tax evasion, and currency structuring.\\n\\n In Cheek v. United States, a tax case, the Court explained\\nCongress’s inclusion of a willfulness requirement for felony\\ntax-evasion as resting partly on the recognition that “[t]he\\nproliferation of statutes and regulations has sometimes made it\\ndifficult for the average citizen to know and comprehend the\\nextent of the duties and obligations imposed by the tax laws.”\\n498 U.S. 192, 199-200 (1991). The Court held that the\\nwillfulness requirement could be defeated by the defendant’s\\ngood-faith belief that he “was not a person required to file a\\nreturn or to pay income taxes and that wages are not taxable\\nincome,” even if that belief was unreasonable. Id. at 203.\\n\\n In Ratzlaf v. United States, the Court interpreted\\n“willfully” in the context of a statute criminalizing “currency\\nstructuring.” 510 U.S. 135, 136, 144 (1994). The anti-\\nstructuring law prohibited a customer from breaking up a large\\nfinancial transaction into multiple smaller ones to avoid\\ntriggering a bank’s legal duty to report transactions over\\n$10,000. Id. at 145. The Court held that “willfulness” meant\\nthat the defendant had to be aware not only of reporting\\nobligations applicable to banks, but of his own distinct\\nobligation not to structure his transactions for the purpose of\\navoiding triggering the bank’s obligation to report. Id. at 137-\\n38. In other words, “[t]o convict Ratzlaf of the crime with\\n\\n 24\\nwhich he was charged . . . the jury had to find he knew the\\nstructuring in which he was engaged was unlawful.” Id. at 149.\\n\\n In Bryan, which concerned unlicensed firearms-dealing,\\nthe Court concluded that “requiring only knowledge that the\\nconduct is unlawful [was] fully consistent” with Congress’s\\npurpose in adding a willfulness requirement to the firearms\\nstatute “to protect law-abiding citizens who might\\ninadvertently violate the law.” 524 U.S. at 195 n.23. Bryan\\ndistinguished the firearms law before it from the “highly\\ntechnical” statutes at issue in Cheek and Ratzlaf, which had\\njustified “‘carv[ing] out an exception to the traditional rule’\\nthat ignorance of the law is no excuse and requir[ing] that the\\ndefendant have knowledge of the law”—that is, the specific\\nlaw he violated. Bryan, 524 U.S. at 194-95 (quoting Cheek,\\n498 U.S. at 200).\\n\\n Following Bryan, most courts of appeals to consider the\\nissue have interpreted the AECA’s requirement that a violation\\nbe “willful” not to require specific awareness of the Munitions\\nList as such. What they have required—as we do here—is\\nproof that defendants knew it was illegal to export the items\\nthey shipped without a license. See United States v. Henry, 888\\nF.3d 589, 598-600 (2d Cir. 2018) (“[W]illfulness requires only\\nthat the defendant know that what he was doing was illegal,\\nand not that he know that his conduct was prohibited under a\\nspecific AECA provision or related regulation.”); United States\\nv. Bishop, 740 F.3d 927, 932-33 (4th Cir. 2014) (government\\nneed not show defendant knew “that the ammunition was\\nspecifically covered by the AECA,” but only that he had\\n“general knowledge of an export’s illegality”); United States v.\\nChi Mak, 683 F.3d 1126, 1138 (9th Cir. 2012) (affirming\\ninstruction that conviction did not require proof that “the\\ndefendant had read, was aware of, or had consulted the specific\\nregulations governing his activities”); United States v. Roth,\\n\\n 25\\n628 F.3d 827, 835 (6th Cir. 2011) (the AECA “does not require\\na defendant to know that the items being exported are on the\\nMunitions List,” but only “knowledge that the underlying\\naction is unlawful”); United States v. Tsai, 954 F.2d 155, 160-\\n62 & n.3 (3d Cir. 1992) (jury need not find that defendant\\n“knew all of the specifics of the law,” but “what it had to find\\nwas that that the defendant knew that he could not export that\\nparticular item”); United States v. Murphy, 852 F.2d 1, 6-7 (1st\\nCir. 1988) (requiring proof that “defendant must know that his\\nconduct in exporting from the United States articles proscribed\\nby the statute is violative of the law,” but rejecting requirement\\nthat he must know specifically that the article was on the\\nMunitions List or that there was a licensing requirement).\\n\\n The Eleventh Circuit in United States v. Wenxia Man\\ncharacterized the willfulness requirement of the AECA as more\\ndemanding than that of the statute in Bryan, but its bottom line\\nis not materially different from that of the other circuits. See\\n891 F.3d 1253, 1268 (11th Cir. 2018). That court hewed to its\\npre-Bryan precedent that “aware[ness] of the generally\\nunlawful nature of [one’s] actions is insufficient,” but was\\nsatisfied with proof “that a conspirator actually knew that it was\\nunlawful to export the [items] and intentionally violated [the]\\nknown legal duty not to export [them].” Id. (internal quotation\\nmarks omitted) (alterations in original) (quoting United States\\nv. Adames, 878 F.2d 1374, 1377 (11th Cir. 1989)). The Fifth\\nCircuit held in United States v. Hernandez that the AECA\\ndemands a finding that the defendant “knew he was unlawfully\\nexporting weapons on the Munitions List,” but that case\\npredated Bryan, Cheek, and Ratzlaf. 662 F.2d 289, 292 (5th\\nCir. 1981). The Seventh Circuit did so more recently, but it,\\ntoo, took no account of those Supreme Court cases. See United\\nStates v. Dobek, 789 F.3d 698, 700 (7th Cir. 2015).\\n\\n 26\\n We hold that the district court’s jury instruction was\\ncorrect insofar as it rejected defendants’ position that the\\nwillfulness requirement in the AECA is akin to those at issue\\nin Cheek and Ratzlaf, and clarified that the government need\\nnot prove defendants knew the specific law their conduct\\nviolated.\\n\\n That does not quite resolve the issue, however, because of\\nambiguity in the meaning of “conduct” in the instruction that a\\nwillful violation of the AECA requires that the defendant\\n“acted with knowledge that his conduct was unlawful.” App.\\n66. More is required where, as here, there is evidence of\\nwillfully unlawful conduct apart from the charged offenses,\\ncreating a risk that the jury may consider any and all evidence\\nof the defendant’s guilty mind—whatever its object—as\\nsupporting willful commission of the charged offense. The\\nwillfulness instruction arguably fell short in not specifying that\\nthe unlawful “conduct” the jury must find the defendants to\\nhave willfully done was the actus reus that violated the AECA:\\nunlawfully exporting the magazines and mount without a\\nlicense. To be sure, the rest of the instruction made clear that\\nthe actus reus of the charged offense was exporting the items\\nwithout a license, and we review jury instructions as a whole.\\nUnited States v. McGill, 815 F.3d 846, 888 (D.C. Cir. 2016).\\nBut we nonetheless detect a potential problem specific to the\\ninstruction’s discussion of willfulness—an issue that is both of\\ncentral importance and easily muddied. It is natural enough\\nthat jurors who are told they may convict upon finding the\\ndefendants did the requisite act, so long as the defendant had a\\nguilty mind, may not parse the object of the guilty mind. To\\nprompt that important precision, the instruction’s definition of\\n“willfully”—the word at issue here—should specify that the\\nactus reus was the only “conduct” to which defendants’ guilty\\nmental state could apply for a finding of willfulness.\\n\\n 27\\n The requirement that the mens rea relate to the charged\\nactus reus is the baseline for any criminal mental standard. See\\nDixon v. United States, 548 U.S. 1, 6 (2006) (“[C]riminal\\nliability is normally based upon the concurrence of two factors,\\nan evil-meaning mind [and] an evil-doing hand.” (second\\nalteration in original) (internal quotation marks omitted)). If a\\nstatute requires that the defendant knowingly perform the act\\nthat violates the law—even where he need not also know that\\nthe act is illegal—then he must knowingly perform the charged\\nactus reus, not some other, uncharged act. See Bryan, 524 U.S.\\nat 193 (“[T]he term ‘knowingly’ merely requires proof of\\nknowledge of the facts that constitute the offense.” (emphasis\\nadded)). So, too, with willfulness. The district court and the\\nparties agree that willfulness requires that the defendant “acted\\nwith knowledge that his conduct was unlawful.” App. 66; see\\nBryan, 524 U.S. at 193. The conduct that he must know was\\nunlawful is the actus reus of the crime with which he is\\ncharged.\\n\\n Because the willfulness instruction required only that the\\ndefendants acted with knowledge that “the conduct” was\\nunlawful, App. 66, there is some chance that the jury convicted\\nbased in part on defendants’ evasiveness in importing to\\nThailand. On retrial, the instruction should make clear that an\\nAECA conviction requires that defendants knew of the\\nunlawfulness of the charged unlicensed export of the items\\nfrom the United States, and that a willfulness finding cannot\\ndraw on evidence that they knew the related, but legally and\\nfactually distinct, import of those items into Thailand was\\nillegal.\\n\\n 28\\n C. Burden’s Limited English Did Not Render His\\n Interrogation Custodial for Miranda Purposes.\\n\\n We review de novo the determination whether Burden was\\nin custody and thus entitled to Miranda warnings before any\\ninterrogation, and we review the underlying factual findings for\\nclear error. United States v. Hallford, 756 F. App’x 1, 5 (D.C.\\nCir. 2018). If the court erroneously admits a non-Mirandized\\nstatement, we must reverse unless the admission was harmless\\nbeyond a reasonable doubt. United States v. Brinson-Scott, 714\\nF.3d 616, 622 (D.C. Cir. 2013).\\n\\n In Miranda, the Court held that “the prosecution may not\\nuse statements, whether exculpatory or inculpatory, stemming\\nfrom custodial interrogation of the defendant unless it\\ndemonstrates the use of procedural safeguards effective to\\nsecure the privilege against self-incrimination.” 384 U.S. at\\n444. Defendants argue that Burden’s limited English abilities\\nrendered the interview “custodial.” We disagree. We see no\\nerror in the district court’s admission of Burden’s non-\\nMirandized statement because Burden was not in custody when\\nthe agents questioned him at his warehouse. His English\\nproficiency was not an “objective circumstance surrounding\\nthe interrogation” that a reasonable officer would have\\nassumed bore on whether Burden felt free to leave the\\ninterview.\\n\\n Miranda warnings are required as a bulwark against the\\ncoercive power of being taken into police custody and\\ninterrogated. “An individual swept from familiar surroundings\\ninto police custody, surrounded by antagonistic forces, and\\nsubjected to . . . techniques of persuasion . . . cannot be\\notherwise than under compulsion to speak.” Miranda, 384\\nU.S. at 461. Police custody is sometimes self-evident, as in\\nMiranda’s own case: The police arrested him and took him to\\n\\n 29\\na police station interrogation room where they questioned him.\\nId. at 491. But a person is also in custody when he is\\n“otherwise deprived of his freedom of action in any significant\\nway.” Id. at 444. In “Miranda case law, ‘custody’ is a term of\\nart that specifies circumstances that are thought generally to\\npresent a serious danger of coercion.” Howes v. Fields, 565\\nU.S. 499, 508 (2012).\\n\\n The Supreme Court has laid out guidelines for the custody\\nanalysis. “In determining whether a person is in custody,”\\ntriggering the duty to give Miranda warnings,\\n\\n the initial step is to ascertain whether, in light of\\n the objective circumstances of the interrogation, a\\n reasonable person [would] have felt he or she was\\n not at liberty to terminate the interrogation and\\n leave. And in order to determine how a suspect\\n would have gauge[d] his freedom of movement,\\n courts must examine all of the circumstances\\n surrounding the interrogation. Relevant factors\\n include the location of the questioning, its duration,\\n statements made during the interview, the presence\\n or absence of physical restraints during the\\n questioning, and the release of the interviewee at\\n the end of the questioning.\\n\\nId. at 509 (alterations in original) (internal quotation marks and\\ncitations omitted).\\n\\n “[W]hether a suspect is ‘in custody’ is an objective\\ninquiry.” J.D.B. v. North Carolina, 564 U.S. 261, 270 (2011).\\nThat means that, while the analysis accounts for “any\\ncircumstance that would have affected how a reasonable person\\nin the suspect’s position would perceive his or her freedom to\\nleave,” it “involves no consideration of the actual mindset of\\nthe particular suspect subjected to police questioning.” Id. at\\n\\n 30\\n271 (internal quotation marks omitted). The benefit of the\\nobjective test is practical: It “avoids burdening police with the\\ntask of anticipating the idiosyncrasies of every individual\\nsuspect and divining how those particular traits affect each\\nperson’s subjective state of mind.” Id.\\n\\n The test, while objective, is also contextual. Because all\\nof the “objective circumstances of the interrogation” must be\\nconsidered, Howes, 565 U.S. at 508, individual characteristics\\nthat have an “objectively discernible relationship to a\\nreasonable person’s understanding of his freedom of action,”\\nsuch as a child suspect’s age, J.D.B., 564 U.S. at 275, must be\\ntaken into account. Under circumstances in which “a\\nreasonable child subjected to police questioning will . . . feel\\npressured to submit when a reasonable adult would feel free to\\ngo[,] . . . courts can account for that reality without doing any\\ndamage to the objective nature of the custody analysis.” Id. at\\n272. That is because age “generates commonsense conclusions\\nabout behavior and perception” that “apply broadly to children\\nas a class.” Id. A reasonable eight-year-old will not necessarily\\nfeel free to leave when a reasonable adult would.\\n\\n Because the test is designed to guide police, a person’s\\nyouth—or analogous circumstances bearing on a reasonable\\nperson’s perception of her freedom to leave—only factors into\\nthe custody analysis where it “was known to the officer at the\\ntime of the interview, or would have been objectively apparent\\nto any reasonable officer.” Id. at 274. The test thus includes a\\ndouble inquiry: whether it would have been apparent to a\\nreasonable officer that a reasonable person in the suspect’s\\nposition would not have felt free to leave. In holding that the\\nsuspect’s youth was an objective circumstance relevant to the\\ncustody analysis, the Supreme Court suggested that other\\npersonal characteristics, like blindness, could be similarly\\nrelevant. Id. at 278. The Court underscored: “Not once have\\n\\n 31\\nwe excluded from the custody analysis a circumstance that we\\ndetermined was relevant and objective, simply to make the\\nfault line between custodial and noncustodial ‘brighter.’” Id.\\nat 280.\\n\\n Under some circumstances, English language capabilities\\nmight have an “objectively discernible relationship to a\\nreasonable person’s understanding of his freedom of action”\\nthat would bear on the custody analysis for purposes of\\nMiranda. Id. at 275. Some courts have accordingly factored\\nlimited English abilities into the custody inquiry. The Eighth\\nCircuit explained that “the ultimate issue is whether a\\nreasonable police officer conducting [an] otherwise non-\\ncustodial interview would have given Miranda warnings\\nbecause he realized that the questioning would be perceived by\\n[the defendant] as custodial due to his limited English language\\nskills.” Thatsaphone v. Weber, 137 F.3d 1041, 1045 (8th Cir.\\n1998). In United States v. Kim, the Ninth Circuit weighed\\nlimited English proficiency (among other factors the court\\ndeemed important) because, under the circumstances, it bore\\non the defendant’s ability to understand whether she was a\\ncriminal suspect. 292 F.3d 969, 977 (9th Cir. 2002). Kim had\\narrived at the store her family owned to find many police cars\\noutside and the door locked; the police let her, but not her\\nhusband, enter, locked the door, did not let her son (who was\\nalready inside) speak to her, and forced her to speak only\\nEnglish, though she and her son informed the officers that she\\ndid not speak English well. Id. at 971-72. The court held that\\nKim’s interrogation was custodial partly because the police\\n“temporarily took over complete control of Kim’s store,\\ncreating ‘a police-dominated atmosphere,’ in which the police\\nkept Kim physically isolated from two family members who\\ncould have provided both moral support and, given her limited\\nEnglish, a more complete understanding of the overall\\nsituation.” Id. at 977. Both Thatsaphone and Kim preceded the\\n\\n 32\\nCourt’s application of the custody analysis to a child in J.D.B.,\\nbut both decisions comport with J.D.B. insofar as they consider\\nwhether a reasonable officer would have been able to discern\\nthat the language limitations of a reasonable person in the\\nsuspect’s position would have contributed to that person’s not\\nfeeling free to leave. Since J.D.B., at least one other court has\\nheld a suspect’s limited English proficiency relevant to\\nJ.D.B.’s objective inquiry. See United States v. Han, 199 F.\\nSupp. 3d 38, 52-54 (D.D.C. 2016).\\n\\n The district court in this case correctly held that a\\nreasonable officer would not have thought that Burden’s\\nlanguage abilities prevented him from feeling free to leave, and\\nthus properly admitted Burden’s non-Mirandized statement.\\nEven though Burden sometimes had trouble formulating\\nresponses and appeared to lack perfect comprehension of all\\nthe questions, the evidence does not suggest that it would have\\nbeen apparent to a reasonable officer that Burden was not\\nunderstanding what was being said.\\n\\n The defendants identify only one moment, when the DHS\\nofficer was explaining the purpose and terms of the interview,\\nwhen they believe that a reasonable officer should have\\nrecognized that Burden’s English skills would affect the\\nperception of a reasonable person in his position as to whether\\nhe was free to leave:\\n\\n UNIDENTIFIED AGENT NO. 1: We are federal\\n agents for the U.S. Government so I have to let you\\n know that you have to be honest with us —\\n\\n PHEERAYUTH BURDEN: Uh-huh.\\n\\n UNIDENTIFIED AGENT NO. 1: — okay? If you\\n don’t want to answer something, you don’t have to\\n answer but you cannot lie to us.\\n\\n 33\\n PHEERAYUTH BURDEN: Okay.\\n\\n UNIDENTIFIED AGENT NO. 1: All right? And\\n you can’t withhold relevant information.\\n\\n PHEERAYUTH BURDEN: Uh-huh.\\n\\n UNIDENTIFIED AGENT NO. 1: If you do, that is\\n a crime.\\n\\n PHEERAYUTH BURDEN: Okay\\n\\n UNIDENTIFIED AGENT NO. 1: Okay?\\n Punishable by up to five years in prison so just\\n please be honest.\\n\\n PHEERAYUTH BURDEN: Yeah.\\n\\nApp. 174. While the agent’s statement was somewhat\\nconfusing, it is not clear from this exchange that Burden was\\nnot comprehending what the agent was saying to him or\\nsomehow believed he could not leave an interview at his own\\nwarehouse that he had agreed to by phone and shown up for of\\nhis own accord.\\n\\n The district court did not refuse to consider Burden’s\\nproficiency entirely, as defendants assert. It properly applied\\nthe custody test by evaluating how a reasonable officer would\\nhave perceived Burden’s comprehension. It noted “that for\\neach of the points that the Special Agent was communicating\\nto Mr. Burden, Mr. Burden said okay or yes or [uh-huh],\\nreflect[ing] responses like that which would give one\\nreasonably the understanding that Mr. Burden understood what\\nwas being said.” App. 430-31.\\n\\n 34\\n The custody question was not otherwise close. The district\\ncourt found that Burden “arrived with his wife at his own work\\nplace, at the time that he had set and there was no evident effort\\nto overcome his will. There was no effort to put him in\\nhandcuffs, no threats, other threats during the course of the\\ninterview.” App. 431. Burden himself “chose where he was\\ngoing to sit, he chose the room in which they were going to\\ntalk,” and he “showed that he knew he could get up, go out,\\nopen the door and talk to somebody outside throughout the\\ncourse of this.” App. 432. One of the two agents who\\nquestioned Burden testified that Burden had chosen where and\\nwhen the interview would take place, S.A. 400, and that the\\nagents wore plainclothes and did not display their badges or\\nweapons, S.A. 407. The interview lasted no more than three\\nhours and Burden left when it was over. In sum, “the location\\nof the questioning, its duration, statements made during the\\ninterview, the presence or absence of physical restraints during\\nthe questioning, . . . the release of the interviewee at the end of\\nthe questioning,” and other factors all support the conclusion\\nthat Burden was not in custody. Howes, 565 U.S. at 509\\n(internal quotation marks and citations omitted). Because\\nBurden’s imperfect English would not have given a reasonable\\nofficer the impression that a reasonable person in Burden’s\\nposition would have believed himself detained, it does not\\nchange that determination here.\\n\\n * * *\\n\\n For the foregoing reasons, we affirm the district court’s\\nadmission of Burden’s non-Mirandized statement to DHS\\nagents. We nonetheless vacate the convictions in view of the\\nerror in the admission of Yindeear-Rom’s deposition\\ntestimony.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4181,"cells":{"id":{"kind":"number","value":4649684,"string":"4,649,684"},"name":{"kind":"string","value":"CALIFORNIA COMMUNITIES AGAINST TOXICS, Et Al., Petitioners v. ENVIRONMENTAL PROTECTION AGENCY and Andrew Wheeler, Administrator, U.S. Environmental Protection Agency, Respondents Air Permitting Forum, Et Al., Intervenors"},"name_abbreviation":{"kind":"string","value":"Cal. Cmty. Against Toxics v. Envtl. Prot. Agency"},"decision_date":{"kind":"string","value":"2019-08-20"},"docket_number":{"kind":"string","value":"18-1085; C/w 18-1095, 18-1096"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Rogers, Silberman, Wilkins\", \"parties\": \"\", \"opinions\": [{\"author\": \"\", \"type\": \"020lead\", \"text\": \"\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/9889116/\", \"author_raw\": \"\"}, {\"author\": \"\", \"type\": \"040dissent\", \"text\": \"\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/9889117/\", \"author_raw\": \"\"}, {\"author\": \"Opinion for the Court filed by Circuit Judge WILKINS\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 1, 2019 Decided August 20, 2019\\n\\n No. 18-1085\\n\\n CALIFORNIA COMMUNITIES AGAINST TOXICS, ET AL.,\\n PETITIONERS\\n\\n v.\\n\\n ENVIRONMENTAL PROTECTION AGENCY AND ANDREW\\n WHEELER, ADMINISTRATOR, U.S. ENVIRONMENTAL\\n PROTECTION AGENCY,\\n RESPONDENTS\\n\\n AIR PERMITTING FORUM, ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 18-1095, 18-1096\\n\\n\\n On Petitions for Review of Action of the\\n United States Environmental Protection Agency\\n\\n\\n Sanjay Narayan argued the cause for Environmental\\nPetitioners. With him on the briefs were James S. Pew, Tomás\\nE. Carbonell, Vickie Patton, Surbhi Sarang, John Walke, Emily\\nDavis, Thomas Zimpleman, and Keri N. Powell.\\n\\n Kavita P. Lesser, Deputy Attorney General, Office of the\\nAttorney General for the State of California, argued the cause\\n\\n 2\\nfor petitioner State of California. With her on the briefs were\\nXavier Becerra, Attorney General, David A. Zonana, Deputy\\nAttorney General, and Jonathan Wiener, Deputy Attorney\\nGeneral.\\n\\n Eric Grant, Attorney, U.S. Department of Justice, argued\\nthe cause for respondents. On the brief were Jeffrey Bossert\\nClark, Assistant Attorney General, Jonathan D. Brightbill,\\nDeputy Assistant Attorney General, and Scott Jordan,\\nAttorney, U.S. Environmental Protection Agency.\\n\\n Shannon S. Broome argued the cause for intervenors-\\nrespondents Air Permitting Forum, et al. With her on the briefs\\nwere Charles H. Knauss, Leslie Sue Ritts, Makram B. Jaber,\\nand Andrew D. Knudsen.\\n\\n David M. Friedland, Leslie A. Hulse, Felicia H. Barnes,\\nSteven P. Lehotsky, Michael B. Schon, and amici curiae\\nAmerican Chemistry Council, et al. in support of respondents.\\n\\n Before: ROGERS and WILKINS, Circuit Judges, and\\nSILBERMAN, Senior Judge.\\n\\n Opinion for the Court filed by Circuit Judge WILKINS.\\n\\n Dissenting Opinion filed by Circuit Judge ROGERS.\\n\\n This case asks us to determine the nature of an agency\\naction, an inquiry that – paradoxically – is quotidian but\\nabstruse. When we are confronted with agency action, the\\nlitany of questions is by now very well-rehearsed: Is it final? Is\\nit ripe? Is it a policy statement? Is it an interpretive rule? Is it a\\nlegislative rule? Despite the clarity of these questions,\\nhowever, predictable answers have eluded courts and\\ncommentators. See, e.g., Perez v. Mortg. Bankers Ass’n., 135\\n\\n 3\\nS. Ct. 1199, 1204 (2015) (describing the question of how to\\ndistinguish between legislative and interpretive rules as “the\\nsource of much scholarly and judicial debate”); Ticor Title Ins.\\nCo. v. FTC, 814 F.2d 731, 745 (D.C. Cir. 1987) (opinion of\\nWilliams, J.) (characterizing the law governing finality and\\nripeness as “chaotic”); Ronald M. Levin, Rulemaking and the\\nGuidance Exemption, 70 ADMIN L. REV. 264, 348 (2018) (“The\\nstandard view among commentators is that [distinguishing\\nbetween legislative and nonlegislative rules] is exceptionally\\nperplexing and incoherent.”). Indeed, the nature of agency\\naction, it seems, is too often in the eye of the beholder. We\\nresolve the instant matter, therefore, with our eye toward the\\n“continuing project” of clarifying this “byzantine” area of the\\nlaw. Nat’l Min. Ass’n v. McCarthy, 758 F.3d 243, 246 (D.C.\\nCir. 2014).\\n\\n The agency action before us is a 2018 memorandum\\n(“Wehrum Memo”) that William L. Wehrum, Assistant\\nAdministrator for the Environmental Protection Agency’s\\n(“EPA”) Office of Air and Radiation, issued to all Regional Air\\nDivision Directors. The Wehrum Memo declares that the plain\\nlanguage of § 112 of the Clean Air Act (“Act” or “CAA”), 42\\nU.S.C. § 7412, compels the conclusion that a source of toxic\\nemissions classified as “major” can reclassify to an “area\\nsource,” and thereby ease its regulatory burden, at any time\\nafter it limits its potential to emit to below the major source\\nthreshold. J.A. 1. The Wehrum Memo states that it supersedes\\na prior 1995 EPA memorandum (“Seitz Memo”) issued by\\nJohn Seitz, then Director of EPA’s Office of Air Quality\\nPlanning and Standards, which interpreted § 112 to mean that\\nonce EPA classifies a source as major, that source can never\\nreclassify to area source status, even if it limits its potential to\\nemit to below the major source threshold. Id.\\n\\n 4\\n Petitioners are the State of California and a group of\\nenvironmental organizations whose citizens and members,\\nrespectively, breathe the air in the vicinity of regulated sources.\\nEPA is the Respondent, and a group of industry organizations\\nhave joined as Intervenor. Petitioners contend that we can and\\nshould review the Wehrum Memo because it is final agency\\naction and prudentially ripe. Moreover, Petitioners argue, the\\nWehrum Memo is a legislative rule, and it is therefore\\nprocedurally defective under the Administrative Procedure Act\\n(“APA”), 5 U.S.C. § 551 et seq., because EPA failed to provide\\nnotice and comment before issuing it, see id. § 553. But even\\nif we hold that the Wehrum Memo is an interpretive rule (for\\nwhich notice and comment is not required, see id.), Petitioners\\ncontend that we still must vacate it because EPA’s\\ninterpretation of § 112 is incorrect. Respondent and Intervenor\\nretort that this Court lacks jurisdiction over the Wehrum Memo\\nbecause it is not final agency action. Alternatively, they argue,\\nwe should not review it because it is not prudentially ripe. If,\\nhowever, we find the Wehrum Memo final and ripe,\\nRespondent and Intervenor assert, we must deny the petitions\\nbecause it is an interpretive rule and is thus procedurally sound,\\nand its interpretation of § 112 is correct.\\n\\n For the reasons explained herein, we hold that the Wehrum\\nMemo is not final agency action, and we dismiss the petitions\\nfor lack of subject matter jurisdiction under the Act. We\\nexpress no opinion as to whether the Wehrum Memo is\\nprudentially ripe, an interpretive rule or a legislative rule, or on\\nthe merits of its interpretation of § 112. In holding that the\\nWehrum Memo is not final, we emphasize two points. First,\\nwhen assessing the nature of an agency action (including\\nwhether it is final), courts should resist the temptation to define\\nthe action by comparing it to superficially similar actions in the\\ncaselaw. Rather, courts should take as their NorthStar the\\nunique constellation of statutes and regulations that govern the\\n\\n 5\\naction at issue. Second, although all legislative rules are final,\\nnot all final rules are legislative, and the finality analysis is\\ntherefore distinct from the test for whether an agency action is\\na legislative rule.\\n\\n I.\\n\\n Because they share a progenitor, a reliable approach to\\nunderstanding a James Baldwin novel is to compare it,\\naccording to a set of criteria, to another work in his oeuvre.\\nIndeed, a thematic reading of Giovanni’s Room is sure to\\ninform such a reading of The Fire Next Time, and vice versa.\\nNot so, however, with respect to the broad set of phenomena\\nwe categorize as agency action. Because few, if any, of them\\nare governed by the exact same combination of statutes and\\nregulations, it is a mistake to assume – even if they appear\\nfacially similar – that they can lend each other definition\\nthrough comparison, or that they are decipherable under a\\ncommon rubric. Rather, to ascertain the nature of an agency\\naction, courts should ground the analysis in the idiosyncratic\\nregime of statutes and regulations that govern it. We have great\\nsympathy for the desire to develop a one-size-fits-all heuristic.\\nSee, e.g., Nat’l Min., 758 F.3d at 251 (“. . . all relevant parties\\nshould instantly be able to tell whether an agency action is a\\nlegislative rule, an interpretive rule, or a general statement of\\npolicy . . . .”). But this desire is perhaps misplaced, since, as we\\nonce said of interpretation itself, agency action is “a chameleon\\nthat takes its color from its context.” American Min. Congress\\nv. EPA, 995 F.2d 1106, 1111 (D.C. Cir. 1993).\\n\\n Accordingly, we turn first to the CAA provisions and EPA\\nregulations that govern the Wehrum Memo.\\n\\n Congress enacted the CAA, 42 U.S.C. § 7401 et seq., to\\n“protect and enhance the quality of the Nation’s air\\n\\n 6\\nresources . . . . ” Id. § 7401(b)(1). Toward this end, § 112\\nrequires EPA to regulate “Hazardous Air Pollutants,” i.e. toxic\\nemissions such as chloroform. Id. § 7412. Congress\\nestablished an initial list of hazardous air pollutants, id.\\n§ 7412(b)(1), but the Act requires EPA to curate it, deleting or\\nadding hazardous air pollutants over time according to certain\\ncriteria, id. § 7412(b)(2)-(3). Based on this list, the Act\\nmandates EPA to create a second list of categories of sources\\nof hazardous air pollutants, id. § 7412(c), like asphalt\\nprocessing plants and industrial dry-cleaning facilities, see\\nRevision of Source Category List under Section 112 of the\\nClean Air Act, 70 Fed. Reg. 37819-01 (June 30, 2005).\\nImportantly, the Act distinguishes between “major” and “area”\\nsources. Id. § 7412(a)(1)-(2). According to the Act’s\\ndefinitional provisions, a major source means any source\\nwithin a listed category that “emits or has the potential to emit\\nconsidering controls, in the aggregate, 10 tons per year or more\\nof any [listed hazardous air pollutant] or 25 tons per year or\\nmore of any combination of [listed hazardous air pollutants].”\\nId. § 7412(a)(1). Area source means “any stationary source of\\n[hazardous air pollutants] that is not a major source.” Id.\\n§ 7412(a)(2).\\n\\n Whether EPA classifies a source as major or area has\\nmajor consequences for both sources of hazardous air\\npollutants, which must comply with emissions standards, and\\nregulatory beneficiaries, who live, work, recreate – and thus\\nregularly breathe the air – near sources of hazardous air\\npollutants. For major sources, the Act requires EPA to establish\\nstringent emissions caps that result in “the maximum degree of\\nreduction in emissions . . . (including a prohibition on such\\nemissions, where achievable).” Id. § 7412(d). EPA refers to\\nthese emissions limitations as “Maximum Achievable Control\\nTechnology” (“MACT”) standards. J.A. 1. The Act mandates\\nthat MACT standards be “no less stringent than the emission\\n\\n 7\\ncontrol that is achieved in practice by the best controlled\\nsimilar source.” Id. § 7412(d)(3). By contrast, for area sources,\\nEPA need not set emissions caps at all, save under limited\\ncircumstances. See id. § 7412(c)(3). Moreover, where the\\nagency chooses to cap emissions for an area source, it may set\\nemissions limits based on “Generally Available Control\\nTechnology” (“GACT”) standards, which are far more lenient\\nthan their MACT counterparts.1\\n\\n Of course, emissions caps are of little use if sources do not\\ncomply with them. Presumably in recognition of this, Congress\\nenacted Title V of the CAA, 42 U.S.C. § 7661 et seq., which\\nmakes it unlawful for a source subject to regulations under the\\nAct – including GACT or MACT standards under § 112 – to\\noperate without a permit, see id. § 7661a(a). Specifically,\\nwithin a year of becoming subject to an obligation under the\\nAct, Title V requires a source to submit a permit application\\nand compliance plan to a state permitting authority. Id.\\n§ 7661b(b)-(c). In addition, a source must certify its\\ncompliance annually and submit to inspection, monitoring, and\\nreporting requirements. Id. § 7661c(a)-(c). A source may apply\\nto modify its permit, 40 C.F.R. § 70.7(e), and state permitting\\nauthorities must provide for public comment and a hearing on\\nall permit applications that they receive, 42 U.S.C. §\\n7661a(b)(6).\\n\\n But what if a state permitting authority issues or denies a\\npermit application on a ground that a regulated source, or a\\n\\n1\\n As we have observed, the Act does not provide any parameters for\\nsetting GACT standards, but its legislative history describes GACT\\nas “‘methods . . . [that] are commercially available and appropriate\\nfor application . . . considering economic impacts and the technical\\ncapabilities of firms to operate and maintain the emissions control\\nsystems.’” U.S. Sugar Corp. v. EPA, 830 F.3d 579, 595 (D.C. Cir.\\n2016) (quoting S. REP. NO. 101-228, at 171 (1989)).\\n\\n 8\\nregulatory beneficiary believes contravenes the Act? Congress\\napparently foresaw this circumstance, too. Indeed, Title V\\nprovides a detailed administrative process that dictates exactly\\nwhen and how regulated sources and regulatory beneficiaries\\nmay seek EPA review of a state permitting authority’s action,\\nand, ultimately, judicial review of EPA action. See id. § 7611d.\\nThe process works as follows. First, state permitting authorities\\nmust submit to EPA all proposed operating permits. Id.\\n§ 7611d(a)(1). If any permit contains a provision that the\\nAdministrator determines is not in compliance with the Act, the\\nAdministrator must object in writing, and provide a statement\\nof reasons for the objection, within forty-five days after\\nreceiving a copy of the proposed permit. Id. § 7661d(b)(1). If,\\nwithin ninety days of an EPA objection, a permitting authority\\nfails to submit a revised permit that satisfies the objection, the\\nAdministrator must issue or deny the permit in accordance with\\nthe Act. Id. § 7661d(c). Notably, refusing to revise a permit to\\nconform with an EPA objection does not expose a permitting\\nauthority to any sort of penalty or liability whatsoever. If the\\nAdministrator does not object in writing within forty-five days\\nof receiving a proposed permit, any person – including a\\nregulated source or a regulatory beneficiary – may, within\\nsixty days after EPA’s forty-five-day objection period expires,\\npetition the Administrator to object. Id. § 7661d(b)(2). The\\nAdministrator must grant or deny such a petition within sixty\\ndays after it is filed. Id.\\n\\n Importantly, for reasons that will become clear, § 7661d\\nspecifies: (1) that “[n]o objection shall be subject to judicial\\nreview until the Administrator takes final action to issue or\\ndeny a permit under this subsection,” id. § 7661d(c); and (2)\\nthat the Administrator’s denial of a petition to object “shall be\\nsubject to judicial review under section 7607,” id.\\n§ 7661d(b)(2). In turn, § 7607 contains the Act’s umbrella\\njudicial review provision, which confers jurisdiction in the\\n\\n 9\\nappropriate circuit for regionally applicable final action of the\\nAdministrator and in this Court for, inter alia, final action of\\nthe Administrator that is “nationally applicable.” Id.\\n§ 7607(b)(1).\\n\\n With an understanding of the major statutory provisions\\nand some of the regulations that govern the Wehrum Memo,\\nwe now provide fuller descriptions of the Wehrum Memo’s\\npredecessor, the Seitz Memo, and the Wehrum Memo itself.\\nWhere appropriate, we take care to note additional applicable\\nCAA provisions and EPA regulations.\\n\\n In 1995, without providing notice and comment, John\\nSeitz – then Director of EPA’s Office of Air Quality Planning\\nand Standards – issued a memorandum to “clarify when a\\nmajor source of [hazardous air pollutants] can become an area\\nsource” under § 112. J.A. 232 (underline in original). A major\\nsource may reclassify to an area source by limiting its potential\\nto emit to below the major source threshold, the Seitz Memo\\nconcluded, only until the first date on which it must comply\\nwith a MACT standard or any other substantive regulatory\\nrequirement under the Act. Id. at 236. The Seitz Memo referred\\nto this policy as “once in, always in.” Id. In other words, under\\nthe Seitz Memo, once EPA classifies a source as major under\\n§ 112 and its first compliance date passes, the source is\\nineligible to reclassify as an area source, even if it takes an\\nenforceable limit on its potential to emit to below the major\\nsource threshold. Despite EPA’s stated intention to do so, see\\nJ.A. 234, the agency never formalized the Seitz Memo through\\nnotice and comment rulemaking. Nevertheless, the Seitz Memo\\nhas remained in effect for nearly twenty-five years.\\n\\n On January 25, 2018, however, EPA announced it was\\nreversing course. That day, William L. Wehrum, Assistant\\nAdministrator for EPA’s Office of Air and Radiation, and\\n\\n 10\\n“principal adviser to the Administrator in matters pertaining to\\nair and radiation programs,” 40 C.F.R. § 1.41, issued a four-\\npage memo to the agency’s Regional Air Division Directors; it\\nannounced that EPA would no longer interpret § 112 in\\naccordance with the Seitz Memo. Indeed, the Wehrum Memo\\nexplains, the agency cannot interpret § 112 in accordance with\\nthe Seitz Memo because the statute’s plain-language “compels\\nthe conclusion” that a major source becomes an area source at\\nsuch time when it takes an enforceable limit on its potential to\\nemit to below the major source threshold. J.A. 1. Congress, the\\nWehrum Memo argues, placed no “temporal limitations” on\\nwhen a major source is eligible to reclassify as an area source.\\nId. at 3. Accordingly, the Wehrum Memo declares that when\\na source previously classified as major limits its potential to\\nemit to below the major source threshold, it “will no longer be\\nsubject either to the major source MACT or other major source\\nrequirements that were applicable to it as a major source under\\nCAA section 112.” Id. at 1. In addition, the Wehrum Memo\\nstates that it “supersedes” the Seitz Memo, id., and it instructs\\nthat “[t]he Regional offices should send this memorandum to\\nstates within their jurisdiction,” id. at 4.\\n\\n II.\\n\\n Before explaining why the Wehrum Memo is not final\\nagency action, we take a moment to clarify the proper test for\\nfinality. In this Court, its contours have become blurred amidst\\nthe “considerable smog,” Ass’n. of Flight Attendants v. Huerta,\\n785 F.3d 710, 717 (D.C. Cir. 2015), enshrouding the related\\nbut separate analysis of whether an agency action is a\\nlegislative rule. In Flight Attendants, for example, we framed\\nthe finality inquiry as asking whether an action is “non-\\nbinding” or a “binding legislative rule,” Flight Attendants, 785\\nF.3d at 716, and we held that the guidance document at issue\\nwas nonfinal because it was “not a legislative rule carrying the\\n\\n 11\\n‘force and effect of law,’” id. (quoting Perez, 135 S. Ct. at\\n1204). Likewise, in National Mining, we opined that in order\\nto analyze whether an action is final, we must first “take a step\\nback” and analyze whether the rule is a legislative rule,\\ninterpretive rule, or general statement of policy. Nat’l Min.\\nAss’n, 758 F.3d at 251-52. The most important factor in this\\nanalysis, we continued, is whether an action has “actual legal\\neffect,” id. at 252, and we held that the action at issue did not\\nand was therefore unreviewable, id. at 252-53.\\n\\n Subsuming the finality analysis within the test for whether\\na rule is legislative is not always inappropriate; if a rule is\\nlegislative it has the force and effect of law, and a legislative\\nrule is thus necessarily final. As the Supreme Court has twice\\nreminded us within the last five years, however, if a rule is final\\nit is not necessarily legislative, and therefore the finality\\nanalysis is distinct from the test for whether an agency action\\nis a legislative rule.\\n\\n In United States Army Corps of Engineers v. Hawkes Co.,\\n136 S. Ct. 1807 (2016), the Court affirmed that the two-prong\\ntest in Bennett v. Spear, 520 U.S. 154 (1997), remains finality’s\\ntouchstone, see Hawkes, 136 S. Ct. at 1813 (quoting Bennett,\\n520 U.S. at 177-78) (“First, the agency action must mark the\\nconsummation of the agency’s decisionmaking process . . . .\\nAnd second, the action must be one by which rights or\\nobligations have been determined, or from which legal\\nconsequences will flow.”). In Hawkes, the question of whether\\nthe agency action at issue was the consummation of the\\nagency’s decisionmaking process was not in dispute.\\nAccordingly, the Court’s analysis focused on whether the\\naction satisfied the second prong of Bennett. Notably, in\\nundertaking this inquiry, the Court neither asked whether the\\naction at issue had the force and effect of law nor made a single\\nmention of legislative rules. Rather, the Court’s inquiry\\n\\n 12\\nfocused on whether the action at issue gave “rise to ‘direct and\\nappreciable legal consequences.’” Hawkes, 136 S. Ct. at 1814\\n(quoting Bennett, 520 U.S. at 178).\\n\\n Perez, too, makes clear that the finality analysis is distinct\\nfrom the test for whether a rule is legislative. There, the Court\\naffirmed the “longstanding recognition that interpretive rules\\ndo not have the force and effect of law.” Perez, 135 S. Ct. at\\n1208 (internal citations omitted). Accordingly, overruling\\nParalyzed Veterans of Am. v. D.C. Arena L.P., 117 F.3d 579\\n(D.C. Cir. 1997), the Court held that the APA does not require\\nan agency to provide notice and comment in amending an\\ninterpretive rule, even if the new rule deviates significantly\\nfrom its predecessor. Id. at 1206. In so holding, the Court\\nreassured regulated entities and regulatory beneficiaries that\\nthey are not without recourse should an agency – perhaps to\\nevade notice and comment – repudiate a longstanding\\ninterpretive rule by way of a second interpretive rule. Id. at\\n1209. In such a circumstance, the Court explained, an affected\\nparty can seek judicial review pursuant to the APA. Id. Because\\nonly final agency action is reviewable under the APA, see 5\\nU.S.C. § 704, Perez thus affirms that interpretive rules can be\\nfinal, and, by implication, that the test for finality is\\nindependent of the analysis for whether an agency action is a\\nlegislative rule rather than an interpretive rule.\\n\\n As commentators explain, maintaining an independent\\nfinality analysis is not merely a theoretical nicety; it has several\\nsalutary effects in practice. For example – as Perez alludes to,\\nsee 135 S. Ct. at 1209 – maintaining a finality analysis that is\\ndistinct from the test for whether a rule is legislative permits\\ncourts to review nonlegislative rules and thus safeguards\\nagainst agencies evading both judicial review and notice and\\ncomment by acting via nonlegislative rules. See William Funk,\\nFinal Agency Action After Hawkes, 11 N.Y.U. J. L. & LIBERTY\\n\\n 13\\n285, 304 (2017). This is especially important when viewed\\nfrom the perspective of regulatory beneficiaries, who are\\ngenerally not parties to enforcement actions, and, therefore,\\nmay only be able to challenge nonlegislative rules via judicial\\nreview. See Nina A. Mendelson, Regulatory Beneficiaries and\\nInformal Agency Policymaking, 92 CORNELL L. REV. 397, 420-\\n24 (2007).\\n\\n Having clarified the proper test for finality, we now apply\\nit to the Wehrum Memo. Consistent with the interpretive\\nmethod we endorse herein, we hew closely to the CAA\\nprovisions and EPA regulations appertaining thereto.\\n\\n Our first question is whether the Wehrum Memo “mark[s]\\nthe consummation of [EPA’s] decisionmaking process.”\\nHawkes, 136 S. Ct. at 1813 (quoting Bennett, 520 U.S. at 177-\\n78). It does. Notably, neither Respondent nor Intervenor offer\\nsubstantive argument to the contrary. They were smart to save\\ntheir ink. The Wehrum Memo unequivocally states that the\\nplain language of § 112 “compels” the legal conclusion that\\nqualifying major sources can reclassify at such time that they\\ntake an enforceable limit on their potential to emit to below the\\nmajor source threshold. J.A. 1. In other words, the Wehrum\\nMemo does not advance what EPA believes is a reasonable\\ninterpretation of § 112; it advances what EPA believes is the\\nonly permissible interpretation of the statute. Moreover, no\\nmere subordinate issued the Wehrum Memo. Far from it. The\\nAssistant Administrator for the Office of Air and Radiation\\nissued it. As discussed, under EPA regulations, he is the\\n“principal advisor to the Administrator in matters pertaining to\\nair and radiation,” see 40 C.F.R. § 1.41, and, as we have held\\npreviously with respect to the Assistant Administrator for the\\nOffice of Air and Radiation, nothing within EPA’s regulations\\nprovides us “reason to question his authority to speak for the\\nEPA.” Her Majesty the Queen in Right of Ontario v. EPA, 912\\n\\n 14\\nF.2d 1525, 1532 (D.C. Cir. 1990) (internal citations omitted).\\nCf. Soundboard Ass’n. v. FTC, 888 F.3d 1261, 1267-69 (D.C.\\nCir. 2018), cert denied 139 S. Ct. 1544, 2019 WL 1590248\\n(Apr. 15, 2019) (Federal Trade Commission (“FTC”) staff\\nopinion letter not consummation of agency’s decisionmaking\\nprocess because FTC regulations expressly delineated between\\nCommission advice and staff advice and provided petitioners\\nopportunity to seek opinion from Commission itself).\\nMoreover, EPA published notice of the Wehrum Memo, and\\nreiterated its principal conclusion, in the Federal Register. See\\n83 Fed. Reg. 5543-01 (Feb. 8, 2018). Accordingly, the\\nWehrum Memo can only reasonably be described as EPA’s last\\nword on when a major source can reclassify to an area source\\nunder § 112.\\n\\n Because the Wehrum Memo satisfies Bennett’s first prong,\\nwe ask next whether it has “direct and appreciable legal\\nconsequences.” Hawkes, 136 S. Ct. at 1814 (quoting Bennett,\\n520 U.S. at 178). Petitioners argue that it does because it\\ncreates a new right – i.e. it allows major sources unable to\\nreclassify to area sources under the Seitz Memo to so\\nreclassify. Cal. Pet’rs’ Br. 17-20. Respondent counters that the\\nWehrum Memo does not change the rights of regulated\\nsources. EPA Br. 26-28. Whether or not a regulated source has\\nthe right to reclassify, Respondent contends, is only determined\\nwithin the Title V permitting process. Id. (citing 42 U.S.C. §\\n7661d).\\n\\n For reasons now explained, we hold that the Wehrum\\nMemo does not have a single direct and appreciable legal\\nconsequence.\\n\\n Hawkes instructs that whether an agency action has direct\\nand appreciable legal consequences is a “‘pragmatic’” inquiry.\\nId. at 1815 (quoting Abbot Labs. v. Gardner, 387 U.S. 136, 149\\n\\n 15\\n(1967)). In characterizing the inquiry as pragmatic, we do not\\ntake the Court to be encouraging some sort of common-sense\\napproach. Quite the opposite. We take it as counseling lower\\ncourts to make Bennett prong-two determinations based on the\\nconcrete consequences an agency action has or does not have\\nas a result of the specific statutes and regulations that govern\\nit. Thus, in Hawkes, the Court held, in part, that the agency\\ndetermination at issue had direct and appreciable legal\\nconsequences because, under the applicable statutes and\\nregulations, if petitioners failed to heed the determination they\\ndid so at the risk of significant criminal and civil penalties. Id.\\nAnd the cases Hawkes relies on as past examples of the\\n“pragmatic approach [the Court] has long taken to finality”\\nhold similarly. Id. (citing and quoting Sackett v. EPA, 566 U.S.\\n120, 126 (2012) (holding that agency action at issue satisfied\\nBennett prong-two because, under the relevant statutes and\\nregulations, it appeared to expose petitioners to double\\npenalties in a future enforcement proceeding and to limit their\\nability to obtain a certain type of permit); Abbot Labs., 387 U.S.\\nat 152 (holding that action at issue had a “sufficiently direct\\nand immediate” impact on petitioners, such that judicial review\\nwas appropriate, because, under the governing statutes and\\nregulations, noncompliance risked “serious criminal and civil\\npenalties”); Frozen Food Express v. United States, 351 U.S.\\n40, 44 (1956) (same)).\\n\\n Quite recently, in Valero Energy Corporation v. EPA, No.\\n18-1028, 2019 WL 2587837 (D.C. Cir. June 25, 2019), we\\naffirmed this approach. At issue there, like here, was whether\\nan EPA guidance document that declared the agency’s\\ninterpretation of a statute was final under the Act. We held that\\nit was not. Assessing it within the context of the Act, we\\nemphasized that: (1) the guidance imposed no obligations,\\nprohibitions, or restrictions; (2) it put no party to the choice\\nbetween costly compliance and the risk of a penalty of any sort;\\n\\n 16\\n(3) EPA acknowledged at oral argument that the guidance had\\nno independent legal authority; and (4) that the Act provided\\nregulated parties a statutory mechanism by which to challenge\\nany EPA action that was premised on the statutory\\ninterpretation that the guidance advanced. Valero, 2019 WL at\\n*3-5.\\n\\n Assessing the Wehrum Memo under Hawkes and in\\naccordance with Valero, we find that it is not final. True, it\\nunequivocally declares that major sources, at such time that\\nthey limit their potential to emit to below the major source\\nthreshold, “will no longer be” subject to MACT standards.\\nJ.A. 1. Viewed within the context of the Act, however, the\\nWehrum Memo is all bark and no bite. As Respondent averred\\ntwice at oral argument, neither EPA nor any regulated source\\ncan rely on the Wehrum Memo within the Title V permitting\\nprocess or in any other proceeding. Oral Arg. 50:15-50:27,\\n1:01:13-1:01:50. In other words, as Respondent concedes,\\nalthough the Wehrum Memo forecasts EPA’s position as to\\n§ 112, it has no independent legal authority. In addition, under\\nthe Act and EPA regulations, a state permitting authority that\\nrefuses to comply with the Wehrum Memo faces no penalty or\\nliability of any sort. Further still, the instant matter does not\\npresent a circumstance where the action at issue may be legally\\nconsequential because its binds agency staff and affected\\nparties have no means (outside of judicial review) by which to\\nchallenge it. To the contrary, the Act contains clear provisions\\npursuant to which: (1) a state permitting authority can refuse to\\napply the Wehrum Memo and seek judicial review if EPA\\nissues a permit over its refusal, id. § 7661d(c); and (2) a\\nregulatory beneficiary can petition EPA to object to a state\\npermitting authority’s application of the Wehrum Memo and\\nseek judicial review if EPA denies the petition, id. §\\n7661d(b)(2).\\n\\n 17\\n Accordingly, although the Wehrum Memo, in no uncertain\\nterms, forecasts EPA’s definitive interpretation of § 112, it has\\nno direct and appreciable legal consequences: neither EPA nor\\nregulated sources can rely on it as independently authoritative\\nin any proceeding; state permitting authorities face no penalty\\nor liability of any sort in ignoring it; and state permitting\\nauthorities and regulatory beneficiaries have clear statutory\\navenues by which to challenge a permitting decision adopting\\nthe reasoning of the Wehrum Memo and seek judicial review\\nif EPA fails to sustain their challenges. Under § 7661d(c), if a\\nstate permitting authority refuses to issue a permit allowing a\\nmajor source to reclassify as an area source, and EPA\\nsubsequently issues such a permit following the reasoning of\\nthe Wehrum Memo, judicial review is appropriate. Under\\n§ 7661d(b)(2), if EPA, following the reasoning of the Wehrum\\nMemo, denies a petition from any person asking the agency to\\nobject to a state permitting authority’s issuance of a permit that\\nallows a major source to reclassify as an area source, judicial\\nreview is appropriate. Indeed, because Congress specified that\\n“[n]o objection shall be subject to review until the\\nAdministrator takes final action to issue or deny a permit under\\nthis subsection,” § 7661d(c), we would contravene\\nCongressional intent if we were to hold that a memo that\\nmerely forecasts a future objection is final agency action and\\nsubject to judicial review at this time.\\n\\n The dissent insists that the Wehrum Memo satisfies\\nBennett’s second prong because it “altered the legal regime.”\\nDis. Op. 12. Indeed, the dissent forewarns, the Wehrum Memo\\n“commands, orders, and dictates [to]” EPA employees, id. at 4,\\nand “state permitting authorities are subject to” the statutory\\ninterpretation it advances, id. Said differently, according to the\\ndissent, because of the Wehrum Memo, sources subject to\\nMACT standards that limit their potential to emit to below the\\nmajor source threshold are now “assured that they will be\\n\\n 18\\nsubject to decreased regulation with EPA’s support.” Id. at 12\\n(emphasis added).\\n\\n While the question is not free from doubt, we respectfully\\ndisagree. As noted above, we must remain laser focused on\\nwhether the Wehrum Memo gives “rise to ‘direct and\\nappreciable legal consequences.’” Hawkes, 136 S. Ct. at 1814\\n(quoting Bennett, 520 U.S. at 178), and when viewed in its\\nspecific regulatory context, it does not. “[M]ajor sources must\\ncomply with technology-based emission standards requiring\\nthe maximum degree of reduction in emissions EPA deems\\nachievable, . . . [and] [i]n order to obtain an operating permit\\nunder title V of the [CAA], major sources must comply with\\nextensive monitoring, reporting and record-keeping\\nrequirements. Nat'l Min. Ass'n v. U.S. E.P.A., 59 F.3d 1351,\\n1353 (D.C. Cir. 1995). Major sources must obtain a permit in\\norder to operate, and unless and until that permit is amended or\\nset aside, the stringent requirements set forth therein must be\\ncomplied with while that equipment is operational. The\\nWehrum Memo itself does not revoke or amend a single\\npermit. As acknowledged by the Ohio environmental\\nauthorities in materials cited by petitioners, “[i]f you want to\\ntake advantage of the new guidance [in the Wehrum Memo],\\nyou will need to submit an application to modify your current\\npermit.” Environmental Pet’rs’ Br., Standing Addendum 0198.\\nAssuredly, although the Wehrum Memo advises EPA\\nemployees of the agency’s position as to § 112, it does not bind\\nstate permitting authorities or assure regulated entities of the\\nability to reclassify. As EPA concedes, EPA Br. 21, 25, in\\nreceiving such an application to modify a permit, a state\\npermitting authority may – with total impunity – ignore the\\nWehrum Memo and deny the application. It is true that the\\nAdministrator must issue a revised permit over the state\\npermitting authority’s protest if he or she believes that the\\nstatute so requires, § 7661d(c), but in such a case, the statute\\n\\n 19\\nexplicitly provides the state permitting authority a mechanism\\nby which to seek judicial review of the Administrator’s action.\\nId. Regardless of whether Congress generally intended to\\nallow pre-enforcement review of guidance documents under\\nsome circumstances in the CAA, here, as described above,\\nCongress specifically directed that judicial review shall not be\\navailable until the Title V permit amendment process reaches a\\nconclusion, see §§7661d(b)(2), 7661d(c). Congress’ explicit\\nunderstanding of finality in this specific statutory context\\ncontrols our consideration of the instant guidance document,\\nwhich pertains to that same permit amendment process.\\n\\n ***\\n\\n Before concluding, we note that we have twice had\\noccasion to ask whether an EPA guidance document that\\nimplicated the Act’s Title V permitting process was final\\nagency action: first in Appalachian Power Co. v. EPA, 208\\nF.3d 1015, 1021 (D.C. Cir. 2000), then in National\\nEnvironmental Development Ass’n v. EPA, 752 F.3d 999 (D.C.\\nCir. 2014). In each, we found that the guidance at issue was\\nfinal. A brief analysis of our reasoning in those cases\\ndemonstrates why the Wehrum Memo is not.\\n\\n In Appalachian Power, at issue was a nineteen-page\\nguidance document relating to certain monitoring requirements\\nfor Title V sources. Appalachian Power, 208 F.3d at 1019-20.\\nIn assessing our jurisdiction over the guidance, we did not\\nconsider § 7611d. Instead, we framed our jurisdiction solely in\\nterms of § 7607(b)(1). See id. at 1021 n. 10 (“Our jurisdiction\\nextends to ‘any . . . nationally applicable . . . final action taken\\nby,’ the EPA ‘Administrator.’”) (quoting 42 U.S.C.\\n§ 7607(b)(1))). We predicated our holding that the guidance\\nwas final on the following findings. First, we found that it\\nrequired state permitting authorities to: (1) “review their\\n\\n 20\\nemission standards and the emission standards EPA has\\npromulgated to determine if the standards provide enough\\nmonitoring;” and (2) “insert additional monitoring\\nrequirements as terms or conditions of a permit . . . if they\\nbelieve existing requirements are inadequate, as measured by\\nEPA’s multi-factor, case-by-case analysis set forth in the\\nGuidance.” Id. at 1022. Second, we found that EPA did not\\ndispute petitioners’ assertion that state permitting authorities\\nwere relying on EPA’s guidance in insisting that regulated\\nsources utilize a monitoring method that was more burdensome\\nthan the monitoring method set out under existing EPA\\nregulations. Id. at 1023 & n.17. Finally, we found that a\\nchallenge to an individual permit applying the guidance would\\nnot be heard in this Court, presumably because we felt any such\\nchallenge would have only regional implications. Id. at 1023 n.\\n18.\\n\\n In National Environmental, the guidance document before\\nus explained that, due to a decision of the Sixth Circuit, EPA\\nwas altering a certain interpretation of its regulations only for\\nTitle V sources located within the Sixth Circuit’s jurisdiction.\\nNat’l Envtl., 752 F.3d at 1003. As in Appalachian Power, in\\nassessing our jurisdiction over the guidance, we asked only\\nwhether it was final under § 7607(b)(1) and made no mention\\nof § 7661d. Id. at 1006. In holding that it was final, we found\\nthat the “finality and legal consequences” of the guidance\\n“were made plain” when EPA “relied on [it]” in approving a\\nFederal Implementation Plan (“FIP”) “involving a company\\nlocated outside the jurisdiction of the Sixth Circuit.” Id. at 1007\\n(citing 78 Fed. Reg. 17836, 17842 & n. 10 (March 22, 2013)).\\nIndeed, within the FIP approval – which is a final, legislative\\nrule carrying the force and effect of law – EPA cited the\\nguidance as the sole authority for the legal conclusion that\\ncertain regulations applied to certain sources located outside of\\n\\n 21\\nthe Sixth Circuit’s jurisdiction. See 78 Fed. Reg. 17836, 17842\\n& n.10 (March 22, 2013)\\n\\n Appalachian Power and National Environmental are thus,\\ncontrary to what the dissent suggests, see Dis. Op. 2-3, distinct\\nfrom the instant matter in a crucial respect. In those cases, we\\nheld that the guidance documents at issue were final under §\\n7607(b)(1), without reference to § 7661d, because EPA and\\nstate permitting authorities wielded them to effectuate legal\\nconsequences. In Appalachian Power, we found that the\\nguidance at issue required state permitting authorities to take at\\nleast two specific actions and that EPA did not deny that state\\npermitting authorities used it to coerce regulated sources to\\nadopt a stricter monitoring method. In National Environmental,\\nwe found that EPA cited the guidance, within a binding FIP\\napproval, as the sole authority in support of a legal conclusion.\\nBy contrast, the Wehrum Memo does not require any entity or\\nperson to do anything, and EPA concedes that it has not, will\\nnot, and cannot rely on it in any proceeding. Accordingly,\\nunlike in Appalachian Power and National Environmental, we\\nhave no basis to conclude, without reference to § 7661d, that\\nwe have jurisdiction over the guidance before us under\\n§ 7607(b)(1). We note, in addition, that in Appalachian Power,\\nwe found that we would lack jurisdiction over challenges to\\npermitting decisions applying the guidance at issue. Here,\\nhowever, any party entitled to review under § 7661d that\\nwishes to challenge an application of the Wehrum Memo in this\\nCourt will be so heard, since the Wehrum Memo’s principal\\nconclusion is nationally applicable. See § 7607(b)(1).\\n\\n In sum, we find that the Wehrum Memo – assessed within\\nthe context of the Act and EPA regulations – is not final agency\\naction, and we dismiss the petitions for lack of subject matter\\njurisdiction under the Act. The Wehrum Memo marks the\\nconsummation of EPA’s decisionmaking process as to when a\\n\\n 22\\nmajor source may reclassify to an area source under § 112. But\\nthe Wehrum Memo does not have direct and appreciable legal\\nconsequences: it does not require anyone to do anything;\\nneither EPA nor regulated sources can rely on it in any\\nproceeding; state permitting authorities face no penalty or\\nliability in ignoring it; state permitting authorities and\\nregulatory beneficiaries have clear statutory avenues by which\\nto challenge it and seek judicial review if EPA refuses to heed\\ntheir challenges; and any such challenges, if so desired, will be\\nheard in this Court.\\n\\n III.\\n\\n To conclude, we note that we are under no illusion that this\\nopinion will be the Rosetta Stone of understanding the nature\\nof agency action. Developing this area of the law is indeed an\\n“important continuing project.” Nat’l Min. Ass’n, 758 F.3d at\\n251. Nonetheless, today we humbly submit our contribution\\ntoward clarifying this somewhat gnarled field of jurisprudence.\\nIn ascertaining the nature of an agency action, we emphasize,\\ncourts should look first to the matrix of statutes and regulations\\ngoverning that specific action. In addition, we offer a gentle\\nreminder that the finality analysis is sui generis, separate and\\ndistinct from the test for whether an agency action is a\\nlegislative rule.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4426937/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge WILKINS\"}, {\"author\": \"ROGERS, Circuit Judge, dissenting\", \"type\": \"dissent\", \"text\": \"ROGERS, Circuit Judge, dissenting: On February 8, 2018,\\nEPA formally announced in the Federal Register that “the plain\\nlanguage of the definitions of ‘major source’. . . and of ‘area\\nsource’ in Section 112 of the [Clean Air Act] compels the\\nconclusion that a major source becomes an area source at such\\ntime that the source takes an enforceable limit on its potential\\nto emit [] hazardous air pollutants [] below the major source\\nthresholds . . . .” 83 Fed. Reg. 5543 (Feb. 8, 2018) (emphasis\\nadded). “In such circumstances, a source that was previously\\nclassified as major . . . will no longer be subject either to the\\nmajor source [maximum achievable control technology] or\\nother major source requirements that were applicable to it as a\\nmajor source under CAA section 112.” Id. Further, EPA stated\\nthis guidance memorandum “supersedes” the prior guidance in\\nthe May 1995 Seitz memorandum barring such\\nreclassifications. Id. The guidance memorandum referred to\\nin the Federal Register Notice was issued under the signature\\nof William L. Wehrum, EPA Assistant Administrator for the\\nOffice of Air and Radiation. Petitioners now seek pre-\\nenforcement review of the Wehrum Memorandum pursuant to\\n42 U.S.C. § 7607(b)(1), contending that the guidance\\nmemorandum is a legislative rule issued without notice and\\ncomment.\\n\\n I.\\n\\n Section 7607(b)(1) provides that this court shall have\\njurisdiction to review nationally applicable “final action taken”\\nby the Administrator of EPA. 42 U.S.C. § 7607(b)(1). The\\nterm “final action” in Section 7607(b)(1) is synonymous with\\n“final agency action” in the Administrative Procedure Act\\n(“APA”), 5 U.S.C. § 704. Whitman v. Am. Trucking Ass’ns,\\nInc., 531 U.S. 457, 478 (2001). The finality inquiry itself is\\ngoverned by the test articulated in Bennett v. Spear, 520 U.S.\\n154 (1997). Sackett v. EPA, 566 U.S. 120, 126–27 (2012). An\\nagency action is final if: (1) the action marks the\\n“consummation of the agency’s decisionmaking process,” and\\n\\n 2\\n(2) the action is one “by which rights or obligations have been\\ndetermined, or from which legal consequences will flow.”\\nBennett, 520 U.S. at 177–78 (internal quotation marks and\\ncitation omitted) (emphases added).\\n\\n The Supreme Court has “characterized the special judicial\\nreview provision of the CAA, 42 U.S.C. § 7607(b), as one of\\nthose statutes that specifically provides for ‘preenforcement’\\nreview.” Whitman, 531 U.S. at 479 (citing Ohio Forestry\\nAss’n, Inc. v. Sierra Club, 523 U.S. 726 (1998)). In addressing\\nripeness, the Court has pointed out that the CAA “permit[s]\\n‘judicial review directly, even before the concrete effects\\nnormally required for APA review are felt.’” Id. at 479–80\\n(quoting Lujan v. National Wildlife Federation, 497 U.S. 871,\\n891 (1990)). This court, in turn, recognized that “Congress has\\nemphatically declared a preference for immediate review with\\nrespect to Clean Air Act rulemaking,” NRDC v. EPA, 643 F.3d\\n311, 320 (D.C. Cir. 2011) (internal quotations and citations\\nomitted), which is what petitioners contend the Wehrum\\nMemorandum is. So understood, the statutory scheme not only\\nallows but encourages pre-enforcement review of final actions\\nsuch as the Wehrum Memorandum.\\n\\n A.\\n The court has repeatedly held that judicial review is\\navailable pursuant to Section 7607(b)(1) for guidance\\ndocuments that bind EPA officials on how to make Title V\\npermitting decisions.\\n\\n In Appalachian Power Co. v. EPA, 208 F.3d 1015, 1020\\n(D.C. Cir. 2000), the court considered a guidance document\\ninstructing that a source’s Title V permit must include periodic\\nmonitoring requirements to ensure compliance with certain\\nfederal or state standards. The guidance document thus\\nreflected “a position [EPA] plans to follow in reviewing State-\\n\\n 3\\nissued permits” and “a position EPA officials in the field are\\nbound to apply.” Id. at 1022. The court explained that the\\nguidance document had legal consequences for both\\nenforcement officials and regulated entities because it “reads\\nlike a ukase. It commands, it requires, it orders, it dictates.”\\nId. at 1023. The court held that the guidance document was a\\nfinal action over which the court had jurisdiction pursuant to\\nSection 7607(b). Id. at 1022–23 & n.10.\\n\\n Also, in National Environmental Development Ass’n’s\\nClean Air Project v. EPA, 752 F.3d 999, 1007 (D.C. Cir. 2014),\\nthis court held that a guidance document on how EPA would\\ndetermine whether groups of activities qualified as a “single\\nstationary source” or multiple sources in Title V permits was a\\nfinal action. The guidance document had legal consequences,\\nthe court explained, because it “provides firm guidance to\\nenforcement officials about how to handle permitting\\ndecisions” and “compels agency officials” to apply certain\\npermitting standards. Id. (emphasis in original). The court\\nheld that the guidance was “final agency action that is subject\\nto judicial review” pursuant to Section 7607(b)(1). Id. at 1006–\\n07.\\n\\n Similarly, in the context of review of state implementation\\nplans required by the CAA, the court held in Natural Resources\\nDefense Council, 643 F.3d at 320, that a guidance document\\nthat “definitively interpreted” a provision of the CAA “altered\\nthe legal regime” because it required EPA officials to consider\\nalternatives to a specific program when reviewing state\\nimplementation plans. The court explained that the guidance\\n“binds EPA regional directors and thus qualifies as final.” Id.\\n\\n In sum, the court has repeatedly held that guidance\\ndocuments, which on their face bind enforcement officials to\\napply a certain standard or interpretation under the CAA,\\n\\n 4\\nincluding in the Title V context, are final actions subject to\\nreview pursuant to Section 7607(b)(1).\\n\\n The Wehrum Memorandum states the law that EPA\\nofficials must apply in Title V permitting. Addressed to EPA\\nRegional Air Division Directors, the Wehrum Memorandum\\n“provides firm guidance to enforcement officials about how to\\nhandle permitting decisions.” Nat. Envmtl. Dev., 752 F.3d at\\n1007. By its express terms, the Wehrum Memorandum\\nunequivocally provides the interpretation of Section 112 that is\\nto be applied by EPA employees. See NRDC, 643 F.3d at 320.\\nThe Wehrum Memorandum explains that the plain text of\\nSection 112 “compels the conclusion that a major source\\nbecomes an area source at such time that the source takes an\\nenforceable limit on its potential to emit . . . below major\\nsource thresholds.” Wehrum Memorandum at 1 (emphasis\\nadded). Referencing its legal consequences, the Wehrum\\nMemorandum instructs that upon taking such a limit on its\\npotential to emit below the major source thresholds, a source\\n“will not be subject thereafter to those requirements applicable\\nto the source as a major source under CAA section 112.” Id. at\\n4 (emphasis added). Like the guidance document in\\nAppalachian Power, 208 F.3d at 1023, the Wehrum\\nMemorandum “reads like a ukase.” It commands, orders, and\\ndictates without caveats or disclaimers about the binding nature\\nof its statutory interpretation. Compare id., with Nat. Mining\\nAss’n v. McCarthy, 758 F.3d 243, 252–53 (D.C. Cir. 2014). It\\nexpressly “supersedes” EPA’s prior interpretation, stating that\\nthe Seitz Memorandum is withdrawn, “effective immediately.”\\nWehrum Memorandum at 1.\\n\\n Under the statutory scheme, state permitting authorities\\nare subject to the statutory interpretation announced in the\\nWehrum Memorandum stating EPA’s unequivocal position.\\nThe Wehrum Memorandum directs EPA enforcement officials\\n\\n 5\\nto send the memorandum to the States and thereby, in light of\\nthe Federal Register Notice, puts States doubly on notice that\\nEPA’s interpretation of Section 112 has changed, effective\\nimmediately. Given the text, structure, and purpose of the\\nCAA, state permitting authorities are not free to ignore EPA’s\\nnew interpretation of Section 112. The statutory scheme is\\nbased on a partnership between federal and state governments,\\nwhereby EPA sets federal standards and States develop\\nimplementation plans to set emissions limitations and\\nstandards to conform to these federal standards. Appalachian\\nPower, 208 F.3d at 1019. “Typically, EPA delegates to the\\nStates its authority to require companies to comply with federal\\nstandards.” Id. The terms and conditions in permits issued\\nunder Title V incorporate the applicable federal standards for\\nindividual sources. Id. Reinforcing that States must act in\\nconformity with the Wehrum Memorandum, the CAA\\nprohibits the Administrator of EPA from approving a state\\nimplementation plan under Title V except “to the extent that\\nthe program meets the requirements of [the CAA].” 42 U.S.C.\\n§ 7661a(d)(1). If a State proposes to issue an individual permit\\nthat does not comply with the CAA requirements, then the\\nAdministrator “shall” object. Id. § 7661d(b)(1). The\\nAdministrator is authorized to modify an individual permit. Id.\\n§ 7661d(e). The CAA even contemplates that a state permitting\\nauthority can be sanctioned for not adequately administering\\nand enforcing a program. Id. § 7661a(i).\\n\\n In sum, by announcing an unequivocal interpretation of\\nwhich federal standards apply to which sources under the CAA,\\n“EPA expects States to fall in line.” Appalachian Power, 208\\nF.3d at 1023. Through the Wehrum Memorandum, EPA has\\ninstructed its employees that the plain text of the CAA includes\\nno temporal limitation on the reclassification of “major\\nsources.” By publicly announcing an unequivocal statement\\nthat the plain text of the CAA “compels” its conclusion, EPA\\n\\n 6\\nhas given States their “marching orders” to allow\\nreclassification of major sources. Id. And States have heeded\\nEPA’s direction. See, e.g., Kuiken Decl. ¶¶ 6, 11 & Att.;\\nMcCloud Decl. ¶¶ 6, 10 & Att.; Gharrity Decl. Att. (Ohio EPA\\npublication providing guidance to regulated entities treating the\\nWehrum Memorandum as binding); see also Standing Add. 43,\\n45, 48, 52–53, 57, 275.\\n\\n Therefore, under this court’s precedent issuance of the\\nWehrum Memorandum is final action subject to judicial review\\npursuant to Section 7607(b)(1) because it provides EPA’s\\nunequivocal interpretation on the reclassification of “major\\nsources,” thereby binding EPA enforcement officials.\\n\\n B.\\n Contrary to the court’s conclusion, the Wehrum\\nMemorandum is final action under the two-prong Bennett v.\\nSpear test. 520 U.S. at 177–78. First, the Wehrum\\nMemorandum marks the consummation of EPA’s\\ndecisionmaking process with respect to its interpretation of\\nwhether Section 112 of the CAA allows major sources to\\nreclassify as area sources at any time. The Wehrum\\nMemorandum is unequivocal — if a major source “takes an\\nenforceable limit on its potential to emit . . . below the major\\nsource thresholds,” the CAA “compels” that the source can\\nreclassify as an area source at that time. Wehrum\\nMemorandum at 1. It states the official position of the EPA\\nAdministrator; in signing the guidance memorandum, the\\nAssistant Administrator for the Office of Air and Radiation was\\nacting on behalf of the Administrator. See 40 C.F.R. § 1.41;\\nHer Majesty the Queen in Right of Ontario v. EPA, 912 F.2d\\n1525, 1532 (D.C. Cir. 1990). Addressed to the Regional Air\\nDivision Directors, it instructs the Regional offices on what\\nSection 112 of the CAA “compels,” and to “send this\\nmemorandum to states within their jurisdiction.” Id. at 4. By\\n\\n 7\\nFederal Register Notice, EPA announced to the public it had\\nabandoned its prior interpretation and now concluded the plain\\ntext of Section 112 imposed no temporal limit on\\nreclassifications by “major sources.” 83 Fed. Reg. at 5543.\\nRegardless of whether EPA may change its position in the\\nfuture, see, e.g., Gen. Elec. Co. v. EPA, 290 F.3d 377, 380\\n(D.C. Cir. 2002); Appalachian Power, 208 F.3d at 1022, the\\nWehrum Memorandum marks EPA’s unequivocal statutory\\ninterpretation of whether “major sources” may, at any time,\\nreclassify under the CAA upon limiting their potential to emit\\nhazardous pollutants.\\n\\n Second, the Wehrum Memorandum is an action “from\\nwhich legal consequences will flow” because it announces a\\nbinding change in the legal regime. Bennett, 520 U.S. at 178\\n(emphasis added); see also U.S. Army Corps of Engineers v.\\nHawkes Co., 136 S. Ct. 1807, 1814–15 (2016); NRDC, 643\\nF.3d at 319–20; Indep. Equip. Dealers Ass’n v. EPA, 372 F.3d\\n420, 427 (D.C. Cir. 2004); Appalachian Power, 208 F.3d at\\n1020–21. The Wehrum Memorandum alters the legal regime\\nby changing the regulatory requirements for any “major\\nsource” that “takes an enforceable limit on its potential to emit\\n. . . below major source thresholds.” Wehrum Memorandum at\\n1. Those sources now have the opportunity to reclassify as area\\nsources at any time by limiting their potential to emit below\\nmajor source thresholds and thereafter will not be subject to the\\nmore onerous major source requirements, such as the\\nMaximum Achievable Control Technology standards.\\n\\n The court’s recent decision in Valero Energy Corp. v.\\nEPA, 927 F.3d 532 (D.C. Cir. 2019) reaffirms that legal\\nconsequences will flow from the Wehrum Memorandum.\\nThere, the court held legal consequences did not flow from a\\nguidance document that interpreted EPA’s duty to conduct\\n“periodic reviews” of renewable fuel standards under 42\\n\\n 8\\nU.S.C. § 7545(o)(11) and explained how EPA’s prior actions\\nfulfilled any statutory duty to conduct periodic reviews. Id. at\\n535. The document did not purport bind EPA to its\\ninterpretation and had no identifiable effect on the regulated\\ncommunity. Id. at 536–37. Here, in contrast, the Wehrum\\nMemorandum announces a binding interpretation that has an\\nidentifiable effect on major sources that take enforceable limits\\non their potential to emit below major source thresholds.\\n\\n EPA’s contrary position, that the Wehrum Memorandum\\nis not final because it has no immediate impact or direct legal\\nconsequences for specific sources, misstates the finality test.\\n“The test for finality . . . is not so narrow — it is met if ‘the\\naction [is] one by which rights or obligations have been\\ndetermined, or from which legal consequences will flow.’”\\nHarris v. FAA, 353 F.3d 1006, 1011 (D.C. Cir. 2004) (quoting\\nBennett, 520 U.S. at 178) (emphasis added). The court’s\\nsuggestion that the Wehrum Memorandum is “all bark and no\\nbite,” Op. 16, ignores its plain text as well as the second clause\\nof the second prong of the Bennett v. Spear test. With EPA’s\\nblessing, legal consequences will flow from the Wehrum\\nMemorandum no later than when “major sources” take\\nenforceable limits on their potential to emit below “major\\nsource” thresholds and obtain new or modified Title V permits.\\nIndeed, such legal consequences have already occurred; EPA\\nacknowledged that at least two “major sources” in Indiana have\\nreclassified as area sources as of filing of the briefs in the\\ninstant appeal, and the Sierra Club has identified numerous\\nother “major sources” that are eligible to reclassify. Resp’t’s\\nBr. 29; Kuiken Decl. ¶ 6 & Att.; McCloud Decl. ¶ 5 & Att.\\n\\n Additionally, the opportunity for judicial review at a later\\ntime has no direct bearing on the availability of pre-\\nenforcement review of the Wehrum Memorandum. Section\\n7661d provides for judicial review under Section 7607 of an\\n\\n 9\\nAdministrator’s objection or denial of a petition to object to a\\nspecific Title V permit for a specific source. 42 U.S.C.\\n§ 7661d(b). Petitioners are not challenging a source-specific\\nobjection. Instead, they seek review of a generally applicable\\nguidance document pursuant to Section 7607(b), which\\nprovides for judicial review of such a general guidance\\ndocument that is a “final action.” Id. § 7607(b)(1). The two\\nprovisions for judicial review serve different purposes.\\nJudicial review of national standards at the start of the\\nregulatory process can ensure that Congress’s intent is being\\ncarried out before States and the regulated community must\\ntake costly implementing actions, while later enforcement\\nreview can ensure compliance with terms and conditions in\\nindividual permits. Nothing in the text, structure, purpose, or\\nlegislative history of the CAA indicates the availability of\\nreview of a decision in a source-specific Title V proceeding\\nunder Section 7661d would preclude pre-enforcement review\\nof a general guidance document under Section 7607(b). That\\nboth exist in the CAA is a rational approach for complex\\nlegislation where Congress intended to bring about significant\\nchanges to the status quo impacting the environment, the\\npublic, and entities emitting hazardous air pollutants. See\\nAppalachian Power, 208 F.3d at 1017; see generally Hon.\\nHenry A. Waxman, An Overview of the Clean Air Act\\nAmendments of 1990, 21 ENVTL. L. 1721, 1723, 1742 (1991).\\nPut otherwise, the provision of judicial review of Title V permit\\ndecisions “in one section of a long and complicated statute” is\\nhardly sufficient to overcome Congress’s decision to provide\\npre-enforcement review. See Sackett, 132 S. Ct. at 1373. Not\\nonly does nothing in the text of Section 7661d override the\\nprovision for pre-enforcement review under Section 7607(b),\\nthe Supreme Court has acknowledged the CAA encourages\\npre-enforcement judicial review. See Whitman, 531 U.S. at\\n479 (quoting Ohio Forestry, 523 U.S. at 737); see also NRDC,\\n643 F.3d at 320.\\n\\n 10\\n\\n Furthermore, Congress’s express purpose in enacting the\\nCAA was “to promote the public health and welfare and the\\nproductive capacity of [the Nation’s] population.” 42 U.S.C.\\n§ 7401(b)(1). Delaying the opportunity for judicial review\\nuntil individual source permit enforcement proceedings could\\neffectively squelch the opportunity for regulatory beneficiaries\\nto obtain judicial review of an agency’s position. See Nina A.\\nMendelson, Regulatory Beneficiaries and Informal Agency\\nPolicymaking, 92 CORNELL L. REV. 397, 420–24 (2007)\\n(“Mendelson”). Title V does provide regulatory beneficiaries\\nthe opportunity to file a petition to object and to seek judicial\\nreview of denial of a petition to object in individual permitting\\nproceedings. 42 U.S.C. § 7661d(b)(2). Yet requiring\\nregulatory beneficiaries to monitor and to file petitions in\\nindividual permit proceedings throughout the United States\\nrequires resources that may constrain beneficiaries’ ability to\\nseek judicial review. See Mendelson at 451–52. Pre-\\nenforcement judicial review of a nationally applicable\\nguidance document, in contrast, is more accessible for\\nregulatory beneficiaries. Precluding pre-enforcement review\\nwould impose a burden Congress has not required.\\n\\n Notably, irrelevant to the finality inquiry is the fact that the\\nWehrum Memorandum is deregulatory rather than regulatory.\\nThis is the fallacy underlying the court’s efforts to distinguish\\nour precedent on the basis that the Wehrum Memorandum does\\nnot require anyone to do anything. See Op. 21. Although the\\nSupreme Court and this court have regularly been confronted\\nwith challenges to regulatory actions as too strong or too weak\\nand held that agency actions that require parties to take certain\\nactions or expose parties to penalties are final, see, e.g.,\\nHawkes, 136 S. Ct. at 1814–15; Sackett, 566 U.S. at 126; Nat.\\nMining Ass’n, 758 F.3d at 252; CSI Aviation Servs., Inc. v.\\nDOT, 637 F.3d 408, 412–13 (D.C. Cir. 2011), the focus of the\\n\\n 11\\ninquiry has been on whether the legal regime has changed. The\\nWehrum Memorandum changed the legal regime by enabling\\ncertain regulated entities to become subject to decreased\\nregulation — an opportunity not clearly available under the\\nCAA, much less under EPA’s prior interpretation. Prior to\\nEPA’s issuance of guidance, enforcement officials had\\ndiscretion to interpret the CAA as either allowing or\\nprohibiting “major source” reclassification after the first\\ncompliance date. See NRDC, 643 F.3d at 319–20. Now that\\ndiscretion has been withdrawn as regulated “major sources” are\\neligible to be reclassified at any time upon taking emissions\\nlimitations.\\n\\n Further, the Supreme Court has held that legal\\nconsequences can flow from the “denial of a safe harbor.”\\nHawkes, 136 S. Ct. at 1814. In Scenic America, Inc. v. DOT,\\n836 F.3d 42, 56 (D.C. Cir. 2016), this court recognized that\\nlegal consequences would flow from a guidance document that\\ncreated a safe harbor whereby digital billboard permits would\\nnot be denied on the basis of violating certain standards. And\\nin determining whether a document was a “rule” under the\\nToxic Substance Control Act in General Electric, 290 F.3d at\\n384–85, this court held that a guidance document that “appears\\nto bind [EPA] to accept applications using a total toxicity factor\\nof 4.0 (mg/kg/day)-1” imposed binding obligations, explaining\\nthat “if the language of the document is such that private parties\\ncan rely on it as a norm or safe harbor by which to shape their\\nactions, it can be binding as a practical matter.” Id. at 383\\n(quoting Robert A. Anthony, Interpretive Rules, Policy\\nStatements, Guidances, Manuals, and the Like—Should\\nFederal Agencies Use Them to Bind the Public?, 41 DUKE L.J.\\n1311, 1328–29 (1992)). The Wehrum Memorandum creates a\\nsafe harbor for “major sources” by removing a prior barrier to\\nreclassification — those sources that take an enforceable limit\\non their potential to emit below the “major source” threshold\\n\\n 12\\nare assured that they will be subject to decreased regulation\\nwith EPA’s support. This safe harbor has a “clear legal effect\\non regulated entities.” See Scenic America, 836 F.3d at 56.\\n\\n For these reasons, the Wehrum Memorandum is final\\naction, reviewable pursuant to Section 7607(b)(1). It is an\\nagency action with the telltale signs of finality — it presents a\\nunequivocal interpretation of requirements under the CAA; it\\nis binding on its face; and it altered the legal regime by\\nproviding an opportunity for “major sources” that take\\nenforceable limits on their potential to emit below the “major\\nsource” thresholds to reclassify as “area sources” at any time.\\n“Once the agency publicly articulates an unequivocal position\\n. . . and expects regulated entities to alter their primary conduct\\nto conform to that position, the agency has voluntarily\\nrelinquished the benefit of postponed judicial review.” Ciba-\\nGeigy Corp. v. EPA, 801 F.3d 430, 436 (D.C. Cir. 1986).\\n\\n II.\\n\\n The question remains whether the Wehrum Memorandum\\nis an agency action ripe for review. To decide whether an\\nagency’s action is ripe for review, courts generally consider the\\n“fitness of the issues for judicial decision” and the “hardship to\\nthe parties of withholding court consideration.” Ohio Forestry,\\n523 U.S. at 733 (quoting Abbott Labs. v. Gardner, 387 U.S.\\n136, 149 (1967)). In Appalachian Power, 208 F.3d at 1023\\nn.18, the court held that a guidance document that reflected\\nEPA’s settled position regarding periodic monitoring\\nrequirements in Title V permits was ripe for review because the\\npropriety of EPA’s statutory interpretation would “not turn on\\nthe specifics of any particular permit.” Id. EPA’s guidance\\ndocument was “national in scope and Congress clearly\\nintended this court to determine the validity of such EPA\\nactions,” see 42 U.S.C. § 7607, yet “[a] challenge to an\\n\\n 13\\nindividual permit would not be heard in this court,”\\nAppalachian Power, 208 F.3d at 1023 n.18.\\n\\n The same is true here. Whether EPA was required, as\\npetitioners contend, to promulgate the Wehrum Memorandum\\nthrough notice-and-comment rulemaking and whether EPA’s\\nstatutory interpretation in the Wehrum Memorandum is proper\\nwill not turn on the specifics of any particular permit. EPA has\\nannounced that “a major source that takes an enforceable limit\\non its [potential to emit] . . . no matter when the source may\\nchoose to take measures to limit its [potential to emit] . . . will\\nnot be subject thereafter to those requirements applicable to the\\nsource as a major source under CAA section 112.” Wehrum\\nMemorandum at 4 (emphasis added). Its guidance is national\\nin scope, as the court looks only to the face of an agency action\\nto determine whether the action is nationally applicable.\\nDalton Trucking, Inc. v. EPA, 808 F.3d 875, 881 (D.C. Cir.\\n2015); Am. Road & Trans. Builders Ass’n v. EPA, 705 F.3d\\n453, 456 (D.C. Cir. 2013). Any objection or denial of a petition\\nto object to a Title V permit would apply solely to the specific\\nsource applying for the Title V permit; inclusion of a general\\nstatutory interpretation that may apply as precedent in future\\nTitle V permit proceedings would not render the action\\nnationally applicable under 42 U.S.C. § 7607(b)(1). See Sierra\\nClub v. EPA, 926 F.3d 844, 849–50 (D.C. Cir. 2019).\\nConcluding that petitioners’ challenges are not ripe until the\\nWehrum Memorandum is applied in an individual Title V\\npermit proceeding would frustrate Congress’s intent that\\n“nationally applicable” actions such as the Wehrum\\nMemorandum be reviewable in this court pursuant to 42 U.S.C.\\n§ 7607(b)(1). Under the court’s approach, challenges would\\ninstead be directed to appropriate regional courts. See Op. 16–\\n17; see e.g., Sierra Club, 926 F.3d at 847–50.\\n\\n 14\\n In any event, petitioners’ challenges are fit for judicial\\nreview because they present purely legal issues. See Nat. Envtl.\\nDev., 752 F.3d at 1008; Gen. Elec., 290 F.3d at 380. Whether\\nSection 112 of the CAA allows “major sources” to reclassify\\nas “area sources” at any time upon taking enforceable limits on\\ntheir potential to emit is a question of statutory interpretation\\nthat will not benefit from further factual development. See\\nOhio Forestry, 523 U.S. at 733. Given EPA’s conclusion that\\nthe plain text “compels” the interpretation in the Wehrum\\nMemorandum, this is not a circumstance in which judicial\\nreview would hinder EPA’s effort to refine its position. See id.\\nat 735. Nor will petitioners’ claims under the APA be affected\\nby further factual development. See Gen. Elec., 290 F.3d at\\n380. In view of Congress’s stated preference for immediate\\nreview under the CAA, NRDC, 643 F.3d at 320, the court need\\nnot consider hardship to the parties of delaying review, see\\nCement Kiln Recycling Coal. v. EPA, 493 F.3d 207, 215 (D.C.\\nCir. 2007); Gen. Elec., 290 F.3d at 381. As noted, the CAA is\\na statute that “permit[s] judicial review directly, even before\\nthe concrete effects normally required for APA review are\\nfelt.” Whitman, 531 U.S. at 479 (quoting Lujan v. Nat’l\\nWildlife Fed’n, 497 U.S, 871, 891 (1990)).\\n\\n III.\\n\\n The APA requires that a legislative rule, which carries the\\n“force and effect of law,” Ass’n of Flight Attendants-CWA,\\nARL-CIO v. Huerta, 785 F.3d 710, 716 (D.C. Cir. 2015)\\n(quoting Perez v. Mortgage Bankers Ass’n, 135 S. Ct. 1199,\\n1204 (2015)), must be promulgated pursuant to notice-and-\\ncomment rulemaking. Id. To determine whether agency action\\ncarries the force and effect of law, the court generally looks to\\nthe actual legal effect (or lack thereof) of the agency action,\\npaying particular attention to the express words used in the\\ndocument. Flight Attendants, 785 F.3d at 717; Nat. Mining,\\n\\n 15\\n758 F.3d at 252. “[A] document that reads like an edict is likely\\nto be binding, while one riddled with caveats is not.” Flight\\nAttendants, 785 F.3d at 717. The court also considers whether\\nthe action was published in the Federal Register or the Code of\\nFederal Regulations, and whether the action has binding effects\\non the agency or private parties. Ctr. for Auto Safety v. Nat.\\nHighway Traffic Safety Admin., 452 F.3d 798, 806–07 (D.C.\\nCir. 2006) (citing Molycorp, Inc. v. EPA, 197 F.3d 543, 545\\n(D.C. Cir. 1999)); see also Flight Attendants, 785 F.3d at 717;\\nNat. Mining, 758 F.3d at 252. An agency’s adoption of a\\nbinding norm that could not be properly promulgated absent\\nthe notice-and-comment rulemaking required by the APA\\n“obviously would reflect final agency action.” Ctr. for Auto\\nSafety, 452 F.3d at 804; see also Flight Attendants, 785 F.3d at\\n716. When an agency action is final because it creates a\\nbinding norm that alters the legal regime, the question of\\nwhether the action is a legislative rule is “easy.” NRDC, 643\\nF.3d at 320.\\n\\n That is the situation here. The Wehrum Memorandum\\nmakes its legal effect clear; it “reads like an edict,” Flight\\nAttendants, 785 F.3d at 717, instructing regional offices that\\nthe “unambiguous language” of Section 112 of the CAA\\n“compels” “major source” reclassifications. Wehrum\\nMemorandum at 1, 3. The document itself contains no\\ndisclaimers or caveats. Upon taking an enforceable limit on\\ntheir potential to emit “below major source thresholds,” major\\nsources “will not be subject thereafter” to “major source”\\nregulations. Id. at 4 (emphasis added). EPA’s Federal Register\\nNotice announced the new interpretation and binds EPA to the\\nchanged legal regime. As such, the Wehrum Memorandum is\\na legislative rule that failed to conform to the APA’s notice-\\nand-comment requirement. Cf. Gen. Elec., 290 F.3d at 385.\\n\\n 16\\n Accordingly, I would grant the petitions for review and\\nvacate the Wehrum Memorandum, and I respectfully dissent.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4426937/\", \"author_raw\": \"ROGERS, Circuit Judge, dissenting\"}]}"},"judge1":{"kind":"string","value":"ROGERS"},"judge2":{"kind":"string","value":"SILBERMAN"},"judge3":{"kind":"string","value":"WILKINS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":4,"string":"4"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":2,"string":"2"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":2,"string":"2"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4649684/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":1,"string":"1"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":1,"string":"1"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: CALIFORNIA COMMUNITIES AGAINST TOXICS, Et Al., Petitioners v. ENVIRONMENTAL PROTECTION AGENCY and Andrew Wheeler, Administrator, U.S. Environmental Protection Agency, Respondents Air Permitting Forum, Et Al., Intervenors\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-20\nOpinion type: majority\nHas dissenting opinion: Yes\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued April 1, 2019 Decided August 20, 2019\n\n No. 18-1085\n\n CALIFORNIA COMMUNITIES AGAINST TOXICS, ET AL.,\n PETITIONERS\n\n v.\n\n ENVIRONMENTAL PROTECTION AGENCY AND ANDREW\n WHEELER, ADMINISTRATOR, U.S. ENVIRONMENTAL\n PROTECTION AGENCY,\n RESPONDENTS\n\n AIR PERMITTING FORUM, ET AL.,\n INTERVENORS\n\n\n Consolidated with 18-1095, 18-1096\n\n\n On Petitions for Review of Action of the\n United States Environmental Protection Agency\n\n\n Sanjay Narayan argued the cause for Environmental\nPetitioners. With him on the briefs were James S. Pew, Tomás\nE. Carbonell, Vickie Patton, Surbhi Sarang, John Walke, Emily\nDavis, Thomas Zimpleman, and Keri N. Powell.\n\n Kavita P. Lesser, Deputy Attorney General, Office of the\nAttorney General for the State of California, argued the cause\n\n 2\nfor petitioner State of California. With her on the briefs were\nXavier Becerra, Attorney General, David A. Zonana, Deputy\nAttorney General, and Jonathan Wiener, Deputy Attorney\nGeneral.\n\n Eric Grant, Attorney, U.S. Department of Justice, argued\nthe cause for respondents. On the brief were Jeffrey Bossert\nClark, Assistant Attorney General, Jonathan D. Brightbill,\nDeputy Assistant Attorney General, and Scott Jordan,\nAttorney, U.S. Environmental Protection Agency.\n\n Shannon S. Broome argued the cause for intervenors-\nrespondents Air Permitting Forum, et al. With her on the briefs\nwere Charles H. Knauss, Leslie Sue Ritts, Makram B. Jaber,\nand Andrew D. Knudsen.\n\n David M. Friedland, Leslie A. Hulse, Felicia H. Barnes,\nSteven P. Lehotsky, Michael B. Schon, and amici curiae\nAmerican Chemistry Council, et al. in support of respondents.\n\n Before: ROGERS and WILKINS, Circuit Judges, and\nSILBERMAN, Senior Judge.\n\n Opinion for the Court filed by Circuit Judge WILKINS.\n\n Dissenting Opinion filed by Circuit Judge ROGERS.\n\n This case asks us to determine the nature of an agency\naction, an inquiry that – paradoxically – is quotidian but\nabstruse. When we are confronted with agency action, the\nlitany of questions is by now very well-rehearsed: Is it final? Is\nit ripe? Is it a policy statement? Is it an interpretive rule? Is it a\nlegislative rule? Despite the clarity of these questions,\nhowever, predictable answers have eluded courts and\ncommentators. See, e.g., Perez v. Mortg. Bankers Ass’n., 135\n\n 3\nS. Ct. 1199, 1204 (2015) (describing the question of how to\ndistinguish between legislative and interpretive rules as “the\nsource of much scholarly and judicial debate”); Ticor Title Ins.\nCo. v. FTC, 814 F.2d 731, 745 (D.C. Cir. 1987) (opinion of\nWilliams, J.) (characterizing the law governing finality and\nripeness as “chaotic”); Ronald M. Levin, Rulemaking and the\nGuidance Exemption, 70 ADMIN L. REV. 264, 348 (2018) (“The\nstandard view among commentators is that [distinguishing\nbetween legislative and nonlegislative rules] is exceptionally\nperplexing and incoherent.”). Indeed, the nature of agency\naction, it seems, is too often in the eye of the beholder. We\nresolve the instant matter, therefore, with our eye toward the\n“continuing project” of clarifying this “byzantine” area of the\nlaw. Nat’l Min. Ass’n v. McCarthy, 758 F.3d 243, 246 (D.C.\nCir. 2014).\n\n The agency action before us is a 2018 memorandum\n(“Wehrum Memo”) that William L. Wehrum, Assistant\nAdministrator for the Environmental Protection Agency’s\n(“EPA”) Office of Air and Radiation, issued to all Regional Air\nDivision Directors. The Wehrum Memo declares that the plain\nlanguage of § 112 of the Clean Air Act (“Act” or “CAA”), 42\nU.S.C. § 7412, compels the conclusion that a source of toxic\nemissions classified as “major” can reclassify to an “area\nsource,” and thereby ease its regulatory burden, at any time\nafter it limits its potential to emit to below the major source\nthreshold. J.A. 1. The Wehrum Memo states that it supersedes\na prior 1995 EPA memorandum (“Seitz Memo”) issued by\nJohn Seitz, then Director of EPA’s Office of Air Quality\nPlanning and Standards, which interpreted § 112 to mean that\nonce EPA classifies a source as major, that source can never\nreclassify to area source status, even if it limits its potential to\nemit to below the major source threshold. Id.\n\n 4\n Petitioners are the State of California and a group of\nenvironmental organizations whose citizens and members,\nrespectively, breathe the air in the vicinity of regulated sources.\nEPA is the Respondent, and a group of industry organizations\nhave joined as Intervenor. Petitioners contend that we can and\nshould review the Wehrum Memo because it is final agency\naction and prudentially ripe. Moreover, Petitioners argue, the\nWehrum Memo is a legislative rule, and it is therefore\nprocedurally defective under the Administrative Procedure Act\n(“APA”), 5 U.S.C. § 551 et seq., because EPA failed to provide\nnotice and comment before issuing it, see id. § 553. But even\nif we hold that the Wehrum Memo is an interpretive rule (for\nwhich notice and comment is not required, see id.), Petitioners\ncontend that we still must vacate it because EPA’s\ninterpretation of § 112 is incorrect. Respondent and Intervenor\nretort that this Court lacks jurisdiction over the Wehrum Memo\nbecause it is not final agency action. Alternatively, they argue,\nwe should not review it because it is not prudentially ripe. If,\nhowever, we find the Wehrum Memo final and ripe,\nRespondent and Intervenor assert, we must deny the petitions\nbecause it is an interpretive rule and is thus procedurally sound,\nand its interpretation of § 112 is correct.\n\n For the reasons explained herein, we hold that the Wehrum\nMemo is not final agency action, and we dismiss the petitions\nfor lack of subject matter jurisdiction under the Act. We\nexpress no opinion as to whether the Wehrum Memo is\nprudentially ripe, an interpretive rule or a legislative rule, or on\nthe merits of its interpretation of § 112. In holding that the\nWehrum Memo is not final, we emphasize two points. First,\nwhen assessing the nature of an agency action (including\nwhether it is final), courts should resist the temptation to define\nthe action by comparing it to superficially similar actions in the\ncaselaw. Rather, courts should take as their NorthStar the\nunique constellation of statutes and regulations that govern the\n\n 5\naction at issue. Second, although all legislative rules are final,\nnot all final rules are legislative, and the finality analysis is\ntherefore distinct from the test for whether an agency action is\na legislative rule.\n\n I.\n\n Because they share a progenitor, a reliable approach to\nunderstanding a James Baldwin novel is to compare it,\naccording to a set of criteria, to another work in his oeuvre.\nIndeed, a thematic reading of Giovanni’s Room is sure to\ninform such a reading of The Fire Next Time, and vice versa.\nNot so, however, with respect to the broad set of phenomena\nwe categorize as agency action. Because few, if any, of them\nare governed by the exact same combination of statutes and\nregulations, it is a mistake to assume – even if they appear\nfacially similar – that they can lend each other definition\nthrough comparison, or that they are decipherable under a\ncommon rubric. Rather, to ascertain the nature of an agency\naction, courts should ground the analysis in the idiosyncratic\nregime of statutes and regulations that govern it. We have great\nsympathy for the desire to develop a one-size-fits-all heuristic.\nSee, e.g., Nat’l Min., 758 F.3d at 251 (“. . . all relevant parties\nshould instantly be able to tell whether an agency action is a\nlegislative rule, an interpretive rule, or a general statement of\npolicy . . . .”). But this desire is perhaps misplaced, since, as we\nonce said of interpretation itself, agency action is “a chameleon\nthat takes its color from its context.” American Min. Congress\nv. EPA, 995 F.2d 1106, 1111 (D.C. Cir. 1993).\n\n Accordingly, we turn first to the CAA provisions and EPA\nregulations that govern the Wehrum Memo.\n\n Congress enacted the CAA, 42 U.S.C. § 7401 et seq., to\n“protect and enhance the quality of the Nation’s air\n\n 6\nresources . . . . ” Id. § 7401(b)(1). Toward this end, § 112\nrequires EPA to regulate “Hazardous Air Pollutants,” i.e. toxic\nemissions such as chloroform. Id. § 7412. Congress\nestablished an initial list of hazardous air pollutants, id.\n§ 7412(b)(1), but the Act requires EPA to curate it, deleting or\nadding hazardous air pollutants over time according to certain\ncriteria, id. § 7412(b)(2)-(3). Based on this list, the Act\nmandates EPA to create a second list of categories of sources\nof hazardous air pollutants, id. § 7412(c), like asphalt\nprocessing plants and industrial dry-cleaning facilities, see\nRevision of Source Category List under Section 112 of the\nClean Air Act, 70 Fed. Reg. 37819-01 (June 30, 2005).\nImportantly, the Act distinguishes between “major” and “area”\nsources. Id. § 7412(a)(1)-(2). According to the Act’s\ndefinitional provisions, a major source means any source\nwithin a listed category that “emits or has the potential to emit\nconsidering controls, in the aggregate, 10 tons per year or more\nof any [listed hazardous air pollutant] or 25 tons per year or\nmore of any combination of [listed hazardous air pollutants].”\nId. § 7412(a)(1). Area source means “any stationary source of\n[hazardous air pollutants] that is not a major source.” Id.\n§ 7412(a)(2).\n\n Whether EPA classifies a source as major or area has\nmajor consequences for both sources of hazardous air\npollutants, which must comply with emissions standards, and\nregulatory beneficiaries, who live, work, recreate – and thus\nregularly breathe the air – near sources of hazardous air\npollutants. For major sources, the Act requires EPA to establish\nstringent emissions caps that result in “the maximum degree of\nreduction in emissions . . . (including a prohibition on such\nemissions, where achievable).” Id. § 7412(d). EPA refers to\nthese emissions limitations as “Maximum Achievable Control\nTechnology” (“MACT”) standards. J.A. 1. The Act mandates\nthat MACT standards be “no less stringent than the emission\n\n 7\ncontrol that is achieved in practice by the best controlled\nsimilar source.” Id. § 7412(d)(3). By contrast, for area sources,\nEPA need not set emissions caps at all, save under limited\ncircumstances. See id. § 7412(c)(3). Moreover, where the\nagency chooses to cap emissions for an area source, it may set\nemissions limits based on “Generally Available Control\nTechnology” (“GACT”) standards, which are far more lenient\nthan their MACT counterparts.1\n\n Of course, emissions caps are of little use if sources do not\ncomply with them. Presumably in recognition of this, Congress\nenacted Title V of the CAA, 42 U.S.C. § 7661 et seq., which\nmakes it unlawful for a source subject to regulations under the\nAct – including GACT or MACT standards under § 112 – to\noperate without a permit, see id. § 7661a(a). Specifically,\nwithin a year of becoming subject to an obligation under the\nAct, Title V requires a source to submit a permit application\nand compliance plan to a state permitting authority. Id.\n§ 7661b(b)-(c). In addition, a source must certify its\ncompliance annually and submit to inspection, monitoring, and\nreporting requirements. Id. § 7661c(a)-(c). A source may apply\nto modify its permit, 40 C.F.R. § 70.7(e), and state permitting\nauthorities must provide for public comment and a hearing on\nall permit applications that they receive, 42 U.S.C. §\n7661a(b)(6).\n\n But what if a state permitting authority issues or denies a\npermit application on a ground that a regulated source, or a\n\n1\n As we have observed, the Act does not provide any parameters for\nsetting GACT standards, but its legislative history describes GACT\nas “‘methods . . . [that] are commercially available and appropriate\nfor application . . . considering economic impacts and the technical\ncapabilities of firms to operate and maintain the emissions control\nsystems.’” U.S. Sugar Corp. v. EPA, 830 F.3d 579, 595 (D.C. Cir.\n2016) (quoting S. REP. NO. 101-228, at 171 (1989)).\n\n 8\nregulatory beneficiary believes contravenes the Act? Congress\napparently foresaw this circumstance, too. Indeed, Title V\nprovides a detailed administrative process that dictates exactly\nwhen and how regulated sources and regulatory beneficiaries\nmay seek EPA review of a state permitting authority’s action,\nand, ultimately, judicial review of EPA action. See id. § 7611d.\nThe process works as follows. First, state permitting authorities\nmust submit to EPA all proposed operating permits. Id.\n§ 7611d(a)(1). If any permit contains a provision that the\nAdministrator determines is not in compliance with the Act, the\nAdministrator must object in writing, and provide a statement\nof reasons for the objection, within forty-five days after\nreceiving a copy of the proposed permit. Id. § 7661d(b)(1). If,\nwithin ninety days of an EPA objection, a permitting authority\nfails to submit a revised permit that satisfies the objection, the\nAdministrator must issue or deny the permit in accordance with\nthe Act. Id. § 7661d(c). Notably, refusing to revise a permit to\nconform with an EPA objection does not expose a permitting\nauthority to any sort of penalty or liability whatsoever. If the\nAdministrator does not object in writing within forty-five days\nof receiving a proposed permit, any person – including a\nregulated source or a regulatory beneficiary – may, within\nsixty days after EPA’s forty-five-day objection period expires,\npetition the Administrator to object. Id. § 7661d(b)(2). The\nAdministrator must grant or deny such a petition within sixty\ndays after it is filed. Id.\n\n Importantly, for reasons that will become clear, § 7661d\nspecifies: (1) that “[n]o objection shall be subject to judicial\nreview until the Administrator takes final action to issue or\ndeny a permit under this subsection,” id. § 7661d(c); and (2)\nthat the Administrator’s denial of a petition to object “shall be\nsubject to judicial review under section 7607,” id.\n§ 7661d(b)(2). In turn, § 7607 contains the Act’s umbrella\njudicial review provision, which confers jurisdiction in the\n\n 9\nappropriate circuit for regionally applicable final action of the\nAdministrator and in this Court for, inter alia, final action of\nthe Administrator that is “nationally applicable.” Id.\n§ 7607(b)(1).\n\n With an understanding of the major statutory provisions\nand some of the regulations that govern the Wehrum Memo,\nwe now provide fuller descriptions of the Wehrum Memo’s\npredecessor, the Seitz Memo, and the Wehrum Memo itself.\nWhere appropriate, we take care to note additional applicable\nCAA provisions and EPA regulations.\n\n In 1995, without providing notice and comment, John\nSeitz – then Director of EPA’s Office of Air Quality Planning\nand Standards – issued a memorandum to “clarify when a\nmajor source of [hazardous air pollutants] can become an area\nsource” under § 112. J.A. 232 (underline in original). A major\nsource may reclassify to an area source by limiting its potential\nto emit to below the major source threshold, the Seitz Memo\nconcluded, only until the first date on which it must comply\nwith a MACT standard or any other substantive regulatory\nrequirement under the Act. Id. at 236. The Seitz Memo referred\nto this policy as “once in, always in.” Id. In other words, under\nthe Seitz Memo, once EPA classifies a source as major under\n§ 112 and its first compliance date passes, the source is\nineligible to reclassify as an area source, even if it takes an\nenforceable limit on its potential to emit to below the major\nsource threshold. Despite EPA’s stated intention to do so, see\nJ.A. 234, the agency never formalized the Seitz Memo through\nnotice and comment rulemaking. Nevertheless, the Seitz Memo\nhas remained in effect for nearly twenty-five years.\n\n On January 25, 2018, however, EPA announced it was\nreversing course. That day, William L. Wehrum, Assistant\nAdministrator for EPA’s Office of Air and Radiation, and\n\n 10\n“principal adviser to the Administrator in matters pertaining to\nair and radiation programs,” 40 C.F.R. § 1.41, issued a four-\npage memo to the agency’s Regional Air Division Directors; it\nannounced that EPA would no longer interpret § 112 in\naccordance with the Seitz Memo. Indeed, the Wehrum Memo\nexplains, the agency cannot interpret § 112 in accordance with\nthe Seitz Memo because the statute’s plain-language “compels\nthe conclusion” that a major source becomes an area source at\nsuch time when it takes an enforceable limit on its potential to\nemit to below the major source threshold. J.A. 1. Congress, the\nWehrum Memo argues, placed no “temporal limitations” on\nwhen a major source is eligible to reclassify as an area source.\nId. at 3. Accordingly, the Wehrum Memo declares that when\na source previously classified as major limits its potential to\nemit to below the major source threshold, it “will no longer be\nsubject either to the major source MACT or other major source\nrequirements that were applicable to it as a major source under\nCAA section 112.” Id. at 1. In addition, the Wehrum Memo\nstates that it “supersedes” the Seitz Memo, id., and it instructs\nthat “[t]he Regional offices should send this memorandum to\nstates within their jurisdiction,” id. at 4.\n\n II.\n\n Before explaining why the Wehrum Memo is not final\nagency action, we take a moment to clarify the proper test for\nfinality. In this Court, its contours have become blurred amidst\nthe “considerable smog,” Ass’n. of Flight Attendants v. Huerta,\n785 F.3d 710, 717 (D.C. Cir. 2015), enshrouding the related\nbut separate analysis of whether an agency action is a\nlegislative rule. In Flight Attendants, for example, we framed\nthe finality inquiry as asking whether an action is “non-\nbinding” or a “binding legislative rule,” Flight Attendants, 785\nF.3d at 716, and we held that the guidance document at issue\nwas nonfinal because it was “not a legislative rule carrying the\n\n 11\n‘force and effect of law,’” id. (quoting Perez, 135 S. Ct. at\n1204). Likewise, in National Mining, we opined that in order\nto analyze whether an action is final, we must first “take a step\nback” and analyze whether the rule is a legislative rule,\ninterpretive rule, or general statement of policy. Nat’l Min.\nAss’n, 758 F.3d at 251-52. The most important factor in this\nanalysis, we continued, is whether an action has “actual legal\neffect,” id. at 252, and we held that the action at issue did not\nand was therefore unreviewable, id. at 252-53.\n\n Subsuming the finality analysis within the test for whether\na rule is legislative is not always inappropriate; if a rule is\nlegislative it has the force and effect of law, and a legislative\nrule is thus necessarily final. As the Supreme Court has twice\nreminded us within the last five years, however, if a rule is final\nit is not necessarily legislative, and therefore the finality\nanalysis is distinct from the test for whether an agency action\nis a legislative rule.\n\n In United States Army Corps of Engineers v. Hawkes Co.,\n136 S. Ct. 1807 (2016), the Court affirmed that the two-prong\ntest in Bennett v. Spear, 520 U.S. 154 (1997), remains finality’s\ntouchstone, see Hawkes, 136 S. Ct. at 1813 (quoting Bennett,\n520 U.S. at 177-78) (“First, the agency action must mark the\nconsummation of the agency’s decisionmaking process . . . .\nAnd second, the action must be one by which rights or\nobligations have been determined, or from which legal\nconsequences will flow.”). In Hawkes, the question of whether\nthe agency action at issue was the consummation of the\nagency’s decisionmaking process was not in dispute.\nAccordingly, the Court’s analysis focused on whether the\naction satisfied the second prong of Bennett. Notably, in\nundertaking this inquiry, the Court neither asked whether the\naction at issue had the force and effect of law nor made a single\nmention of legislative rules. Rather, the Court’s inquiry\n\n 12\nfocused on whether the action at issue gave “rise to ‘direct and\nappreciable legal consequences.’” Hawkes, 136 S. Ct. at 1814\n(quoting Bennett, 520 U.S. at 178).\n\n Perez, too, makes clear that the finality analysis is distinct\nfrom the test for whether a rule is legislative. There, the Court\naffirmed the “longstanding recognition that interpretive rules\ndo not have the force and effect of law.” Perez, 135 S. Ct. at\n1208 (internal citations omitted). Accordingly, overruling\nParalyzed Veterans of Am. v. D.C. Arena L.P., 117 F.3d 579\n(D.C. Cir. 1997), the Court held that the APA does not require\nan agency to provide notice and comment in amending an\ninterpretive rule, even if the new rule deviates significantly\nfrom its predecessor. Id. at 1206. In so holding, the Court\nreassured regulated entities and regulatory beneficiaries that\nthey are not without recourse should an agency – perhaps to\nevade notice and comment – repudiate a longstanding\ninterpretive rule by way of a second interpretive rule. Id. at\n1209. In such a circumstance, the Court explained, an affected\nparty can seek judicial review pursuant to the APA. Id. Because\nonly final agency action is reviewable under the APA, see 5\nU.S.C. § 704, Perez thus affirms that interpretive rules can be\nfinal, and, by implication, that the test for finality is\nindependent of the analysis for whether an agency action is a\nlegislative rule rather than an interpretive rule.\n\n As commentators explain, maintaining an independent\nfinality analysis is not merely a theoretical nicety; it has several\nsalutary effects in practice. For example – as Perez alludes to,\nsee 135 S. Ct. at 1209 – maintaining a finality analysis that is\ndistinct from the test for whether a rule is legislative permits\ncourts to review nonlegislative rules and thus safeguards\nagainst agencies evading both judicial review and notice and\ncomment by acting via nonlegislative rules. See William Funk,\nFinal Agency Action After Hawkes, 11 N.Y.U. J. L. & LIBERTY\n\n 13\n285, 304 (2017). This is especially important when viewed\nfrom the perspective of regulatory beneficiaries, who are\ngenerally not parties to enforcement actions, and, therefore,\nmay only be able to challenge nonlegislative rules via judicial\nreview. See Nina A. Mendelson, Regulatory Beneficiaries and\nInformal Agency Policymaking, 92 CORNELL L. REV. 397, 420-\n24 (2007).\n\n Having clarified the proper test for finality, we now apply\nit to the Wehrum Memo. Consistent with the interpretive\nmethod we endorse herein, we hew closely to the CAA\nprovisions and EPA regulations appertaining thereto.\n\n Our first question is whether the Wehrum Memo “mark[s]\nthe consummation of [EPA’s] decisionmaking process.”\nHawkes, 136 S. Ct. at 1813 (quoting Bennett, 520 U.S. at 177-\n78). It does. Notably, neither Respondent nor Intervenor offer\nsubstantive argument to the contrary. They were smart to save\ntheir ink. The Wehrum Memo unequivocally states that the\nplain language of § 112 “compels” the legal conclusion that\nqualifying major sources can reclassify at such time that they\ntake an enforceable limit on their potential to emit to below the\nmajor source threshold. J.A. 1. In other words, the Wehrum\nMemo does not advance what EPA believes is a reasonable\ninterpretation of § 112; it advances what EPA believes is the\nonly permissible interpretation of the statute. Moreover, no\nmere subordinate issued the Wehrum Memo. Far from it. The\nAssistant Administrator for the Office of Air and Radiation\nissued it. As discussed, under EPA regulations, he is the\n“principal advisor to the Administrator in matters pertaining to\nair and radiation,” see 40 C.F.R. § 1.41, and, as we have held\npreviously with respect to the Assistant Administrator for the\nOffice of Air and Radiation, nothing within EPA’s regulations\nprovides us “reason to question his authority to speak for the\nEPA.” Her Majesty the Queen in Right of Ontario v. EPA, 912\n\n 14\nF.2d 1525, 1532 (D.C. Cir. 1990) (internal citations omitted).\nCf. Soundboard Ass’n. v. FTC, 888 F.3d 1261, 1267-69 (D.C.\nCir. 2018), cert denied 139 S. Ct. 1544, 2019 WL 1590248\n(Apr. 15, 2019) (Federal Trade Commission (“FTC”) staff\nopinion letter not consummation of agency’s decisionmaking\nprocess because FTC regulations expressly delineated between\nCommission advice and staff advice and provided petitioners\nopportunity to seek opinion from Commission itself).\nMoreover, EPA published notice of the Wehrum Memo, and\nreiterated its principal conclusion, in the Federal Register. See\n83 Fed. Reg. 5543-01 (Feb. 8, 2018). Accordingly, the\nWehrum Memo can only reasonably be described as EPA’s last\nword on when a major source can reclassify to an area source\nunder § 112.\n\n Because the Wehrum Memo satisfies Bennett’s first prong,\nwe ask next whether it has “direct and appreciable legal\nconsequences.” Hawkes, 136 S. Ct. at 1814 (quoting Bennett,\n520 U.S. at 178). Petitioners argue that it does because it\ncreates a new right – i.e. it allows major sources unable to\nreclassify to area sources under the Seitz Memo to so\nreclassify. Cal. Pet’rs’ Br. 17-20. Respondent counters that the\nWehrum Memo does not change the rights of regulated\nsources. EPA Br. 26-28. Whether or not a regulated source has\nthe right to reclassify, Respondent contends, is only determined\nwithin the Title V permitting process. Id. (citing 42 U.S.C. §\n7661d).\n\n For reasons now explained, we hold that the Wehrum\nMemo does not have a single direct and appreciable legal\nconsequence.\n\n Hawkes instructs that whether an agency action has direct\nand appreciable legal consequences is a “‘pragmatic’” inquiry.\nId. at 1815 (quoting Abbot Labs. v. Gardner, 387 U.S. 136, 149\n\n 15\n(1967)). In characterizing the inquiry as pragmatic, we do not\ntake the Court to be encouraging some sort of common-sense\napproach. Quite the opposite. We take it as counseling lower\ncourts to make Bennett prong-two determinations based on the\nconcrete consequences an agency action has or does not have\nas a result of the specific statutes and regulations that govern\nit. Thus, in Hawkes, the Court held, in part, that the agency\ndetermination at issue had direct and appreciable legal\nconsequences because, under the applicable statutes and\nregulations, if petitioners failed to heed the determination they\ndid so at the risk of significant criminal and civil penalties. Id.\nAnd the cases Hawkes relies on as past examples of the\n“pragmatic approach [the Court] has long taken to finality”\nhold similarly. Id. (citing and quoting Sackett v. EPA, 566 U.S.\n120, 126 (2012) (holding that agency action at issue satisfied\nBennett prong-two because, under the relevant statutes and\nregulations, it appeared to expose petitioners to double\npenalties in a future enforcement proceeding and to limit their\nability to obtain a certain type of permit); Abbot Labs., 387 U.S.\nat 152 (holding that action at issue had a “sufficiently direct\nand immediate” impact on petitioners, such that judicial review\nwas appropriate, because, under the governing statutes and\nregulations, noncompliance risked “serious criminal and civil\npenalties”); Frozen Food Express v. United States, 351 U.S.\n40, 44 (1956) (same)).\n\n Quite recently, in Valero Energy Corporation v. EPA, No.\n18-1028, 2019 WL 2587837 (D.C. Cir. June 25, 2019), we\naffirmed this approach. At issue there, like here, was whether\nan EPA guidance document that declared the agency’s\ninterpretation of a statute was final under the Act. We held that\nit was not. Assessing it within the context of the Act, we\nemphasized that: (1) the guidance imposed no obligations,\nprohibitions, or restrictions; (2) it put no party to the choice\nbetween costly compliance and the risk of a penalty of any sort;\n\n 16\n(3) EPA acknowledged at oral argument that the guidance had\nno independent legal authority; and (4) that the Act provided\nregulated parties a statutory mechanism by which to challenge\nany EPA action that was premised on the statutory\ninterpretation that the guidance advanced. Valero, 2019 WL at\n*3-5.\n\n Assessing the Wehrum Memo under Hawkes and in\naccordance with Valero, we find that it is not final. True, it\nunequivocally declares that major sources, at such time that\nthey limit their potential to emit to below the major source\nthreshold, “will no longer be” subject to MACT standards.\nJ.A. 1. Viewed within the context of the Act, however, the\nWehrum Memo is all bark and no bite. As Respondent averred\ntwice at oral argument, neither EPA nor any regulated source\ncan rely on the Wehrum Memo within the Title V permitting\nprocess or in any other proceeding. Oral Arg. 50:15-50:27,\n1:01:13-1:01:50. In other words, as Respondent concedes,\nalthough the Wehrum Memo forecasts EPA’s position as to\n§ 112, it has no independent legal authority. In addition, under\nthe Act and EPA regulations, a state permitting authority that\nrefuses to comply with the Wehrum Memo faces no penalty or\nliability of any sort. Further still, the instant matter does not\npresent a circumstance where the action at issue may be legally\nconsequential because its binds agency staff and affected\nparties have no means (outside of judicial review) by which to\nchallenge it. To the contrary, the Act contains clear provisions\npursuant to which: (1) a state permitting authority can refuse to\napply the Wehrum Memo and seek judicial review if EPA\nissues a permit over its refusal, id. § 7661d(c); and (2) a\nregulatory beneficiary can petition EPA to object to a state\npermitting authority’s application of the Wehrum Memo and\nseek judicial review if EPA denies the petition, id. §\n7661d(b)(2).\n\n 17\n Accordingly, although the Wehrum Memo, in no uncertain\nterms, forecasts EPA’s definitive interpretation of § 112, it has\nno direct and appreciable legal consequences: neither EPA nor\nregulated sources can rely on it as independently authoritative\nin any proceeding; state permitting authorities face no penalty\nor liability of any sort in ignoring it; and state permitting\nauthorities and regulatory beneficiaries have clear statutory\navenues by which to challenge a permitting decision adopting\nthe reasoning of the Wehrum Memo and seek judicial review\nif EPA fails to sustain their challenges. Under § 7661d(c), if a\nstate permitting authority refuses to issue a permit allowing a\nmajor source to reclassify as an area source, and EPA\nsubsequently issues such a permit following the reasoning of\nthe Wehrum Memo, judicial review is appropriate. Under\n§ 7661d(b)(2), if EPA, following the reasoning of the Wehrum\nMemo, denies a petition from any person asking the agency to\nobject to a state permitting authority’s issuance of a permit that\nallows a major source to reclassify as an area source, judicial\nreview is appropriate. Indeed, because Congress specified that\n“[n]o objection shall be subject to review until the\nAdministrator takes final action to issue or deny a permit under\nthis subsection,” § 7661d(c), we would contravene\nCongressional intent if we were to hold that a memo that\nmerely forecasts a future objection is final agency action and\nsubject to judicial review at this time.\n\n The dissent insists that the Wehrum Memo satisfies\nBennett’s second prong because it “altered the legal regime.”\nDis. Op. 12. Indeed, the dissent forewarns, the Wehrum Memo\n“commands, orders, and dictates [to]” EPA employees, id. at 4,\nand “state permitting authorities are subject to” the statutory\ninterpretation it advances, id. Said differently, according to the\ndissent, because of the Wehrum Memo, sources subject to\nMACT standards that limit their potential to emit to below the\nmajor source threshold are now “assured that they will be\n\n 18\nsubject to decreased regulation with EPA’s support.” Id. at 12\n(emphasis added).\n\n While the question is not free from doubt, we respectfully\ndisagree. As noted above, we must remain laser focused on\nwhether the Wehrum Memo gives “rise to ‘direct and\nappreciable legal consequences.’” Hawkes, 136 S. Ct. at 1814\n(quoting Bennett, 520 U.S. at 178), and when viewed in its\nspecific regulatory context, it does not. “[M]ajor sources must\ncomply with technology-based emission standards requiring\nthe maximum degree of reduction in emissions EPA deems\nachievable, . . . [and] [i]n order to obtain an operating permit\nunder title V of the [CAA], major sources must comply with\nextensive monitoring, reporting and record-keeping\nrequirements. Nat'l Min. Ass'n v. U.S. E.P.A., 59 F.3d 1351,\n1353 (D.C. Cir. 1995). Major sources must obtain a permit in\norder to operate, and unless and until that permit is amended or\nset aside, the stringent requirements set forth therein must be\ncomplied with while that equipment is operational. The\nWehrum Memo itself does not revoke or amend a single\npermit. As acknowledged by the Ohio environmental\nauthorities in materials cited by petitioners, “[i]f you want to\ntake advantage of the new guidance [in the Wehrum Memo],\nyou will need to submit an application to modify your current\npermit.” Environmental Pet’rs’ Br., Standing Addendum 0198.\nAssuredly, although the Wehrum Memo advises EPA\nemployees of the agency’s position as to § 112, it does not bind\nstate permitting authorities or assure regulated entities of the\nability to reclassify. As EPA concedes, EPA Br. 21, 25, in\nreceiving such an application to modify a permit, a state\npermitting authority may – with total impunity – ignore the\nWehrum Memo and deny the application. It is true that the\nAdministrator must issue a revised permit over the state\npermitting authority’s protest if he or she believes that the\nstatute so requires, § 7661d(c), but in such a case, the statute\n\n 19\nexplicitly provides the state permitting authority a mechanism\nby which to seek judicial review of the Administrator’s action.\nId. Regardless of whether Congress generally intended to\nallow pre-enforcement review of guidance documents under\nsome circumstances in the CAA, here, as described above,\nCongress specifically directed that judicial review shall not be\navailable until the Title V permit amendment process reaches a\nconclusion, see §§7661d(b)(2), 7661d(c). Congress’ explicit\nunderstanding of finality in this specific statutory context\ncontrols our consideration of the instant guidance document,\nwhich pertains to that same permit amendment process.\n\n ***\n\n Before concluding, we note that we have twice had\noccasion to ask whether an EPA guidance document that\nimplicated the Act’s Title V permitting process was final\nagency action: first in Appalachian Power Co. v. EPA, 208\nF.3d 1015, 1021 (D.C. Cir. 2000), then in National\nEnvironmental Development Ass’n v. EPA, 752 F.3d 999 (D.C.\nCir. 2014). In each, we found that the guidance at issue was\nfinal. A brief analysis of our reasoning in those cases\ndemonstrates why the Wehrum Memo is not.\n\n In Appalachian Power, at issue was a nineteen-page\nguidance document relating to certain monitoring requirements\nfor Title V sources. Appalachian Power, 208 F.3d at 1019-20.\nIn assessing our jurisdiction over the guidance, we did not\nconsider § 7611d. Instead, we framed our jurisdiction solely in\nterms of § 7607(b)(1). See id. at 1021 n. 10 (“Our jurisdiction\nextends to ‘any . . . nationally applicable . . . final action taken\nby,’ the EPA ‘Administrator.’”) (quoting 42 U.S.C.\n§ 7607(b)(1))). We predicated our holding that the guidance\nwas final on the following findings. First, we found that it\nrequired state permitting authorities to: (1) “review their\n\n 20\nemission standards and the emission standards EPA has\npromulgated to determine if the standards provide enough\nmonitoring;” and (2) “insert additional monitoring\nrequirements as terms or conditions of a permit . . . if they\nbelieve existing requirements are inadequate, as measured by\nEPA’s multi-factor, case-by-case analysis set forth in the\nGuidance.” Id. at 1022. Second, we found that EPA did not\ndispute petitioners’ assertion that state permitting authorities\nwere relying on EPA’s guidance in insisting that regulated\nsources utilize a monitoring method that was more burdensome\nthan the monitoring method set out under existing EPA\nregulations. Id. at 1023 & n.17. Finally, we found that a\nchallenge to an individual permit applying the guidance would\nnot be heard in this Court, presumably because we felt any such\nchallenge would have only regional implications. Id. at 1023 n.\n18.\n\n In National Environmental, the guidance document before\nus explained that, due to a decision of the Sixth Circuit, EPA\nwas altering a certain interpretation of its regulations only for\nTitle V sources located within the Sixth Circuit’s jurisdiction.\nNat’l Envtl., 752 F.3d at 1003. As in Appalachian Power, in\nassessing our jurisdiction over the guidance, we asked only\nwhether it was final under § 7607(b)(1) and made no mention\nof § 7661d. Id. at 1006. In holding that it was final, we found\nthat the “finality and legal consequences” of the guidance\n“were made plain” when EPA “relied on [it]” in approving a\nFederal Implementation Plan (“FIP”) “involving a company\nlocated outside the jurisdiction of the Sixth Circuit.” Id. at 1007\n(citing 78 Fed. Reg. 17836, 17842 & n. 10 (March 22, 2013)).\nIndeed, within the FIP approval – which is a final, legislative\nrule carrying the force and effect of law – EPA cited the\nguidance as the sole authority for the legal conclusion that\ncertain regulations applied to certain sources located outside of\n\n 21\nthe Sixth Circuit’s jurisdiction. See 78 Fed. Reg. 17836, 17842\n& n.10 (March 22, 2013)\n\n Appalachian Power and National Environmental are thus,\ncontrary to what the dissent suggests, see Dis. Op. 2-3, distinct\nfrom the instant matter in a crucial respect. In those cases, we\nheld that the guidance documents at issue were final under §\n7607(b)(1), without reference to § 7661d, because EPA and\nstate permitting authorities wielded them to effectuate legal\nconsequences. In Appalachian Power, we found that the\nguidance at issue required state permitting authorities to take at\nleast two specific actions and that EPA did not deny that state\npermitting authorities used it to coerce regulated sources to\nadopt a stricter monitoring method. In National Environmental,\nwe found that EPA cited the guidance, within a binding FIP\napproval, as the sole authority in support of a legal conclusion.\nBy contrast, the Wehrum Memo does not require any entity or\nperson to do anything, and EPA concedes that it has not, will\nnot, and cannot rely on it in any proceeding. Accordingly,\nunlike in Appalachian Power and National Environmental, we\nhave no basis to conclude, without reference to § 7661d, that\nwe have jurisdiction over the guidance before us under\n§ 7607(b)(1). We note, in addition, that in Appalachian Power,\nwe found that we would lack jurisdiction over challenges to\npermitting decisions applying the guidance at issue. Here,\nhowever, any party entitled to review under § 7661d that\nwishes to challenge an application of the Wehrum Memo in this\nCourt will be so heard, since the Wehrum Memo’s principal\nconclusion is nationally applicable. See § 7607(b)(1).\n\n In sum, we find that the Wehrum Memo – assessed within\nthe context of the Act and EPA regulations – is not final agency\naction, and we dismiss the petitions for lack of subject matter\njurisdiction under the Act. The Wehrum Memo marks the\nconsummation of EPA’s decisionmaking process as to when a\n\n 22\nmajor source may reclassify to an area source under § 112. But\nthe Wehrum Memo does not have direct and appreciable legal\nconsequences: it does not require anyone to do anything;\nneither EPA nor regulated sources can rely on it in any\nproceeding; state permitting authorities face no penalty or\nliability in ignoring it; state permitting authorities and\nregulatory beneficiaries have clear statutory avenues by which\nto challenge it and seek judicial review if EPA refuses to heed\ntheir challenges; and any such challenges, if so desired, will be\nheard in this Court.\n\n III.\n\n To conclude, we note that we are under no illusion that this\nopinion will be the Rosetta Stone of understanding the nature\nof agency action. Developing this area of the law is indeed an\n“important continuing project.” Nat’l Min. Ass’n, 758 F.3d at\n251. Nonetheless, today we humbly submit our contribution\ntoward clarifying this somewhat gnarled field of jurisprudence.\nIn ascertaining the nature of an agency action, we emphasize,\ncourts should look first to the matrix of statutes and regulations\ngoverning that specific action. In addition, we offer a gentle\nreminder that the finality analysis is sui generis, separate and\ndistinct from the test for whether an agency action is a\nlegislative rule.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: CALIFORNIA COMMUNITIES AGAINST TOXICS, Et Al., Petitioners v. ENVIRONMENTAL PROTECTION AGENCY and Andrew Wheeler, Administrator, U.S. Environmental Protection Agency, Respondents Air Permitting Forum, Et Al., Intervenors\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-20\\nOpinion type: majority\\nHas dissenting opinion: Yes\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 1, 2019 Decided August 20, 2019\\n\\n No. 18-1085\\n\\n CALIFORNIA COMMUNITIES AGAINST TOXICS, ET AL.,\\n PETITIONERS\\n\\n v.\\n\\n ENVIRONMENTAL PROTECTION AGENCY AND ANDREW\\n WHEELER, ADMINISTRATOR, U.S. ENVIRONMENTAL\\n PROTECTION AGENCY,\\n RESPONDENTS\\n\\n AIR PERMITTING FORUM, ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 18-1095, 18-1096\\n\\n\\n On Petitions for Review of Action of the\\n United States Environmental Protection Agency\\n\\n\\n Sanjay Narayan argued the cause for Environmental\\nPetitioners. With him on the briefs were James S. Pew, Tomás\\nE. Carbonell, Vickie Patton, Surbhi Sarang, John Walke, Emily\\nDavis, Thomas Zimpleman, and Keri N. Powell.\\n\\n Kavita P. Lesser, Deputy Attorney General, Office of the\\nAttorney General for the State of California, argued the cause\\n\\n 2\\nfor petitioner State of California. With her on the briefs were\\nXavier Becerra, Attorney General, David A. Zonana, Deputy\\nAttorney General, and Jonathan Wiener, Deputy Attorney\\nGeneral.\\n\\n Eric Grant, Attorney, U.S. Department of Justice, argued\\nthe cause for respondents. On the brief were Jeffrey Bossert\\nClark, Assistant Attorney General, Jonathan D. Brightbill,\\nDeputy Assistant Attorney General, and Scott Jordan,\\nAttorney, U.S. Environmental Protection Agency.\\n\\n Shannon S. Broome argued the cause for intervenors-\\nrespondents Air Permitting Forum, et al. With her on the briefs\\nwere Charles H. Knauss, Leslie Sue Ritts, Makram B. Jaber,\\nand Andrew D. Knudsen.\\n\\n David M. Friedland, Leslie A. Hulse, Felicia H. Barnes,\\nSteven P. Lehotsky, Michael B. Schon, and amici curiae\\nAmerican Chemistry Council, et al. in support of respondents.\\n\\n Before: ROGERS and WILKINS, Circuit Judges, and\\nSILBERMAN, Senior Judge.\\n\\n Opinion for the Court filed by Circuit Judge WILKINS.\\n\\n Dissenting Opinion filed by Circuit Judge ROGERS.\\n\\n This case asks us to determine the nature of an agency\\naction, an inquiry that – paradoxically – is quotidian but\\nabstruse. When we are confronted with agency action, the\\nlitany of questions is by now very well-rehearsed: Is it final? Is\\nit ripe? Is it a policy statement? Is it an interpretive rule? Is it a\\nlegislative rule? Despite the clarity of these questions,\\nhowever, predictable answers have eluded courts and\\ncommentators. See, e.g., Perez v. Mortg. Bankers Ass’n., 135\\n\\n 3\\nS. Ct. 1199, 1204 (2015) (describing the question of how to\\ndistinguish between legislative and interpretive rules as “the\\nsource of much scholarly and judicial debate”); Ticor Title Ins.\\nCo. v. FTC, 814 F.2d 731, 745 (D.C. Cir. 1987) (opinion of\\nWilliams, J.) (characterizing the law governing finality and\\nripeness as “chaotic”); Ronald M. Levin, Rulemaking and the\\nGuidance Exemption, 70 ADMIN L. REV. 264, 348 (2018) (“The\\nstandard view among commentators is that [distinguishing\\nbetween legislative and nonlegislative rules] is exceptionally\\nperplexing and incoherent.”). Indeed, the nature of agency\\naction, it seems, is too often in the eye of the beholder. We\\nresolve the instant matter, therefore, with our eye toward the\\n“continuing project” of clarifying this “byzantine” area of the\\nlaw. Nat’l Min. Ass’n v. McCarthy, 758 F.3d 243, 246 (D.C.\\nCir. 2014).\\n\\n The agency action before us is a 2018 memorandum\\n(“Wehrum Memo”) that William L. Wehrum, Assistant\\nAdministrator for the Environmental Protection Agency’s\\n(“EPA”) Office of Air and Radiation, issued to all Regional Air\\nDivision Directors. The Wehrum Memo declares that the plain\\nlanguage of § 112 of the Clean Air Act (“Act” or “CAA”), 42\\nU.S.C. § 7412, compels the conclusion that a source of toxic\\nemissions classified as “major” can reclassify to an “area\\nsource,” and thereby ease its regulatory burden, at any time\\nafter it limits its potential to emit to below the major source\\nthreshold. J.A. 1. The Wehrum Memo states that it supersedes\\na prior 1995 EPA memorandum (“Seitz Memo”) issued by\\nJohn Seitz, then Director of EPA’s Office of Air Quality\\nPlanning and Standards, which interpreted § 112 to mean that\\nonce EPA classifies a source as major, that source can never\\nreclassify to area source status, even if it limits its potential to\\nemit to below the major source threshold. Id.\\n\\n 4\\n Petitioners are the State of California and a group of\\nenvironmental organizations whose citizens and members,\\nrespectively, breathe the air in the vicinity of regulated sources.\\nEPA is the Respondent, and a group of industry organizations\\nhave joined as Intervenor. Petitioners contend that we can and\\nshould review the Wehrum Memo because it is final agency\\naction and prudentially ripe. Moreover, Petitioners argue, the\\nWehrum Memo is a legislative rule, and it is therefore\\nprocedurally defective under the Administrative Procedure Act\\n(“APA”), 5 U.S.C. § 551 et seq., because EPA failed to provide\\nnotice and comment before issuing it, see id. § 553. But even\\nif we hold that the Wehrum Memo is an interpretive rule (for\\nwhich notice and comment is not required, see id.), Petitioners\\ncontend that we still must vacate it because EPA’s\\ninterpretation of § 112 is incorrect. Respondent and Intervenor\\nretort that this Court lacks jurisdiction over the Wehrum Memo\\nbecause it is not final agency action. Alternatively, they argue,\\nwe should not review it because it is not prudentially ripe. If,\\nhowever, we find the Wehrum Memo final and ripe,\\nRespondent and Intervenor assert, we must deny the petitions\\nbecause it is an interpretive rule and is thus procedurally sound,\\nand its interpretation of § 112 is correct.\\n\\n For the reasons explained herein, we hold that the Wehrum\\nMemo is not final agency action, and we dismiss the petitions\\nfor lack of subject matter jurisdiction under the Act. We\\nexpress no opinion as to whether the Wehrum Memo is\\nprudentially ripe, an interpretive rule or a legislative rule, or on\\nthe merits of its interpretation of § 112. In holding that the\\nWehrum Memo is not final, we emphasize two points. First,\\nwhen assessing the nature of an agency action (including\\nwhether it is final), courts should resist the temptation to define\\nthe action by comparing it to superficially similar actions in the\\ncaselaw. Rather, courts should take as their NorthStar the\\nunique constellation of statutes and regulations that govern the\\n\\n 5\\naction at issue. Second, although all legislative rules are final,\\nnot all final rules are legislative, and the finality analysis is\\ntherefore distinct from the test for whether an agency action is\\na legislative rule.\\n\\n I.\\n\\n Because they share a progenitor, a reliable approach to\\nunderstanding a James Baldwin novel is to compare it,\\naccording to a set of criteria, to another work in his oeuvre.\\nIndeed, a thematic reading of Giovanni’s Room is sure to\\ninform such a reading of The Fire Next Time, and vice versa.\\nNot so, however, with respect to the broad set of phenomena\\nwe categorize as agency action. Because few, if any, of them\\nare governed by the exact same combination of statutes and\\nregulations, it is a mistake to assume – even if they appear\\nfacially similar – that they can lend each other definition\\nthrough comparison, or that they are decipherable under a\\ncommon rubric. Rather, to ascertain the nature of an agency\\naction, courts should ground the analysis in the idiosyncratic\\nregime of statutes and regulations that govern it. We have great\\nsympathy for the desire to develop a one-size-fits-all heuristic.\\nSee, e.g., Nat’l Min., 758 F.3d at 251 (“. . . all relevant parties\\nshould instantly be able to tell whether an agency action is a\\nlegislative rule, an interpretive rule, or a general statement of\\npolicy . . . .”). But this desire is perhaps misplaced, since, as we\\nonce said of interpretation itself, agency action is “a chameleon\\nthat takes its color from its context.” American Min. Congress\\nv. EPA, 995 F.2d 1106, 1111 (D.C. Cir. 1993).\\n\\n Accordingly, we turn first to the CAA provisions and EPA\\nregulations that govern the Wehrum Memo.\\n\\n Congress enacted the CAA, 42 U.S.C. § 7401 et seq., to\\n“protect and enhance the quality of the Nation’s air\\n\\n 6\\nresources . . . . ” Id. § 7401(b)(1). Toward this end, § 112\\nrequires EPA to regulate “Hazardous Air Pollutants,” i.e. toxic\\nemissions such as chloroform. Id. § 7412. Congress\\nestablished an initial list of hazardous air pollutants, id.\\n§ 7412(b)(1), but the Act requires EPA to curate it, deleting or\\nadding hazardous air pollutants over time according to certain\\ncriteria, id. § 7412(b)(2)-(3). Based on this list, the Act\\nmandates EPA to create a second list of categories of sources\\nof hazardous air pollutants, id. § 7412(c), like asphalt\\nprocessing plants and industrial dry-cleaning facilities, see\\nRevision of Source Category List under Section 112 of the\\nClean Air Act, 70 Fed. Reg. 37819-01 (June 30, 2005).\\nImportantly, the Act distinguishes between “major” and “area”\\nsources. Id. § 7412(a)(1)-(2). According to the Act’s\\ndefinitional provisions, a major source means any source\\nwithin a listed category that “emits or has the potential to emit\\nconsidering controls, in the aggregate, 10 tons per year or more\\nof any [listed hazardous air pollutant] or 25 tons per year or\\nmore of any combination of [listed hazardous air pollutants].”\\nId. § 7412(a)(1). Area source means “any stationary source of\\n[hazardous air pollutants] that is not a major source.” Id.\\n§ 7412(a)(2).\\n\\n Whether EPA classifies a source as major or area has\\nmajor consequences for both sources of hazardous air\\npollutants, which must comply with emissions standards, and\\nregulatory beneficiaries, who live, work, recreate – and thus\\nregularly breathe the air – near sources of hazardous air\\npollutants. For major sources, the Act requires EPA to establish\\nstringent emissions caps that result in “the maximum degree of\\nreduction in emissions . . . (including a prohibition on such\\nemissions, where achievable).” Id. § 7412(d). EPA refers to\\nthese emissions limitations as “Maximum Achievable Control\\nTechnology” (“MACT”) standards. J.A. 1. The Act mandates\\nthat MACT standards be “no less stringent than the emission\\n\\n 7\\ncontrol that is achieved in practice by the best controlled\\nsimilar source.” Id. § 7412(d)(3). By contrast, for area sources,\\nEPA need not set emissions caps at all, save under limited\\ncircumstances. See id. § 7412(c)(3). Moreover, where the\\nagency chooses to cap emissions for an area source, it may set\\nemissions limits based on “Generally Available Control\\nTechnology” (“GACT”) standards, which are far more lenient\\nthan their MACT counterparts.1\\n\\n Of course, emissions caps are of little use if sources do not\\ncomply with them. Presumably in recognition of this, Congress\\nenacted Title V of the CAA, 42 U.S.C. § 7661 et seq., which\\nmakes it unlawful for a source subject to regulations under the\\nAct – including GACT or MACT standards under § 112 – to\\noperate without a permit, see id. § 7661a(a). Specifically,\\nwithin a year of becoming subject to an obligation under the\\nAct, Title V requires a source to submit a permit application\\nand compliance plan to a state permitting authority. Id.\\n§ 7661b(b)-(c). In addition, a source must certify its\\ncompliance annually and submit to inspection, monitoring, and\\nreporting requirements. Id. § 7661c(a)-(c). A source may apply\\nto modify its permit, 40 C.F.R. § 70.7(e), and state permitting\\nauthorities must provide for public comment and a hearing on\\nall permit applications that they receive, 42 U.S.C. §\\n7661a(b)(6).\\n\\n But what if a state permitting authority issues or denies a\\npermit application on a ground that a regulated source, or a\\n\\n1\\n As we have observed, the Act does not provide any parameters for\\nsetting GACT standards, but its legislative history describes GACT\\nas “‘methods . . . [that] are commercially available and appropriate\\nfor application . . . considering economic impacts and the technical\\ncapabilities of firms to operate and maintain the emissions control\\nsystems.’” U.S. Sugar Corp. v. EPA, 830 F.3d 579, 595 (D.C. Cir.\\n2016) (quoting S. REP. NO. 101-228, at 171 (1989)).\\n\\n 8\\nregulatory beneficiary believes contravenes the Act? Congress\\napparently foresaw this circumstance, too. Indeed, Title V\\nprovides a detailed administrative process that dictates exactly\\nwhen and how regulated sources and regulatory beneficiaries\\nmay seek EPA review of a state permitting authority’s action,\\nand, ultimately, judicial review of EPA action. See id. § 7611d.\\nThe process works as follows. First, state permitting authorities\\nmust submit to EPA all proposed operating permits. Id.\\n§ 7611d(a)(1). If any permit contains a provision that the\\nAdministrator determines is not in compliance with the Act, the\\nAdministrator must object in writing, and provide a statement\\nof reasons for the objection, within forty-five days after\\nreceiving a copy of the proposed permit. Id. § 7661d(b)(1). If,\\nwithin ninety days of an EPA objection, a permitting authority\\nfails to submit a revised permit that satisfies the objection, the\\nAdministrator must issue or deny the permit in accordance with\\nthe Act. Id. § 7661d(c). Notably, refusing to revise a permit to\\nconform with an EPA objection does not expose a permitting\\nauthority to any sort of penalty or liability whatsoever. If the\\nAdministrator does not object in writing within forty-five days\\nof receiving a proposed permit, any person – including a\\nregulated source or a regulatory beneficiary – may, within\\nsixty days after EPA’s forty-five-day objection period expires,\\npetition the Administrator to object. Id. § 7661d(b)(2). The\\nAdministrator must grant or deny such a petition within sixty\\ndays after it is filed. Id.\\n\\n Importantly, for reasons that will become clear, § 7661d\\nspecifies: (1) that “[n]o objection shall be subject to judicial\\nreview until the Administrator takes final action to issue or\\ndeny a permit under this subsection,” id. § 7661d(c); and (2)\\nthat the Administrator’s denial of a petition to object “shall be\\nsubject to judicial review under section 7607,” id.\\n§ 7661d(b)(2). In turn, § 7607 contains the Act’s umbrella\\njudicial review provision, which confers jurisdiction in the\\n\\n 9\\nappropriate circuit for regionally applicable final action of the\\nAdministrator and in this Court for, inter alia, final action of\\nthe Administrator that is “nationally applicable.” Id.\\n§ 7607(b)(1).\\n\\n With an understanding of the major statutory provisions\\nand some of the regulations that govern the Wehrum Memo,\\nwe now provide fuller descriptions of the Wehrum Memo’s\\npredecessor, the Seitz Memo, and the Wehrum Memo itself.\\nWhere appropriate, we take care to note additional applicable\\nCAA provisions and EPA regulations.\\n\\n In 1995, without providing notice and comment, John\\nSeitz – then Director of EPA’s Office of Air Quality Planning\\nand Standards – issued a memorandum to “clarify when a\\nmajor source of [hazardous air pollutants] can become an area\\nsource” under § 112. J.A. 232 (underline in original). A major\\nsource may reclassify to an area source by limiting its potential\\nto emit to below the major source threshold, the Seitz Memo\\nconcluded, only until the first date on which it must comply\\nwith a MACT standard or any other substantive regulatory\\nrequirement under the Act. Id. at 236. The Seitz Memo referred\\nto this policy as “once in, always in.” Id. In other words, under\\nthe Seitz Memo, once EPA classifies a source as major under\\n§ 112 and its first compliance date passes, the source is\\nineligible to reclassify as an area source, even if it takes an\\nenforceable limit on its potential to emit to below the major\\nsource threshold. Despite EPA’s stated intention to do so, see\\nJ.A. 234, the agency never formalized the Seitz Memo through\\nnotice and comment rulemaking. Nevertheless, the Seitz Memo\\nhas remained in effect for nearly twenty-five years.\\n\\n On January 25, 2018, however, EPA announced it was\\nreversing course. That day, William L. Wehrum, Assistant\\nAdministrator for EPA’s Office of Air and Radiation, and\\n\\n 10\\n“principal adviser to the Administrator in matters pertaining to\\nair and radiation programs,” 40 C.F.R. § 1.41, issued a four-\\npage memo to the agency’s Regional Air Division Directors; it\\nannounced that EPA would no longer interpret § 112 in\\naccordance with the Seitz Memo. Indeed, the Wehrum Memo\\nexplains, the agency cannot interpret § 112 in accordance with\\nthe Seitz Memo because the statute’s plain-language “compels\\nthe conclusion” that a major source becomes an area source at\\nsuch time when it takes an enforceable limit on its potential to\\nemit to below the major source threshold. J.A. 1. Congress, the\\nWehrum Memo argues, placed no “temporal limitations” on\\nwhen a major source is eligible to reclassify as an area source.\\nId. at 3. Accordingly, the Wehrum Memo declares that when\\na source previously classified as major limits its potential to\\nemit to below the major source threshold, it “will no longer be\\nsubject either to the major source MACT or other major source\\nrequirements that were applicable to it as a major source under\\nCAA section 112.” Id. at 1. In addition, the Wehrum Memo\\nstates that it “supersedes” the Seitz Memo, id., and it instructs\\nthat “[t]he Regional offices should send this memorandum to\\nstates within their jurisdiction,” id. at 4.\\n\\n II.\\n\\n Before explaining why the Wehrum Memo is not final\\nagency action, we take a moment to clarify the proper test for\\nfinality. In this Court, its contours have become blurred amidst\\nthe “considerable smog,” Ass’n. of Flight Attendants v. Huerta,\\n785 F.3d 710, 717 (D.C. Cir. 2015), enshrouding the related\\nbut separate analysis of whether an agency action is a\\nlegislative rule. In Flight Attendants, for example, we framed\\nthe finality inquiry as asking whether an action is “non-\\nbinding” or a “binding legislative rule,” Flight Attendants, 785\\nF.3d at 716, and we held that the guidance document at issue\\nwas nonfinal because it was “not a legislative rule carrying the\\n\\n 11\\n‘force and effect of law,’” id. (quoting Perez, 135 S. Ct. at\\n1204). Likewise, in National Mining, we opined that in order\\nto analyze whether an action is final, we must first “take a step\\nback” and analyze whether the rule is a legislative rule,\\ninterpretive rule, or general statement of policy. Nat’l Min.\\nAss’n, 758 F.3d at 251-52. The most important factor in this\\nanalysis, we continued, is whether an action has “actual legal\\neffect,” id. at 252, and we held that the action at issue did not\\nand was therefore unreviewable, id. at 252-53.\\n\\n Subsuming the finality analysis within the test for whether\\na rule is legislative is not always inappropriate; if a rule is\\nlegislative it has the force and effect of law, and a legislative\\nrule is thus necessarily final. As the Supreme Court has twice\\nreminded us within the last five years, however, if a rule is final\\nit is not necessarily legislative, and therefore the finality\\nanalysis is distinct from the test for whether an agency action\\nis a legislative rule.\\n\\n In United States Army Corps of Engineers v. Hawkes Co.,\\n136 S. Ct. 1807 (2016), the Court affirmed that the two-prong\\ntest in Bennett v. Spear, 520 U.S. 154 (1997), remains finality’s\\ntouchstone, see Hawkes, 136 S. Ct. at 1813 (quoting Bennett,\\n520 U.S. at 177-78) (“First, the agency action must mark the\\nconsummation of the agency’s decisionmaking process . . . .\\nAnd second, the action must be one by which rights or\\nobligations have been determined, or from which legal\\nconsequences will flow.”). In Hawkes, the question of whether\\nthe agency action at issue was the consummation of the\\nagency’s decisionmaking process was not in dispute.\\nAccordingly, the Court’s analysis focused on whether the\\naction satisfied the second prong of Bennett. Notably, in\\nundertaking this inquiry, the Court neither asked whether the\\naction at issue had the force and effect of law nor made a single\\nmention of legislative rules. Rather, the Court’s inquiry\\n\\n 12\\nfocused on whether the action at issue gave “rise to ‘direct and\\nappreciable legal consequences.’” Hawkes, 136 S. Ct. at 1814\\n(quoting Bennett, 520 U.S. at 178).\\n\\n Perez, too, makes clear that the finality analysis is distinct\\nfrom the test for whether a rule is legislative. There, the Court\\naffirmed the “longstanding recognition that interpretive rules\\ndo not have the force and effect of law.” Perez, 135 S. Ct. at\\n1208 (internal citations omitted). Accordingly, overruling\\nParalyzed Veterans of Am. v. D.C. Arena L.P., 117 F.3d 579\\n(D.C. Cir. 1997), the Court held that the APA does not require\\nan agency to provide notice and comment in amending an\\ninterpretive rule, even if the new rule deviates significantly\\nfrom its predecessor. Id. at 1206. In so holding, the Court\\nreassured regulated entities and regulatory beneficiaries that\\nthey are not without recourse should an agency – perhaps to\\nevade notice and comment – repudiate a longstanding\\ninterpretive rule by way of a second interpretive rule. Id. at\\n1209. In such a circumstance, the Court explained, an affected\\nparty can seek judicial review pursuant to the APA. Id. Because\\nonly final agency action is reviewable under the APA, see 5\\nU.S.C. § 704, Perez thus affirms that interpretive rules can be\\nfinal, and, by implication, that the test for finality is\\nindependent of the analysis for whether an agency action is a\\nlegislative rule rather than an interpretive rule.\\n\\n As commentators explain, maintaining an independent\\nfinality analysis is not merely a theoretical nicety; it has several\\nsalutary effects in practice. For example – as Perez alludes to,\\nsee 135 S. Ct. at 1209 – maintaining a finality analysis that is\\ndistinct from the test for whether a rule is legislative permits\\ncourts to review nonlegislative rules and thus safeguards\\nagainst agencies evading both judicial review and notice and\\ncomment by acting via nonlegislative rules. See William Funk,\\nFinal Agency Action After Hawkes, 11 N.Y.U. J. L. & LIBERTY\\n\\n 13\\n285, 304 (2017). This is especially important when viewed\\nfrom the perspective of regulatory beneficiaries, who are\\ngenerally not parties to enforcement actions, and, therefore,\\nmay only be able to challenge nonlegislative rules via judicial\\nreview. See Nina A. Mendelson, Regulatory Beneficiaries and\\nInformal Agency Policymaking, 92 CORNELL L. REV. 397, 420-\\n24 (2007).\\n\\n Having clarified the proper test for finality, we now apply\\nit to the Wehrum Memo. Consistent with the interpretive\\nmethod we endorse herein, we hew closely to the CAA\\nprovisions and EPA regulations appertaining thereto.\\n\\n Our first question is whether the Wehrum Memo “mark[s]\\nthe consummation of [EPA’s] decisionmaking process.”\\nHawkes, 136 S. Ct. at 1813 (quoting Bennett, 520 U.S. at 177-\\n78). It does. Notably, neither Respondent nor Intervenor offer\\nsubstantive argument to the contrary. They were smart to save\\ntheir ink. The Wehrum Memo unequivocally states that the\\nplain language of § 112 “compels” the legal conclusion that\\nqualifying major sources can reclassify at such time that they\\ntake an enforceable limit on their potential to emit to below the\\nmajor source threshold. J.A. 1. In other words, the Wehrum\\nMemo does not advance what EPA believes is a reasonable\\ninterpretation of § 112; it advances what EPA believes is the\\nonly permissible interpretation of the statute. Moreover, no\\nmere subordinate issued the Wehrum Memo. Far from it. The\\nAssistant Administrator for the Office of Air and Radiation\\nissued it. As discussed, under EPA regulations, he is the\\n“principal advisor to the Administrator in matters pertaining to\\nair and radiation,” see 40 C.F.R. § 1.41, and, as we have held\\npreviously with respect to the Assistant Administrator for the\\nOffice of Air and Radiation, nothing within EPA’s regulations\\nprovides us “reason to question his authority to speak for the\\nEPA.” Her Majesty the Queen in Right of Ontario v. EPA, 912\\n\\n 14\\nF.2d 1525, 1532 (D.C. Cir. 1990) (internal citations omitted).\\nCf. Soundboard Ass’n. v. FTC, 888 F.3d 1261, 1267-69 (D.C.\\nCir. 2018), cert denied 139 S. Ct. 1544, 2019 WL 1590248\\n(Apr. 15, 2019) (Federal Trade Commission (“FTC”) staff\\nopinion letter not consummation of agency’s decisionmaking\\nprocess because FTC regulations expressly delineated between\\nCommission advice and staff advice and provided petitioners\\nopportunity to seek opinion from Commission itself).\\nMoreover, EPA published notice of the Wehrum Memo, and\\nreiterated its principal conclusion, in the Federal Register. See\\n83 Fed. Reg. 5543-01 (Feb. 8, 2018). Accordingly, the\\nWehrum Memo can only reasonably be described as EPA’s last\\nword on when a major source can reclassify to an area source\\nunder § 112.\\n\\n Because the Wehrum Memo satisfies Bennett’s first prong,\\nwe ask next whether it has “direct and appreciable legal\\nconsequences.” Hawkes, 136 S. Ct. at 1814 (quoting Bennett,\\n520 U.S. at 178). Petitioners argue that it does because it\\ncreates a new right – i.e. it allows major sources unable to\\nreclassify to area sources under the Seitz Memo to so\\nreclassify. Cal. Pet’rs’ Br. 17-20. Respondent counters that the\\nWehrum Memo does not change the rights of regulated\\nsources. EPA Br. 26-28. Whether or not a regulated source has\\nthe right to reclassify, Respondent contends, is only determined\\nwithin the Title V permitting process. Id. (citing 42 U.S.C. §\\n7661d).\\n\\n For reasons now explained, we hold that the Wehrum\\nMemo does not have a single direct and appreciable legal\\nconsequence.\\n\\n Hawkes instructs that whether an agency action has direct\\nand appreciable legal consequences is a “‘pragmatic’” inquiry.\\nId. at 1815 (quoting Abbot Labs. v. Gardner, 387 U.S. 136, 149\\n\\n 15\\n(1967)). In characterizing the inquiry as pragmatic, we do not\\ntake the Court to be encouraging some sort of common-sense\\napproach. Quite the opposite. We take it as counseling lower\\ncourts to make Bennett prong-two determinations based on the\\nconcrete consequences an agency action has or does not have\\nas a result of the specific statutes and regulations that govern\\nit. Thus, in Hawkes, the Court held, in part, that the agency\\ndetermination at issue had direct and appreciable legal\\nconsequences because, under the applicable statutes and\\nregulations, if petitioners failed to heed the determination they\\ndid so at the risk of significant criminal and civil penalties. Id.\\nAnd the cases Hawkes relies on as past examples of the\\n“pragmatic approach [the Court] has long taken to finality”\\nhold similarly. Id. (citing and quoting Sackett v. EPA, 566 U.S.\\n120, 126 (2012) (holding that agency action at issue satisfied\\nBennett prong-two because, under the relevant statutes and\\nregulations, it appeared to expose petitioners to double\\npenalties in a future enforcement proceeding and to limit their\\nability to obtain a certain type of permit); Abbot Labs., 387 U.S.\\nat 152 (holding that action at issue had a “sufficiently direct\\nand immediate” impact on petitioners, such that judicial review\\nwas appropriate, because, under the governing statutes and\\nregulations, noncompliance risked “serious criminal and civil\\npenalties”); Frozen Food Express v. United States, 351 U.S.\\n40, 44 (1956) (same)).\\n\\n Quite recently, in Valero Energy Corporation v. EPA, No.\\n18-1028, 2019 WL 2587837 (D.C. Cir. June 25, 2019), we\\naffirmed this approach. At issue there, like here, was whether\\nan EPA guidance document that declared the agency’s\\ninterpretation of a statute was final under the Act. We held that\\nit was not. Assessing it within the context of the Act, we\\nemphasized that: (1) the guidance imposed no obligations,\\nprohibitions, or restrictions; (2) it put no party to the choice\\nbetween costly compliance and the risk of a penalty of any sort;\\n\\n 16\\n(3) EPA acknowledged at oral argument that the guidance had\\nno independent legal authority; and (4) that the Act provided\\nregulated parties a statutory mechanism by which to challenge\\nany EPA action that was premised on the statutory\\ninterpretation that the guidance advanced. Valero, 2019 WL at\\n*3-5.\\n\\n Assessing the Wehrum Memo under Hawkes and in\\naccordance with Valero, we find that it is not final. True, it\\nunequivocally declares that major sources, at such time that\\nthey limit their potential to emit to below the major source\\nthreshold, “will no longer be” subject to MACT standards.\\nJ.A. 1. Viewed within the context of the Act, however, the\\nWehrum Memo is all bark and no bite. As Respondent averred\\ntwice at oral argument, neither EPA nor any regulated source\\ncan rely on the Wehrum Memo within the Title V permitting\\nprocess or in any other proceeding. Oral Arg. 50:15-50:27,\\n1:01:13-1:01:50. In other words, as Respondent concedes,\\nalthough the Wehrum Memo forecasts EPA’s position as to\\n§ 112, it has no independent legal authority. In addition, under\\nthe Act and EPA regulations, a state permitting authority that\\nrefuses to comply with the Wehrum Memo faces no penalty or\\nliability of any sort. Further still, the instant matter does not\\npresent a circumstance where the action at issue may be legally\\nconsequential because its binds agency staff and affected\\nparties have no means (outside of judicial review) by which to\\nchallenge it. To the contrary, the Act contains clear provisions\\npursuant to which: (1) a state permitting authority can refuse to\\napply the Wehrum Memo and seek judicial review if EPA\\nissues a permit over its refusal, id. § 7661d(c); and (2) a\\nregulatory beneficiary can petition EPA to object to a state\\npermitting authority’s application of the Wehrum Memo and\\nseek judicial review if EPA denies the petition, id. §\\n7661d(b)(2).\\n\\n 17\\n Accordingly, although the Wehrum Memo, in no uncertain\\nterms, forecasts EPA’s definitive interpretation of § 112, it has\\nno direct and appreciable legal consequences: neither EPA nor\\nregulated sources can rely on it as independently authoritative\\nin any proceeding; state permitting authorities face no penalty\\nor liability of any sort in ignoring it; and state permitting\\nauthorities and regulatory beneficiaries have clear statutory\\navenues by which to challenge a permitting decision adopting\\nthe reasoning of the Wehrum Memo and seek judicial review\\nif EPA fails to sustain their challenges. Under § 7661d(c), if a\\nstate permitting authority refuses to issue a permit allowing a\\nmajor source to reclassify as an area source, and EPA\\nsubsequently issues such a permit following the reasoning of\\nthe Wehrum Memo, judicial review is appropriate. Under\\n§ 7661d(b)(2), if EPA, following the reasoning of the Wehrum\\nMemo, denies a petition from any person asking the agency to\\nobject to a state permitting authority’s issuance of a permit that\\nallows a major source to reclassify as an area source, judicial\\nreview is appropriate. Indeed, because Congress specified that\\n“[n]o objection shall be subject to review until the\\nAdministrator takes final action to issue or deny a permit under\\nthis subsection,” § 7661d(c), we would contravene\\nCongressional intent if we were to hold that a memo that\\nmerely forecasts a future objection is final agency action and\\nsubject to judicial review at this time.\\n\\n The dissent insists that the Wehrum Memo satisfies\\nBennett’s second prong because it “altered the legal regime.”\\nDis. Op. 12. Indeed, the dissent forewarns, the Wehrum Memo\\n“commands, orders, and dictates [to]” EPA employees, id. at 4,\\nand “state permitting authorities are subject to” the statutory\\ninterpretation it advances, id. Said differently, according to the\\ndissent, because of the Wehrum Memo, sources subject to\\nMACT standards that limit their potential to emit to below the\\nmajor source threshold are now “assured that they will be\\n\\n 18\\nsubject to decreased regulation with EPA’s support.” Id. at 12\\n(emphasis added).\\n\\n While the question is not free from doubt, we respectfully\\ndisagree. As noted above, we must remain laser focused on\\nwhether the Wehrum Memo gives “rise to ‘direct and\\nappreciable legal consequences.’” Hawkes, 136 S. Ct. at 1814\\n(quoting Bennett, 520 U.S. at 178), and when viewed in its\\nspecific regulatory context, it does not. “[M]ajor sources must\\ncomply with technology-based emission standards requiring\\nthe maximum degree of reduction in emissions EPA deems\\nachievable, . . . [and] [i]n order to obtain an operating permit\\nunder title V of the [CAA], major sources must comply with\\nextensive monitoring, reporting and record-keeping\\nrequirements. Nat'l Min. Ass'n v. U.S. E.P.A., 59 F.3d 1351,\\n1353 (D.C. Cir. 1995). Major sources must obtain a permit in\\norder to operate, and unless and until that permit is amended or\\nset aside, the stringent requirements set forth therein must be\\ncomplied with while that equipment is operational. The\\nWehrum Memo itself does not revoke or amend a single\\npermit. As acknowledged by the Ohio environmental\\nauthorities in materials cited by petitioners, “[i]f you want to\\ntake advantage of the new guidance [in the Wehrum Memo],\\nyou will need to submit an application to modify your current\\npermit.” Environmental Pet’rs’ Br., Standing Addendum 0198.\\nAssuredly, although the Wehrum Memo advises EPA\\nemployees of the agency’s position as to § 112, it does not bind\\nstate permitting authorities or assure regulated entities of the\\nability to reclassify. As EPA concedes, EPA Br. 21, 25, in\\nreceiving such an application to modify a permit, a state\\npermitting authority may – with total impunity – ignore the\\nWehrum Memo and deny the application. It is true that the\\nAdministrator must issue a revised permit over the state\\npermitting authority’s protest if he or she believes that the\\nstatute so requires, § 7661d(c), but in such a case, the statute\\n\\n 19\\nexplicitly provides the state permitting authority a mechanism\\nby which to seek judicial review of the Administrator’s action.\\nId. Regardless of whether Congress generally intended to\\nallow pre-enforcement review of guidance documents under\\nsome circumstances in the CAA, here, as described above,\\nCongress specifically directed that judicial review shall not be\\navailable until the Title V permit amendment process reaches a\\nconclusion, see §§7661d(b)(2), 7661d(c). Congress’ explicit\\nunderstanding of finality in this specific statutory context\\ncontrols our consideration of the instant guidance document,\\nwhich pertains to that same permit amendment process.\\n\\n ***\\n\\n Before concluding, we note that we have twice had\\noccasion to ask whether an EPA guidance document that\\nimplicated the Act’s Title V permitting process was final\\nagency action: first in Appalachian Power Co. v. EPA, 208\\nF.3d 1015, 1021 (D.C. Cir. 2000), then in National\\nEnvironmental Development Ass’n v. EPA, 752 F.3d 999 (D.C.\\nCir. 2014). In each, we found that the guidance at issue was\\nfinal. A brief analysis of our reasoning in those cases\\ndemonstrates why the Wehrum Memo is not.\\n\\n In Appalachian Power, at issue was a nineteen-page\\nguidance document relating to certain monitoring requirements\\nfor Title V sources. Appalachian Power, 208 F.3d at 1019-20.\\nIn assessing our jurisdiction over the guidance, we did not\\nconsider § 7611d. Instead, we framed our jurisdiction solely in\\nterms of § 7607(b)(1). See id. at 1021 n. 10 (“Our jurisdiction\\nextends to ‘any . . . nationally applicable . . . final action taken\\nby,’ the EPA ‘Administrator.’”) (quoting 42 U.S.C.\\n§ 7607(b)(1))). We predicated our holding that the guidance\\nwas final on the following findings. First, we found that it\\nrequired state permitting authorities to: (1) “review their\\n\\n 20\\nemission standards and the emission standards EPA has\\npromulgated to determine if the standards provide enough\\nmonitoring;” and (2) “insert additional monitoring\\nrequirements as terms or conditions of a permit . . . if they\\nbelieve existing requirements are inadequate, as measured by\\nEPA’s multi-factor, case-by-case analysis set forth in the\\nGuidance.” Id. at 1022. Second, we found that EPA did not\\ndispute petitioners’ assertion that state permitting authorities\\nwere relying on EPA’s guidance in insisting that regulated\\nsources utilize a monitoring method that was more burdensome\\nthan the monitoring method set out under existing EPA\\nregulations. Id. at 1023 & n.17. Finally, we found that a\\nchallenge to an individual permit applying the guidance would\\nnot be heard in this Court, presumably because we felt any such\\nchallenge would have only regional implications. Id. at 1023 n.\\n18.\\n\\n In National Environmental, the guidance document before\\nus explained that, due to a decision of the Sixth Circuit, EPA\\nwas altering a certain interpretation of its regulations only for\\nTitle V sources located within the Sixth Circuit’s jurisdiction.\\nNat’l Envtl., 752 F.3d at 1003. As in Appalachian Power, in\\nassessing our jurisdiction over the guidance, we asked only\\nwhether it was final under § 7607(b)(1) and made no mention\\nof § 7661d. Id. at 1006. In holding that it was final, we found\\nthat the “finality and legal consequences” of the guidance\\n“were made plain” when EPA “relied on [it]” in approving a\\nFederal Implementation Plan (“FIP”) “involving a company\\nlocated outside the jurisdiction of the Sixth Circuit.” Id. at 1007\\n(citing 78 Fed. Reg. 17836, 17842 & n. 10 (March 22, 2013)).\\nIndeed, within the FIP approval – which is a final, legislative\\nrule carrying the force and effect of law – EPA cited the\\nguidance as the sole authority for the legal conclusion that\\ncertain regulations applied to certain sources located outside of\\n\\n 21\\nthe Sixth Circuit’s jurisdiction. See 78 Fed. Reg. 17836, 17842\\n& n.10 (March 22, 2013)\\n\\n Appalachian Power and National Environmental are thus,\\ncontrary to what the dissent suggests, see Dis. Op. 2-3, distinct\\nfrom the instant matter in a crucial respect. In those cases, we\\nheld that the guidance documents at issue were final under §\\n7607(b)(1), without reference to § 7661d, because EPA and\\nstate permitting authorities wielded them to effectuate legal\\nconsequences. In Appalachian Power, we found that the\\nguidance at issue required state permitting authorities to take at\\nleast two specific actions and that EPA did not deny that state\\npermitting authorities used it to coerce regulated sources to\\nadopt a stricter monitoring method. In National Environmental,\\nwe found that EPA cited the guidance, within a binding FIP\\napproval, as the sole authority in support of a legal conclusion.\\nBy contrast, the Wehrum Memo does not require any entity or\\nperson to do anything, and EPA concedes that it has not, will\\nnot, and cannot rely on it in any proceeding. Accordingly,\\nunlike in Appalachian Power and National Environmental, we\\nhave no basis to conclude, without reference to § 7661d, that\\nwe have jurisdiction over the guidance before us under\\n§ 7607(b)(1). We note, in addition, that in Appalachian Power,\\nwe found that we would lack jurisdiction over challenges to\\npermitting decisions applying the guidance at issue. Here,\\nhowever, any party entitled to review under § 7661d that\\nwishes to challenge an application of the Wehrum Memo in this\\nCourt will be so heard, since the Wehrum Memo’s principal\\nconclusion is nationally applicable. See § 7607(b)(1).\\n\\n In sum, we find that the Wehrum Memo – assessed within\\nthe context of the Act and EPA regulations – is not final agency\\naction, and we dismiss the petitions for lack of subject matter\\njurisdiction under the Act. The Wehrum Memo marks the\\nconsummation of EPA’s decisionmaking process as to when a\\n\\n 22\\nmajor source may reclassify to an area source under § 112. But\\nthe Wehrum Memo does not have direct and appreciable legal\\nconsequences: it does not require anyone to do anything;\\nneither EPA nor regulated sources can rely on it in any\\nproceeding; state permitting authorities face no penalty or\\nliability in ignoring it; state permitting authorities and\\nregulatory beneficiaries have clear statutory avenues by which\\nto challenge it and seek judicial review if EPA refuses to heed\\ntheir challenges; and any such challenges, if so desired, will be\\nheard in this Court.\\n\\n III.\\n\\n To conclude, we note that we are under no illusion that this\\nopinion will be the Rosetta Stone of understanding the nature\\nof agency action. Developing this area of the law is indeed an\\n“important continuing project.” Nat’l Min. Ass’n, 758 F.3d at\\n251. Nonetheless, today we humbly submit our contribution\\ntoward clarifying this somewhat gnarled field of jurisprudence.\\nIn ascertaining the nature of an agency action, we emphasize,\\ncourts should look first to the matrix of statutes and regulations\\ngoverning that specific action. In addition, we offer a gentle\\nreminder that the finality analysis is sui generis, separate and\\ndistinct from the test for whether an agency action is a\\nlegislative rule.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4182,"cells":{"id":{"kind":"number","value":4649685,"string":"4,649,685"},"name":{"kind":"string","value":"AMERICAN BANKERS ASSOCIATION, Appellee v. NATIONAL CREDIT UNION ADMINISTRATION, Appellant"},"name_abbreviation":{"kind":"string","value":"American Bankers Association v. National Credit Union Administration"},"decision_date":{"kind":"string","value":"2019-08-20"},"docket_number":{"kind":"string","value":"18-5154; C/w 18-5181"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Henderson, Pillard, Wilkins\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge WILKINS\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 16, 2019 Decided August 20, 2019\\n\\n No. 18-5154\\n\\n AMERICAN BANKERS ASSOCIATION,\\n APPELLEE\\n\\n v.\\n\\n NATIONAL CREDIT UNION ADMINISTRATION,\\n APPELLANT\\n\\n\\n Consolidated with 18-5181\\n\\n\\n Appeals from the United States District Court\\n for the District of Columbia\\n (No. 1:16-cv-02394)\\n\\n\\n Daniel Aguilar, Attorney, U.S. Department of Justice,\\nargued the cause for Appellant-Cross-Appellee. With him on\\nthe briefs was Mark B. Stern, Attorney.\\n\\n Allison Jones Rushing was on the brief for amici curiae\\nCredit Union National Association, et al. in support of\\nAppellant-Cross-Appellee. Nicholas G. Gamse entered an\\nappearance.\\n\\f 2\\n Steven D. Gordon was on the brief for amici curiae State\\nBankers Associations in support Appellee-Cross-Appellant\\nAmerican Bankers Association.\\n\\n Robert A. Long Jr. argued the cause for Appellee-Cross-\\nAppellant. With him on the briefs were Andrew J. Soukup,\\nPhilip Levitz, and Lauren Moxley.\\n\\n Before: HENDERSON, PILLARD, and WILKINS, Circuit\\nJudges.\\n\\n Opinion for the Court filed by Circuit Judge WILKINS.\\n\\n WILKINS, Circuit Judge: Longstanding principles of\\nadministrative law teach us to give federal agencies breathing\\nroom when they make policy and “resolv[e] the struggle\\nbetween competing views of the public interest.” Chevron,\\nU.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 866\\n(1984). And because many policy decisions merge with legal\\nones, Chevron requires us frequently to sustain agency\\ninterpretations of certain federal statutes. Congress often\\nexpects agencies, with their political accountability, “bod[ies]\\nof experience[,] and informed judgment,” to make sound\\ninterpretive choices “with the force of law.” United States v.\\nMead Corp., 533 U.S. 218, 227, 229 (2001) (citation omitted).\\n\\n Congress expressly tasked the National Credit Union\\nAdministration (NCUA) with making such choices in defining\\nthe reach of federal credit unions. Since the Great Depression,\\nCongress has maintained a “system of federal credit unions\\nthat . . . provide credit at reasonable rates” and banking\\nservices to “people of ‘small means.’” First Nat’l Bank & Tr.\\nCo. v. NCUA (First Nat’l Bank I), 988 F.2d 1272, 1274 (D.C.\\nCir. 1993) (citation omitted), aff’d, 522 U.S. 479 (1998).\\nAlthough a private bank may solicit and welcome customers\\n\\f 3\\nfrom anywhere, Congress has limited whom these federal\\nfinancial institutions may serve. For instance, certain\\ninstitutions called “community credit unions” may cover\\nindividuals and entities only within a preapproved\\ngeographical area. The credit union will not receive a federal\\ncharter (and thus cannot start operations) unless it first proffers\\na geographical coverage area and the NCUA accepts the\\nproposal. Congress explicitly assigns the agency the task of\\ncreating vetting standards.\\n\\n Exercising its expressly delegated power, the NCUA has\\npromulgated a final rule that makes it easier for community\\ncredit unions to expand their geographical coverage and thus to\\nreach more potential members. Representing competitors to\\nthe credit unions, the American Bankers Association\\n(Association) has challenged the NCUA’s new rule as neither\\n“in accordance with law” nor within “statutory jurisdiction.” 5\\nU.S.C. § 706(2)(A), (C). The District Court vacated significant\\nportions of the rule, deeming them to be based on unreasonable\\nagency interpretations of the Federal Credit Union Act (Act),\\nPub. L. No. 73-467, 48 Stat. 1216 (1934) (codified as amended\\nat 12 U.S.C. §§ 1751 to 1795k). See Am. Bankers Ass’n v.\\nNCUA, 306 F. Supp. 3d 44, 61, 69-70 (D.D.C. 2018).\\n\\n We appreciate the District Court’s conclusions, made after\\na thoughtful analysis of the Act. But we ultimately disagree\\nwith many of them. In this facial challenge, we review the rule\\nnot as armchair bankers or geographers, but rather as lay judges\\ncognizant that Congress expressly delegated certain policy\\nchoices to the NCUA. After considering the Act’s text,\\npurpose, and legislative history, we hold the agency’s policy\\nchoices “entirely appropriate” for the most part. Chevron, 467\\nU.S. at 865. We therefore sustain the bulk of the rule. Still, we\\ndo not rubber-stamp this regulation. We remand, without\\nvacating, one portion for further consideration of the\\n\\f 4\\ndiscriminatory impact it might have on poor and minority\\nurban residents.\\n\\n I.\\n\\n A.\\n\\n The nation’s credit unions started in the early twentieth\\ncentury “as a populist mechanism designed to empower\\nfarmers against bad loans.” Mehrsa Baradaran, How the Poor\\nGot Cut Out of Banking, 62 EMORY L.J. 483, 500 (2013).\\nWalloped by crop failures and the Great Depression, farmers\\nseeking credit became not only increasingly suspicious of\\ntraditional bankers, who “disregard[ed]” poor individuals and\\nstayed in the big cities, but also fearful of loan sharks, “who\\nwould extract ‘up to a thousand percent’ in interest rates.” Id.\\nat 500-01 (quoting 80 CONG. REC. 6752 (1936) (statement of\\nRep. Lundeen)). The farmers thus began to build their own\\ncredit networks.\\n\\n In a national grassroots campaign, farmers created\\nlocalized, non-profit “credit groups” collecting funds from and\\nloaning small sums to one another at low interest rates. See id.\\nat 501-02. The success of any such self-help institution\\n“hinge[s] on the interpersonal dynamics of its members:\\nLenders must be able to evaluate the ability and willingness of\\npotential borrowers to pay back their loans and borrowers must\\nfeel obligated to pay back those loans.” Wendy Cassity, Note,\\nThe Case for a Credit Union Community Reinvestment Act, 100\\nCOLUM. L. REV. 331, 337 (2000); see also First Nat’l Bank &\\nTr. Co. v. NCUA (First Nat’l Bank II), 90 F.3d 525, 526 (D.C.\\nCir. 1996), aff’d, 522 U.S. 479 (1998).\\n\\n By 1934, individuals had organized about 3,000 local\\ncredit unions, with about 750,000 members. See 80 CONG.\\n\\f 5\\nREC. at 6753. Recognizing the success of credit unions at the\\nstate level, Congress created a federal system that year by\\npassing the Act. Legislators worried that “usurious money\\nlending . . . obviously destroy[ed] vast totals of buying power\\n[once held by] . . . the average worker.” H.R. REP. NO. 73-\\n2021, at 1-2 (1934); see also S. REP. NO. 73-555, at 1 (1934).\\nCongress touted the Act’s ability to “make more available to\\npeople of small means credit for provident purposes.” H.R.\\nREP. NO. 73-2021, at 1; see also S. REP. NO. 73-555, at 1.\\n\\n Credit unions multiplied over the ensuing decades. By\\n1970, Congress created an independent agency to supervise\\nfederal credit unions: the NCUA. See Pub. L. No. 91-468, 84\\nStat. 994 (1970) (codified as amended in scattered sections of\\n12 U.S.C.); see also Swan v. Clinton, 100 F.3d 973, 974 (D.C.\\nCir. 1996) (noting that Congress “entrusted” the agency with\\n“the responsibility of overseeing” federal credit unions).\\nLegislators thought that the agency would be “more responsive\\nto the needs of credit unions” and would “provide more flexible\\nand innovative regulation” than prior government agencies,\\nwhich did not have federal credit unions as their sole focus. S.\\nREP. NO. 91-518, at 3 (1969).\\n\\n The NCUA faced its first major crisis at the end of the\\n1970s. After years of economic decline in several industrial\\nsectors, federal credit unions tied to those business sectors\\nbegan to suffer. The resulting liquidation of numerous credit\\nunions “threaten[ed] ‘the safety and soundness of the federal\\ncredit union system.’” Cassity, supra, at 338-39 (footnote\\nomitted). Reacting to the emergency, the NCUA in 1981\\npromulgated a groundbreaking rule that loosened a major size\\nlimitation on certain federal credit unions. Almost\\nimmediately, those financial institutions grew in membership.\\n\\f 6\\n Meanwhile, credit unions became “caught up in the\\nbroader changes in banking and faced internal as well as\\nexternal pressure to compete with [private] banks and seek\\nhigher profits.” Baradaran, supra, at 505. Unlike credit\\nunions, private, for-profit banks were “owned by equity holders\\nwho may not necessarily be customers (depositors or\\nborrowers),” and they did “not have similar membership and\\ncommercial lending restrictions” as credit unions. DARRYL E.\\nGETTER, CONG. RESEARCH SERV., IF11048, INTRODUCTION TO\\nBANK REGULATION: CREDIT UNIONS AND COMMUNITY BANKS:\\nA COMPARISON 1 (2018). To remain viable, credit unions\\n“started to focus on attracting more customers and expanding\\nthe industry.” Baradaran, supra, at 505. As part of that\\nstrategy, many consolidated through mergers. And private\\nbanks soon treated credit unions as serious competitors,\\nseeking to curb their growth. See NCUA v. First Nat’l Bank &\\nTr. Co. (First Nat’l Bank III), 522 U.S. 479, 485 (1998); First\\nNat’l Bank I, 988 F.2d at 1276.\\n\\n In 1998, the banking industry successfully challenged as\\ncontrary to the Act the 1981 rule that had eased size limitations\\nfor certain federal credit unions. See First Nat’l Bank III, 522\\nU.S. at 503. Congress swiftly responded. In less than six\\nmonths, legislators amended the Act, superseding the holding\\nin First National Bank III, loosening size limitations on certain\\nfederal credit unions, and adding other reforms. See Credit\\nUnion Membership Access Act, Pub. L. No. 105-219, 112 Stat.\\n913 (1998) (codified as amended in scattered sections of 12\\nU.S.C.). Partly because of the 1998 amendments and related\\nNCUA regulations, credit unions continued to merge and grow\\nin membership. Now, more than 61 million customers perform\\ntheir banking services at about 3,400 federal credit unions. See\\n2018 NAT’L CREDIT UNION ADMIN. ANN. REP. 192.\\n\\f 7\\n B.\\n\\n Federal credit unions pool funds from – and give loans\\nto – their members and other credit-union entities. 12 U.S.C.\\n§ 1757(5), (6). A credit union’s members, whether individual\\nor corporate, must come from the credit union’s membership\\n“field,” id. § 1753(5), which is based on a shared occupation,\\nassociation, or geographical area. Members receive regular\\ndividends. Id. § 1763. Congress has shielded federal credit\\nunions from federal corporate income taxes and most state and\\nlocal taxes, but members must pay taxes on their dividends.\\nSee JAMES M. BICKLEY, CONG. RESEARCH SERV., 97-548 E,\\nSHOULD CREDIT UNIONS BE TAXED? 3-5 (2005).\\n\\n To create a federal credit union, at least seven individuals\\nmust present a proposed charter and pay a fee to the NCUA.\\nSee MICHAEL P. MALLOY, BANKING LAW & REGULATION\\n§ 2.04 (2d ed. 2019). In the application, the organizers must\\npledge to deposit funds for shares in the institution and must\\ndescribe the credit union’s proposed membership field. 12\\nU.S.C. § 1753(3), (5). The NCUA must approve the charter\\nbefore the institution may start. See id. § 1754. The agency\\nwill complete an “appropriate investigation” and determine the\\n“general character and fitness” of the organizers, the\\n“economic advisability of establishing” the credit union, and\\nthe “conform[ity]” of proposal details with the Act. Id.\\n\\n The Act governs two types of federal credit unions:\\n“common-bond” credit unions and “community” credit unions.\\nSee id. § 1759(b). This case deals with the latter category. The\\n1934 version of the Act required a community credit union’s\\nmembership field to reflect a particular geographical area – to\\nwit, “a well-defined neighborhood, community, or rural\\ndistrict.” § 9, 48 Stat. at 1219. As amended in 1998, the Act\\nprovides that membership for a community credit union “shall\\n\\f 8\\nbe limited to . . . [p]ersons or organizations within a well-\\ndefined local community, neighborhood, or rural district.” 12\\nU.S.C. § 1759(b) (emphasis added). The 1998 version calls on\\nthe NCUA to “prescribe, by regulation, a definition for the term\\n‘well-defined local community, neighborhood, or rural\\ndistrict.’” Id. § 1759(g)(1). Thus, under the new regime,\\nindividuals seeking to organize a new community credit union\\n(or alter an existing one) must commit to serving members\\nwithin the NCUA’s contemporaneous definition of “local\\ncommunity, neighborhood, or rural district.” See S. REP. NO.\\n105-193, at 4, 8 (1998); H.R. REP. NO. 105-472, at 21 (1998).\\nAs part of their application to the NCUA, they must provide a\\nproposed description of the precise geographical area that the\\ncredit union would serve.\\n\\n Since 1998, there has been “dramatic growth” in the\\nnumber of community credit unions. U.S. GOV’T\\nACCOUNTABILITY OFF., GAO-07-29, CREDIT UNIONS:\\nGREATER TRANSPARENCY NEEDED ON WHO CREDIT UNIONS\\nSERVE AND ON SENIOR EXECUTIVE COMPENSATION\\nARRANGEMENTS 4 (2006). Despite a 11-percent drop in the\\nnumber of federal credit unions from 2000 to 2005, community\\ncredit unions doubled to 1,115. Id. at 4, 12. Meanwhile, the\\namount of assets in community credit unions quadrupled to\\n$104 billion. Id. at 4.\\n\\n C.\\n\\n On December 7, 2016, the NCUA amended its\\nmembership-field rules for community credit unions. See\\nChartering and Field of Membership Manual, 81 Fed. Reg.\\n88,412 (Dec. 7, 2016). Several changes rely on two terms\\ndevised by the Office of Management and Budget (OMB) and\\nbased on data collected by the Census Bureau (Census): “Core\\nBased Statistical Areas” and “Combined Statistical Areas.”\\n\\f 9\\n\\n The OMB has designated numerous regions around the\\ncountry as Core Based Statistical Areas, which comprise at\\nleast one urban cluster, or core, of 10,000 or more people and\\nadjacent counties with substantial commuting ties to that core.\\nSee U.S. CENSUS BUREAU, GEOGRAPHICAL PROGRAM,\\nGLOSSARY, https://www.census.gov/programs-surveys/geogra\\nphy/about/glossary.html. In layman’s terms, a Core Based\\nStatistical Area is a city or town and its suburbs.\\n\\n Meanwhile, a Combined Statistical Area is a conglomerate\\nof two or more adjoining Core Based Statistical Areas, each of\\nwhich has substantial commuting ties with at least one other\\nCore Based Statistical Area in the group. Id. Essentially, a\\nCombined Statistical Area is a regional hub with urban centers\\nconnected by commuting patterns. Combined Statistical Areas\\nmay “reflect broader social and economic interactions, such as\\nwholesaling, commodity distribution, and weekend recreation\\nactivities.” OFFICE OF MGMT. & BUDGET, EXEC. OFFICE OF THE\\nPRESIDENT, OMB BULL. NO. 15-01, REVISED DELINEATIONS OF\\nMETROPOLITAN STATISTICAL AREAS, MICROPOLITAN\\nSTATISTICAL AREAS, AND COMBINED STATISTICAL AREAS, AND\\nGUIDANCE ON USES OF THE DELINEATIONS OF THESE AREAS\\napp. 2-3 (2015).\\n\\n Relevant here, the 2016 rule made two changes to the\\nNCUA’s definition of the term “local community” under\\n§ 1759(b)(3) and one to that of “rural district.” The changes\\naffect what proposed membership areas satisfy the\\ngeographical limitation imposed by the Act.\\n\\n The first change to the “local community” definition\\ninvolves Combined Statistical Areas. A proposed area\\nqualifies as a local community if it encompasses the whole or\\na portion of a Combined Statistical Area and does not exceed a\\n\\f 10\\ndesignated population limit. See 81 Fed. Reg. at 88,440. The\\nNCUA has set that cap at 2.5 million people.\\n\\n The second change involves Core Based Statistical Areas.\\nThe parties agree that all or part of a Core Based Statistical\\nArea may qualify as a local community so long as it does not\\nexceed the population limit. But since 2010, the NCUA\\nrequired such a membership area to include the urban core.\\nThe new rule no longer requires that the core be included in the\\nlocal community that a credit union proposes to serve. See id.\\nat 88,413, 88,440.\\n\\n As for the “rural district” definition, the new rule increases\\nthe population cap for valid rural districts from 250,000 people\\n(or 3 percent of the population of the state where most eligible\\nresidents are located) to 1 million people. See id. at 88,416,\\n88,440. The new population limit works with two other\\nconstraints set by the rule: (1) an outer geographical limit on\\nhow far a rural district may extend past the borders of the credit\\nunion’s headquarters state; and (2) a requirement either that\\nmost eligible residents reside in Census-designated rural areas,\\nor that the population density of the proposed district equals\\n100 or fewer people per square mile. See id. at 88,440.\\n\\n D.\\n\\n On the day the NCUA published the rule, the Association\\nfiled this injunctive and declaratory action in the District Court.\\nThe Association claimed that the three changes described\\nabove were not only arbitrary and capricious under the\\nAdministrative Procedure Act (APA), Pub. L. No. 79-404, 60\\nStat. 237 (1946) (codified as 5 U.S.C. §§ 701-06), but also\\nunreasonable and entitled to no deference under Chevron. The\\nagency and Association filed cross-motions for summary\\n\\f 11\\njudgment. On March 29, 2018, the District Court granted both\\nmotions in part and denied them in part.\\n\\n The court made three relevant holdings. First, it rejected\\nas unreasonable the qualification of certain Combined\\nStatistical Areas as local communities. See Am. Bankers Ass’n,\\n306 F. Supp. 3d at 61. Second, it sustained as well-reasoned\\nthe elimination of the core requirement from the Core Based\\nStatistical Area a credit union proposes to serve as its local\\ncommunity. Id. at 64-65. Third, it rejected as unreasonable the\\nincreased population cap for rural districts. Id. at 69-70. (It\\nalso sustained a separate portion of the rule, which the\\nAssociation does not challenge here.)\\n\\n The NCUA and Association timely appealed. We have\\nappellate jurisdiction under 28 U.S.C. § 1291.\\n\\n II.\\n\\n At the outset, we must assure ourselves of our subject\\nmatter jurisdiction over the appellate proceeding. See, e.g.,\\nUnited States v. Gooch, 842 F.3d 1274, 1277 (D.C. Cir. 2016).\\n\\n In their original briefing, the parties failed to apprise us of\\na rule that was promulgated while this appeal was pending and\\nthat changed membership-field requirements for community\\ncredit unions. See Chartering and Field of Membership, 83\\nFed. Reg. 30,289 (June 28, 2018). The 2018 rule eliminated\\nthe portion of the 2016 rule allowing Combined Statistical\\nAreas to qualify as local communities. Compare 12 C.F.R. pt.\\n701, app. B, ch. 2 § V.A.2 (2018), with id. (2019). The 2018\\nrule preamble did not specifically discuss the removal but\\nconcluded that “[a]ny modification in th[e] final rule is\\nconsistent with the District Court decision” in this action. 83\\nFed. Reg. at 30,291.\\n\\f 12\\n\\n “Under the mootness doctrine, we cannot decide a case if\\n‘events have so transpired that the decision will neither\\npresently affect the parties’ rights nor have a more-than-\\nspeculative chance of affecting them in the future.’” Reid v.\\nHurwitz, 920 F.3d 828, 832 (D.C. Cir. 2019) (citation omitted).\\nThe same principle applies to individual claims. See Tucson\\nMed. Ctr. v. Sullivan, 947 F.2d 971, 977 (D.C. Cir. 1991).\\nAccordingly, if a rule under review (or a portion of it) is\\nsuperseded or amended during an appeal, the proceeding (or\\nrelevant part) might be moot. See, e.g., Am. Bankers Ass’n v.\\nNCUA, 271 F.3d 262, 274 (D.C. Cir. 2001). We therefore\\nrequested supplemental briefing on the issue of appellate\\njurisdiction.\\n\\n Based on the government’s submission and\\nrepresentations at oral argument, we hold that the portion of the\\nappeal related to Combined Statistical Areas is not moot.\\n“[T]he mere power to [reinstitute] a challenged law is not a\\nsufficient basis on which a court can conclude that” a challenge\\nremains live. Nat’l Black Police Ass’n v. District of Columbia,\\n108 F.3d 346, 349 (D.C. Cir. 1997). The government has stated\\nthat, if we reverse the relevant part of the District Court’s\\ndecision, all three members of the NCUA’s board intend to\\nreinstitute the Combined Statistical Area portion of the 2016\\nrule if we reverse the relevant part of the District Court’s\\ndecision. See Decl. of Michael McKenna ¶ 3, ECF No.\\n1781123; Oral Arg. Recording 11:33-48. Accordingly, City of\\nMesquite v. Aladdin’s Castle, Inc., 455 U.S. 283 (1982),\\ngoverns this case. In Aladdin’s Castle, the Supreme Court\\ndeemed the controversy live in part because the government\\nhad announced “an intention” to restore the rule under\\nchallenge if the lower-court decision were vacated. Id. at 289\\n& n.11; see also Am. Bankers Ass’n, 271 F.3d at 274. The\\nNCUA’s submission and representations evince such an\\n\\f 13\\nintention here, and the Association – which bears the “heavy\\nburden” of proving mootness, Friends of the Earth, Inc. v.\\nLaidlaw Envtl. Servs. (TOC), Inc., 528 U.S. 167, 189 (2000)\\n(citation omitted) – offers no evidence to the contrary.\\n\\n The Association attempts to distinguish Aladdin’s Castle\\non two grounds, but neither sways us. First, the Association\\nnotes that the city government in Aladdin’s Castle said it would\\nreenact “precisely the same” law, see 455 U.S. at 289, but that\\nin this case any future notice-and-comment proceedings might\\nproduce a different “local community” definition, perhaps not\\neven relying on Combined Statistical Areas. After all, the\\nagency must keep a “flexible and open-minded attitude” during\\nthe process. See Nat’l Tour Brokers Ass’n v. United States, 591\\nF.2d 896, 902 (D.C. Cir. 1978). We grant that commenters\\nmight convince the NCUA to change its mind; mann tracht,\\nund Gott lacht. But that strikes us as speculative as public input\\nconvincing Mesquite legislators to enact a different law. We\\ndo not see Aladdin’s Castle turning on such conjecture.\\nInstead, we see a live dispute because there is “no certainty”\\nthat the NCUA will forego reinstating the same Combined\\nStatistical Area definition. Aladdin’s Castle, 455 U.S. at 289;\\nsee also Trinity Lutheran Church of Columbia, Inc. v. Comer,\\n137 S. Ct. 2012, 2019 n.1 (2017) (holding that the court must\\nfind “absolutely clear that the allegedly wrongful behavior\\ncould not reasonably be expected to recur” (emphasis added)\\n(quoting Friends of the Earth, 528 U.S. at 189)). After all, the\\nNCUA continues to defend the definition here. See Knox v.\\nServ. Emps. Int’l Union, Local 100, 567 U.S. 298, 307 (2012).\\n\\n Second, the Association observes that both sides in\\nAladdin’s Castle urged the Supreme Court to treat their dispute\\nas live. In contrast, only the NCUA seeks to proceed here; the\\nAssociation would prefer to wait until the agency reinstitutes\\nthe rule. But the existence or absence of jurisdiction does not\\n\\f 14\\nturn on which parties challenge or defend it. Cf. Bender v.\\nWilliamsport Area Sch. Dist., 475 U.S. 534, 541 (1986).\\nBecause the NCUA remains bound by the lower-court\\njudgment, the present injury renders irrelevant the\\nAssociation’s preference. “Jurisdiction existing,” our duty to\\ndecide the appeal “is ‘virtually unflagging.’” Sprint\\nCommc’ns, Inc. v. Jacobs, 571 U.S. 69, 77 (2013) (citation\\nomitted).\\n\\n In short, we may review the challenge to the rule change\\ninvolving Combined Statistical Areas. We see no jurisdictional\\nissues with the rest of the appeal. We thus turn to the merits.\\n\\n III.\\n\\n We review de novo the District Court’s rulings on\\nsummary judgment. See Loan Syndications & Trading Ass’n\\nv. SEC, 882 F.3d 220, 222 (D.C. Cir. 2018). We review the\\nadministrative record and give “no particular deference” to the\\nDistrict Court’s views. Oceana, Inc. v. Ross, 920 F.3d 855, 860\\n(D.C. Cir. 2019) (citation omitted).\\n\\n The APA governs this suit. In relevant part, the statute\\nprovides that we “decide all relevant questions of law” and\\n“interpret . . . statutory provisions.” 5 U.S.C. § 706. We\\nordinarily set aside agency actions that are either “arbitrary,\\ncapricious, an abuse of discretion, or otherwise not in\\naccordance with law,” id. § 706(2)(A), or “in excess of\\nstatutory jurisdiction, authority, or limitations, or short of\\nstatutory right,” id. § 706(2)(C).\\n\\n We review the agency rule in accordance with the familiar\\nChevron doctrine, a two-prong test for determining whether an\\nagency “has stayed within the bounds of its statutory authority”\\nwhen issuing its action. City of Arlington v. FCC, 569 U.S.\\n\\f 15\\n290, 297 (2013) (emphasis omitted). At the first step, we\\ndetermine “whether Congress has directly spoken to the precise\\nquestion at issue,” and we “give effect” to any “unambiguously\\nexpressed intent.” Chevron, 467 U.S. at 842-43 & n.9.\\n\\n If we glean no such unambiguous intent, we turn to the\\nsecond step and determine “whether the agency’s answer” to\\nthe question “is based on a permissible construction of the\\nstatute.” Id. at 843. By arriving at the second step, we have\\nconcluded that Congress either explicitly or implicitly\\ndelegated to the agency the lawmaking authority to clarify the\\nstatute. We presume that Congress would not authorize the\\npromulgation of an “[im]permissible construction.” Chevron,\\n467 U.S. at 843. Accordingly, we will set aside agency actions\\nbased on such a construction, because they are either “not in\\naccordance with law” or “in excess of statutory jurisdiction.”\\n5 U.S.C. § 706(2)(A), (C); see also CSX Transp., Inc. v.\\nSurface Transp. Bd., 754 F.3d 1056, 1063 (D.C. Cir. 2014);\\nAss’n of Privacy Sector Colls. & Univs. v. Duncan, 681 F.3d\\n427, 441 (D.C. Cir. 2012).\\n\\n IV.\\n\\n For all three challenges, the first step of the Chevron\\nanalysis proceeds in the same way. “We begin our analysis, as\\nalways, with the statutory text.” Tesoro Alaska Co. v. FERC,\\n778 F.3d 1034, 1038 (D.C. Cir. 2015). Congress having\\nexpressly assigned the NCUA the power to define the\\nchallenged terms, see 12 U.S.C. § 1759(g)(1), we may proceed\\nto Chevron’s second prong without further analysis, see U.S.\\nTelecom Ass’n v. FCC, 825 F.3d 674, 717 (D.C. Cir. 2016);\\nRush Univ. Med. Ctr. v. Burwell, 763 F.3d 754, 760 (7th Cir.\\n2014); Comm’r v. Pepsi-Cola Niagara Bottling Corp., 399\\nF.2d 390, 393 (2d Cir. 1968) (Friendly, J.) (“When Congress\\nhas used a general term and has empowered an administrator\\n\\f 16\\nto define it, the courts must respect his construction if this is\\nwithin the range of reason.”).\\n\\n An express delegation of definitional power “necessarily\\nsuggests that Congress did not intend the [terms] to be applied\\nin [their] plain meaning sense,” Women Involved in Farm\\nEcon. v. U.S. Dep’t of Agric., 876 F.2d 994, 1000 (D.C. Cir.\\n1989), that they are not “self-defining,” id., and that the agency\\n“enjoy[s] broad discretion” in how to define them, Lindeen v.\\nSEC, 825 F.3d 646, 653 (D.C. Cir. 2016). In Chevron terms,\\nCongress through explicit language “has directly spoken to the\\nprecise question” of whether the identified terms must carry\\ncertain meanings. 467 U.S. at 842-43. The answer is no. See\\nBuongiorno v. Sullivan, 912 F.2d 504, 509 (D.C. Cir. 1990)\\n(Thomas, J.) (“When Congress expressly delegates the\\nauthority to fill a gap in a statute, Congress speaks, in effect,\\ndirectly, and says, succinctly, that it wants the agency to\\nannotate its words.”). To hold otherwise at the first Chevron\\nstep would “undermine” the ability of Congress to delegate\\ndefinitional power. Rush Univ. Med. Ctr., 763 F.3d at 760.\\n\\n Consequently, we turn to whether the NCUA’s definitions\\nare “based on a permissible construction of the statute.”\\nChevron, 467 U.S. at 843.\\n\\n V.\\n\\n Agency interpretations promulgated to fill an explicit\\nlegislative gap “are given controlling weight unless they are\\narbitrary, capricious, or manifestly contrary to the statute.”\\nChevron, 467 U.S. at 844; see also Pharm. Research & Mfrs.\\nof Am. v. FTC, 790 F.3d 198, 204 (D.C. Cir. 2015).\\n\\n Under arbitrary and capricious review, “we may not\\nsubstitute our own judgment for that” of the agency. FERC v.\\n\\f 17\\nElec. Power Supply Ass’n, 136 S. Ct. 760, 782 (2016); accord\\nDep’t of Commerce v. New York, 139 S. Ct. 2551, 2569 (2019).\\nStill, we are “not a ‘rubber stamp,’” Oceana, 920 F.3d at 863;\\n“the agency must examine the relevant data and articulate a\\nsatisfactory explanation for its action,” Motor Vehicle Mfrs.\\nAss’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co. (State\\nFarm), 463 U.S. 29, 43 (1983). A rule is arbitrary and\\ncapricious if (1) the agency “has relied on factors which\\nCongress has not intended it to consider”; (2) the agency\\n“entirely failed to consider an important aspect of the\\nproblem”; (3) the agency’s explanation “runs counter to the\\nevidence before the agency”; or (4) the explanation “is so\\nimplausible that it could not be ascribed to a difference in view\\nor the product of agency expertise.” Id.\\n\\n In turn, we assess three definitional changes in the 2016\\nrule: (1) qualifying Combined Statistical Areas as local\\ncommunities; (2) eliminating the core requirement for local\\ncommunities based on Core Based Statistical Areas; and\\n(3) raising the population cap for rural districts. We sustain the\\nfirst and third amendments in full. As for the second, we hold\\nthat it is rationally related to the Act’s text and purposes, but\\nthat it is insufficiently explained.\\n\\n A.\\n\\n The District Court rejected the first change because it\\napproved certain Combined Statistical Areas “no matter how\\ngeographically dispersed and unconnected” the “members may\\nbe.” Am. Bankers Ass’n, 306 F. Supp. 3d at 61. To the court,\\nthe approval did not fit with the term “local community,”\\nwhich, in its view, “encompass[es] an area no larger than a\\ncounty.” Id. at 58, 61. We respectfully disagree.\\n\\f 18\\n The NCUA possesses vast discretion to define terms\\nbecause Congress expressly has given it such power. But the\\nauthority is not boundless. The agency must craft a reasonable\\ndefinition consistent with the Act’s text and purposes; that is\\ncentral to the review we apply at Chevron’s second step. Here,\\nthe NCUA’s definition meets the standard.\\n\\n We first focus on the text. Congress introduced the phrase\\n“local community” in the 1998 amendments. The word\\n“community” had a broad scope at the time. It meant not only\\n“society at large” but also a “body of individuals organized into\\na unit or manifesting usu[ally] with awareness of some\\nunifying trait.” See Community, WEBSTER’S THIRD\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE\\nUNABRIDGED 460 (1993). The group could be “united by\\nhistorical consciousness or . . . common social, economic, and\\npolitical interests.” Id. But the unifying trait could also be\\nsimply “living in a particular place or region.” Id.\\n\\n The NCUA has recognized that the modifier “local”\\n“reflects congressional intent that it takes ‘a more circumspect\\nand restricted approach to chartering community credit\\nunions.’” Am. Bankers Ass’n, 271 F.3d at 273 (citation\\nomitted); see also Oral Arg. Recording 3:16-18. Indeed,\\nCongress made clear its intention to “modif[y]” the “current\\nlaw regarding community credit unions” by adding the word\\n“local” to community. S. REP. NO. 105-193, at 6-7.\\n\\n Insertion of the modifier “local” before “community”\\nimplies that the community “relate[s] to” a “particular limited\\ndistrict” or is “confined to a particular place.” Local,\\nWEBSTER’S THIRD INTERNATIONAL DICTIONARY OF THE\\nENGLISH LANGUAGE UNABRIDGED 1327 (1993). But that place\\nneed not be the size of a county, as the District Court held. The\\nSupreme Court has recognized that the geographical areas\\n\\f 19\\n“need not be small.” First Nat’l Bank III, 522 U.S. at 492. And\\nif Congress wanted a local community to correspond to a\\nparticular geographical unit, such as a county, “‘it easily could\\nhave written’ that limitation explicitly.” NCUA Br. 25-26\\n(quoting Ali v. Fed. Bureau of Prisons, 552 U.S. 214, 227\\n(2008)); see also Credit Union Nat’l Ass’n Amicus Br. 15. The\\nNCUA sensibly reads the term “local” to mean simply that the\\ncommunity, regardless of shape or size, should be neither\\n“broad” nor “general.” Local, SUPRA, at 1327.\\n\\n To be clear, we do not hold today that the NCUA must\\nconsider only bonds and social connections as understood in\\n1998. The parties agree, and thus we assume, that the NCUA,\\ndespite its expressly delegated authority, must adopt a\\ndefinition consistent with what the term “local community”\\nmeant in 1998, the time of its adoption.\\n\\n After consulting state statutes and invoking the canon of\\nnoscitur a sociis, the District Court developed a rather size-\\nrestrictive meaning for the phrase. See Am. Bankers Ass’n, 306\\nF. Supp. 3d at 57-59. But we do not think that defining a “local\\ncommunity” to refer to an area larger than a county is an\\nunreasonable interpretation of the Act’s text.\\n\\n We receive little guidance from the state statutes in effect\\nin 1998. Indeed, some of the statutes considered “local\\ncommunities” to be quite large. See, e.g., ALASKA STAT.\\n§ 18.66.990(7) (1998) (defining “local community entity” as “a\\ncity or borough or other political subdivision of the state, a\\nnonprofit organization, or a combination of these” (emphasis\\nadded)); see also NCUA Br. 25 n.4.\\n\\n We also reject the District Court’s invocation of the\\nnoscitur a sociis canon. When several terms “are associated in\\na context suggesting that [they] have something in common,\\n\\f 20\\nthey should be assigned a permissible meaning that makes\\nthem similar.” ANTONIN SCALIA & BRYAN A. GARNER,\\nREADING LAW: THE INTERPRETATION OF LEGAL TEXTS 195\\n(2012). But the “substantive connection, or fit, between” the\\nterms here – local community, neighborhood, and rural\\ndistrict – is “not so tight or so self-evident.” Graham Cty. Soil\\n& Water Conservation Dist. v. United States ex rel. Wilson, 559\\nU.S. 280, 288 (2010). Only the word “neighborhood” has a\\ndictionary definition clearly suggesting a region of small size;\\nthe phrase “local community” does not, as we have explained\\nabove, and neither does “rural district,” as we explain below.\\nA size restriction would impermissibly “submerge[]” the\\nindependent “character” of the latter two terms. Babbitt v.\\nSweet Home Chapter of Cmtys. for a Great Or. (Sweet Home),\\n515 U.S. 687, 702 (1995) (citation omitted).\\n\\n The Association also points to other textual indicators.\\nBut contrary to what it suggests, see Am. Bankers Ass’n Br.\\n27-28, we see nothing in the record suggesting that “local\\ncommunity” is a term of art. The Association also says the\\nusage of word “local” in two other federal statutes indicates\\nthat rules permitting coverage areas of larger than a county\\nwould be manifestly contrary to the Act. See id. at 32; see also\\n15 U.S.C. § 2203 (1998) (defining “local” as “city, town,\\ncounty, . . . or other political subdivision of a State”); 18 U.S.C.\\n§ 666 (1998) (same). (The Association pointed to other laws,\\nbut they did not exist in 1998. See National Defense\\nAuthorization Act for Fiscal Year 2010, Pub. L. No. 111-84,\\n§ 4703(b), 123 Stat. 2190, 2837 (2009); PRIME Act Grants, 66\\nFed. Reg. 29,010 (May 29, 2001).) True enough. But the two\\nstatutes address materially distinct issues – fire prevention and\\nembezzlement – and thus do not indicate that the term “local”\\nmust imply a size limitation here. See Envtl. Def. v. Duke\\nEnergy Corp., 549 U.S. 561, 574 (2007) (“[M]ost words have\\n\\f 21\\ndifferent shades of meaning and consequently may be variously\\nconstrued . . . in different statutes . . . .” (citation omitted)).\\n\\n In addition to being consistent with the Act’s text, the\\nCombined Statistical Area definition rationally advances the\\nAct’s underlying purposes. In the 1998 amendments, Congress\\nmade two relevant findings about purpose. First, legislators\\nfound “essential” to the credit-union system a “meaningful\\naffinity and bond among members, manifested by a\\ncommonality of routine interaction[;] shared and related work\\nexperiences, interests, or activities[;] or the maintenance of an\\notherwise well-understood sense of cohesion or identity.” § 2,\\n112 Stat. at 914. Second, Congress highlighted the importance\\nof “credit union safety and soundness,” because a credit union\\non firm financial footing “will enhance the public benefit that\\ncitizens receive.” Id. The legislative history also confirms the\\nimportance of common bonds, see S. REP. NO. 73-555, at 2; see\\nalso H.R. REP. NO. 105-472, at 12; H.R. REP. NO. 73-2021, at\\n2; 144 CONG. REC. S9094 (daily ed. July 28, 1998) (statement\\nof Sen. Mikulski); id. at S8971-72 (daily ed. July 24, 1998)\\n(statement of Sen. Moseley-Braun), and economic “integrity,”\\nS. REP. NO. 105-193, at 3; see also 144 CONG. REC. S9094\\n(statement of Sen. Mikulski); id. at S8972 (statement of Sen.\\nMoseley-Braun), to federal credit unions.\\n\\n We recognize that there may be some tension between the\\nAct’s principal purposes: A credit union with exceedingly\\nclose ties among its members is unlikely to have a large enough\\ncustomer base to thrive economically. To the extent that such\\ntension exists, the Act leaves to the NCUA to strike a\\nreasonable balance. Congress was well aware that a viable\\ncredit union might serve a relatively large geographical area.\\nSee, e.g., H.R. REP. NO. 73-2021, at 2 (“[T]here are cases in\\nwhich communities and organizations cross State lines.”).\\n\\f 22\\n The NCUA did just that in promulgating the Combined\\nStatistical Area definition. That definition allows for larger\\ncommunity credit unions; the decision is consistent with\\ndecades of history promoting the economic viability of credit\\nunions in the face of banks and other competing financial\\ninstitutions. Nonetheless, the NCUA struck a balance by\\nensuring that members within the local community maintain\\nsomewhat of a commuter relationship with each other. As the\\nAssociation even acknowledges, commuting patterns “may\\nsometimes serve as a proxy for community interaction.” Am.\\nBankers Ass’n Br. 38 n.25. We see nothing irrational about\\nadopting the factor as a proxy for the common bond\\ncontemplated by Congress. Perhaps we would have made a\\ndifferent call had we been the policymakers. Perhaps we would\\nhave sought a tighter bond. Or perhaps we would not have\\nprioritized credit-union growth. See Iowa Bankers Ass’n\\nAmicus Br. 24-25. But we must “refrain from substituting\\n[our] own interstitial lawmaking for that” of the agency.\\nHousehold Credit Servs., Inc. v. Pfennig, 541 U.S. 232, 244\\n(2004) (citation omitted).\\n\\n The NCUA also reasonably explained its amendment to\\nthe “local community” definition. To begin with, the agency\\nreasonably circumscribed the size of a local community under\\nthe Combined Statistical Area rule by imposing a 2.5 million-\\nperson population limit. Used by the OMB in analogous\\ncircumstances, the cap is a “logical breaking point in terms of\\ncommunity cohesiveness with respect to a multijurisdictional\\narea.” Chartering and Field of Membership for Federal Credit\\nUnions, 75 Fed. Reg. 36,257, 36,259 (June 25, 2010).\\nMoreover, the NCUA reasonably relied on its prior experience\\nwith Core Based Statistical Areas, which no party disputes can\\nserve as local communities under the Act. The agency noted\\nthat the average geographic size of Combined Statistical Areas\\nwith populations of up to 2.5 million people is 4,553 square\\n\\f 23\\nmiles, and that the average size of the Core Based Statistical\\nAreas approved by the NCUA as local communities is 4,572\\nsquare miles. See 81 Fed. Reg. at 88,414-15. Given that\\nsimilarity in size, the NCUA reasonably considered adopted its\\npopulation-limited Combined Statistical Areas standard.\\n\\n Finally, the NCUA’s definition does not readily create\\ngeneral, widely dispersed regions. Cf. First Nat’l Bank III, 522\\nU.S. at 502 (indicating that community credit unions may not\\nbe “composed of members from an unlimited number of\\nunrelated geographical units”). Combined Statistical Areas are\\ngeographical units well-accepted within the government. See\\n81 Fed. Reg. at 88,414. Because they essentially are regional\\nhubs, the Combined Statistical Areas concentrate around\\ncentral locations. The OMB carved out the areas so that their\\nconstituent parts share commuting ties and they reflect\\n“broader social and economic interactions.” OFFICE OF MGMT.\\n& BUDGET, SUPRA, at app. 2-3. The NCUA rationally believed\\nthat such “real-world interconnections” would qualify as the\\ntype of mutual bonds suggested by the term “local\\ncommunity.” Credit Union Nat’l Ass’n Amicus Br. 9. Thus,\\nthe agency reasonably determined that Combined Statistical\\nAreas “simply unif[y], as a single community,” already\\nconnected neighboring regions. See 81 Fed. Reg. at 88,415.\\n\\n The Association raises a potpourri of objections to the\\nNCUA’s decision-making. See Am. Bankers Ass’n Br. 33-48.\\nVirtually all of its gripes are forfeited because it failed first to\\nraise them to the agency, see, e.g., Koretoff v. Vilsack, 707 F.3d\\n394, 397-98 (D.C. Cir. 2013) (per curiam), or lack merit\\nbecause they involve outdated “local community” definitions,\\nwhich either did not allow for the qualification of certain\\ngeographical areas as local communities or lacked a population\\ncap of 2.5 million people.\\n\\f 24\\n But one of the complaints is “deserving of sustained\\nconsideration.” Ortiz v. United States, 138 S. Ct. 2165, 2173\\n(2018). The Association contends that some large Combined\\nStatistical Areas are so sprawling that the NCUA’s definition,\\nwhich treats a 2.5-million-person portion of them as a local\\ncommunity, must be unreasonable. Recall that a Combined\\nStatistical Area is a regional hub with more than one urban\\ncenter. Under the OMB’s technical specifications, each center\\nneed not be connected to every other by commuting ties; rather,\\neach need only have ties to one other center within the hub. As\\nthe Association colorfully puts it, the OMB may designate as a\\nCombined Statistical Area a mere “daisy chain” of urban\\ncenters “that are linked to their neighbors but have nothing to\\nwith those at the other end of the chain.” See Am. Bankers\\nAss’n Br. 36 (citation omitted). Theoretically, the Association\\ncontinues, certain 2.5-million-person communities might bring\\ntogether parts of different urban centers with no connection\\nwith one another. The Association also suggests that the rule\\nmight permit local communities comprising non-contiguous\\nportions of a Combined Statistical Area. See id. at 39.\\n\\n We understand the Association’s argument to be attacking\\nthe Combined Statistical Area definition as unreasonable. To\\nthe Association, the NCUA failed sufficiently to consider the\\npotential for the rule to create unreasonable results. One\\nhypothetical application disturbed the District Court here: the\\nprospect that, within the Combined Statistical Area including\\nthe District of Columbia and counties from three states\\n(Maryland, Pennsylvania, and Virginia), one could create a\\nlocal community bringing together Doylesburg, Pennsylvania,\\nand Partlow, Virginia, towns located 200 miles apart. See Am.\\nBankers Ass’n, 306 F. Supp. 3d at 59-60.\\n\\n We might well agree with the District Court that the\\napproval of such a geographical area would contravene the Act.\\n\\f 25\\nBut even so, the Association would need much more to mount\\nits facial pre-enforcement challenge in this case. As the\\nSupreme Court repeatedly has held, “the fact that petitioner can\\npoint to a hypothetical case in which the rule might lead to an\\narbitrary result does not render the rule” facially invalid. Am.\\nHosp. Ass’n v. NLRB, 499 U.S. 606, 619 (1991); see also EPA\\nv. EME Homer City Generation, L.P. (EME Homer), 572 U.S.\\n489, 524 (2014) (“The possibility that the rule, in uncommon\\nparticular applications, might exceed [the agency]’s statutory\\nauthority does not warrant judicial condemnation of the rule in\\nits entirety.”); INS v. Nat’l Ctr. for Immigrants’ Rights, Inc.,\\n502 U.S. 183, 188 (1991) (“That the regulation may be invalid\\nas applied in s[ome] cases . . . does not mean that the regulation\\nis facially invalid because it is without statutory authority.”);\\ncf. Barnhart v. Thomas, 540 U.S. 20, 29 (2003) (“Virtually\\nevery legal (or other) rule has imperfect applications in\\nparticular circumstances.”).\\n\\n Here, the Association’s complaint and the District Court’s\\naccompanying worry strike us as too conjectural. The NCUA\\nmust assess the “economic advisability of establishing” the\\nproposed credit union before approving it, 12 U.S.C. § 1754,\\nand as part of the assessment, the organizers must propose a\\n“realistic” business plan showing how the institution and its\\nbranches would serve all members in the local community, see\\n12 C.F.R. pt. 701, app. B, ch. 1 § IV.D. The Association has\\nfailed to demonstrate the plausibility of a local community that\\nis defined like the hypothetical narrow, multi-state strip and\\naccompanies a realistic business plan. And if the agency were\\nto receive and approve such an application, a petitioner can\\nmake an as-applied challenge. See, e.g., EME Homer, 572 U.S.\\nat 523-24; Buongiorno, 912 F.2d at 510. The Association has\\nsucceeded in such challenges in the past. See, e.g., Am.\\nBankers Ass’n v. NCUA, No. 1:05-CV-2247, 2008 WL\\n2857678, at *1 (M.D. Pa. July 21, 2008); Am. Bankers Ass’n v.\\n\\f 26\\nNCUA, 347 F. Supp. 2d 1061, 1074 (D. Utah 2004). Thus, we\\nreject the facial attack on the amended “local community”\\ndefinition involving Combined Statistical Areas.\\n\\n B.\\n\\n We turn to the next rule change. The District Court upheld\\nthe eliminated core requirement for a local community based\\non a Core Based Statistical Area. The court acknowledged that\\ndefining a local community without its urban core “does not\\nalter the . . . common bond” shared by the members in the\\nremainder. Am. Bankers Ass’n, 306 F. Supp. 3d at 62-63.\\nMeanwhile, the court accepted the agency’s response to the\\nclaim that the new definition encouraged redlining – a broad\\ncategory of lending practices with negative impact on “areas\\nwhere low-income and minority populations are concentrated.”\\nId. at 65 (quoting 81 Fed. Reg. at 88,413).\\n\\n Like the District Court, we hold that the eliminated core\\nrequirement is consistent with the Act’s text and purposes.\\nStill, we see merit in the Association’s redlining argument and\\nthus hold the definitional change to be arbitrary and capricious.\\n\\n 1.\\n\\n We will sustain the eliminated core requirement if it\\nreflects a reasonable interpretation of “local community” that\\nis rationally related to the dual purposes of promoting credit-\\nunion growth and ensuring some cohesion among members. It\\ndoes. Omission of the urban core from proposed geographical\\narea will permit community credit unions to reach new\\nmembers in the suburban parts of the Core Based Statistical\\nArea and thus to maintain a healthy membership. Because the\\nsuburbs under the OMB’s definition have substantial\\ncommuting ties to the urban cluster, they all will be “within a\\n\\f 27\\nfeasible commuting radius” and thus “share[] at least some\\ngeographic ties.” Am. Bankers Ass’n, 306 F. Supp. 3d at 64.\\nThose bonds do not disappear if the local community lacks the\\ncore. The Association seeks greater “assurance[s]” of a\\nmeaningful bond among members, Am. Bankers Ass’n Br. 66-\\n67; see also Am. Bankers Ass’n Reply 11, but we do not\\nreplace the agency’s policy judgment with ours.\\n\\n We also do not believe that the eliminated core\\nrequirement would create sprawling, boundless geographical\\nregions. No one disputes that, as a general matter, a\\nmembership area comprising an intact Core Based Statistical\\nArea will satisfy any definition of “local community.” If the\\nlocal community with the core poses no problem, we fail to see\\nhow a local community without one would. And even as to\\nCore Based Statistical Areas that do not qualify as local\\ncommunities (because they have populations of more than 2.5\\nmillion), the geographical ties ensure that the proposed\\nmembership area will still be contained within the boundaries\\nof a single, well-recognized metropolitan region. A single\\nregion is not what concerned the Supreme Court in First\\nNational Bank III. See 522 U.S. at 502.\\n\\n The Association objects to the expansive hypothetical\\nmembership fields, highlighting two Core Based Statistical\\nAreas whose populations exceed the NCUA’s cap of 2.5\\nmillion. The Association asserts that the rule change “makes it\\nmuch easier to unite far-distant edges” of those sprawling areas\\nin a single membership area. Am. Bankers Ass’n Reply 9. Fair\\npoint. But economic realities do not make it plausible that\\norganizers would propose such a local community or that the\\nNCUA would approve it. Like the Combined Statistical Area\\ndefinition, the eliminated core requirement does not become\\nfacially infirm because of farfetched hypotheticals. To the\\nextent they occur in the future, troublesome rule applications\\n\\f 28\\nmight be subject to as-applied challenges. See Am. Bankers\\nAss’n, 271 F.3d at 267.\\n\\n 2\\n\\n Despite the eliminated core requirement’s consistency\\nwith the Act, we cannot sustain the definitional change because\\nthe NCUA has not adequately explained it.\\n\\n The Association contends that the rule change “effectively\\nallows credit unions to engage in ‘redlining’ by denying service\\nto urban areas with large numbers of minority and lower-\\nincome residents.” Am. Bankers Ass’n Br. 65. The NCUA\\nattempted to address this concern in the preamble. At first\\nblush, the agency’s statements appear persuasive. Still, the\\nAssociation persuasively argues that the response fails to\\n“consider an important aspect” of the redlining issue or is\\notherwise “so implausible” as to be unreasonable. State Farm,\\n463 U.S. at 43. Thus, the eliminated core requirement is\\narbitrary and capricious.\\n\\n During the notice-and-comment proceedings, the\\nAssociation warned against redlining and objected that\\ncommunity credit unions could now “serv[e] wealthier\\nsuburban counties and exclud[e] markets containing low-\\nincome and minority communities that reside in the core area.”\\nLetter from James Chessen, Exec. Vice President & Chief\\nEconomist, Am. Bankers Ass’n, to Gerard S. Poliquin, Sec’y\\nof the Board, Nat’l Credit Union Admin. (Feb 5, 2016), Am.\\nBankers Ass’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C.\\nfiled Aug. 23, 2017), ECF No. 26-1 at 228. Fairly read, the\\nAssociation’s objection is not to traditional redlining, which\\nencompasses the refusal to make loans in low-income or\\nminority neighborhoods within a service area, see Baradaran,\\nsupra, at 494; see also Cassity, supra, at 348, 355. Federal\\n\\f 29\\ncredit unions “cannot ‘redline’” in the traditional sense because\\n“everyone in a credit union’s ‘community’ is a member who is\\neligible to take advantage of credit union services” and credit\\nunions “by definition cannot reach beyond their member\\ncommunities to offer credit to the general public.” Cassity,\\nsupra, at 355. But a community credit union can engage in\\nmore unconventional redlining practices: “gerrymander[ing] to\\ncreate its own community of exclusively higher-income\\nmembers.” Id. at 359. We think it evident that the Association\\nwas focusing on such gerrymandering.\\n\\n In response, the NCUA acknowledged that it originally\\nkept the core requirement as a benefit to “low-income and\\nunderserved populations.” 81 Fed. Reg. at 88,413. Some\\ncommenters criticized the core requirement as failing to\\nachieve that goal; it caused community credit unions “to\\nsacrifice service to other areas” within the Core Based\\nStatistical Area, id., and such service arguably could benefit\\npoor and minority customers residing outside the core, see id.\\nat 88,414 (remarking on the importance of credit unions\\n“providing financial services to low income and underserved\\npopulations without regard to where they are located within a\\ncommunity, i.e., beyond its ‘core area’”).\\n\\n Still, the agency dismissed the redlining concern on other\\ngrounds, pointing to its “supervisory process to assess [credit-\\nunion] management’s efforts to offer service to the entire\\ncommunity [the credit union] seeks to serve.” Id. The NCUA\\nfocused on two aspects of its process. First, the agency touted\\nits “annual evaluation,” which “encompasses [the credit\\nunion]’s implementation of its business and marketing plans.”\\nId. Citing its “[e]xperience” as support, the agency identified\\nthe evaluation as a “more effective means” than a core\\nrequirement to “ensur[e] that the low-income and underserved\\npopulations are fairly served.” Id. Indeed, prior evaluations\\n\\f 30\\nconfirm that the agency has had “success in providing financial\\nservices to low-income and underserved populations without\\nregard to where they are located within” the membership area.\\nId. at 88,414. Second, the NCUA noted its “mandate to\\nconsider member complaints alleging discriminatory practices\\naffecting low-income and underserved populations, such as\\nredlining, and to respond as necessary when such practices are\\nshown to exist.” Id.\\n\\n Both aspects of the NCUA’s supervisory process fail to\\naddress the redlining issue raised by the Association. The\\nannual evaluation process might be an adequate response to\\ntraditional redlining, because it might ensure that community\\ncredit unions adequately serve poor and minority residents\\nliving in their local communities. But we do not see how it\\nfixes gerrymandering or the potential discriminatory economic\\nimpact on urban residents. See State Farm, 463 U.S. at 43\\n(requiring the agency to consider relevant factors). The annual\\nevaluation process necessarily does not come into effect until\\nthe NCUA already has approved the charter, business plan, and\\nproposed local community. See Iowa Bankers Ass’n Amicus\\nBr. 12. And nothing in the record suggests that business plans\\nmay focus on residents residing outside the finalized\\nmembership area; in fact, the law forbids federal credit unions\\nfrom serving those residents.\\n\\n The complaint process also does not work in the\\ngerrymandering context. As the preamble points out,\\ncomplaints are raised by the membership, which would not\\ninclude the affected urban residents because of the rule change.\\nIt seems quite implausible, absent some contrary evidence the\\nagency failed to detail, that members will file grievances based\\non gerrymandering harms suffered by residents outside the\\ncoverage area. See State Farm, 463 U.S. at 43 (rejecting\\nimplausible explanations).\\n\\f 31\\n\\n The NCUA attempts to defend the rule change by offering\\na buffet of other potential rationales in its briefing and at oral\\nargument. They all fail because the agency did not adopt them\\nwhen promulgating the rule change. See State Farm, 463 U.S.\\nat 43 (noting that courts may not “supply a reasoned basis for\\nthe agency’s action that the agency itself has not given”\\n(quoting SEC v. Chenery Corp., 332 U.S. 194, 196 (1947))).\\nWe pause, though, to caution the NCUA about two proffered\\nreasons. At oral argument, the government counsel suggested\\nthat the agency may reject proposed local communities if it\\nsuspects they discriminate against residents in the urban core.\\nSee Oral Arg. Recording 5:28-49, 13:43-14:09. But current\\nreviewing guidelines do not indicate that the agency looks for\\nsuch discrimination. See 12 C.F.R. pt. 701, app. B, ch. 1\\n§ VII.A. And the NCUA made the opposite representations to\\nthe District Court. See Transcript of Motion Hearing at 48:23-\\n49:7 (“[District Court]: . . . If a credit union comes to the\\nagency and says I want to serve X area, either in a rural district\\nor a combined statistical area, and they meet the definition, the\\nagency has no authority to reject that application, as long as the\\ncredit union can demonstrate that they can serve the\\narea? [NCUA]: . . . I think that’s probably right, your\\nHonor.”), Am. Bankers Ass’n v. NCUA, No. 1:16-cv-02394-\\nDLF (D.D.C. filed Mar. 27, 2018), ECF No. 33. The\\ngovernment counsel also suggested that community credit\\nunions already cover the vast majority of urban cores. See Oral\\nArg. Recording 8:10-22, 12:30-41; see also id. at 7:39-8:01; cf.\\nNCUA Reply 27-28. Perhaps, but the current record does not\\nsupport the assertion.\\n\\f 32\\n C.\\n\\n Finally, the District Court rejected the increased\\npopulation cap for rural districts. The court worried that a rural\\ndistrict satisfying the new, higher limit could be too large or\\ncould contain “numerous urban centers.” Am. Bankers Ass’n,\\n306 F. Supp. 3d at 69. Because the rule qualified certain\\ndistricts “no matter [their] geographic size or the number or\\nsize of metropolitan areas falling within [their] proposed\\nboundaries,” the court held that the NCUA’s new definition is\\nunreasonable. Id. at 69-70. But in our view, the new “rural\\ndistrict” definition is reasonable.\\n\\n As suggested above, we assume that the NCUA must\\nadopt definitions consistent with the statutory terms as\\nunderstood in 1934, not today. The terms “rural” and “district”\\ndo not connote specific population or geographical constraints.\\nSee First Nat’l Bank III, 522 U.S. at 492 (noting that markets\\n“need not be small”). A “district” means a “portion of a state,\\ncounty, country, town, or city . . . made for administrative,\\nelectoral, or other purposes.” District, WEBSTER’S NEW\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 649\\n(2d ed. 1934) (emphasis added). The district may be “of\\nundefined extent.” Id. Meanwhile, “rural” indicates a\\n“country” or “agricultur[al]” lifestyle, as opposed to that of a\\n“city or town.” Rural, WEBSTER’S NEW INTERNATIONAL\\nDICTIONARY OF THE ENGLISH LANGUAGE 1861 (2d ed. 1934).\\n“Rural” lands generally “resembl[e]” the “country[side].” Id.\\n\\n Nothing about the 1-million-person cap prevents the rural\\ndistrict from resembling the countryside. And one of the\\nunchallenged restrictions helps provide a rural character to\\nsuch districts. Either most residents in the proposed district\\nlive on rural land, or the population density is 100 or fewer\\npeople per square mile. See 81 Fed. Reg. at 88,440.\\n\\f 33\\nAccordingly, even if most residents live on urban areas in the\\nrural district, those areas will be surrounded by rural land.\\nThat’s because 100 people per square mile – or one person for\\nevery six or so acres – is a rural population density.\\n\\n As for size, the contemporaneous definition of “district”\\nreassures us that rural districts may cross state lines. And a\\nsecond unchallenged restriction assures that the 1-million-\\nperson cap will not support gigantic or straggly rural districts.\\nAs the rule explains, the boundaries of a community credit\\nunion’s district may not “exceed the outer boundaries of the\\nstates” immediately surrounding the state where the proposed\\ncredit union would have its headquarters. 81 Fed. Reg. at\\n88,440. Thus, even though the population density of the entire\\nUnited States is less than 100 people per square mile, see\\nNCUA Br. 56 n.44; Oral Arg. Recording 39:34-38, the\\ngeographical limitation forces a rural district to be much\\nsmaller. We are confident that such districts will not be so\\nenormous as to amount to federations of “unrelated geographic\\nunits.” See First Nat’l Bank III, 522 U.S. at 502.\\n\\n In arguing that the amended “rural district” definition is\\nunreasonable, the Association relies heavily on the interpretive\\nanalysis performed by the District Court. See Am. Bankers\\nAss’n, 306 F. Supp. 3d at 67-69. The court’s core contention\\nis that we must construe “rural” and “district” together as a\\nspecialized phrase that meant, in 1934, an “area[] much smaller\\nthan a state.” Id. at 68. For support, the District Court not only\\nconsulted two 1934-era dictionaries, results of a Westlaw\\nsearch of contemporaneous opinions, and a database containing\\nhistorical uses of the phrase, but also invoked the noscitur a\\nsociis canon. See id. at 67-69.\\n\\n The proffered definitional evidence is pretty thin. The\\n1934-era dictionaries described what the specialized phrase\\n\\f 34\\n“rural district” meant in Great Britain, not in the much larger\\nand more expansive United States, which by the 1930s\\nencompassed forty-eight continental states. See, e.g., Rural\\nDistrict, WEBSTER’S NEW INTERNATIONAL DICTIONARY OF THE\\nENGLISH LANGUAGE 1861 (1934) (“in England, a subdivision\\nof an administrative county embracing . . . county parishes”).\\n\\n The proffered Westlaw search results included scores of\\nfederal and state judicial and administrative opinions. See\\nPlaintiff’s Supplemental Memorandum app. 6, Am. Bankers\\nAss’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C. filed Mar.\\n16, 2018), ECF No. 32-6. But we find little of use in those\\ndocuments. Many of those opinions do not discuss size or\\npopulation. See, e.g., Nicolai v. Wis. Power & Light Co., 269\\nN.W. 281, 282-83 (Wis. 1936). Those that do involve specific\\nrural districts whose sizes were between a town and a state.\\nSee, e.g., Sarther Grocery Co. v Comm’r, 22 B.T.A. 1273, 1274\\n(1931). But none declares that rural districts by definition are\\nrestricted to any particular size or population.\\n\\n The District Court next turned to the Corpus of Historical\\nAmerican Usage, a free and public online database. See Am.\\nBankers Ass’n, 306 F. Supp. 3d at 68 & n.5. The database notes\\n197 mentions of the phrase in the first half of the twentieth\\ncentury and only 37 in the second. See Search Results for\\nRURAL DISTRICTS and RURAL DISTRICT, CORPUS OF\\nHISTORICAL AMERICAN ENGLISH, https://www.english-\\ncorpora.org/coha/ (follow “List” hyperlink; then search\\nmatching strings field for “RURAL DISTRICT”). The\\nperceived drop-off led the District Court to determine that,\\n“even if rural district does not carry meaning distinct from its\\nindividual words today, it did in 1934.” Am. Bankers Ass’n,\\n306 F. Supp. 3d at 68.\\n\\f 35\\n Although federal courts may use “crude[]” searches on\\ndatabases to learn of ambiguities in a statutory term,\\nMuscarello v. United States, 524 U.S. 125, 129-30 (1998), the\\nsearch here did not suffice to show that the agency’s definition\\nwas unreasonable. Much more is required to cabin the\\nagency’s discretion. A search of a commercial database,\\nProQuest, reveals that the phrase appeared at least 500 times in\\nthe second half of the century. We are not confident that the\\nproffered evidence establishes a particular historical trend.\\n\\n The District Court lastly invoked noscitur a sociis. See\\nAm. Bankers Ass’n, 306 F. Supp. 3d at 68-69. The canon is\\ninapposite. Recall that the term “rural district” was listed with\\n“community,” not “local community,” in 1934. At the time,\\n“community” could refer to “[s]ociety at large” or a\\n“commonwealth or state.” Community, WEBSTER’S NEW\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 452\\n(1934). Indeed, we said that the word is “too indefinite to be\\nused for purposes of exact measurement in terms of acres or\\nsquare miles.” Lukens Steel Co. v. Perkins, 107 F.2d 627, 631\\n(D.C. Cir. 1939), rev’d sub nom., Perkins v. Lukens Steel Co.,\\n310 U.S. 113 (1940). Thus, we do not see size or population\\nas a connection between the terms.\\n\\n The increased population cap is consistent with not only\\nthe Act’s text but also its purposes. For instance, as the\\npreamble noted, the expanded definition allows community\\ncredit unions in rural districts to reach more “persons of modest\\nmeans who may reside” in those areas. 81 Fed. Reg. at 88,416;\\nsee also § 2(1), (2), 112 Stat. at 913.\\n\\n And the change is neither arbitrary nor capricious. The\\nnew rule explains that it provides a “level of operating\\nefficiencies and scale” making rural districts attractive options\\nfor prospective credit unions. 81 Fed. Reg. at 88,416. “[A]\\n\\f 36\\nsufficiently large population base is essential to enable” them\\n“to offer financial services economically.” Id. at 88,417. The\\nNCUA also chose the 1-million figure based on prior\\nexperience. The agency noted that it had approved of eight\\ndistricts with populations of more than 250,000, and that one\\nof them already had exceeded 1 million. See id. “Having set a\\n1 million precedent in one state,” the agency felt justified in\\nhaving a “fixed 1 million population cap for the other 49\\nstates.” Id. (emphasis added). We cannot say this policy\\nchoice lacks rational explanation.\\n\\n The Association says the NCUA failed to consider prior\\nagency decisions. See Am. Bankers Ass’n Br. 61. But we see\\nno discrepancy and thus summarily reject the objection. The\\nAssociation also turns to troubling hypothetical examples of\\nrural districts with unruly shapes and those with dense urban\\nareas such as Denver, Colorado. See NCUA Br. 55-61; Oral\\nArg. Recording 37:55-38:05. Again, such implausible outliers\\ndo not impugn the rule’s general reasonableness.\\n\\n VI.\\n\\n We now consider the remedy. To sum up, we hold that\\ndefining certain Combined Statistical Areas or portions of them\\nas local communities and raising the population cap for rural\\ndistricts are consistent with the Act and reasonably explained.\\nThus, we sustain both aspects of the challenged rule. We also\\nleave undisturbed the portion of the District Court’s opinion\\nthat the parties do not contest on appeal.\\n\\n But we deem the eliminated core requirement to be\\narbitrary and capricious. When a rule is contrary to law, the\\n“ordinary practice is to vacate” it. United Steel v. Mine Safety\\n& Health Admin., 925 F.3d 1279, 1287 (D.C. Cir. 2019); see\\nalso 5 U.S.C. § 706(2) (noting that the “reviewing court\\n\\f 37\\nshall . . . set aside” unlawful agency action). But in “rare\\ncases,” we will opt instead to remand without vacating the rule,\\nso that the agency can “correct its errors.” United Steel, 925\\nF.3d at 1287; see also 5 U.S.C. § 702 (“Nothing\\nherein . . . affects . . . the power or duty of the court to . . . deny\\nrelief on any . . . appropriate . . . equitable ground . . . .”); Oglala\\nSioux Tribe v. U.S. Nuclear Regulatory Comm’n, 896 F.3d 520,\\n536 (D.C. Cir. 2018). In considering whether to adopt the latter\\nequitable remedy, we balance “(1) the seriousness of the\\ndeficiencies of the action, that is, how likely it is the agency\\nwill be able to justify its decision on remand; and (2) the\\ndisruptive consequences of vacatur.” United Steel, 925 F.3d at\\n1287 (citation omitted). A strong showing of one factor may\\nobviate the need to find a similar showing of the other. See,\\ne.g., Fox Television Stations, Inc. v. FCC, 280 F.3d 1027, 1049\\n(D.C. Cir. 2002) (holding that, because the agency likely could\\njustify its action on remand, the potential for disruption was\\n“only barely relevant”); Allied-Signal, Inc. v. U.S. Nuclear\\nRegulatory Comm’n, 988 F.2d 146, 152 (D.C. Cir. 1993)\\n(determining that, because vacatur would give regulated parties\\na “peculiar windfall,” the meager chance of justifying the\\naction was given “little weight” in the remedial analysis).\\n\\n The NCUA has not requested remand without vacatur in\\nthis case. But because we have a “duty” to ensure the propriety\\nof the APA remedy, 5 U.S.C. § 702, we hold that we have the\\ndiscretion to raise the issue sua sponte, cf. Igonia v. Califano,\\n568 F.2d 1383, 1387 (D.C. Cir. 1977); Gaddis v. Dixie Realty\\nCo., 420 F.2d 245, 247 (D.C. Cir. 1969) (per curiam). Remand\\nwithout vacatur is appropriate here.\\n\\n We conclude that the NCUA might be able to offer a\\nsatisfactory reason on remand. And as for disruptive effect, we\\nperceive a substantial likelihood that vacating the rule could\\nmake it more difficult for some poor and minority suburban\\n\\f 38\\nresidents to receive adequate financial services. Even\\ntemporarily depriving these members of the opportunity is\\ninequitable, because it would “set back” the Act’s objective of\\noffering financial services to people of small means. See Nat’l\\nRes. Def. Council v. EPA, 489 F.3d 1364, 1374 (D.C. Cir.\\n2007); see also Advocates for Highway & Auto Safety v. Fed.\\nMotor Carrier Safety Admin., 429 F.3d 1136, 1152 (D.C. Cir.\\n2005) (declining to vacate rule because, in the interim, it “may\\ndo some good, if it does anything at all”).\\n\\n Given the potential for sufficient justification and the\\nsubstantial likelihood of disruptive effect, we have a rare case\\nin which vacatur is inappropriate. See United Steel, 925 F.3d\\nat 1287. Thus, we remand without vacating the relevant part\\nof the 2016 rule. We trust that the agency will act\\nexpediently. See EME Homer City Generation, L.P. v. EPA,\\n795 F.3d 118, 132 (D.C. Cir. 2015).\\n\\n * * *\\n\\n In short, we reverse the challenged portions of the District\\nCourt’s summary judgment order. With respect to the\\nqualification of certain Combined Statistical Areas as local\\ncommunities and the increased population cap for rural\\ndistricts, we direct the District Court to issue summary\\njudgment in favor of the NCUA. As to the elimination of the\\nurban-core requirement for local communities based on Core\\nBased Statistical Areas, we direct the District Court to issue\\nsummary judgment in favor of the Association and to remand,\\nwithout vacating, the relevant portion of the 2016 rule for\\nfurther explanation.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4426938/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge WILKINS\"}]}"},"judge1":{"kind":"string","value":"HENDERSON"},"judge2":{"kind":"string","value":"PILLARD"},"judge3":{"kind":"string","value":"WILKINS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4649685/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: AMERICAN BANKERS ASSOCIATION, Appellee v. NATIONAL CREDIT UNION ADMINISTRATION, Appellant\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-20\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued April 16, 2019 Decided August 20, 2019\n\n No. 18-5154\n\n AMERICAN BANKERS ASSOCIATION,\n APPELLEE\n\n v.\n\n NATIONAL CREDIT UNION ADMINISTRATION,\n APPELLANT\n\n\n Consolidated with 18-5181\n\n\n Appeals from the United States District Court\n for the District of Columbia\n (No. 1:16-cv-02394)\n\n\n Daniel Aguilar, Attorney, U.S. Department of Justice,\nargued the cause for Appellant-Cross-Appellee. With him on\nthe briefs was Mark B. Stern, Attorney.\n\n Allison Jones Rushing was on the brief for amici curiae\nCredit Union National Association, et al. in support of\nAppellant-Cross-Appellee. Nicholas G. Gamse entered an\nappearance.\n\f 2\n Steven D. Gordon was on the brief for amici curiae State\nBankers Associations in support Appellee-Cross-Appellant\nAmerican Bankers Association.\n\n Robert A. Long Jr. argued the cause for Appellee-Cross-\nAppellant. With him on the briefs were Andrew J. Soukup,\nPhilip Levitz, and Lauren Moxley.\n\n Before: HENDERSON, PILLARD, and WILKINS, Circuit\nJudges.\n\n Opinion for the Court filed by Circuit Judge WILKINS.\n\n WILKINS, Circuit Judge: Longstanding principles of\nadministrative law teach us to give federal agencies breathing\nroom when they make policy and “resolv[e] the struggle\nbetween competing views of the public interest.” Chevron,\nU.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 866\n(1984). And because many policy decisions merge with legal\nones, Chevron requires us frequently to sustain agency\ninterpretations of certain federal statutes. Congress often\nexpects agencies, with their political accountability, “bod[ies]\nof experience[,] and informed judgment,” to make sound\ninterpretive choices “with the force of law.” United States v.\nMead Corp., 533 U.S. 218, 227, 229 (2001) (citation omitted).\n\n Congress expressly tasked the National Credit Union\nAdministration (NCUA) with making such choices in defining\nthe reach of federal credit unions. Since the Great Depression,\nCongress has maintained a “system of federal credit unions\nthat . . . provide credit at reasonable rates” and banking\nservices to “people of ‘small means.’” First Nat’l Bank & Tr.\nCo. v. NCUA (First Nat’l Bank I), 988 F.2d 1272, 1274 (D.C.\nCir. 1993) (citation omitted), aff’d, 522 U.S. 479 (1998).\nAlthough a private bank may solicit and welcome customers\n\f 3\nfrom anywhere, Congress has limited whom these federal\nfinancial institutions may serve. For instance, certain\ninstitutions called “community credit unions” may cover\nindividuals and entities only within a preapproved\ngeographical area. The credit union will not receive a federal\ncharter (and thus cannot start operations) unless it first proffers\na geographical coverage area and the NCUA accepts the\nproposal. Congress explicitly assigns the agency the task of\ncreating vetting standards.\n\n Exercising its expressly delegated power, the NCUA has\npromulgated a final rule that makes it easier for community\ncredit unions to expand their geographical coverage and thus to\nreach more potential members. Representing competitors to\nthe credit unions, the American Bankers Association\n(Association) has challenged the NCUA’s new rule as neither\n“in accordance with law” nor within “statutory jurisdiction.” 5\nU.S.C. § 706(2)(A), (C). The District Court vacated significant\nportions of the rule, deeming them to be based on unreasonable\nagency interpretations of the Federal Credit Union Act (Act),\nPub. L. No. 73-467, 48 Stat. 1216 (1934) (codified as amended\nat 12 U.S.C. §§ 1751 to 1795k). See Am. Bankers Ass’n v.\nNCUA, 306 F. Supp. 3d 44, 61, 69-70 (D.D.C. 2018).\n\n We appreciate the District Court’s conclusions, made after\na thoughtful analysis of the Act. But we ultimately disagree\nwith many of them. In this facial challenge, we review the rule\nnot as armchair bankers or geographers, but rather as lay judges\ncognizant that Congress expressly delegated certain policy\nchoices to the NCUA. After considering the Act’s text,\npurpose, and legislative history, we hold the agency’s policy\nchoices “entirely appropriate” for the most part. Chevron, 467\nU.S. at 865. We therefore sustain the bulk of the rule. Still, we\ndo not rubber-stamp this regulation. We remand, without\nvacating, one portion for further consideration of the\n\f 4\ndiscriminatory impact it might have on poor and minority\nurban residents.\n\n I.\n\n A.\n\n The nation’s credit unions started in the early twentieth\ncentury “as a populist mechanism designed to empower\nfarmers against bad loans.” Mehrsa Baradaran, How the Poor\nGot Cut Out of Banking, 62 EMORY L.J. 483, 500 (2013).\nWalloped by crop failures and the Great Depression, farmers\nseeking credit became not only increasingly suspicious of\ntraditional bankers, who “disregard[ed]” poor individuals and\nstayed in the big cities, but also fearful of loan sharks, “who\nwould extract ‘up to a thousand percent’ in interest rates.” Id.\nat 500-01 (quoting 80 CONG. REC. 6752 (1936) (statement of\nRep. Lundeen)). The farmers thus began to build their own\ncredit networks.\n\n In a national grassroots campaign, farmers created\nlocalized, non-profit “credit groups” collecting funds from and\nloaning small sums to one another at low interest rates. See id.\nat 501-02. The success of any such self-help institution\n“hinge[s] on the interpersonal dynamics of its members:\nLenders must be able to evaluate the ability and willingness of\npotential borrowers to pay back their loans and borrowers must\nfeel obligated to pay back those loans.” Wendy Cassity, Note,\nThe Case for a Credit Union Community Reinvestment Act, 100\nCOLUM. L. REV. 331, 337 (2000); see also First Nat’l Bank &\nTr. Co. v. NCUA (First Nat’l Bank II), 90 F.3d 525, 526 (D.C.\nCir. 1996), aff’d, 522 U.S. 479 (1998).\n\n By 1934, individuals had organized about 3,000 local\ncredit unions, with about 750,000 members. See 80 CONG.\n\f 5\nREC. at 6753. Recognizing the success of credit unions at the\nstate level, Congress created a federal system that year by\npassing the Act. Legislators worried that “usurious money\nlending . . . obviously destroy[ed] vast totals of buying power\n[once held by] . . . the average worker.” H.R. REP. NO. 73-\n2021, at 1-2 (1934); see also S. REP. NO. 73-555, at 1 (1934).\nCongress touted the Act’s ability to “make more available to\npeople of small means credit for provident purposes.” H.R.\nREP. NO. 73-2021, at 1; see also S. REP. NO. 73-555, at 1.\n\n Credit unions multiplied over the ensuing decades. By\n1970, Congress created an independent agency to supervise\nfederal credit unions: the NCUA. See Pub. L. No. 91-468, 84\nStat. 994 (1970) (codified as amended in scattered sections of\n12 U.S.C.); see also Swan v. Clinton, 100 F.3d 973, 974 (D.C.\nCir. 1996) (noting that Congress “entrusted” the agency with\n“the responsibility of overseeing” federal credit unions).\nLegislators thought that the agency would be “more responsive\nto the needs of credit unions” and would “provide more flexible\nand innovative regulation” than prior government agencies,\nwhich did not have federal credit unions as their sole focus. S.\nREP. NO. 91-518, at 3 (1969).\n\n The NCUA faced its first major crisis at the end of the\n1970s. After years of economic decline in several industrial\nsectors, federal credit unions tied to those business sectors\nbegan to suffer. The resulting liquidation of numerous credit\nunions “threaten[ed] ‘the safety and soundness of the federal\ncredit union system.’” Cassity, supra, at 338-39 (footnote\nomitted). Reacting to the emergency, the NCUA in 1981\npromulgated a groundbreaking rule that loosened a major size\nlimitation on certain federal credit unions. Almost\nimmediately, those financial institutions grew in membership.\n\f 6\n Meanwhile, credit unions became “caught up in the\nbroader changes in banking and faced internal as well as\nexternal pressure to compete with [private] banks and seek\nhigher profits.” Baradaran, supra, at 505. Unlike credit\nunions, private, for-profit banks were “owned by equity holders\nwho may not necessarily be customers (depositors or\nborrowers),” and they did “not have similar membership and\ncommercial lending restrictions” as credit unions. DARRYL E.\nGETTER, CONG. RESEARCH SERV., IF11048, INTRODUCTION TO\nBANK REGULATION: CREDIT UNIONS AND COMMUNITY BANKS:\nA COMPARISON 1 (2018). To remain viable, credit unions\n“started to focus on attracting more customers and expanding\nthe industry.” Baradaran, supra, at 505. As part of that\nstrategy, many consolidated through mergers. And private\nbanks soon treated credit unions as serious competitors,\nseeking to curb their growth. See NCUA v. First Nat’l Bank &\nTr. Co. (First Nat’l Bank III), 522 U.S. 479, 485 (1998); First\nNat’l Bank I, 988 F.2d at 1276.\n\n In 1998, the banking industry successfully challenged as\ncontrary to the Act the 1981 rule that had eased size limitations\nfor certain federal credit unions. See First Nat’l Bank III, 522\nU.S. at 503. Congress swiftly responded. In less than six\nmonths, legislators amended the Act, superseding the holding\nin First National Bank III, loosening size limitations on certain\nfederal credit unions, and adding other reforms. See Credit\nUnion Membership Access Act, Pub. L. No. 105-219, 112 Stat.\n913 (1998) (codified as amended in scattered sections of 12\nU.S.C.). Partly because of the 1998 amendments and related\nNCUA regulations, credit unions continued to merge and grow\nin membership. Now, more than 61 million customers perform\ntheir banking services at about 3,400 federal credit unions. See\n2018 NAT’L CREDIT UNION ADMIN. ANN. REP. 192.\n\f 7\n B.\n\n Federal credit unions pool funds from – and give loans\nto – their members and other credit-union entities. 12 U.S.C.\n§ 1757(5), (6). A credit union’s members, whether individual\nor corporate, must come from the credit union’s membership\n“field,” id. § 1753(5), which is based on a shared occupation,\nassociation, or geographical area. Members receive regular\ndividends. Id. § 1763. Congress has shielded federal credit\nunions from federal corporate income taxes and most state and\nlocal taxes, but members must pay taxes on their dividends.\nSee JAMES M. BICKLEY, CONG. RESEARCH SERV., 97-548 E,\nSHOULD CREDIT UNIONS BE TAXED? 3-5 (2005).\n\n To create a federal credit union, at least seven individuals\nmust present a proposed charter and pay a fee to the NCUA.\nSee MICHAEL P. MALLOY, BANKING LAW & REGULATION\n§ 2.04 (2d ed. 2019). In the application, the organizers must\npledge to deposit funds for shares in the institution and must\ndescribe the credit union’s proposed membership field. 12\nU.S.C. § 1753(3), (5). The NCUA must approve the charter\nbefore the institution may start. See id. § 1754. The agency\nwill complete an “appropriate investigation” and determine the\n“general character and fitness” of the organizers, the\n“economic advisability of establishing” the credit union, and\nthe “conform[ity]” of proposal details with the Act. Id.\n\n The Act governs two types of federal credit unions:\n“common-bond” credit unions and “community” credit unions.\nSee id. § 1759(b). This case deals with the latter category. The\n1934 version of the Act required a community credit union’s\nmembership field to reflect a particular geographical area – to\nwit, “a well-defined neighborhood, community, or rural\ndistrict.” § 9, 48 Stat. at 1219. As amended in 1998, the Act\nprovides that membership for a community credit union “shall\n\f 8\nbe limited to . . . [p]ersons or organizations within a well-\ndefined local community, neighborhood, or rural district.” 12\nU.S.C. § 1759(b) (emphasis added). The 1998 version calls on\nthe NCUA to “prescribe, by regulation, a definition for the term\n‘well-defined local community, neighborhood, or rural\ndistrict.’” Id. § 1759(g)(1). Thus, under the new regime,\nindividuals seeking to organize a new community credit union\n(or alter an existing one) must commit to serving members\nwithin the NCUA’s contemporaneous definition of “local\ncommunity, neighborhood, or rural district.” See S. REP. NO.\n105-193, at 4, 8 (1998); H.R. REP. NO. 105-472, at 21 (1998).\nAs part of their application to the NCUA, they must provide a\nproposed description of the precise geographical area that the\ncredit union would serve.\n\n Since 1998, there has been “dramatic growth” in the\nnumber of community credit unions. U.S. GOV’T\nACCOUNTABILITY OFF., GAO-07-29, CREDIT UNIONS:\nGREATER TRANSPARENCY NEEDED ON WHO CREDIT UNIONS\nSERVE AND ON SENIOR EXECUTIVE COMPENSATION\nARRANGEMENTS 4 (2006). Despite a 11-percent drop in the\nnumber of federal credit unions from 2000 to 2005, community\ncredit unions doubled to 1,115. Id. at 4, 12. Meanwhile, the\namount of assets in community credit unions quadrupled to\n$104 billion. Id. at 4.\n\n C.\n\n On December 7, 2016, the NCUA amended its\nmembership-field rules for community credit unions. See\nChartering and Field of Membership Manual, 81 Fed. Reg.\n88,412 (Dec. 7, 2016). Several changes rely on two terms\ndevised by the Office of Management and Budget (OMB) and\nbased on data collected by the Census Bureau (Census): “Core\nBased Statistical Areas” and “Combined Statistical Areas.”\n\f 9\n\n The OMB has designated numerous regions around the\ncountry as Core Based Statistical Areas, which comprise at\nleast one urban cluster, or core, of 10,000 or more people and\nadjacent counties with substantial commuting ties to that core.\nSee U.S. CENSUS BUREAU, GEOGRAPHICAL PROGRAM,\nGLOSSARY, https://www.census.gov/programs-surveys/geogra\nphy/about/glossary.html. In layman’s terms, a Core Based\nStatistical Area is a city or town and its suburbs.\n\n Meanwhile, a Combined Statistical Area is a conglomerate\nof two or more adjoining Core Based Statistical Areas, each of\nwhich has substantial commuting ties with at least one other\nCore Based Statistical Area in the group. Id. Essentially, a\nCombined Statistical Area is a regional hub with urban centers\nconnected by commuting patterns. Combined Statistical Areas\nmay “reflect broader social and economic interactions, such as\nwholesaling, commodity distribution, and weekend recreation\nactivities.” OFFICE OF MGMT. & BUDGET, EXEC. OFFICE OF THE\nPRESIDENT, OMB BULL. NO. 15-01, REVISED DELINEATIONS OF\nMETROPOLITAN STATISTICAL AREAS, MICROPOLITAN\nSTATISTICAL AREAS, AND COMBINED STATISTICAL AREAS, AND\nGUIDANCE ON USES OF THE DELINEATIONS OF THESE AREAS\napp. 2-3 (2015).\n\n Relevant here, the 2016 rule made two changes to the\nNCUA’s definition of the term “local community” under\n§ 1759(b)(3) and one to that of “rural district.” The changes\naffect what proposed membership areas satisfy the\ngeographical limitation imposed by the Act.\n\n The first change to the “local community” definition\ninvolves Combined Statistical Areas. A proposed area\nqualifies as a local community if it encompasses the whole or\na portion of a Combined Statistical Area and does not exceed a\n\f 10\ndesignated population limit. See 81 Fed. Reg. at 88,440. The\nNCUA has set that cap at 2.5 million people.\n\n The second change involves Core Based Statistical Areas.\nThe parties agree that all or part of a Core Based Statistical\nArea may qualify as a local community so long as it does not\nexceed the population limit. But since 2010, the NCUA\nrequired such a membership area to include the urban core.\nThe new rule no longer requires that the core be included in the\nlocal community that a credit union proposes to serve. See id.\nat 88,413, 88,440.\n\n As for the “rural district” definition, the new rule increases\nthe population cap for valid rural districts from 250,000 people\n(or 3 percent of the population of the state where most eligible\nresidents are located) to 1 million people. See id. at 88,416,\n88,440. The new population limit works with two other\nconstraints set by the rule: (1) an outer geographical limit on\nhow far a rural district may extend past the borders of the credit\nunion’s headquarters state; and (2) a requirement either that\nmost eligible residents reside in Census-designated rural areas,\nor that the population density of the proposed district equals\n100 or fewer people per square mile. See id. at 88,440.\n\n D.\n\n On the day the NCUA published the rule, the Association\nfiled this injunctive and declaratory action in the District Court.\nThe Association claimed that the three changes described\nabove were not only arbitrary and capricious under the\nAdministrative Procedure Act (APA), Pub. L. No. 79-404, 60\nStat. 237 (1946) (codified as 5 U.S.C. §§ 701-06), but also\nunreasonable and entitled to no deference under Chevron. The\nagency and Association filed cross-motions for summary\n\f 11\njudgment. On March 29, 2018, the District Court granted both\nmotions in part and denied them in part.\n\n The court made three relevant holdings. First, it rejected\nas unreasonable the qualification of certain Combined\nStatistical Areas as local communities. See Am. Bankers Ass’n,\n306 F. Supp. 3d at 61. Second, it sustained as well-reasoned\nthe elimination of the core requirement from the Core Based\nStatistical Area a credit union proposes to serve as its local\ncommunity. Id. at 64-65. Third, it rejected as unreasonable the\nincreased population cap for rural districts. Id. at 69-70. (It\nalso sustained a separate portion of the rule, which the\nAssociation does not challenge here.)\n\n The NCUA and Association timely appealed. We have\nappellate jurisdiction under 28 U.S.C. § 1291.\n\n II.\n\n At the outset, we must assure ourselves of our subject\nmatter jurisdiction over the appellate proceeding. See, e.g.,\nUnited States v. Gooch, 842 F.3d 1274, 1277 (D.C. Cir. 2016).\n\n In their original briefing, the parties failed to apprise us of\na rule that was promulgated while this appeal was pending and\nthat changed membership-field requirements for community\ncredit unions. See Chartering and Field of Membership, 83\nFed. Reg. 30,289 (June 28, 2018). The 2018 rule eliminated\nthe portion of the 2016 rule allowing Combined Statistical\nAreas to qualify as local communities. Compare 12 C.F.R. pt.\n701, app. B, ch. 2 § V.A.2 (2018), with id. (2019). The 2018\nrule preamble did not specifically discuss the removal but\nconcluded that “[a]ny modification in th[e] final rule is\nconsistent with the District Court decision” in this action. 83\nFed. Reg. at 30,291.\n\f 12\n\n “Under the mootness doctrine, we cannot decide a case if\n‘events have so transpired that the decision will neither\npresently affect the parties’ rights nor have a more-than-\nspeculative chance of affecting them in the future.’” Reid v.\nHurwitz, 920 F.3d 828, 832 (D.C. Cir. 2019) (citation omitted).\nThe same principle applies to individual claims. See Tucson\nMed. Ctr. v. Sullivan, 947 F.2d 971, 977 (D.C. Cir. 1991).\nAccordingly, if a rule under review (or a portion of it) is\nsuperseded or amended during an appeal, the proceeding (or\nrelevant part) might be moot. See, e.g., Am. Bankers Ass’n v.\nNCUA, 271 F.3d 262, 274 (D.C. Cir. 2001). We therefore\nrequested supplemental briefing on the issue of appellate\njurisdiction.\n\n Based on the government’s submission and\nrepresentations at oral argument, we hold that the portion of the\nappeal related to Combined Statistical Areas is not moot.\n“[T]he mere power to [reinstitute] a challenged law is not a\nsufficient basis on which a court can conclude that” a challenge\nremains live. Nat’l Black Police Ass’n v. District of Columbia,\n108 F.3d 346, 349 (D.C. Cir. 1997). The government has stated\nthat, if we reverse the relevant part of the District Court’s\ndecision, all three members of the NCUA’s board intend to\nreinstitute the Combined Statistical Area portion of the 2016\nrule if we reverse the relevant part of the District Court’s\ndecision. See Decl. of Michael McKenna ¶ 3, ECF No.\n1781123; Oral Arg. Recording 11:33-48. Accordingly, City of\nMesquite v. Aladdin’s Castle, Inc., 455 U.S. 283 (1982),\ngoverns this case. In Aladdin’s Castle, the Supreme Court\ndeemed the controversy live in part because the government\nhad announced “an intention” to restore the rule under\nchallenge if the lower-court decision were vacated. Id. at 289\n& n.11; see also Am. Bankers Ass’n, 271 F.3d at 274. The\nNCUA’s submission and representations evince such an\n\f 13\nintention here, and the Association – which bears the “heavy\nburden” of proving mootness, Friends of the Earth, Inc. v.\nLaidlaw Envtl. Servs. (TOC), Inc., 528 U.S. 167, 189 (2000)\n(citation omitted) – offers no evidence to the contrary.\n\n The Association attempts to distinguish Aladdin’s Castle\non two grounds, but neither sways us. First, the Association\nnotes that the city government in Aladdin’s Castle said it would\nreenact “precisely the same” law, see 455 U.S. at 289, but that\nin this case any future notice-and-comment proceedings might\nproduce a different “local community” definition, perhaps not\neven relying on Combined Statistical Areas. After all, the\nagency must keep a “flexible and open-minded attitude” during\nthe process. See Nat’l Tour Brokers Ass’n v. United States, 591\nF.2d 896, 902 (D.C. Cir. 1978). We grant that commenters\nmight convince the NCUA to change its mind; mann tracht,\nund Gott lacht. But that strikes us as speculative as public input\nconvincing Mesquite legislators to enact a different law. We\ndo not see Aladdin’s Castle turning on such conjecture.\nInstead, we see a live dispute because there is “no certainty”\nthat the NCUA will forego reinstating the same Combined\nStatistical Area definition. Aladdin’s Castle, 455 U.S. at 289;\nsee also Trinity Lutheran Church of Columbia, Inc. v. Comer,\n137 S. Ct. 2012, 2019 n.1 (2017) (holding that the court must\nfind “absolutely clear that the allegedly wrongful behavior\ncould not reasonably be expected to recur” (emphasis added)\n(quoting Friends of the Earth, 528 U.S. at 189)). After all, the\nNCUA continues to defend the definition here. See Knox v.\nServ. Emps. Int’l Union, Local 100, 567 U.S. 298, 307 (2012).\n\n Second, the Association observes that both sides in\nAladdin’s Castle urged the Supreme Court to treat their dispute\nas live. In contrast, only the NCUA seeks to proceed here; the\nAssociation would prefer to wait until the agency reinstitutes\nthe rule. But the existence or absence of jurisdiction does not\n\f 14\nturn on which parties challenge or defend it. Cf. Bender v.\nWilliamsport Area Sch. Dist., 475 U.S. 534, 541 (1986).\nBecause the NCUA remains bound by the lower-court\njudgment, the present injury renders irrelevant the\nAssociation’s preference. “Jurisdiction existing,” our duty to\ndecide the appeal “is ‘virtually unflagging.’” Sprint\nCommc’ns, Inc. v. Jacobs, 571 U.S. 69, 77 (2013) (citation\nomitted).\n\n In short, we may review the challenge to the rule change\ninvolving Combined Statistical Areas. We see no jurisdictional\nissues with the rest of the appeal. We thus turn to the merits.\n\n III.\n\n We review de novo the District Court’s rulings on\nsummary judgment. See Loan Syndications & Trading Ass’n\nv. SEC, 882 F.3d 220, 222 (D.C. Cir. 2018). We review the\nadministrative record and give “no particular deference” to the\nDistrict Court’s views. Oceana, Inc. v. Ross, 920 F.3d 855, 860\n(D.C. Cir. 2019) (citation omitted).\n\n The APA governs this suit. In relevant part, the statute\nprovides that we “decide all relevant questions of law” and\n“interpret . . . statutory provisions.” 5 U.S.C. § 706. We\nordinarily set aside agency actions that are either “arbitrary,\ncapricious, an abuse of discretion, or otherwise not in\naccordance with law,” id. § 706(2)(A), or “in excess of\nstatutory jurisdiction, authority, or limitations, or short of\nstatutory right,” id. § 706(2)(C).\n\n We review the agency rule in accordance with the familiar\nChevron doctrine, a two-prong test for determining whether an\nagency “has stayed within the bounds of its statutory authority”\nwhen issuing its action. City of Arlington v. FCC, 569 U.S.\n\f 15\n290, 297 (2013) (emphasis omitted). At the first step, we\ndetermine “whether Congress has directly spoken to the precise\nquestion at issue,” and we “give effect” to any “unambiguously\nexpressed intent.” Chevron, 467 U.S. at 842-43 & n.9.\n\n If we glean no such unambiguous intent, we turn to the\nsecond step and determine “whether the agency’s answer” to\nthe question “is based on a permissible construction of the\nstatute.” Id. at 843. By arriving at the second step, we have\nconcluded that Congress either explicitly or implicitly\ndelegated to the agency the lawmaking authority to clarify the\nstatute. We presume that Congress would not authorize the\npromulgation of an “[im]permissible construction.” Chevron,\n467 U.S. at 843. Accordingly, we will set aside agency actions\nbased on such a construction, because they are either “not in\naccordance with law” or “in excess of statutory jurisdiction.”\n5 U.S.C. § 706(2)(A), (C); see also CSX Transp., Inc. v.\nSurface Transp. Bd., 754 F.3d 1056, 1063 (D.C. Cir. 2014);\nAss’n of Privacy Sector Colls. & Univs. v. Duncan, 681 F.3d\n427, 441 (D.C. Cir. 2012).\n\n IV.\n\n For all three challenges, the first step of the Chevron\nanalysis proceeds in the same way. “We begin our analysis, as\nalways, with the statutory text.” Tesoro Alaska Co. v. FERC,\n778 F.3d 1034, 1038 (D.C. Cir. 2015). Congress having\nexpressly assigned the NCUA the power to define the\nchallenged terms, see 12 U.S.C. § 1759(g)(1), we may proceed\nto Chevron’s second prong without further analysis, see U.S.\nTelecom Ass’n v. FCC, 825 F.3d 674, 717 (D.C. Cir. 2016);\nRush Univ. Med. Ctr. v. Burwell, 763 F.3d 754, 760 (7th Cir.\n2014); Comm’r v. Pepsi-Cola Niagara Bottling Corp., 399\nF.2d 390, 393 (2d Cir. 1968) (Friendly, J.) (“When Congress\nhas used a general term and has empowered an administrator\n\f 16\nto define it, the courts must respect his construction if this is\nwithin the range of reason.”).\n\n An express delegation of definitional power “necessarily\nsuggests that Congress did not intend the [terms] to be applied\nin [their] plain meaning sense,” Women Involved in Farm\nEcon. v. U.S. Dep’t of Agric., 876 F.2d 994, 1000 (D.C. Cir.\n1989), that they are not “self-defining,” id., and that the agency\n“enjoy[s] broad discretion” in how to define them, Lindeen v.\nSEC, 825 F.3d 646, 653 (D.C. Cir. 2016). In Chevron terms,\nCongress through explicit language “has directly spoken to the\nprecise question” of whether the identified terms must carry\ncertain meanings. 467 U.S. at 842-43. The answer is no. See\nBuongiorno v. Sullivan, 912 F.2d 504, 509 (D.C. Cir. 1990)\n(Thomas, J.) (“When Congress expressly delegates the\nauthority to fill a gap in a statute, Congress speaks, in effect,\ndirectly, and says, succinctly, that it wants the agency to\nannotate its words.”). To hold otherwise at the first Chevron\nstep would “undermine” the ability of Congress to delegate\ndefinitional power. Rush Univ. Med. Ctr., 763 F.3d at 760.\n\n Consequently, we turn to whether the NCUA’s definitions\nare “based on a permissible construction of the statute.”\nChevron, 467 U.S. at 843.\n\n V.\n\n Agency interpretations promulgated to fill an explicit\nlegislative gap “are given controlling weight unless they are\narbitrary, capricious, or manifestly contrary to the statute.”\nChevron, 467 U.S. at 844; see also Pharm. Research & Mfrs.\nof Am. v. FTC, 790 F.3d 198, 204 (D.C. Cir. 2015).\n\n Under arbitrary and capricious review, “we may not\nsubstitute our own judgment for that” of the agency. FERC v.\n\f 17\nElec. Power Supply Ass’n, 136 S. Ct. 760, 782 (2016); accord\nDep’t of Commerce v. New York, 139 S. Ct. 2551, 2569 (2019).\nStill, we are “not a ‘rubber stamp,’” Oceana, 920 F.3d at 863;\n“the agency must examine the relevant data and articulate a\nsatisfactory explanation for its action,” Motor Vehicle Mfrs.\nAss’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co. (State\nFarm), 463 U.S. 29, 43 (1983). A rule is arbitrary and\ncapricious if (1) the agency “has relied on factors which\nCongress has not intended it to consider”; (2) the agency\n“entirely failed to consider an important aspect of the\nproblem”; (3) the agency’s explanation “runs counter to the\nevidence before the agency”; or (4) the explanation “is so\nimplausible that it could not be ascribed to a difference in view\nor the product of agency expertise.” Id.\n\n In turn, we assess three definitional changes in the 2016\nrule: (1) qualifying Combined Statistical Areas as local\ncommunities; (2) eliminating the core requirement for local\ncommunities based on Core Based Statistical Areas; and\n(3) raising the population cap for rural districts. We sustain the\nfirst and third amendments in full. As for the second, we hold\nthat it is rationally related to the Act’s text and purposes, but\nthat it is insufficiently explained.\n\n A.\n\n The District Court rejected the first change because it\napproved certain Combined Statistical Areas “no matter how\ngeographically dispersed and unconnected” the “members may\nbe.” Am. Bankers Ass’n, 306 F. Supp. 3d at 61. To the court,\nthe approval did not fit with the term “local community,”\nwhich, in its view, “encompass[es] an area no larger than a\ncounty.” Id. at 58, 61. We respectfully disagree.\n\f 18\n The NCUA possesses vast discretion to define terms\nbecause Congress expressly has given it such power. But the\nauthority is not boundless. The agency must craft a reasonable\ndefinition consistent with the Act’s text and purposes; that is\ncentral to the review we apply at Chevron’s second step. Here,\nthe NCUA’s definition meets the standard.\n\n We first focus on the text. Congress introduced the phrase\n“local community” in the 1998 amendments. The word\n“community” had a broad scope at the time. It meant not only\n“society at large” but also a “body of individuals organized into\na unit or manifesting usu[ally] with awareness of some\nunifying trait.” See Community, WEBSTER’S THIRD\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE\nUNABRIDGED 460 (1993). The group could be “united by\nhistorical consciousness or . . . common social, economic, and\npolitical interests.” Id. But the unifying trait could also be\nsimply “living in a particular place or region.” Id.\n\n The NCUA has recognized that the modifier “local”\n“reflects congressional intent that it takes ‘a more circumspect\nand restricted approach to chartering community credit\nunions.’” Am. Bankers Ass’n, 271 F.3d at 273 (citation\nomitted); see also Oral Arg. Recording 3:16-18. Indeed,\nCongress made clear its intention to “modif[y]” the “current\nlaw regarding community credit unions” by adding the word\n“local” to community. S. REP. NO. 105-193, at 6-7.\n\n Insertion of the modifier “local” before “community”\nimplies that the community “relate[s] to” a “particular limited\ndistrict” or is “confined to a particular place.” Local,\nWEBSTER’S THIRD INTERNATIONAL DICTIONARY OF THE\nENGLISH LANGUAGE UNABRIDGED 1327 (1993). But that place\nneed not be the size of a county, as the District Court held. The\nSupreme Court has recognized that the geographical areas\n\f 19\n“need not be small.” First Nat’l Bank III, 522 U.S. at 492. And\nif Congress wanted a local community to correspond to a\nparticular geographical unit, such as a county, “‘it easily could\nhave written’ that limitation explicitly.” NCUA Br. 25-26\n(quoting Ali v. Fed. Bureau of Prisons, 552 U.S. 214, 227\n(2008)); see also Credit Union Nat’l Ass’n Amicus Br. 15. The\nNCUA sensibly reads the term “local” to mean simply that the\ncommunity, regardless of shape or size, should be neither\n“broad” nor “general.” Local, SUPRA, at 1327.\n\n To be clear, we do not hold today that the NCUA must\nconsider only bonds and social connections as understood in\n1998. The parties agree, and thus we assume, that the NCUA,\ndespite its expressly delegated authority, must adopt a\ndefinition consistent with what the term “local community”\nmeant in 1998, the time of its adoption.\n\n After consulting state statutes and invoking the canon of\nnoscitur a sociis, the District Court developed a rather size-\nrestrictive meaning for the phrase. See Am. Bankers Ass’n, 306\nF. Supp. 3d at 57-59. But we do not think that defining a “local\ncommunity” to refer to an area larger than a county is an\nunreasonable interpretation of the Act’s text.\n\n We receive little guidance from the state statutes in effect\nin 1998. Indeed, some of the statutes considered “local\ncommunities” to be quite large. See, e.g., ALASKA STAT.\n§ 18.66.990(7) (1998) (defining “local community entity” as “a\ncity or borough or other political subdivision of the state, a\nnonprofit organization, or a combination of these” (emphasis\nadded)); see also NCUA Br. 25 n.4.\n\n We also reject the District Court’s invocation of the\nnoscitur a sociis canon. When several terms “are associated in\na context suggesting that [they] have something in common,\n\f 20\nthey should be assigned a permissible meaning that makes\nthem similar.” ANTONIN SCALIA & BRYAN A. GARNER,\nREADING LAW: THE INTERPRETATION OF LEGAL TEXTS 195\n(2012). But the “substantive connection, or fit, between” the\nterms here – local community, neighborhood, and rural\ndistrict – is “not so tight or so self-evident.” Graham Cty. Soil\n& Water Conservation Dist. v. United States ex rel. Wilson, 559\nU.S. 280, 288 (2010). Only the word “neighborhood” has a\ndictionary definition clearly suggesting a region of small size;\nthe phrase “local community” does not, as we have explained\nabove, and neither does “rural district,” as we explain below.\nA size restriction would impermissibly “submerge[]” the\nindependent “character” of the latter two terms. Babbitt v.\nSweet Home Chapter of Cmtys. for a Great Or. (Sweet Home),\n515 U.S. 687, 702 (1995) (citation omitted).\n\n The Association also points to other textual indicators.\nBut contrary to what it suggests, see Am. Bankers Ass’n Br.\n27-28, we see nothing in the record suggesting that “local\ncommunity” is a term of art. The Association also says the\nusage of word “local” in two other federal statutes indicates\nthat rules permitting coverage areas of larger than a county\nwould be manifestly contrary to the Act. See id. at 32; see also\n15 U.S.C. § 2203 (1998) (defining “local” as “city, town,\ncounty, . . . or other political subdivision of a State”); 18 U.S.C.\n§ 666 (1998) (same). (The Association pointed to other laws,\nbut they did not exist in 1998. See National Defense\nAuthorization Act for Fiscal Year 2010, Pub. L. No. 111-84,\n§ 4703(b), 123 Stat. 2190, 2837 (2009); PRIME Act Grants, 66\nFed. Reg. 29,010 (May 29, 2001).) True enough. But the two\nstatutes address materially distinct issues – fire prevention and\nembezzlement – and thus do not indicate that the term “local”\nmust imply a size limitation here. See Envtl. Def. v. Duke\nEnergy Corp., 549 U.S. 561, 574 (2007) (“[M]ost words have\n\f 21\ndifferent shades of meaning and consequently may be variously\nconstrued . . . in different statutes . . . .” (citation omitted)).\n\n In addition to being consistent with the Act’s text, the\nCombined Statistical Area definition rationally advances the\nAct’s underlying purposes. In the 1998 amendments, Congress\nmade two relevant findings about purpose. First, legislators\nfound “essential” to the credit-union system a “meaningful\naffinity and bond among members, manifested by a\ncommonality of routine interaction[;] shared and related work\nexperiences, interests, or activities[;] or the maintenance of an\notherwise well-understood sense of cohesion or identity.” § 2,\n112 Stat. at 914. Second, Congress highlighted the importance\nof “credit union safety and soundness,” because a credit union\non firm financial footing “will enhance the public benefit that\ncitizens receive.” Id. The legislative history also confirms the\nimportance of common bonds, see S. REP. NO. 73-555, at 2; see\nalso H.R. REP. NO. 105-472, at 12; H.R. REP. NO. 73-2021, at\n2; 144 CONG. REC. S9094 (daily ed. July 28, 1998) (statement\nof Sen. Mikulski); id. at S8971-72 (daily ed. July 24, 1998)\n(statement of Sen. Moseley-Braun), and economic “integrity,”\nS. REP. NO. 105-193, at 3; see also 144 CONG. REC. S9094\n(statement of Sen. Mikulski); id. at S8972 (statement of Sen.\nMoseley-Braun), to federal credit unions.\n\n We recognize that there may be some tension between the\nAct’s principal purposes: A credit union with exceedingly\nclose ties among its members is unlikely to have a large enough\ncustomer base to thrive economically. To the extent that such\ntension exists, the Act leaves to the NCUA to strike a\nreasonable balance. Congress was well aware that a viable\ncredit union might serve a relatively large geographical area.\nSee, e.g., H.R. REP. NO. 73-2021, at 2 (“[T]here are cases in\nwhich communities and organizations cross State lines.”).\n\f 22\n The NCUA did just that in promulgating the Combined\nStatistical Area definition. That definition allows for larger\ncommunity credit unions; the decision is consistent with\ndecades of history promoting the economic viability of credit\nunions in the face of banks and other competing financial\ninstitutions. Nonetheless, the NCUA struck a balance by\nensuring that members within the local community maintain\nsomewhat of a commuter relationship with each other. As the\nAssociation even acknowledges, commuting patterns “may\nsometimes serve as a proxy for community interaction.” Am.\nBankers Ass’n Br. 38 n.25. We see nothing irrational about\nadopting the factor as a proxy for the common bond\ncontemplated by Congress. Perhaps we would have made a\ndifferent call had we been the policymakers. Perhaps we would\nhave sought a tighter bond. Or perhaps we would not have\nprioritized credit-union growth. See Iowa Bankers Ass’n\nAmicus Br. 24-25. But we must “refrain from substituting\n[our] own interstitial lawmaking for that” of the agency.\nHousehold Credit Servs., Inc. v. Pfennig, 541 U.S. 232, 244\n(2004) (citation omitted).\n\n The NCUA also reasonably explained its amendment to\nthe “local community” definition. To begin with, the agency\nreasonably circumscribed the size of a local community under\nthe Combined Statistical Area rule by imposing a 2.5 million-\nperson population limit. Used by the OMB in analogous\ncircumstances, the cap is a “logical breaking point in terms of\ncommunity cohesiveness with respect to a multijurisdictional\narea.” Chartering and Field of Membership for Federal Credit\nUnions, 75 Fed. Reg. 36,257, 36,259 (June 25, 2010).\nMoreover, the NCUA reasonably relied on its prior experience\nwith Core Based Statistical Areas, which no party disputes can\nserve as local communities under the Act. The agency noted\nthat the average geographic size of Combined Statistical Areas\nwith populations of up to 2.5 million people is 4,553 square\n\f 23\nmiles, and that the average size of the Core Based Statistical\nAreas approved by the NCUA as local communities is 4,572\nsquare miles. See 81 Fed. Reg. at 88,414-15. Given that\nsimilarity in size, the NCUA reasonably considered adopted its\npopulation-limited Combined Statistical Areas standard.\n\n Finally, the NCUA’s definition does not readily create\ngeneral, widely dispersed regions. Cf. First Nat’l Bank III, 522\nU.S. at 502 (indicating that community credit unions may not\nbe “composed of members from an unlimited number of\nunrelated geographical units”). Combined Statistical Areas are\ngeographical units well-accepted within the government. See\n81 Fed. Reg. at 88,414. Because they essentially are regional\nhubs, the Combined Statistical Areas concentrate around\ncentral locations. The OMB carved out the areas so that their\nconstituent parts share commuting ties and they reflect\n“broader social and economic interactions.” OFFICE OF MGMT.\n& BUDGET, SUPRA, at app. 2-3. The NCUA rationally believed\nthat such “real-world interconnections” would qualify as the\ntype of mutual bonds suggested by the term “local\ncommunity.” Credit Union Nat’l Ass’n Amicus Br. 9. Thus,\nthe agency reasonably determined that Combined Statistical\nAreas “simply unif[y], as a single community,” already\nconnected neighboring regions. See 81 Fed. Reg. at 88,415.\n\n The Association raises a potpourri of objections to the\nNCUA’s decision-making. See Am. Bankers Ass’n Br. 33-48.\nVirtually all of its gripes are forfeited because it failed first to\nraise them to the agency, see, e.g., Koretoff v. Vilsack, 707 F.3d\n394, 397-98 (D.C. Cir. 2013) (per curiam), or lack merit\nbecause they involve outdated “local community” definitions,\nwhich either did not allow for the qualification of certain\ngeographical areas as local communities or lacked a population\ncap of 2.5 million people.\n\f 24\n But one of the complaints is “deserving of sustained\nconsideration.” Ortiz v. United States, 138 S. Ct. 2165, 2173\n(2018). The Association contends that some large Combined\nStatistical Areas are so sprawling that the NCUA’s definition,\nwhich treats a 2.5-million-person portion of them as a local\ncommunity, must be unreasonable. Recall that a Combined\nStatistical Area is a regional hub with more than one urban\ncenter. Under the OMB’s technical specifications, each center\nneed not be connected to every other by commuting ties; rather,\neach need only have ties to one other center within the hub. As\nthe Association colorfully puts it, the OMB may designate as a\nCombined Statistical Area a mere “daisy chain” of urban\ncenters “that are linked to their neighbors but have nothing to\nwith those at the other end of the chain.” See Am. Bankers\nAss’n Br. 36 (citation omitted). Theoretically, the Association\ncontinues, certain 2.5-million-person communities might bring\ntogether parts of different urban centers with no connection\nwith one another. The Association also suggests that the rule\nmight permit local communities comprising non-contiguous\nportions of a Combined Statistical Area. See id. at 39.\n\n We understand the Association’s argument to be attacking\nthe Combined Statistical Area definition as unreasonable. To\nthe Association, the NCUA failed sufficiently to consider the\npotential for the rule to create unreasonable results. One\nhypothetical application disturbed the District Court here: the\nprospect that, within the Combined Statistical Area including\nthe District of Columbia and counties from three states\n(Maryland, Pennsylvania, and Virginia), one could create a\nlocal community bringing together Doylesburg, Pennsylvania,\nand Partlow, Virginia, towns located 200 miles apart. See Am.\nBankers Ass’n, 306 F. Supp. 3d at 59-60.\n\n We might well agree with the District Court that the\napproval of such a geographical area would contravene the Act.\n\f 25\nBut even so, the Association would need much more to mount\nits facial pre-enforcement challenge in this case. As the\nSupreme Court repeatedly has held, “the fact that petitioner can\npoint to a hypothetical case in which the rule might lead to an\narbitrary result does not render the rule” facially invalid. Am.\nHosp. Ass’n v. NLRB, 499 U.S. 606, 619 (1991); see also EPA\nv. EME Homer City Generation, L.P. (EME Homer), 572 U.S.\n489, 524 (2014) (“The possibility that the rule, in uncommon\nparticular applications, might exceed [the agency]’s statutory\nauthority does not warrant judicial condemnation of the rule in\nits entirety.”); INS v. Nat’l Ctr. for Immigrants’ Rights, Inc.,\n502 U.S. 183, 188 (1991) (“That the regulation may be invalid\nas applied in s[ome] cases . . . does not mean that the regulation\nis facially invalid because it is without statutory authority.”);\ncf. Barnhart v. Thomas, 540 U.S. 20, 29 (2003) (“Virtually\nevery legal (or other) rule has imperfect applications in\nparticular circumstances.”).\n\n Here, the Association’s complaint and the District Court’s\naccompanying worry strike us as too conjectural. The NCUA\nmust assess the “economic advisability of establishing” the\nproposed credit union before approving it, 12 U.S.C. § 1754,\nand as part of the assessment, the organizers must propose a\n“realistic” business plan showing how the institution and its\nbranches would serve all members in the local community, see\n12 C.F.R. pt. 701, app. B, ch. 1 § IV.D. The Association has\nfailed to demonstrate the plausibility of a local community that\nis defined like the hypothetical narrow, multi-state strip and\naccompanies a realistic business plan. And if the agency were\nto receive and approve such an application, a petitioner can\nmake an as-applied challenge. See, e.g., EME Homer, 572 U.S.\nat 523-24; Buongiorno, 912 F.2d at 510. The Association has\nsucceeded in such challenges in the past. See, e.g., Am.\nBankers Ass’n v. NCUA, No. 1:05-CV-2247, 2008 WL\n2857678, at *1 (M.D. Pa. July 21, 2008); Am. Bankers Ass’n v.\n\f 26\nNCUA, 347 F. Supp. 2d 1061, 1074 (D. Utah 2004). Thus, we\nreject the facial attack on the amended “local community”\ndefinition involving Combined Statistical Areas.\n\n B.\n\n We turn to the next rule change. The District Court upheld\nthe eliminated core requirement for a local community based\non a Core Based Statistical Area. The court acknowledged that\ndefining a local community without its urban core “does not\nalter the . . . common bond” shared by the members in the\nremainder. Am. Bankers Ass’n, 306 F. Supp. 3d at 62-63.\nMeanwhile, the court accepted the agency’s response to the\nclaim that the new definition encouraged redlining – a broad\ncategory of lending practices with negative impact on “areas\nwhere low-income and minority populations are concentrated.”\nId. at 65 (quoting 81 Fed. Reg. at 88,413).\n\n Like the District Court, we hold that the eliminated core\nrequirement is consistent with the Act’s text and purposes.\nStill, we see merit in the Association’s redlining argument and\nthus hold the definitional change to be arbitrary and capricious.\n\n 1.\n\n We will sustain the eliminated core requirement if it\nreflects a reasonable interpretation of “local community” that\nis rationally related to the dual purposes of promoting credit-\nunion growth and ensuring some cohesion among members. It\ndoes. Omission of the urban core from proposed geographical\narea will permit community credit unions to reach new\nmembers in the suburban parts of the Core Based Statistical\nArea and thus to maintain a healthy membership. Because the\nsuburbs under the OMB’s definition have substantial\ncommuting ties to the urban cluster, they all will be “within a\n\f 27\nfeasible commuting radius” and thus “share[] at least some\ngeographic ties.” Am. Bankers Ass’n, 306 F. Supp. 3d at 64.\nThose bonds do not disappear if the local community lacks the\ncore. The Association seeks greater “assurance[s]” of a\nmeaningful bond among members, Am. Bankers Ass’n Br. 66-\n67; see also Am. Bankers Ass’n Reply 11, but we do not\nreplace the agency’s policy judgment with ours.\n\n We also do not believe that the eliminated core\nrequirement would create sprawling, boundless geographical\nregions. No one disputes that, as a general matter, a\nmembership area comprising an intact Core Based Statistical\nArea will satisfy any definition of “local community.” If the\nlocal community with the core poses no problem, we fail to see\nhow a local community without one would. And even as to\nCore Based Statistical Areas that do not qualify as local\ncommunities (because they have populations of more than 2.5\nmillion), the geographical ties ensure that the proposed\nmembership area will still be contained within the boundaries\nof a single, well-recognized metropolitan region. A single\nregion is not what concerned the Supreme Court in First\nNational Bank III. See 522 U.S. at 502.\n\n The Association objects to the expansive hypothetical\nmembership fields, highlighting two Core Based Statistical\nAreas whose populations exceed the NCUA’s cap of 2.5\nmillion. The Association asserts that the rule change “makes it\nmuch easier to unite far-distant edges” of those sprawling areas\nin a single membership area. Am. Bankers Ass’n Reply 9. Fair\npoint. But economic realities do not make it plausible that\norganizers would propose such a local community or that the\nNCUA would approve it. Like the Combined Statistical Area\ndefinition, the eliminated core requirement does not become\nfacially infirm because of farfetched hypotheticals. To the\nextent they occur in the future, troublesome rule applications\n\f 28\nmight be subject to as-applied challenges. See Am. Bankers\nAss’n, 271 F.3d at 267.\n\n 2\n\n Despite the eliminated core requirement’s consistency\nwith the Act, we cannot sustain the definitional change because\nthe NCUA has not adequately explained it.\n\n The Association contends that the rule change “effectively\nallows credit unions to engage in ‘redlining’ by denying service\nto urban areas with large numbers of minority and lower-\nincome residents.” Am. Bankers Ass’n Br. 65. The NCUA\nattempted to address this concern in the preamble. At first\nblush, the agency’s statements appear persuasive. Still, the\nAssociation persuasively argues that the response fails to\n“consider an important aspect” of the redlining issue or is\notherwise “so implausible” as to be unreasonable. State Farm,\n463 U.S. at 43. Thus, the eliminated core requirement is\narbitrary and capricious.\n\n During the notice-and-comment proceedings, the\nAssociation warned against redlining and objected that\ncommunity credit unions could now “serv[e] wealthier\nsuburban counties and exclud[e] markets containing low-\nincome and minority communities that reside in the core area.”\nLetter from James Chessen, Exec. Vice President & Chief\nEconomist, Am. Bankers Ass’n, to Gerard S. Poliquin, Sec’y\nof the Board, Nat’l Credit Union Admin. (Feb 5, 2016), Am.\nBankers Ass’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C.\nfiled Aug. 23, 2017), ECF No. 26-1 at 228. Fairly read, the\nAssociation’s objection is not to traditional redlining, which\nencompasses the refusal to make loans in low-income or\nminority neighborhoods within a service area, see Baradaran,\nsupra, at 494; see also Cassity, supra, at 348, 355. Federal\n\f 29\ncredit unions “cannot ‘redline’” in the traditional sense because\n“everyone in a credit union’s ‘community’ is a member who is\neligible to take advantage of credit union services” and credit\nunions “by definition cannot reach beyond their member\ncommunities to offer credit to the general public.” Cassity,\nsupra, at 355. But a community credit union can engage in\nmore unconventional redlining practices: “gerrymander[ing] to\ncreate its own community of exclusively higher-income\nmembers.” Id. at 359. We think it evident that the Association\nwas focusing on such gerrymandering.\n\n In response, the NCUA acknowledged that it originally\nkept the core requirement as a benefit to “low-income and\nunderserved populations.” 81 Fed. Reg. at 88,413. Some\ncommenters criticized the core requirement as failing to\nachieve that goal; it caused community credit unions “to\nsacrifice service to other areas” within the Core Based\nStatistical Area, id., and such service arguably could benefit\npoor and minority customers residing outside the core, see id.\nat 88,414 (remarking on the importance of credit unions\n“providing financial services to low income and underserved\npopulations without regard to where they are located within a\ncommunity, i.e., beyond its ‘core area’”).\n\n Still, the agency dismissed the redlining concern on other\ngrounds, pointing to its “supervisory process to assess [credit-\nunion] management’s efforts to offer service to the entire\ncommunity [the credit union] seeks to serve.” Id. The NCUA\nfocused on two aspects of its process. First, the agency touted\nits “annual evaluation,” which “encompasses [the credit\nunion]’s implementation of its business and marketing plans.”\nId. Citing its “[e]xperience” as support, the agency identified\nthe evaluation as a “more effective means” than a core\nrequirement to “ensur[e] that the low-income and underserved\npopulations are fairly served.” Id. Indeed, prior evaluations\n\f 30\nconfirm that the agency has had “success in providing financial\nservices to low-income and underserved populations without\nregard to where they are located within” the membership area.\nId. at 88,414. Second, the NCUA noted its “mandate to\nconsider member complaints alleging discriminatory practices\naffecting low-income and underserved populations, such as\nredlining, and to respond as necessary when such practices are\nshown to exist.” Id.\n\n Both aspects of the NCUA’s supervisory process fail to\naddress the redlining issue raised by the Association. The\nannual evaluation process might be an adequate response to\ntraditional redlining, because it might ensure that community\ncredit unions adequately serve poor and minority residents\nliving in their local communities. But we do not see how it\nfixes gerrymandering or the potential discriminatory economic\nimpact on urban residents. See State Farm, 463 U.S. at 43\n(requiring the agency to consider relevant factors). The annual\nevaluation process necessarily does not come into effect until\nthe NCUA already has approved the charter, business plan, and\nproposed local community. See Iowa Bankers Ass’n Amicus\nBr. 12. And nothing in the record suggests that business plans\nmay focus on residents residing outside the finalized\nmembership area; in fact, the law forbids federal credit unions\nfrom serving those residents.\n\n The complaint process also does not work in the\ngerrymandering context. As the preamble points out,\ncomplaints are raised by the membership, which would not\ninclude the affected urban residents because of the rule change.\nIt seems quite implausible, absent some contrary evidence the\nagency failed to detail, that members will file grievances based\non gerrymandering harms suffered by residents outside the\ncoverage area. See State Farm, 463 U.S. at 43 (rejecting\nimplausible explanations).\n\f 31\n\n The NCUA attempts to defend the rule change by offering\na buffet of other potential rationales in its briefing and at oral\nargument. They all fail because the agency did not adopt them\nwhen promulgating the rule change. See State Farm, 463 U.S.\nat 43 (noting that courts may not “supply a reasoned basis for\nthe agency’s action that the agency itself has not given”\n(quoting SEC v. Chenery Corp., 332 U.S. 194, 196 (1947))).\nWe pause, though, to caution the NCUA about two proffered\nreasons. At oral argument, the government counsel suggested\nthat the agency may reject proposed local communities if it\nsuspects they discriminate against residents in the urban core.\nSee Oral Arg. Recording 5:28-49, 13:43-14:09. But current\nreviewing guidelines do not indicate that the agency looks for\nsuch discrimination. See 12 C.F.R. pt. 701, app. B, ch. 1\n§ VII.A. And the NCUA made the opposite representations to\nthe District Court. See Transcript of Motion Hearing at 48:23-\n49:7 (“[District Court]: . . . If a credit union comes to the\nagency and says I want to serve X area, either in a rural district\nor a combined statistical area, and they meet the definition, the\nagency has no authority to reject that application, as long as the\ncredit union can demonstrate that they can serve the\narea? [NCUA]: . . . I think that’s probably right, your\nHonor.”), Am. Bankers Ass’n v. NCUA, No. 1:16-cv-02394-\nDLF (D.D.C. filed Mar. 27, 2018), ECF No. 33. The\ngovernment counsel also suggested that community credit\nunions already cover the vast majority of urban cores. See Oral\nArg. Recording 8:10-22, 12:30-41; see also id. at 7:39-8:01; cf.\nNCUA Reply 27-28. Perhaps, but the current record does not\nsupport the assertion.\n\f 32\n C.\n\n Finally, the District Court rejected the increased\npopulation cap for rural districts. The court worried that a rural\ndistrict satisfying the new, higher limit could be too large or\ncould contain “numerous urban centers.” Am. Bankers Ass’n,\n306 F. Supp. 3d at 69. Because the rule qualified certain\ndistricts “no matter [their] geographic size or the number or\nsize of metropolitan areas falling within [their] proposed\nboundaries,” the court held that the NCUA’s new definition is\nunreasonable. Id. at 69-70. But in our view, the new “rural\ndistrict” definition is reasonable.\n\n As suggested above, we assume that the NCUA must\nadopt definitions consistent with the statutory terms as\nunderstood in 1934, not today. The terms “rural” and “district”\ndo not connote specific population or geographical constraints.\nSee First Nat’l Bank III, 522 U.S. at 492 (noting that markets\n“need not be small”). A “district” means a “portion of a state,\ncounty, country, town, or city . . . made for administrative,\nelectoral, or other purposes.” District, WEBSTER’S NEW\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 649\n(2d ed. 1934) (emphasis added). The district may be “of\nundefined extent.” Id. Meanwhile, “rural” indicates a\n“country” or “agricultur[al]” lifestyle, as opposed to that of a\n“city or town.” Rural, WEBSTER’S NEW INTERNATIONAL\nDICTIONARY OF THE ENGLISH LANGUAGE 1861 (2d ed. 1934).\n“Rural” lands generally “resembl[e]” the “country[side].” Id.\n\n Nothing about the 1-million-person cap prevents the rural\ndistrict from resembling the countryside. And one of the\nunchallenged restrictions helps provide a rural character to\nsuch districts. Either most residents in the proposed district\nlive on rural land, or the population density is 100 or fewer\npeople per square mile. See 81 Fed. Reg. at 88,440.\n\f 33\nAccordingly, even if most residents live on urban areas in the\nrural district, those areas will be surrounded by rural land.\nThat’s because 100 people per square mile – or one person for\nevery six or so acres – is a rural population density.\n\n As for size, the contemporaneous definition of “district”\nreassures us that rural districts may cross state lines. And a\nsecond unchallenged restriction assures that the 1-million-\nperson cap will not support gigantic or straggly rural districts.\nAs the rule explains, the boundaries of a community credit\nunion’s district may not “exceed the outer boundaries of the\nstates” immediately surrounding the state where the proposed\ncredit union would have its headquarters. 81 Fed. Reg. at\n88,440. Thus, even though the population density of the entire\nUnited States is less than 100 people per square mile, see\nNCUA Br. 56 n.44; Oral Arg. Recording 39:34-38, the\ngeographical limitation forces a rural district to be much\nsmaller. We are confident that such districts will not be so\nenormous as to amount to federations of “unrelated geographic\nunits.” See First Nat’l Bank III, 522 U.S. at 502.\n\n In arguing that the amended “rural district” definition is\nunreasonable, the Association relies heavily on the interpretive\nanalysis performed by the District Court. See Am. Bankers\nAss’n, 306 F. Supp. 3d at 67-69. The court’s core contention\nis that we must construe “rural” and “district” together as a\nspecialized phrase that meant, in 1934, an “area[] much smaller\nthan a state.” Id. at 68. For support, the District Court not only\nconsulted two 1934-era dictionaries, results of a Westlaw\nsearch of contemporaneous opinions, and a database containing\nhistorical uses of the phrase, but also invoked the noscitur a\nsociis canon. See id. at 67-69.\n\n The proffered definitional evidence is pretty thin. The\n1934-era dictionaries described what the specialized phrase\n\f 34\n“rural district” meant in Great Britain, not in the much larger\nand more expansive United States, which by the 1930s\nencompassed forty-eight continental states. See, e.g., Rural\nDistrict, WEBSTER’S NEW INTERNATIONAL DICTIONARY OF THE\nENGLISH LANGUAGE 1861 (1934) (“in England, a subdivision\nof an administrative county embracing . . . county parishes”).\n\n The proffered Westlaw search results included scores of\nfederal and state judicial and administrative opinions. See\nPlaintiff’s Supplemental Memorandum app. 6, Am. Bankers\nAss’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C. filed Mar.\n16, 2018), ECF No. 32-6. But we find little of use in those\ndocuments. Many of those opinions do not discuss size or\npopulation. See, e.g., Nicolai v. Wis. Power & Light Co., 269\nN.W. 281, 282-83 (Wis. 1936). Those that do involve specific\nrural districts whose sizes were between a town and a state.\nSee, e.g., Sarther Grocery Co. v Comm’r, 22 B.T.A. 1273, 1274\n(1931). But none declares that rural districts by definition are\nrestricted to any particular size or population.\n\n The District Court next turned to the Corpus of Historical\nAmerican Usage, a free and public online database. See Am.\nBankers Ass’n, 306 F. Supp. 3d at 68 & n.5. The database notes\n197 mentions of the phrase in the first half of the twentieth\ncentury and only 37 in the second. See Search Results for\nRURAL DISTRICTS and RURAL DISTRICT, CORPUS OF\nHISTORICAL AMERICAN ENGLISH, https://www.english-\ncorpora.org/coha/ (follow “List” hyperlink; then search\nmatching strings field for “RURAL DISTRICT”). The\nperceived drop-off led the District Court to determine that,\n“even if rural district does not carry meaning distinct from its\nindividual words today, it did in 1934.” Am. Bankers Ass’n,\n306 F. Supp. 3d at 68.\n\f 35\n Although federal courts may use “crude[]” searches on\ndatabases to learn of ambiguities in a statutory term,\nMuscarello v. United States, 524 U.S. 125, 129-30 (1998), the\nsearch here did not suffice to show that the agency’s definition\nwas unreasonable. Much more is required to cabin the\nagency’s discretion. A search of a commercial database,\nProQuest, reveals that the phrase appeared at least 500 times in\nthe second half of the century. We are not confident that the\nproffered evidence establishes a particular historical trend.\n\n The District Court lastly invoked noscitur a sociis. See\nAm. Bankers Ass’n, 306 F. Supp. 3d at 68-69. The canon is\ninapposite. Recall that the term “rural district” was listed with\n“community,” not “local community,” in 1934. At the time,\n“community” could refer to “[s]ociety at large” or a\n“commonwealth or state.” Community, WEBSTER’S NEW\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 452\n(1934). Indeed, we said that the word is “too indefinite to be\nused for purposes of exact measurement in terms of acres or\nsquare miles.” Lukens Steel Co. v. Perkins, 107 F.2d 627, 631\n(D.C. Cir. 1939), rev’d sub nom., Perkins v. Lukens Steel Co.,\n310 U.S. 113 (1940). Thus, we do not see size or population\nas a connection between the terms.\n\n The increased population cap is consistent with not only\nthe Act’s text but also its purposes. For instance, as the\npreamble noted, the expanded definition allows community\ncredit unions in rural districts to reach more “persons of modest\nmeans who may reside” in those areas. 81 Fed. Reg. at 88,416;\nsee also § 2(1), (2), 112 Stat. at 913.\n\n And the change is neither arbitrary nor capricious. The\nnew rule explains that it provides a “level of operating\nefficiencies and scale” making rural districts attractive options\nfor prospective credit unions. 81 Fed. Reg. at 88,416. “[A]\n\f 36\nsufficiently large population base is essential to enable” them\n“to offer financial services economically.” Id. at 88,417. The\nNCUA also chose the 1-million figure based on prior\nexperience. The agency noted that it had approved of eight\ndistricts with populations of more than 250,000, and that one\nof them already had exceeded 1 million. See id. “Having set a\n1 million precedent in one state,” the agency felt justified in\nhaving a “fixed 1 million population cap for the other 49\nstates.” Id. (emphasis added). We cannot say this policy\nchoice lacks rational explanation.\n\n The Association says the NCUA failed to consider prior\nagency decisions. See Am. Bankers Ass’n Br. 61. But we see\nno discrepancy and thus summarily reject the objection. The\nAssociation also turns to troubling hypothetical examples of\nrural districts with unruly shapes and those with dense urban\nareas such as Denver, Colorado. See NCUA Br. 55-61; Oral\nArg. Recording 37:55-38:05. Again, such implausible outliers\ndo not impugn the rule’s general reasonableness.\n\n VI.\n\n We now consider the remedy. To sum up, we hold that\ndefining certain Combined Statistical Areas or portions of them\nas local communities and raising the population cap for rural\ndistricts are consistent with the Act and reasonably explained.\nThus, we sustain both aspects of the challenged rule. We also\nleave undisturbed the portion of the District Court’s opinion\nthat the parties do not contest on appeal.\n\n But we deem the eliminated core requirement to be\narbitrary and capricious. When a rule is contrary to law, the\n“ordinary practice is to vacate” it. United Steel v. Mine Safety\n& Health Admin., 925 F.3d 1279, 1287 (D.C. Cir. 2019); see\nalso 5 U.S.C. § 706(2) (noting that the “reviewing court\n\f 37\nshall . . . set aside” unlawful agency action). But in “rare\ncases,” we will opt instead to remand without vacating the rule,\nso that the agency can “correct its errors.” United Steel, 925\nF.3d at 1287; see also 5 U.S.C. § 702 (“Nothing\nherein . . . affects . . . the power or duty of the court to . . . deny\nrelief on any . . . appropriate . . . equitable ground . . . .”); Oglala\nSioux Tribe v. U.S. Nuclear Regulatory Comm’n, 896 F.3d 520,\n536 (D.C. Cir. 2018). In considering whether to adopt the latter\nequitable remedy, we balance “(1) the seriousness of the\ndeficiencies of the action, that is, how likely it is the agency\nwill be able to justify its decision on remand; and (2) the\ndisruptive consequences of vacatur.” United Steel, 925 F.3d at\n1287 (citation omitted). A strong showing of one factor may\nobviate the need to find a similar showing of the other. See,\ne.g., Fox Television Stations, Inc. v. FCC, 280 F.3d 1027, 1049\n(D.C. Cir. 2002) (holding that, because the agency likely could\njustify its action on remand, the potential for disruption was\n“only barely relevant”); Allied-Signal, Inc. v. U.S. Nuclear\nRegulatory Comm’n, 988 F.2d 146, 152 (D.C. Cir. 1993)\n(determining that, because vacatur would give regulated parties\na “peculiar windfall,” the meager chance of justifying the\naction was given “little weight” in the remedial analysis).\n\n The NCUA has not requested remand without vacatur in\nthis case. But because we have a “duty” to ensure the propriety\nof the APA remedy, 5 U.S.C. § 702, we hold that we have the\ndiscretion to raise the issue sua sponte, cf. Igonia v. Califano,\n568 F.2d 1383, 1387 (D.C. Cir. 1977); Gaddis v. Dixie Realty\nCo., 420 F.2d 245, 247 (D.C. Cir. 1969) (per curiam). Remand\nwithout vacatur is appropriate here.\n\n We conclude that the NCUA might be able to offer a\nsatisfactory reason on remand. And as for disruptive effect, we\nperceive a substantial likelihood that vacating the rule could\nmake it more difficult for some poor and minority suburban\n\f 38\nresidents to receive adequate financial services. Even\ntemporarily depriving these members of the opportunity is\ninequitable, because it would “set back” the Act’s objective of\noffering financial services to people of small means. See Nat’l\nRes. Def. Council v. EPA, 489 F.3d 1364, 1374 (D.C. Cir.\n2007); see also Advocates for Highway & Auto Safety v. Fed.\nMotor Carrier Safety Admin., 429 F.3d 1136, 1152 (D.C. Cir.\n2005) (declining to vacate rule because, in the interim, it “may\ndo some good, if it does anything at all”).\n\n Given the potential for sufficient justification and the\nsubstantial likelihood of disruptive effect, we have a rare case\nin which vacatur is inappropriate. See United Steel, 925 F.3d\nat 1287. Thus, we remand without vacating the relevant part\nof the 2016 rule. We trust that the agency will act\nexpediently. See EME Homer City Generation, L.P. v. EPA,\n795 F.3d 118, 132 (D.C. Cir. 2015).\n\n * * *\n\n In short, we reverse the challenged portions of the District\nCourt’s summary judgment order. With respect to the\nqualification of certain Combined Statistical Areas as local\ncommunities and the increased population cap for rural\ndistricts, we direct the District Court to issue summary\njudgment in favor of the NCUA. As to the elimination of the\nurban-core requirement for local communities based on Core\nBased Statistical Areas, we direct the District Court to issue\nsummary judgment in favor of the Association and to remand,\nwithout vacating, the relevant portion of the 2016 rule for\nfurther explanation.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: AMERICAN BANKERS ASSOCIATION, Appellee v. NATIONAL CREDIT UNION ADMINISTRATION, Appellant\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-20\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 16, 2019 Decided August 20, 2019\\n\\n No. 18-5154\\n\\n AMERICAN BANKERS ASSOCIATION,\\n APPELLEE\\n\\n v.\\n\\n NATIONAL CREDIT UNION ADMINISTRATION,\\n APPELLANT\\n\\n\\n Consolidated with 18-5181\\n\\n\\n Appeals from the United States District Court\\n for the District of Columbia\\n (No. 1:16-cv-02394)\\n\\n\\n Daniel Aguilar, Attorney, U.S. Department of Justice,\\nargued the cause for Appellant-Cross-Appellee. With him on\\nthe briefs was Mark B. Stern, Attorney.\\n\\n Allison Jones Rushing was on the brief for amici curiae\\nCredit Union National Association, et al. in support of\\nAppellant-Cross-Appellee. Nicholas G. Gamse entered an\\nappearance.\\n\\f 2\\n Steven D. Gordon was on the brief for amici curiae State\\nBankers Associations in support Appellee-Cross-Appellant\\nAmerican Bankers Association.\\n\\n Robert A. Long Jr. argued the cause for Appellee-Cross-\\nAppellant. With him on the briefs were Andrew J. Soukup,\\nPhilip Levitz, and Lauren Moxley.\\n\\n Before: HENDERSON, PILLARD, and WILKINS, Circuit\\nJudges.\\n\\n Opinion for the Court filed by Circuit Judge WILKINS.\\n\\n WILKINS, Circuit Judge: Longstanding principles of\\nadministrative law teach us to give federal agencies breathing\\nroom when they make policy and “resolv[e] the struggle\\nbetween competing views of the public interest.” Chevron,\\nU.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 866\\n(1984). And because many policy decisions merge with legal\\nones, Chevron requires us frequently to sustain agency\\ninterpretations of certain federal statutes. Congress often\\nexpects agencies, with their political accountability, “bod[ies]\\nof experience[,] and informed judgment,” to make sound\\ninterpretive choices “with the force of law.” United States v.\\nMead Corp., 533 U.S. 218, 227, 229 (2001) (citation omitted).\\n\\n Congress expressly tasked the National Credit Union\\nAdministration (NCUA) with making such choices in defining\\nthe reach of federal credit unions. Since the Great Depression,\\nCongress has maintained a “system of federal credit unions\\nthat . . . provide credit at reasonable rates” and banking\\nservices to “people of ‘small means.’” First Nat’l Bank & Tr.\\nCo. v. NCUA (First Nat’l Bank I), 988 F.2d 1272, 1274 (D.C.\\nCir. 1993) (citation omitted), aff’d, 522 U.S. 479 (1998).\\nAlthough a private bank may solicit and welcome customers\\n\\f 3\\nfrom anywhere, Congress has limited whom these federal\\nfinancial institutions may serve. For instance, certain\\ninstitutions called “community credit unions” may cover\\nindividuals and entities only within a preapproved\\ngeographical area. The credit union will not receive a federal\\ncharter (and thus cannot start operations) unless it first proffers\\na geographical coverage area and the NCUA accepts the\\nproposal. Congress explicitly assigns the agency the task of\\ncreating vetting standards.\\n\\n Exercising its expressly delegated power, the NCUA has\\npromulgated a final rule that makes it easier for community\\ncredit unions to expand their geographical coverage and thus to\\nreach more potential members. Representing competitors to\\nthe credit unions, the American Bankers Association\\n(Association) has challenged the NCUA’s new rule as neither\\n“in accordance with law” nor within “statutory jurisdiction.” 5\\nU.S.C. § 706(2)(A), (C). The District Court vacated significant\\nportions of the rule, deeming them to be based on unreasonable\\nagency interpretations of the Federal Credit Union Act (Act),\\nPub. L. No. 73-467, 48 Stat. 1216 (1934) (codified as amended\\nat 12 U.S.C. §§ 1751 to 1795k). See Am. Bankers Ass’n v.\\nNCUA, 306 F. Supp. 3d 44, 61, 69-70 (D.D.C. 2018).\\n\\n We appreciate the District Court’s conclusions, made after\\na thoughtful analysis of the Act. But we ultimately disagree\\nwith many of them. In this facial challenge, we review the rule\\nnot as armchair bankers or geographers, but rather as lay judges\\ncognizant that Congress expressly delegated certain policy\\nchoices to the NCUA. After considering the Act’s text,\\npurpose, and legislative history, we hold the agency’s policy\\nchoices “entirely appropriate” for the most part. Chevron, 467\\nU.S. at 865. We therefore sustain the bulk of the rule. Still, we\\ndo not rubber-stamp this regulation. We remand, without\\nvacating, one portion for further consideration of the\\n\\f 4\\ndiscriminatory impact it might have on poor and minority\\nurban residents.\\n\\n I.\\n\\n A.\\n\\n The nation’s credit unions started in the early twentieth\\ncentury “as a populist mechanism designed to empower\\nfarmers against bad loans.” Mehrsa Baradaran, How the Poor\\nGot Cut Out of Banking, 62 EMORY L.J. 483, 500 (2013).\\nWalloped by crop failures and the Great Depression, farmers\\nseeking credit became not only increasingly suspicious of\\ntraditional bankers, who “disregard[ed]” poor individuals and\\nstayed in the big cities, but also fearful of loan sharks, “who\\nwould extract ‘up to a thousand percent’ in interest rates.” Id.\\nat 500-01 (quoting 80 CONG. REC. 6752 (1936) (statement of\\nRep. Lundeen)). The farmers thus began to build their own\\ncredit networks.\\n\\n In a national grassroots campaign, farmers created\\nlocalized, non-profit “credit groups” collecting funds from and\\nloaning small sums to one another at low interest rates. See id.\\nat 501-02. The success of any such self-help institution\\n“hinge[s] on the interpersonal dynamics of its members:\\nLenders must be able to evaluate the ability and willingness of\\npotential borrowers to pay back their loans and borrowers must\\nfeel obligated to pay back those loans.” Wendy Cassity, Note,\\nThe Case for a Credit Union Community Reinvestment Act, 100\\nCOLUM. L. REV. 331, 337 (2000); see also First Nat’l Bank &\\nTr. Co. v. NCUA (First Nat’l Bank II), 90 F.3d 525, 526 (D.C.\\nCir. 1996), aff’d, 522 U.S. 479 (1998).\\n\\n By 1934, individuals had organized about 3,000 local\\ncredit unions, with about 750,000 members. See 80 CONG.\\n\\f 5\\nREC. at 6753. Recognizing the success of credit unions at the\\nstate level, Congress created a federal system that year by\\npassing the Act. Legislators worried that “usurious money\\nlending . . . obviously destroy[ed] vast totals of buying power\\n[once held by] . . . the average worker.” H.R. REP. NO. 73-\\n2021, at 1-2 (1934); see also S. REP. NO. 73-555, at 1 (1934).\\nCongress touted the Act’s ability to “make more available to\\npeople of small means credit for provident purposes.” H.R.\\nREP. NO. 73-2021, at 1; see also S. REP. NO. 73-555, at 1.\\n\\n Credit unions multiplied over the ensuing decades. By\\n1970, Congress created an independent agency to supervise\\nfederal credit unions: the NCUA. See Pub. L. No. 91-468, 84\\nStat. 994 (1970) (codified as amended in scattered sections of\\n12 U.S.C.); see also Swan v. Clinton, 100 F.3d 973, 974 (D.C.\\nCir. 1996) (noting that Congress “entrusted” the agency with\\n“the responsibility of overseeing” federal credit unions).\\nLegislators thought that the agency would be “more responsive\\nto the needs of credit unions” and would “provide more flexible\\nand innovative regulation” than prior government agencies,\\nwhich did not have federal credit unions as their sole focus. S.\\nREP. NO. 91-518, at 3 (1969).\\n\\n The NCUA faced its first major crisis at the end of the\\n1970s. After years of economic decline in several industrial\\nsectors, federal credit unions tied to those business sectors\\nbegan to suffer. The resulting liquidation of numerous credit\\nunions “threaten[ed] ‘the safety and soundness of the federal\\ncredit union system.’” Cassity, supra, at 338-39 (footnote\\nomitted). Reacting to the emergency, the NCUA in 1981\\npromulgated a groundbreaking rule that loosened a major size\\nlimitation on certain federal credit unions. Almost\\nimmediately, those financial institutions grew in membership.\\n\\f 6\\n Meanwhile, credit unions became “caught up in the\\nbroader changes in banking and faced internal as well as\\nexternal pressure to compete with [private] banks and seek\\nhigher profits.” Baradaran, supra, at 505. Unlike credit\\nunions, private, for-profit banks were “owned by equity holders\\nwho may not necessarily be customers (depositors or\\nborrowers),” and they did “not have similar membership and\\ncommercial lending restrictions” as credit unions. DARRYL E.\\nGETTER, CONG. RESEARCH SERV., IF11048, INTRODUCTION TO\\nBANK REGULATION: CREDIT UNIONS AND COMMUNITY BANKS:\\nA COMPARISON 1 (2018). To remain viable, credit unions\\n“started to focus on attracting more customers and expanding\\nthe industry.” Baradaran, supra, at 505. As part of that\\nstrategy, many consolidated through mergers. And private\\nbanks soon treated credit unions as serious competitors,\\nseeking to curb their growth. See NCUA v. First Nat’l Bank &\\nTr. Co. (First Nat’l Bank III), 522 U.S. 479, 485 (1998); First\\nNat’l Bank I, 988 F.2d at 1276.\\n\\n In 1998, the banking industry successfully challenged as\\ncontrary to the Act the 1981 rule that had eased size limitations\\nfor certain federal credit unions. See First Nat’l Bank III, 522\\nU.S. at 503. Congress swiftly responded. In less than six\\nmonths, legislators amended the Act, superseding the holding\\nin First National Bank III, loosening size limitations on certain\\nfederal credit unions, and adding other reforms. See Credit\\nUnion Membership Access Act, Pub. L. No. 105-219, 112 Stat.\\n913 (1998) (codified as amended in scattered sections of 12\\nU.S.C.). Partly because of the 1998 amendments and related\\nNCUA regulations, credit unions continued to merge and grow\\nin membership. Now, more than 61 million customers perform\\ntheir banking services at about 3,400 federal credit unions. See\\n2018 NAT’L CREDIT UNION ADMIN. ANN. REP. 192.\\n\\f 7\\n B.\\n\\n Federal credit unions pool funds from – and give loans\\nto – their members and other credit-union entities. 12 U.S.C.\\n§ 1757(5), (6). A credit union’s members, whether individual\\nor corporate, must come from the credit union’s membership\\n“field,” id. § 1753(5), which is based on a shared occupation,\\nassociation, or geographical area. Members receive regular\\ndividends. Id. § 1763. Congress has shielded federal credit\\nunions from federal corporate income taxes and most state and\\nlocal taxes, but members must pay taxes on their dividends.\\nSee JAMES M. BICKLEY, CONG. RESEARCH SERV., 97-548 E,\\nSHOULD CREDIT UNIONS BE TAXED? 3-5 (2005).\\n\\n To create a federal credit union, at least seven individuals\\nmust present a proposed charter and pay a fee to the NCUA.\\nSee MICHAEL P. MALLOY, BANKING LAW & REGULATION\\n§ 2.04 (2d ed. 2019). In the application, the organizers must\\npledge to deposit funds for shares in the institution and must\\ndescribe the credit union’s proposed membership field. 12\\nU.S.C. § 1753(3), (5). The NCUA must approve the charter\\nbefore the institution may start. See id. § 1754. The agency\\nwill complete an “appropriate investigation” and determine the\\n“general character and fitness” of the organizers, the\\n“economic advisability of establishing” the credit union, and\\nthe “conform[ity]” of proposal details with the Act. Id.\\n\\n The Act governs two types of federal credit unions:\\n“common-bond” credit unions and “community” credit unions.\\nSee id. § 1759(b). This case deals with the latter category. The\\n1934 version of the Act required a community credit union’s\\nmembership field to reflect a particular geographical area – to\\nwit, “a well-defined neighborhood, community, or rural\\ndistrict.” § 9, 48 Stat. at 1219. As amended in 1998, the Act\\nprovides that membership for a community credit union “shall\\n\\f 8\\nbe limited to . . . [p]ersons or organizations within a well-\\ndefined local community, neighborhood, or rural district.” 12\\nU.S.C. § 1759(b) (emphasis added). The 1998 version calls on\\nthe NCUA to “prescribe, by regulation, a definition for the term\\n‘well-defined local community, neighborhood, or rural\\ndistrict.’” Id. § 1759(g)(1). Thus, under the new regime,\\nindividuals seeking to organize a new community credit union\\n(or alter an existing one) must commit to serving members\\nwithin the NCUA’s contemporaneous definition of “local\\ncommunity, neighborhood, or rural district.” See S. REP. NO.\\n105-193, at 4, 8 (1998); H.R. REP. NO. 105-472, at 21 (1998).\\nAs part of their application to the NCUA, they must provide a\\nproposed description of the precise geographical area that the\\ncredit union would serve.\\n\\n Since 1998, there has been “dramatic growth” in the\\nnumber of community credit unions. U.S. GOV’T\\nACCOUNTABILITY OFF., GAO-07-29, CREDIT UNIONS:\\nGREATER TRANSPARENCY NEEDED ON WHO CREDIT UNIONS\\nSERVE AND ON SENIOR EXECUTIVE COMPENSATION\\nARRANGEMENTS 4 (2006). Despite a 11-percent drop in the\\nnumber of federal credit unions from 2000 to 2005, community\\ncredit unions doubled to 1,115. Id. at 4, 12. Meanwhile, the\\namount of assets in community credit unions quadrupled to\\n$104 billion. Id. at 4.\\n\\n C.\\n\\n On December 7, 2016, the NCUA amended its\\nmembership-field rules for community credit unions. See\\nChartering and Field of Membership Manual, 81 Fed. Reg.\\n88,412 (Dec. 7, 2016). Several changes rely on two terms\\ndevised by the Office of Management and Budget (OMB) and\\nbased on data collected by the Census Bureau (Census): “Core\\nBased Statistical Areas” and “Combined Statistical Areas.”\\n\\f 9\\n\\n The OMB has designated numerous regions around the\\ncountry as Core Based Statistical Areas, which comprise at\\nleast one urban cluster, or core, of 10,000 or more people and\\nadjacent counties with substantial commuting ties to that core.\\nSee U.S. CENSUS BUREAU, GEOGRAPHICAL PROGRAM,\\nGLOSSARY, https://www.census.gov/programs-surveys/geogra\\nphy/about/glossary.html. In layman’s terms, a Core Based\\nStatistical Area is a city or town and its suburbs.\\n\\n Meanwhile, a Combined Statistical Area is a conglomerate\\nof two or more adjoining Core Based Statistical Areas, each of\\nwhich has substantial commuting ties with at least one other\\nCore Based Statistical Area in the group. Id. Essentially, a\\nCombined Statistical Area is a regional hub with urban centers\\nconnected by commuting patterns. Combined Statistical Areas\\nmay “reflect broader social and economic interactions, such as\\nwholesaling, commodity distribution, and weekend recreation\\nactivities.” OFFICE OF MGMT. & BUDGET, EXEC. OFFICE OF THE\\nPRESIDENT, OMB BULL. NO. 15-01, REVISED DELINEATIONS OF\\nMETROPOLITAN STATISTICAL AREAS, MICROPOLITAN\\nSTATISTICAL AREAS, AND COMBINED STATISTICAL AREAS, AND\\nGUIDANCE ON USES OF THE DELINEATIONS OF THESE AREAS\\napp. 2-3 (2015).\\n\\n Relevant here, the 2016 rule made two changes to the\\nNCUA’s definition of the term “local community” under\\n§ 1759(b)(3) and one to that of “rural district.” The changes\\naffect what proposed membership areas satisfy the\\ngeographical limitation imposed by the Act.\\n\\n The first change to the “local community” definition\\ninvolves Combined Statistical Areas. A proposed area\\nqualifies as a local community if it encompasses the whole or\\na portion of a Combined Statistical Area and does not exceed a\\n\\f 10\\ndesignated population limit. See 81 Fed. Reg. at 88,440. The\\nNCUA has set that cap at 2.5 million people.\\n\\n The second change involves Core Based Statistical Areas.\\nThe parties agree that all or part of a Core Based Statistical\\nArea may qualify as a local community so long as it does not\\nexceed the population limit. But since 2010, the NCUA\\nrequired such a membership area to include the urban core.\\nThe new rule no longer requires that the core be included in the\\nlocal community that a credit union proposes to serve. See id.\\nat 88,413, 88,440.\\n\\n As for the “rural district” definition, the new rule increases\\nthe population cap for valid rural districts from 250,000 people\\n(or 3 percent of the population of the state where most eligible\\nresidents are located) to 1 million people. See id. at 88,416,\\n88,440. The new population limit works with two other\\nconstraints set by the rule: (1) an outer geographical limit on\\nhow far a rural district may extend past the borders of the credit\\nunion’s headquarters state; and (2) a requirement either that\\nmost eligible residents reside in Census-designated rural areas,\\nor that the population density of the proposed district equals\\n100 or fewer people per square mile. See id. at 88,440.\\n\\n D.\\n\\n On the day the NCUA published the rule, the Association\\nfiled this injunctive and declaratory action in the District Court.\\nThe Association claimed that the three changes described\\nabove were not only arbitrary and capricious under the\\nAdministrative Procedure Act (APA), Pub. L. No. 79-404, 60\\nStat. 237 (1946) (codified as 5 U.S.C. §§ 701-06), but also\\nunreasonable and entitled to no deference under Chevron. The\\nagency and Association filed cross-motions for summary\\n\\f 11\\njudgment. On March 29, 2018, the District Court granted both\\nmotions in part and denied them in part.\\n\\n The court made three relevant holdings. First, it rejected\\nas unreasonable the qualification of certain Combined\\nStatistical Areas as local communities. See Am. Bankers Ass’n,\\n306 F. Supp. 3d at 61. Second, it sustained as well-reasoned\\nthe elimination of the core requirement from the Core Based\\nStatistical Area a credit union proposes to serve as its local\\ncommunity. Id. at 64-65. Third, it rejected as unreasonable the\\nincreased population cap for rural districts. Id. at 69-70. (It\\nalso sustained a separate portion of the rule, which the\\nAssociation does not challenge here.)\\n\\n The NCUA and Association timely appealed. We have\\nappellate jurisdiction under 28 U.S.C. § 1291.\\n\\n II.\\n\\n At the outset, we must assure ourselves of our subject\\nmatter jurisdiction over the appellate proceeding. See, e.g.,\\nUnited States v. Gooch, 842 F.3d 1274, 1277 (D.C. Cir. 2016).\\n\\n In their original briefing, the parties failed to apprise us of\\na rule that was promulgated while this appeal was pending and\\nthat changed membership-field requirements for community\\ncredit unions. See Chartering and Field of Membership, 83\\nFed. Reg. 30,289 (June 28, 2018). The 2018 rule eliminated\\nthe portion of the 2016 rule allowing Combined Statistical\\nAreas to qualify as local communities. Compare 12 C.F.R. pt.\\n701, app. B, ch. 2 § V.A.2 (2018), with id. (2019). The 2018\\nrule preamble did not specifically discuss the removal but\\nconcluded that “[a]ny modification in th[e] final rule is\\nconsistent with the District Court decision” in this action. 83\\nFed. Reg. at 30,291.\\n\\f 12\\n\\n “Under the mootness doctrine, we cannot decide a case if\\n‘events have so transpired that the decision will neither\\npresently affect the parties’ rights nor have a more-than-\\nspeculative chance of affecting them in the future.’” Reid v.\\nHurwitz, 920 F.3d 828, 832 (D.C. Cir. 2019) (citation omitted).\\nThe same principle applies to individual claims. See Tucson\\nMed. Ctr. v. Sullivan, 947 F.2d 971, 977 (D.C. Cir. 1991).\\nAccordingly, if a rule under review (or a portion of it) is\\nsuperseded or amended during an appeal, the proceeding (or\\nrelevant part) might be moot. See, e.g., Am. Bankers Ass’n v.\\nNCUA, 271 F.3d 262, 274 (D.C. Cir. 2001). We therefore\\nrequested supplemental briefing on the issue of appellate\\njurisdiction.\\n\\n Based on the government’s submission and\\nrepresentations at oral argument, we hold that the portion of the\\nappeal related to Combined Statistical Areas is not moot.\\n“[T]he mere power to [reinstitute] a challenged law is not a\\nsufficient basis on which a court can conclude that” a challenge\\nremains live. Nat’l Black Police Ass’n v. District of Columbia,\\n108 F.3d 346, 349 (D.C. Cir. 1997). The government has stated\\nthat, if we reverse the relevant part of the District Court’s\\ndecision, all three members of the NCUA’s board intend to\\nreinstitute the Combined Statistical Area portion of the 2016\\nrule if we reverse the relevant part of the District Court’s\\ndecision. See Decl. of Michael McKenna ¶ 3, ECF No.\\n1781123; Oral Arg. Recording 11:33-48. Accordingly, City of\\nMesquite v. Aladdin’s Castle, Inc., 455 U.S. 283 (1982),\\ngoverns this case. In Aladdin’s Castle, the Supreme Court\\ndeemed the controversy live in part because the government\\nhad announced “an intention” to restore the rule under\\nchallenge if the lower-court decision were vacated. Id. at 289\\n& n.11; see also Am. Bankers Ass’n, 271 F.3d at 274. The\\nNCUA’s submission and representations evince such an\\n\\f 13\\nintention here, and the Association – which bears the “heavy\\nburden” of proving mootness, Friends of the Earth, Inc. v.\\nLaidlaw Envtl. Servs. (TOC), Inc., 528 U.S. 167, 189 (2000)\\n(citation omitted) – offers no evidence to the contrary.\\n\\n The Association attempts to distinguish Aladdin’s Castle\\non two grounds, but neither sways us. First, the Association\\nnotes that the city government in Aladdin’s Castle said it would\\nreenact “precisely the same” law, see 455 U.S. at 289, but that\\nin this case any future notice-and-comment proceedings might\\nproduce a different “local community” definition, perhaps not\\neven relying on Combined Statistical Areas. After all, the\\nagency must keep a “flexible and open-minded attitude” during\\nthe process. See Nat’l Tour Brokers Ass’n v. United States, 591\\nF.2d 896, 902 (D.C. Cir. 1978). We grant that commenters\\nmight convince the NCUA to change its mind; mann tracht,\\nund Gott lacht. But that strikes us as speculative as public input\\nconvincing Mesquite legislators to enact a different law. We\\ndo not see Aladdin’s Castle turning on such conjecture.\\nInstead, we see a live dispute because there is “no certainty”\\nthat the NCUA will forego reinstating the same Combined\\nStatistical Area definition. Aladdin’s Castle, 455 U.S. at 289;\\nsee also Trinity Lutheran Church of Columbia, Inc. v. Comer,\\n137 S. Ct. 2012, 2019 n.1 (2017) (holding that the court must\\nfind “absolutely clear that the allegedly wrongful behavior\\ncould not reasonably be expected to recur” (emphasis added)\\n(quoting Friends of the Earth, 528 U.S. at 189)). After all, the\\nNCUA continues to defend the definition here. See Knox v.\\nServ. Emps. Int’l Union, Local 100, 567 U.S. 298, 307 (2012).\\n\\n Second, the Association observes that both sides in\\nAladdin’s Castle urged the Supreme Court to treat their dispute\\nas live. In contrast, only the NCUA seeks to proceed here; the\\nAssociation would prefer to wait until the agency reinstitutes\\nthe rule. But the existence or absence of jurisdiction does not\\n\\f 14\\nturn on which parties challenge or defend it. Cf. Bender v.\\nWilliamsport Area Sch. Dist., 475 U.S. 534, 541 (1986).\\nBecause the NCUA remains bound by the lower-court\\njudgment, the present injury renders irrelevant the\\nAssociation’s preference. “Jurisdiction existing,” our duty to\\ndecide the appeal “is ‘virtually unflagging.’” Sprint\\nCommc’ns, Inc. v. Jacobs, 571 U.S. 69, 77 (2013) (citation\\nomitted).\\n\\n In short, we may review the challenge to the rule change\\ninvolving Combined Statistical Areas. We see no jurisdictional\\nissues with the rest of the appeal. We thus turn to the merits.\\n\\n III.\\n\\n We review de novo the District Court’s rulings on\\nsummary judgment. See Loan Syndications & Trading Ass’n\\nv. SEC, 882 F.3d 220, 222 (D.C. Cir. 2018). We review the\\nadministrative record and give “no particular deference” to the\\nDistrict Court’s views. Oceana, Inc. v. Ross, 920 F.3d 855, 860\\n(D.C. Cir. 2019) (citation omitted).\\n\\n The APA governs this suit. In relevant part, the statute\\nprovides that we “decide all relevant questions of law” and\\n“interpret . . . statutory provisions.” 5 U.S.C. § 706. We\\nordinarily set aside agency actions that are either “arbitrary,\\ncapricious, an abuse of discretion, or otherwise not in\\naccordance with law,” id. § 706(2)(A), or “in excess of\\nstatutory jurisdiction, authority, or limitations, or short of\\nstatutory right,” id. § 706(2)(C).\\n\\n We review the agency rule in accordance with the familiar\\nChevron doctrine, a two-prong test for determining whether an\\nagency “has stayed within the bounds of its statutory authority”\\nwhen issuing its action. City of Arlington v. FCC, 569 U.S.\\n\\f 15\\n290, 297 (2013) (emphasis omitted). At the first step, we\\ndetermine “whether Congress has directly spoken to the precise\\nquestion at issue,” and we “give effect” to any “unambiguously\\nexpressed intent.” Chevron, 467 U.S. at 842-43 & n.9.\\n\\n If we glean no such unambiguous intent, we turn to the\\nsecond step and determine “whether the agency’s answer” to\\nthe question “is based on a permissible construction of the\\nstatute.” Id. at 843. By arriving at the second step, we have\\nconcluded that Congress either explicitly or implicitly\\ndelegated to the agency the lawmaking authority to clarify the\\nstatute. We presume that Congress would not authorize the\\npromulgation of an “[im]permissible construction.” Chevron,\\n467 U.S. at 843. Accordingly, we will set aside agency actions\\nbased on such a construction, because they are either “not in\\naccordance with law” or “in excess of statutory jurisdiction.”\\n5 U.S.C. § 706(2)(A), (C); see also CSX Transp., Inc. v.\\nSurface Transp. Bd., 754 F.3d 1056, 1063 (D.C. Cir. 2014);\\nAss’n of Privacy Sector Colls. & Univs. v. Duncan, 681 F.3d\\n427, 441 (D.C. Cir. 2012).\\n\\n IV.\\n\\n For all three challenges, the first step of the Chevron\\nanalysis proceeds in the same way. “We begin our analysis, as\\nalways, with the statutory text.” Tesoro Alaska Co. v. FERC,\\n778 F.3d 1034, 1038 (D.C. Cir. 2015). Congress having\\nexpressly assigned the NCUA the power to define the\\nchallenged terms, see 12 U.S.C. § 1759(g)(1), we may proceed\\nto Chevron’s second prong without further analysis, see U.S.\\nTelecom Ass’n v. FCC, 825 F.3d 674, 717 (D.C. Cir. 2016);\\nRush Univ. Med. Ctr. v. Burwell, 763 F.3d 754, 760 (7th Cir.\\n2014); Comm’r v. Pepsi-Cola Niagara Bottling Corp., 399\\nF.2d 390, 393 (2d Cir. 1968) (Friendly, J.) (“When Congress\\nhas used a general term and has empowered an administrator\\n\\f 16\\nto define it, the courts must respect his construction if this is\\nwithin the range of reason.”).\\n\\n An express delegation of definitional power “necessarily\\nsuggests that Congress did not intend the [terms] to be applied\\nin [their] plain meaning sense,” Women Involved in Farm\\nEcon. v. U.S. Dep’t of Agric., 876 F.2d 994, 1000 (D.C. Cir.\\n1989), that they are not “self-defining,” id., and that the agency\\n“enjoy[s] broad discretion” in how to define them, Lindeen v.\\nSEC, 825 F.3d 646, 653 (D.C. Cir. 2016). In Chevron terms,\\nCongress through explicit language “has directly spoken to the\\nprecise question” of whether the identified terms must carry\\ncertain meanings. 467 U.S. at 842-43. The answer is no. See\\nBuongiorno v. Sullivan, 912 F.2d 504, 509 (D.C. Cir. 1990)\\n(Thomas, J.) (“When Congress expressly delegates the\\nauthority to fill a gap in a statute, Congress speaks, in effect,\\ndirectly, and says, succinctly, that it wants the agency to\\nannotate its words.”). To hold otherwise at the first Chevron\\nstep would “undermine” the ability of Congress to delegate\\ndefinitional power. Rush Univ. Med. Ctr., 763 F.3d at 760.\\n\\n Consequently, we turn to whether the NCUA’s definitions\\nare “based on a permissible construction of the statute.”\\nChevron, 467 U.S. at 843.\\n\\n V.\\n\\n Agency interpretations promulgated to fill an explicit\\nlegislative gap “are given controlling weight unless they are\\narbitrary, capricious, or manifestly contrary to the statute.”\\nChevron, 467 U.S. at 844; see also Pharm. Research & Mfrs.\\nof Am. v. FTC, 790 F.3d 198, 204 (D.C. Cir. 2015).\\n\\n Under arbitrary and capricious review, “we may not\\nsubstitute our own judgment for that” of the agency. FERC v.\\n\\f 17\\nElec. Power Supply Ass’n, 136 S. Ct. 760, 782 (2016); accord\\nDep’t of Commerce v. New York, 139 S. Ct. 2551, 2569 (2019).\\nStill, we are “not a ‘rubber stamp,’” Oceana, 920 F.3d at 863;\\n“the agency must examine the relevant data and articulate a\\nsatisfactory explanation for its action,” Motor Vehicle Mfrs.\\nAss’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co. (State\\nFarm), 463 U.S. 29, 43 (1983). A rule is arbitrary and\\ncapricious if (1) the agency “has relied on factors which\\nCongress has not intended it to consider”; (2) the agency\\n“entirely failed to consider an important aspect of the\\nproblem”; (3) the agency’s explanation “runs counter to the\\nevidence before the agency”; or (4) the explanation “is so\\nimplausible that it could not be ascribed to a difference in view\\nor the product of agency expertise.” Id.\\n\\n In turn, we assess three definitional changes in the 2016\\nrule: (1) qualifying Combined Statistical Areas as local\\ncommunities; (2) eliminating the core requirement for local\\ncommunities based on Core Based Statistical Areas; and\\n(3) raising the population cap for rural districts. We sustain the\\nfirst and third amendments in full. As for the second, we hold\\nthat it is rationally related to the Act’s text and purposes, but\\nthat it is insufficiently explained.\\n\\n A.\\n\\n The District Court rejected the first change because it\\napproved certain Combined Statistical Areas “no matter how\\ngeographically dispersed and unconnected” the “members may\\nbe.” Am. Bankers Ass’n, 306 F. Supp. 3d at 61. To the court,\\nthe approval did not fit with the term “local community,”\\nwhich, in its view, “encompass[es] an area no larger than a\\ncounty.” Id. at 58, 61. We respectfully disagree.\\n\\f 18\\n The NCUA possesses vast discretion to define terms\\nbecause Congress expressly has given it such power. But the\\nauthority is not boundless. The agency must craft a reasonable\\ndefinition consistent with the Act’s text and purposes; that is\\ncentral to the review we apply at Chevron’s second step. Here,\\nthe NCUA’s definition meets the standard.\\n\\n We first focus on the text. Congress introduced the phrase\\n“local community” in the 1998 amendments. The word\\n“community” had a broad scope at the time. It meant not only\\n“society at large” but also a “body of individuals organized into\\na unit or manifesting usu[ally] with awareness of some\\nunifying trait.” See Community, WEBSTER’S THIRD\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE\\nUNABRIDGED 460 (1993). The group could be “united by\\nhistorical consciousness or . . . common social, economic, and\\npolitical interests.” Id. But the unifying trait could also be\\nsimply “living in a particular place or region.” Id.\\n\\n The NCUA has recognized that the modifier “local”\\n“reflects congressional intent that it takes ‘a more circumspect\\nand restricted approach to chartering community credit\\nunions.’” Am. Bankers Ass’n, 271 F.3d at 273 (citation\\nomitted); see also Oral Arg. Recording 3:16-18. Indeed,\\nCongress made clear its intention to “modif[y]” the “current\\nlaw regarding community credit unions” by adding the word\\n“local” to community. S. REP. NO. 105-193, at 6-7.\\n\\n Insertion of the modifier “local” before “community”\\nimplies that the community “relate[s] to” a “particular limited\\ndistrict” or is “confined to a particular place.” Local,\\nWEBSTER’S THIRD INTERNATIONAL DICTIONARY OF THE\\nENGLISH LANGUAGE UNABRIDGED 1327 (1993). But that place\\nneed not be the size of a county, as the District Court held. The\\nSupreme Court has recognized that the geographical areas\\n\\f 19\\n“need not be small.” First Nat’l Bank III, 522 U.S. at 492. And\\nif Congress wanted a local community to correspond to a\\nparticular geographical unit, such as a county, “‘it easily could\\nhave written’ that limitation explicitly.” NCUA Br. 25-26\\n(quoting Ali v. Fed. Bureau of Prisons, 552 U.S. 214, 227\\n(2008)); see also Credit Union Nat’l Ass’n Amicus Br. 15. The\\nNCUA sensibly reads the term “local” to mean simply that the\\ncommunity, regardless of shape or size, should be neither\\n“broad” nor “general.” Local, SUPRA, at 1327.\\n\\n To be clear, we do not hold today that the NCUA must\\nconsider only bonds and social connections as understood in\\n1998. The parties agree, and thus we assume, that the NCUA,\\ndespite its expressly delegated authority, must adopt a\\ndefinition consistent with what the term “local community”\\nmeant in 1998, the time of its adoption.\\n\\n After consulting state statutes and invoking the canon of\\nnoscitur a sociis, the District Court developed a rather size-\\nrestrictive meaning for the phrase. See Am. Bankers Ass’n, 306\\nF. Supp. 3d at 57-59. But we do not think that defining a “local\\ncommunity” to refer to an area larger than a county is an\\nunreasonable interpretation of the Act’s text.\\n\\n We receive little guidance from the state statutes in effect\\nin 1998. Indeed, some of the statutes considered “local\\ncommunities” to be quite large. See, e.g., ALASKA STAT.\\n§ 18.66.990(7) (1998) (defining “local community entity” as “a\\ncity or borough or other political subdivision of the state, a\\nnonprofit organization, or a combination of these” (emphasis\\nadded)); see also NCUA Br. 25 n.4.\\n\\n We also reject the District Court’s invocation of the\\nnoscitur a sociis canon. When several terms “are associated in\\na context suggesting that [they] have something in common,\\n\\f 20\\nthey should be assigned a permissible meaning that makes\\nthem similar.” ANTONIN SCALIA & BRYAN A. GARNER,\\nREADING LAW: THE INTERPRETATION OF LEGAL TEXTS 195\\n(2012). But the “substantive connection, or fit, between” the\\nterms here – local community, neighborhood, and rural\\ndistrict – is “not so tight or so self-evident.” Graham Cty. Soil\\n& Water Conservation Dist. v. United States ex rel. Wilson, 559\\nU.S. 280, 288 (2010). Only the word “neighborhood” has a\\ndictionary definition clearly suggesting a region of small size;\\nthe phrase “local community” does not, as we have explained\\nabove, and neither does “rural district,” as we explain below.\\nA size restriction would impermissibly “submerge[]” the\\nindependent “character” of the latter two terms. Babbitt v.\\nSweet Home Chapter of Cmtys. for a Great Or. (Sweet Home),\\n515 U.S. 687, 702 (1995) (citation omitted).\\n\\n The Association also points to other textual indicators.\\nBut contrary to what it suggests, see Am. Bankers Ass’n Br.\\n27-28, we see nothing in the record suggesting that “local\\ncommunity” is a term of art. The Association also says the\\nusage of word “local” in two other federal statutes indicates\\nthat rules permitting coverage areas of larger than a county\\nwould be manifestly contrary to the Act. See id. at 32; see also\\n15 U.S.C. § 2203 (1998) (defining “local” as “city, town,\\ncounty, . . . or other political subdivision of a State”); 18 U.S.C.\\n§ 666 (1998) (same). (The Association pointed to other laws,\\nbut they did not exist in 1998. See National Defense\\nAuthorization Act for Fiscal Year 2010, Pub. L. No. 111-84,\\n§ 4703(b), 123 Stat. 2190, 2837 (2009); PRIME Act Grants, 66\\nFed. Reg. 29,010 (May 29, 2001).) True enough. But the two\\nstatutes address materially distinct issues – fire prevention and\\nembezzlement – and thus do not indicate that the term “local”\\nmust imply a size limitation here. See Envtl. Def. v. Duke\\nEnergy Corp., 549 U.S. 561, 574 (2007) (“[M]ost words have\\n\\f 21\\ndifferent shades of meaning and consequently may be variously\\nconstrued . . . in different statutes . . . .” (citation omitted)).\\n\\n In addition to being consistent with the Act’s text, the\\nCombined Statistical Area definition rationally advances the\\nAct’s underlying purposes. In the 1998 amendments, Congress\\nmade two relevant findings about purpose. First, legislators\\nfound “essential” to the credit-union system a “meaningful\\naffinity and bond among members, manifested by a\\ncommonality of routine interaction[;] shared and related work\\nexperiences, interests, or activities[;] or the maintenance of an\\notherwise well-understood sense of cohesion or identity.” § 2,\\n112 Stat. at 914. Second, Congress highlighted the importance\\nof “credit union safety and soundness,” because a credit union\\non firm financial footing “will enhance the public benefit that\\ncitizens receive.” Id. The legislative history also confirms the\\nimportance of common bonds, see S. REP. NO. 73-555, at 2; see\\nalso H.R. REP. NO. 105-472, at 12; H.R. REP. NO. 73-2021, at\\n2; 144 CONG. REC. S9094 (daily ed. July 28, 1998) (statement\\nof Sen. Mikulski); id. at S8971-72 (daily ed. July 24, 1998)\\n(statement of Sen. Moseley-Braun), and economic “integrity,”\\nS. REP. NO. 105-193, at 3; see also 144 CONG. REC. S9094\\n(statement of Sen. Mikulski); id. at S8972 (statement of Sen.\\nMoseley-Braun), to federal credit unions.\\n\\n We recognize that there may be some tension between the\\nAct’s principal purposes: A credit union with exceedingly\\nclose ties among its members is unlikely to have a large enough\\ncustomer base to thrive economically. To the extent that such\\ntension exists, the Act leaves to the NCUA to strike a\\nreasonable balance. Congress was well aware that a viable\\ncredit union might serve a relatively large geographical area.\\nSee, e.g., H.R. REP. NO. 73-2021, at 2 (“[T]here are cases in\\nwhich communities and organizations cross State lines.”).\\n\\f 22\\n The NCUA did just that in promulgating the Combined\\nStatistical Area definition. That definition allows for larger\\ncommunity credit unions; the decision is consistent with\\ndecades of history promoting the economic viability of credit\\nunions in the face of banks and other competing financial\\ninstitutions. Nonetheless, the NCUA struck a balance by\\nensuring that members within the local community maintain\\nsomewhat of a commuter relationship with each other. As the\\nAssociation even acknowledges, commuting patterns “may\\nsometimes serve as a proxy for community interaction.” Am.\\nBankers Ass’n Br. 38 n.25. We see nothing irrational about\\nadopting the factor as a proxy for the common bond\\ncontemplated by Congress. Perhaps we would have made a\\ndifferent call had we been the policymakers. Perhaps we would\\nhave sought a tighter bond. Or perhaps we would not have\\nprioritized credit-union growth. See Iowa Bankers Ass’n\\nAmicus Br. 24-25. But we must “refrain from substituting\\n[our] own interstitial lawmaking for that” of the agency.\\nHousehold Credit Servs., Inc. v. Pfennig, 541 U.S. 232, 244\\n(2004) (citation omitted).\\n\\n The NCUA also reasonably explained its amendment to\\nthe “local community” definition. To begin with, the agency\\nreasonably circumscribed the size of a local community under\\nthe Combined Statistical Area rule by imposing a 2.5 million-\\nperson population limit. Used by the OMB in analogous\\ncircumstances, the cap is a “logical breaking point in terms of\\ncommunity cohesiveness with respect to a multijurisdictional\\narea.” Chartering and Field of Membership for Federal Credit\\nUnions, 75 Fed. Reg. 36,257, 36,259 (June 25, 2010).\\nMoreover, the NCUA reasonably relied on its prior experience\\nwith Core Based Statistical Areas, which no party disputes can\\nserve as local communities under the Act. The agency noted\\nthat the average geographic size of Combined Statistical Areas\\nwith populations of up to 2.5 million people is 4,553 square\\n\\f 23\\nmiles, and that the average size of the Core Based Statistical\\nAreas approved by the NCUA as local communities is 4,572\\nsquare miles. See 81 Fed. Reg. at 88,414-15. Given that\\nsimilarity in size, the NCUA reasonably considered adopted its\\npopulation-limited Combined Statistical Areas standard.\\n\\n Finally, the NCUA’s definition does not readily create\\ngeneral, widely dispersed regions. Cf. First Nat’l Bank III, 522\\nU.S. at 502 (indicating that community credit unions may not\\nbe “composed of members from an unlimited number of\\nunrelated geographical units”). Combined Statistical Areas are\\ngeographical units well-accepted within the government. See\\n81 Fed. Reg. at 88,414. Because they essentially are regional\\nhubs, the Combined Statistical Areas concentrate around\\ncentral locations. The OMB carved out the areas so that their\\nconstituent parts share commuting ties and they reflect\\n“broader social and economic interactions.” OFFICE OF MGMT.\\n& BUDGET, SUPRA, at app. 2-3. The NCUA rationally believed\\nthat such “real-world interconnections” would qualify as the\\ntype of mutual bonds suggested by the term “local\\ncommunity.” Credit Union Nat’l Ass’n Amicus Br. 9. Thus,\\nthe agency reasonably determined that Combined Statistical\\nAreas “simply unif[y], as a single community,” already\\nconnected neighboring regions. See 81 Fed. Reg. at 88,415.\\n\\n The Association raises a potpourri of objections to the\\nNCUA’s decision-making. See Am. Bankers Ass’n Br. 33-48.\\nVirtually all of its gripes are forfeited because it failed first to\\nraise them to the agency, see, e.g., Koretoff v. Vilsack, 707 F.3d\\n394, 397-98 (D.C. Cir. 2013) (per curiam), or lack merit\\nbecause they involve outdated “local community” definitions,\\nwhich either did not allow for the qualification of certain\\ngeographical areas as local communities or lacked a population\\ncap of 2.5 million people.\\n\\f 24\\n But one of the complaints is “deserving of sustained\\nconsideration.” Ortiz v. United States, 138 S. Ct. 2165, 2173\\n(2018). The Association contends that some large Combined\\nStatistical Areas are so sprawling that the NCUA’s definition,\\nwhich treats a 2.5-million-person portion of them as a local\\ncommunity, must be unreasonable. Recall that a Combined\\nStatistical Area is a regional hub with more than one urban\\ncenter. Under the OMB’s technical specifications, each center\\nneed not be connected to every other by commuting ties; rather,\\neach need only have ties to one other center within the hub. As\\nthe Association colorfully puts it, the OMB may designate as a\\nCombined Statistical Area a mere “daisy chain” of urban\\ncenters “that are linked to their neighbors but have nothing to\\nwith those at the other end of the chain.” See Am. Bankers\\nAss’n Br. 36 (citation omitted). Theoretically, the Association\\ncontinues, certain 2.5-million-person communities might bring\\ntogether parts of different urban centers with no connection\\nwith one another. The Association also suggests that the rule\\nmight permit local communities comprising non-contiguous\\nportions of a Combined Statistical Area. See id. at 39.\\n\\n We understand the Association’s argument to be attacking\\nthe Combined Statistical Area definition as unreasonable. To\\nthe Association, the NCUA failed sufficiently to consider the\\npotential for the rule to create unreasonable results. One\\nhypothetical application disturbed the District Court here: the\\nprospect that, within the Combined Statistical Area including\\nthe District of Columbia and counties from three states\\n(Maryland, Pennsylvania, and Virginia), one could create a\\nlocal community bringing together Doylesburg, Pennsylvania,\\nand Partlow, Virginia, towns located 200 miles apart. See Am.\\nBankers Ass’n, 306 F. Supp. 3d at 59-60.\\n\\n We might well agree with the District Court that the\\napproval of such a geographical area would contravene the Act.\\n\\f 25\\nBut even so, the Association would need much more to mount\\nits facial pre-enforcement challenge in this case. As the\\nSupreme Court repeatedly has held, “the fact that petitioner can\\npoint to a hypothetical case in which the rule might lead to an\\narbitrary result does not render the rule” facially invalid. Am.\\nHosp. Ass’n v. NLRB, 499 U.S. 606, 619 (1991); see also EPA\\nv. EME Homer City Generation, L.P. (EME Homer), 572 U.S.\\n489, 524 (2014) (“The possibility that the rule, in uncommon\\nparticular applications, might exceed [the agency]’s statutory\\nauthority does not warrant judicial condemnation of the rule in\\nits entirety.”); INS v. Nat’l Ctr. for Immigrants’ Rights, Inc.,\\n502 U.S. 183, 188 (1991) (“That the regulation may be invalid\\nas applied in s[ome] cases . . . does not mean that the regulation\\nis facially invalid because it is without statutory authority.”);\\ncf. Barnhart v. Thomas, 540 U.S. 20, 29 (2003) (“Virtually\\nevery legal (or other) rule has imperfect applications in\\nparticular circumstances.”).\\n\\n Here, the Association’s complaint and the District Court’s\\naccompanying worry strike us as too conjectural. The NCUA\\nmust assess the “economic advisability of establishing” the\\nproposed credit union before approving it, 12 U.S.C. § 1754,\\nand as part of the assessment, the organizers must propose a\\n“realistic” business plan showing how the institution and its\\nbranches would serve all members in the local community, see\\n12 C.F.R. pt. 701, app. B, ch. 1 § IV.D. The Association has\\nfailed to demonstrate the plausibility of a local community that\\nis defined like the hypothetical narrow, multi-state strip and\\naccompanies a realistic business plan. And if the agency were\\nto receive and approve such an application, a petitioner can\\nmake an as-applied challenge. See, e.g., EME Homer, 572 U.S.\\nat 523-24; Buongiorno, 912 F.2d at 510. The Association has\\nsucceeded in such challenges in the past. See, e.g., Am.\\nBankers Ass’n v. NCUA, No. 1:05-CV-2247, 2008 WL\\n2857678, at *1 (M.D. Pa. July 21, 2008); Am. Bankers Ass’n v.\\n\\f 26\\nNCUA, 347 F. Supp. 2d 1061, 1074 (D. Utah 2004). Thus, we\\nreject the facial attack on the amended “local community”\\ndefinition involving Combined Statistical Areas.\\n\\n B.\\n\\n We turn to the next rule change. The District Court upheld\\nthe eliminated core requirement for a local community based\\non a Core Based Statistical Area. The court acknowledged that\\ndefining a local community without its urban core “does not\\nalter the . . . common bond” shared by the members in the\\nremainder. Am. Bankers Ass’n, 306 F. Supp. 3d at 62-63.\\nMeanwhile, the court accepted the agency’s response to the\\nclaim that the new definition encouraged redlining – a broad\\ncategory of lending practices with negative impact on “areas\\nwhere low-income and minority populations are concentrated.”\\nId. at 65 (quoting 81 Fed. Reg. at 88,413).\\n\\n Like the District Court, we hold that the eliminated core\\nrequirement is consistent with the Act’s text and purposes.\\nStill, we see merit in the Association’s redlining argument and\\nthus hold the definitional change to be arbitrary and capricious.\\n\\n 1.\\n\\n We will sustain the eliminated core requirement if it\\nreflects a reasonable interpretation of “local community” that\\nis rationally related to the dual purposes of promoting credit-\\nunion growth and ensuring some cohesion among members. It\\ndoes. Omission of the urban core from proposed geographical\\narea will permit community credit unions to reach new\\nmembers in the suburban parts of the Core Based Statistical\\nArea and thus to maintain a healthy membership. Because the\\nsuburbs under the OMB’s definition have substantial\\ncommuting ties to the urban cluster, they all will be “within a\\n\\f 27\\nfeasible commuting radius” and thus “share[] at least some\\ngeographic ties.” Am. Bankers Ass’n, 306 F. Supp. 3d at 64.\\nThose bonds do not disappear if the local community lacks the\\ncore. The Association seeks greater “assurance[s]” of a\\nmeaningful bond among members, Am. Bankers Ass’n Br. 66-\\n67; see also Am. Bankers Ass’n Reply 11, but we do not\\nreplace the agency’s policy judgment with ours.\\n\\n We also do not believe that the eliminated core\\nrequirement would create sprawling, boundless geographical\\nregions. No one disputes that, as a general matter, a\\nmembership area comprising an intact Core Based Statistical\\nArea will satisfy any definition of “local community.” If the\\nlocal community with the core poses no problem, we fail to see\\nhow a local community without one would. And even as to\\nCore Based Statistical Areas that do not qualify as local\\ncommunities (because they have populations of more than 2.5\\nmillion), the geographical ties ensure that the proposed\\nmembership area will still be contained within the boundaries\\nof a single, well-recognized metropolitan region. A single\\nregion is not what concerned the Supreme Court in First\\nNational Bank III. See 522 U.S. at 502.\\n\\n The Association objects to the expansive hypothetical\\nmembership fields, highlighting two Core Based Statistical\\nAreas whose populations exceed the NCUA’s cap of 2.5\\nmillion. The Association asserts that the rule change “makes it\\nmuch easier to unite far-distant edges” of those sprawling areas\\nin a single membership area. Am. Bankers Ass’n Reply 9. Fair\\npoint. But economic realities do not make it plausible that\\norganizers would propose such a local community or that the\\nNCUA would approve it. Like the Combined Statistical Area\\ndefinition, the eliminated core requirement does not become\\nfacially infirm because of farfetched hypotheticals. To the\\nextent they occur in the future, troublesome rule applications\\n\\f 28\\nmight be subject to as-applied challenges. See Am. Bankers\\nAss’n, 271 F.3d at 267.\\n\\n 2\\n\\n Despite the eliminated core requirement’s consistency\\nwith the Act, we cannot sustain the definitional change because\\nthe NCUA has not adequately explained it.\\n\\n The Association contends that the rule change “effectively\\nallows credit unions to engage in ‘redlining’ by denying service\\nto urban areas with large numbers of minority and lower-\\nincome residents.” Am. Bankers Ass’n Br. 65. The NCUA\\nattempted to address this concern in the preamble. At first\\nblush, the agency’s statements appear persuasive. Still, the\\nAssociation persuasively argues that the response fails to\\n“consider an important aspect” of the redlining issue or is\\notherwise “so implausible” as to be unreasonable. State Farm,\\n463 U.S. at 43. Thus, the eliminated core requirement is\\narbitrary and capricious.\\n\\n During the notice-and-comment proceedings, the\\nAssociation warned against redlining and objected that\\ncommunity credit unions could now “serv[e] wealthier\\nsuburban counties and exclud[e] markets containing low-\\nincome and minority communities that reside in the core area.”\\nLetter from James Chessen, Exec. Vice President & Chief\\nEconomist, Am. Bankers Ass’n, to Gerard S. Poliquin, Sec’y\\nof the Board, Nat’l Credit Union Admin. (Feb 5, 2016), Am.\\nBankers Ass’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C.\\nfiled Aug. 23, 2017), ECF No. 26-1 at 228. Fairly read, the\\nAssociation’s objection is not to traditional redlining, which\\nencompasses the refusal to make loans in low-income or\\nminority neighborhoods within a service area, see Baradaran,\\nsupra, at 494; see also Cassity, supra, at 348, 355. Federal\\n\\f 29\\ncredit unions “cannot ‘redline’” in the traditional sense because\\n“everyone in a credit union’s ‘community’ is a member who is\\neligible to take advantage of credit union services” and credit\\nunions “by definition cannot reach beyond their member\\ncommunities to offer credit to the general public.” Cassity,\\nsupra, at 355. But a community credit union can engage in\\nmore unconventional redlining practices: “gerrymander[ing] to\\ncreate its own community of exclusively higher-income\\nmembers.” Id. at 359. We think it evident that the Association\\nwas focusing on such gerrymandering.\\n\\n In response, the NCUA acknowledged that it originally\\nkept the core requirement as a benefit to “low-income and\\nunderserved populations.” 81 Fed. Reg. at 88,413. Some\\ncommenters criticized the core requirement as failing to\\nachieve that goal; it caused community credit unions “to\\nsacrifice service to other areas” within the Core Based\\nStatistical Area, id., and such service arguably could benefit\\npoor and minority customers residing outside the core, see id.\\nat 88,414 (remarking on the importance of credit unions\\n“providing financial services to low income and underserved\\npopulations without regard to where they are located within a\\ncommunity, i.e., beyond its ‘core area’”).\\n\\n Still, the agency dismissed the redlining concern on other\\ngrounds, pointing to its “supervisory process to assess [credit-\\nunion] management’s efforts to offer service to the entire\\ncommunity [the credit union] seeks to serve.” Id. The NCUA\\nfocused on two aspects of its process. First, the agency touted\\nits “annual evaluation,” which “encompasses [the credit\\nunion]’s implementation of its business and marketing plans.”\\nId. Citing its “[e]xperience” as support, the agency identified\\nthe evaluation as a “more effective means” than a core\\nrequirement to “ensur[e] that the low-income and underserved\\npopulations are fairly served.” Id. Indeed, prior evaluations\\n\\f 30\\nconfirm that the agency has had “success in providing financial\\nservices to low-income and underserved populations without\\nregard to where they are located within” the membership area.\\nId. at 88,414. Second, the NCUA noted its “mandate to\\nconsider member complaints alleging discriminatory practices\\naffecting low-income and underserved populations, such as\\nredlining, and to respond as necessary when such practices are\\nshown to exist.” Id.\\n\\n Both aspects of the NCUA’s supervisory process fail to\\naddress the redlining issue raised by the Association. The\\nannual evaluation process might be an adequate response to\\ntraditional redlining, because it might ensure that community\\ncredit unions adequately serve poor and minority residents\\nliving in their local communities. But we do not see how it\\nfixes gerrymandering or the potential discriminatory economic\\nimpact on urban residents. See State Farm, 463 U.S. at 43\\n(requiring the agency to consider relevant factors). The annual\\nevaluation process necessarily does not come into effect until\\nthe NCUA already has approved the charter, business plan, and\\nproposed local community. See Iowa Bankers Ass’n Amicus\\nBr. 12. And nothing in the record suggests that business plans\\nmay focus on residents residing outside the finalized\\nmembership area; in fact, the law forbids federal credit unions\\nfrom serving those residents.\\n\\n The complaint process also does not work in the\\ngerrymandering context. As the preamble points out,\\ncomplaints are raised by the membership, which would not\\ninclude the affected urban residents because of the rule change.\\nIt seems quite implausible, absent some contrary evidence the\\nagency failed to detail, that members will file grievances based\\non gerrymandering harms suffered by residents outside the\\ncoverage area. See State Farm, 463 U.S. at 43 (rejecting\\nimplausible explanations).\\n\\f 31\\n\\n The NCUA attempts to defend the rule change by offering\\na buffet of other potential rationales in its briefing and at oral\\nargument. They all fail because the agency did not adopt them\\nwhen promulgating the rule change. See State Farm, 463 U.S.\\nat 43 (noting that courts may not “supply a reasoned basis for\\nthe agency’s action that the agency itself has not given”\\n(quoting SEC v. Chenery Corp., 332 U.S. 194, 196 (1947))).\\nWe pause, though, to caution the NCUA about two proffered\\nreasons. At oral argument, the government counsel suggested\\nthat the agency may reject proposed local communities if it\\nsuspects they discriminate against residents in the urban core.\\nSee Oral Arg. Recording 5:28-49, 13:43-14:09. But current\\nreviewing guidelines do not indicate that the agency looks for\\nsuch discrimination. See 12 C.F.R. pt. 701, app. B, ch. 1\\n§ VII.A. And the NCUA made the opposite representations to\\nthe District Court. See Transcript of Motion Hearing at 48:23-\\n49:7 (“[District Court]: . . . If a credit union comes to the\\nagency and says I want to serve X area, either in a rural district\\nor a combined statistical area, and they meet the definition, the\\nagency has no authority to reject that application, as long as the\\ncredit union can demonstrate that they can serve the\\narea? [NCUA]: . . . I think that’s probably right, your\\nHonor.”), Am. Bankers Ass’n v. NCUA, No. 1:16-cv-02394-\\nDLF (D.D.C. filed Mar. 27, 2018), ECF No. 33. The\\ngovernment counsel also suggested that community credit\\nunions already cover the vast majority of urban cores. See Oral\\nArg. Recording 8:10-22, 12:30-41; see also id. at 7:39-8:01; cf.\\nNCUA Reply 27-28. Perhaps, but the current record does not\\nsupport the assertion.\\n\\f 32\\n C.\\n\\n Finally, the District Court rejected the increased\\npopulation cap for rural districts. The court worried that a rural\\ndistrict satisfying the new, higher limit could be too large or\\ncould contain “numerous urban centers.” Am. Bankers Ass’n,\\n306 F. Supp. 3d at 69. Because the rule qualified certain\\ndistricts “no matter [their] geographic size or the number or\\nsize of metropolitan areas falling within [their] proposed\\nboundaries,” the court held that the NCUA’s new definition is\\nunreasonable. Id. at 69-70. But in our view, the new “rural\\ndistrict” definition is reasonable.\\n\\n As suggested above, we assume that the NCUA must\\nadopt definitions consistent with the statutory terms as\\nunderstood in 1934, not today. The terms “rural” and “district”\\ndo not connote specific population or geographical constraints.\\nSee First Nat’l Bank III, 522 U.S. at 492 (noting that markets\\n“need not be small”). A “district” means a “portion of a state,\\ncounty, country, town, or city . . . made for administrative,\\nelectoral, or other purposes.” District, WEBSTER’S NEW\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 649\\n(2d ed. 1934) (emphasis added). The district may be “of\\nundefined extent.” Id. Meanwhile, “rural” indicates a\\n“country” or “agricultur[al]” lifestyle, as opposed to that of a\\n“city or town.” Rural, WEBSTER’S NEW INTERNATIONAL\\nDICTIONARY OF THE ENGLISH LANGUAGE 1861 (2d ed. 1934).\\n“Rural” lands generally “resembl[e]” the “country[side].” Id.\\n\\n Nothing about the 1-million-person cap prevents the rural\\ndistrict from resembling the countryside. And one of the\\nunchallenged restrictions helps provide a rural character to\\nsuch districts. Either most residents in the proposed district\\nlive on rural land, or the population density is 100 or fewer\\npeople per square mile. See 81 Fed. Reg. at 88,440.\\n\\f 33\\nAccordingly, even if most residents live on urban areas in the\\nrural district, those areas will be surrounded by rural land.\\nThat’s because 100 people per square mile – or one person for\\nevery six or so acres – is a rural population density.\\n\\n As for size, the contemporaneous definition of “district”\\nreassures us that rural districts may cross state lines. And a\\nsecond unchallenged restriction assures that the 1-million-\\nperson cap will not support gigantic or straggly rural districts.\\nAs the rule explains, the boundaries of a community credit\\nunion’s district may not “exceed the outer boundaries of the\\nstates” immediately surrounding the state where the proposed\\ncredit union would have its headquarters. 81 Fed. Reg. at\\n88,440. Thus, even though the population density of the entire\\nUnited States is less than 100 people per square mile, see\\nNCUA Br. 56 n.44; Oral Arg. Recording 39:34-38, the\\ngeographical limitation forces a rural district to be much\\nsmaller. We are confident that such districts will not be so\\nenormous as to amount to federations of “unrelated geographic\\nunits.” See First Nat’l Bank III, 522 U.S. at 502.\\n\\n In arguing that the amended “rural district” definition is\\nunreasonable, the Association relies heavily on the interpretive\\nanalysis performed by the District Court. See Am. Bankers\\nAss’n, 306 F. Supp. 3d at 67-69. The court’s core contention\\nis that we must construe “rural” and “district” together as a\\nspecialized phrase that meant, in 1934, an “area[] much smaller\\nthan a state.” Id. at 68. For support, the District Court not only\\nconsulted two 1934-era dictionaries, results of a Westlaw\\nsearch of contemporaneous opinions, and a database containing\\nhistorical uses of the phrase, but also invoked the noscitur a\\nsociis canon. See id. at 67-69.\\n\\n The proffered definitional evidence is pretty thin. The\\n1934-era dictionaries described what the specialized phrase\\n\\f 34\\n“rural district” meant in Great Britain, not in the much larger\\nand more expansive United States, which by the 1930s\\nencompassed forty-eight continental states. See, e.g., Rural\\nDistrict, WEBSTER’S NEW INTERNATIONAL DICTIONARY OF THE\\nENGLISH LANGUAGE 1861 (1934) (“in England, a subdivision\\nof an administrative county embracing . . . county parishes”).\\n\\n The proffered Westlaw search results included scores of\\nfederal and state judicial and administrative opinions. See\\nPlaintiff’s Supplemental Memorandum app. 6, Am. Bankers\\nAss’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C. filed Mar.\\n16, 2018), ECF No. 32-6. But we find little of use in those\\ndocuments. Many of those opinions do not discuss size or\\npopulation. See, e.g., Nicolai v. Wis. Power & Light Co., 269\\nN.W. 281, 282-83 (Wis. 1936). Those that do involve specific\\nrural districts whose sizes were between a town and a state.\\nSee, e.g., Sarther Grocery Co. v Comm’r, 22 B.T.A. 1273, 1274\\n(1931). But none declares that rural districts by definition are\\nrestricted to any particular size or population.\\n\\n The District Court next turned to the Corpus of Historical\\nAmerican Usage, a free and public online database. See Am.\\nBankers Ass’n, 306 F. Supp. 3d at 68 & n.5. The database notes\\n197 mentions of the phrase in the first half of the twentieth\\ncentury and only 37 in the second. See Search Results for\\nRURAL DISTRICTS and RURAL DISTRICT, CORPUS OF\\nHISTORICAL AMERICAN ENGLISH, https://www.english-\\ncorpora.org/coha/ (follow “List” hyperlink; then search\\nmatching strings field for “RURAL DISTRICT”). The\\nperceived drop-off led the District Court to determine that,\\n“even if rural district does not carry meaning distinct from its\\nindividual words today, it did in 1934.” Am. Bankers Ass’n,\\n306 F. Supp. 3d at 68.\\n\\f 35\\n Although federal courts may use “crude[]” searches on\\ndatabases to learn of ambiguities in a statutory term,\\nMuscarello v. United States, 524 U.S. 125, 129-30 (1998), the\\nsearch here did not suffice to show that the agency’s definition\\nwas unreasonable. Much more is required to cabin the\\nagency’s discretion. A search of a commercial database,\\nProQuest, reveals that the phrase appeared at least 500 times in\\nthe second half of the century. We are not confident that the\\nproffered evidence establishes a particular historical trend.\\n\\n The District Court lastly invoked noscitur a sociis. See\\nAm. Bankers Ass’n, 306 F. Supp. 3d at 68-69. The canon is\\ninapposite. Recall that the term “rural district” was listed with\\n“community,” not “local community,” in 1934. At the time,\\n“community” could refer to “[s]ociety at large” or a\\n“commonwealth or state.” Community, WEBSTER’S NEW\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 452\\n(1934). Indeed, we said that the word is “too indefinite to be\\nused for purposes of exact measurement in terms of acres or\\nsquare miles.” Lukens Steel Co. v. Perkins, 107 F.2d 627, 631\\n(D.C. Cir. 1939), rev’d sub nom., Perkins v. Lukens Steel Co.,\\n310 U.S. 113 (1940). Thus, we do not see size or population\\nas a connection between the terms.\\n\\n The increased population cap is consistent with not only\\nthe Act’s text but also its purposes. For instance, as the\\npreamble noted, the expanded definition allows community\\ncredit unions in rural districts to reach more “persons of modest\\nmeans who may reside” in those areas. 81 Fed. Reg. at 88,416;\\nsee also § 2(1), (2), 112 Stat. at 913.\\n\\n And the change is neither arbitrary nor capricious. The\\nnew rule explains that it provides a “level of operating\\nefficiencies and scale” making rural districts attractive options\\nfor prospective credit unions. 81 Fed. Reg. at 88,416. “[A]\\n\\f 36\\nsufficiently large population base is essential to enable” them\\n“to offer financial services economically.” Id. at 88,417. The\\nNCUA also chose the 1-million figure based on prior\\nexperience. The agency noted that it had approved of eight\\ndistricts with populations of more than 250,000, and that one\\nof them already had exceeded 1 million. See id. “Having set a\\n1 million precedent in one state,” the agency felt justified in\\nhaving a “fixed 1 million population cap for the other 49\\nstates.” Id. (emphasis added). We cannot say this policy\\nchoice lacks rational explanation.\\n\\n The Association says the NCUA failed to consider prior\\nagency decisions. See Am. Bankers Ass’n Br. 61. But we see\\nno discrepancy and thus summarily reject the objection. The\\nAssociation also turns to troubling hypothetical examples of\\nrural districts with unruly shapes and those with dense urban\\nareas such as Denver, Colorado. See NCUA Br. 55-61; Oral\\nArg. Recording 37:55-38:05. Again, such implausible outliers\\ndo not impugn the rule’s general reasonableness.\\n\\n VI.\\n\\n We now consider the remedy. To sum up, we hold that\\ndefining certain Combined Statistical Areas or portions of them\\nas local communities and raising the population cap for rural\\ndistricts are consistent with the Act and reasonably explained.\\nThus, we sustain both aspects of the challenged rule. We also\\nleave undisturbed the portion of the District Court’s opinion\\nthat the parties do not contest on appeal.\\n\\n But we deem the eliminated core requirement to be\\narbitrary and capricious. When a rule is contrary to law, the\\n“ordinary practice is to vacate” it. United Steel v. Mine Safety\\n& Health Admin., 925 F.3d 1279, 1287 (D.C. Cir. 2019); see\\nalso 5 U.S.C. § 706(2) (noting that the “reviewing court\\n\\f 37\\nshall . . . set aside” unlawful agency action). But in “rare\\ncases,” we will opt instead to remand without vacating the rule,\\nso that the agency can “correct its errors.” United Steel, 925\\nF.3d at 1287; see also 5 U.S.C. § 702 (“Nothing\\nherein . . . affects . . . the power or duty of the court to . . . deny\\nrelief on any . . . appropriate . . . equitable ground . . . .”); Oglala\\nSioux Tribe v. U.S. Nuclear Regulatory Comm’n, 896 F.3d 520,\\n536 (D.C. Cir. 2018). In considering whether to adopt the latter\\nequitable remedy, we balance “(1) the seriousness of the\\ndeficiencies of the action, that is, how likely it is the agency\\nwill be able to justify its decision on remand; and (2) the\\ndisruptive consequences of vacatur.” United Steel, 925 F.3d at\\n1287 (citation omitted). A strong showing of one factor may\\nobviate the need to find a similar showing of the other. See,\\ne.g., Fox Television Stations, Inc. v. FCC, 280 F.3d 1027, 1049\\n(D.C. Cir. 2002) (holding that, because the agency likely could\\njustify its action on remand, the potential for disruption was\\n“only barely relevant”); Allied-Signal, Inc. v. U.S. Nuclear\\nRegulatory Comm’n, 988 F.2d 146, 152 (D.C. Cir. 1993)\\n(determining that, because vacatur would give regulated parties\\na “peculiar windfall,” the meager chance of justifying the\\naction was given “little weight” in the remedial analysis).\\n\\n The NCUA has not requested remand without vacatur in\\nthis case. But because we have a “duty” to ensure the propriety\\nof the APA remedy, 5 U.S.C. § 702, we hold that we have the\\ndiscretion to raise the issue sua sponte, cf. Igonia v. Califano,\\n568 F.2d 1383, 1387 (D.C. Cir. 1977); Gaddis v. Dixie Realty\\nCo., 420 F.2d 245, 247 (D.C. Cir. 1969) (per curiam). Remand\\nwithout vacatur is appropriate here.\\n\\n We conclude that the NCUA might be able to offer a\\nsatisfactory reason on remand. And as for disruptive effect, we\\nperceive a substantial likelihood that vacating the rule could\\nmake it more difficult for some poor and minority suburban\\n\\f 38\\nresidents to receive adequate financial services. Even\\ntemporarily depriving these members of the opportunity is\\ninequitable, because it would “set back” the Act’s objective of\\noffering financial services to people of small means. See Nat’l\\nRes. Def. Council v. EPA, 489 F.3d 1364, 1374 (D.C. Cir.\\n2007); see also Advocates for Highway & Auto Safety v. Fed.\\nMotor Carrier Safety Admin., 429 F.3d 1136, 1152 (D.C. Cir.\\n2005) (declining to vacate rule because, in the interim, it “may\\ndo some good, if it does anything at all”).\\n\\n Given the potential for sufficient justification and the\\nsubstantial likelihood of disruptive effect, we have a rare case\\nin which vacatur is inappropriate. See United Steel, 925 F.3d\\nat 1287. Thus, we remand without vacating the relevant part\\nof the 2016 rule. We trust that the agency will act\\nexpediently. See EME Homer City Generation, L.P. v. EPA,\\n795 F.3d 118, 132 (D.C. Cir. 2015).\\n\\n * * *\\n\\n In short, we reverse the challenged portions of the District\\nCourt’s summary judgment order. With respect to the\\nqualification of certain Combined Statistical Areas as local\\ncommunities and the increased population cap for rural\\ndistricts, we direct the District Court to issue summary\\njudgment in favor of the NCUA. As to the elimination of the\\nurban-core requirement for local communities based on Core\\nBased Statistical Areas, we direct the District Court to issue\\nsummary judgment in favor of the Association and to remand,\\nwithout vacating, the relevant portion of the 2016 rule for\\nfurther explanation.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4183,"cells":{"id":{"kind":"number","value":4655416,"string":"4,655,416"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"American Bankers Association v. National Credit Union Administration"},"decision_date":{"kind":"string","value":"2019-08-20"},"docket_number":{"kind":"string","value":"18-5154"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before: HENDERSON, PILLARD, and WILKINS, Circuit Judges.\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge WILKINS\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 16, 2019 Decided August 20, 2019\\n\\n No. 18-5154\\n\\n AMERICAN BANKERS ASSOCIATION,\\n APPELLEE\\n\\n v.\\n\\n NATIONAL CREDIT UNION ADMINISTRATION,\\n APPELLANT\\n\\n\\n Consolidated with 18-5181\\n\\n\\n Appeals from the United States District Court\\n for the District of Columbia\\n (No. 1:16-cv-02394)\\n\\n\\n Daniel Aguilar, Attorney, U.S. Department of Justice,\\nargued the cause for Appellant-Cross-Appellee. With him on\\nthe briefs was Mark B. Stern, Attorney.\\n\\n Allison Jones Rushing was on the brief for amici curiae\\nCredit Union National Association, et al. in support of\\nAppellant-Cross-Appellee. Nicholas G. Gamse entered an\\nappearance.\\n\\f 2\\n Steven D. Gordon was on the brief for amici curiae State\\nBankers Associations in support Appellee-Cross-Appellant\\nAmerican Bankers Association.\\n\\n Robert A. Long Jr. argued the cause for Appellee-Cross-\\nAppellant. With him on the briefs were Andrew J. Soukup,\\nPhilip Levitz, and Lauren Moxley.\\n\\n Before: HENDERSON, PILLARD, and WILKINS, Circuit\\nJudges.\\n\\n Opinion for the Court filed by Circuit Judge WILKINS.\\n\\n WILKINS, Circuit Judge: Longstanding principles of\\nadministrative law teach us to give federal agencies breathing\\nroom when they make policy and “resolv[e] the struggle\\nbetween competing views of the public interest.” Chevron,\\nU.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 866\\n(1984). And because many policy decisions merge with legal\\nones, Chevron requires us frequently to sustain agency\\ninterpretations of certain federal statutes. Congress often\\nexpects agencies, with their political accountability, “bod[ies]\\nof experience[,] and informed judgment,” to make sound\\ninterpretive choices “with the force of law.” United States v.\\nMead Corp., 533 U.S. 218, 227, 229 (2001) (citation omitted).\\n\\n Congress expressly tasked the National Credit Union\\nAdministration (NCUA) with making such choices in defining\\nthe reach of federal credit unions. Since the Great Depression,\\nCongress has maintained a “system of federal credit unions\\nthat . . . provide credit at reasonable rates” and banking\\nservices to “people of ‘small means.’” First Nat’l Bank & Tr.\\nCo. v. NCUA (First Nat’l Bank I), 988 F.2d 1272, 1274 (D.C.\\nCir. 1993) (citation omitted), aff’d, 522 U.S. 479 (1998).\\nAlthough a private bank may solicit and welcome customers\\n\\f 3\\nfrom anywhere, Congress has limited whom these federal\\nfinancial institutions may serve. For instance, certain\\ninstitutions called “community credit unions” may cover\\nindividuals and entities only within a preapproved\\ngeographical area. The credit union will not receive a federal\\ncharter (and thus cannot start operations) unless it first proffers\\na geographical coverage area and the NCUA accepts the\\nproposal. Congress explicitly assigns the agency the task of\\ncreating vetting standards.\\n\\n Exercising its expressly delegated power, the NCUA has\\npromulgated a final rule that makes it easier for community\\ncredit unions to expand their geographical coverage and thus to\\nreach more potential members. Representing competitors to\\nthe credit unions, the American Bankers Association\\n(Association) has challenged the NCUA’s new rule as neither\\n“in accordance with law” nor within “statutory jurisdiction.” 5\\nU.S.C. § 706(2)(A), (C). The District Court vacated significant\\nportions of the rule, deeming them to be based on unreasonable\\nagency interpretations of the Federal Credit Union Act (Act),\\nPub. L. No. 73-467, 48 Stat. 1216 (1934) (codified as amended\\nat 12 U.S.C. §§ 1751 to 1795k). See Am. Bankers Ass’n v.\\nNCUA, 306 F. Supp. 3d 44, 61, 69-70 (D.D.C. 2018).\\n\\n We appreciate the District Court’s conclusions, made after\\na thoughtful analysis of the Act. But we ultimately disagree\\nwith many of them. In this facial challenge, we review the rule\\nnot as armchair bankers or geographers, but rather as lay judges\\ncognizant that Congress expressly delegated certain policy\\nchoices to the NCUA. After considering the Act’s text,\\npurpose, and legislative history, we hold the agency’s policy\\nchoices “entirely appropriate” for the most part. Chevron, 467\\nU.S. at 865. We therefore sustain the bulk of the rule. Still, we\\ndo not rubber-stamp this regulation. We remand, without\\nvacating, one portion for further consideration of the\\n\\f 4\\ndiscriminatory impact it might have on poor and minority\\nurban residents.\\n\\n I.\\n\\n A.\\n\\n The nation’s credit unions started in the early twentieth\\ncentury “as a populist mechanism designed to empower\\nfarmers against bad loans.” Mehrsa Baradaran, How the Poor\\nGot Cut Out of Banking, 62 EMORY L.J. 483, 500 (2013).\\nWalloped by crop failures and the Great Depression, farmers\\nseeking credit became not only increasingly suspicious of\\ntraditional bankers, who “disregard[ed]” poor individuals and\\nstayed in the big cities, but also fearful of loan sharks, “who\\nwould extract ‘up to a thousand percent’ in interest rates.” Id.\\nat 500-01 (quoting 80 CONG. REC. 6752 (1936) (statement of\\nRep. Lundeen)). The farmers thus began to build their own\\ncredit networks.\\n\\n In a national grassroots campaign, farmers created\\nlocalized, non-profit “credit groups” collecting funds from and\\nloaning small sums to one another at low interest rates. See id.\\nat 501-02. The success of any such self-help institution\\n“hinge[s] on the interpersonal dynamics of its members:\\nLenders must be able to evaluate the ability and willingness of\\npotential borrowers to pay back their loans and borrowers must\\nfeel obligated to pay back those loans.” Wendy Cassity, Note,\\nThe Case for a Credit Union Community Reinvestment Act, 100\\nCOLUM. L. REV. 331, 337 (2000); see also First Nat’l Bank &\\nTr. Co. v. NCUA (First Nat’l Bank II), 90 F.3d 525, 526 (D.C.\\nCir. 1996), aff’d, 522 U.S. 479 (1998).\\n\\n By 1934, individuals had organized about 3,000 local\\ncredit unions, with about 750,000 members. See 80 CONG.\\n\\f 5\\nREC. at 6753. Recognizing the success of credit unions at the\\nstate level, Congress created a federal system that year by\\npassing the Act. Legislators worried that “usurious money\\nlending . . . obviously destroy[ed] vast totals of buying power\\n[once held by] . . . the average worker.” H.R. REP. NO. 73-\\n2021, at 1-2 (1934); see also S. REP. NO. 73-555, at 1 (1934).\\nCongress touted the Act’s ability to “make more available to\\npeople of small means credit for provident purposes.” H.R.\\nREP. NO. 73-2021, at 1; see also S. REP. NO. 73-555, at 1.\\n\\n Credit unions multiplied over the ensuing decades. By\\n1970, Congress created an independent agency to supervise\\nfederal credit unions: the NCUA. See Pub. L. No. 91-468, 84\\nStat. 994 (1970) (codified as amended in scattered sections of\\n12 U.S.C.); see also Swan v. Clinton, 100 F.3d 973, 974 (D.C.\\nCir. 1996) (noting that Congress “entrusted” the agency with\\n“the responsibility of overseeing” federal credit unions).\\nLegislators thought that the agency would be “more responsive\\nto the needs of credit unions” and would “provide more flexible\\nand innovative regulation” than prior government agencies,\\nwhich did not have federal credit unions as their sole focus. S.\\nREP. NO. 91-518, at 3 (1969).\\n\\n The NCUA faced its first major crisis at the end of the\\n1970s. After years of economic decline in several industrial\\nsectors, federal credit unions tied to those business sectors\\nbegan to suffer. The resulting liquidation of numerous credit\\nunions “threaten[ed] ‘the safety and soundness of the federal\\ncredit union system.’” Cassity, supra, at 338-39 (footnote\\nomitted). Reacting to the emergency, the NCUA in 1981\\npromulgated a groundbreaking rule that loosened a major size\\nlimitation on certain federal credit unions. Almost\\nimmediately, those financial institutions grew in membership.\\n\\f 6\\n Meanwhile, credit unions became “caught up in the\\nbroader changes in banking and faced internal as well as\\nexternal pressure to compete with [private] banks and seek\\nhigher profits.” Baradaran, supra, at 505. Unlike credit\\nunions, private, for-profit banks were “owned by equity holders\\nwho may not necessarily be customers (depositors or\\nborrowers),” and they did “not have similar membership and\\ncommercial lending restrictions” as credit unions. DARRYL E.\\nGETTER, CONG. RESEARCH SERV., IF11048, INTRODUCTION TO\\nBANK REGULATION: CREDIT UNIONS AND COMMUNITY BANKS:\\nA COMPARISON 1 (2018). To remain viable, credit unions\\n“started to focus on attracting more customers and expanding\\nthe industry.” Baradaran, supra, at 505. As part of that\\nstrategy, many consolidated through mergers. And private\\nbanks soon treated credit unions as serious competitors,\\nseeking to curb their growth. See NCUA v. First Nat’l Bank &\\nTr. Co. (First Nat’l Bank III), 522 U.S. 479, 485 (1998); First\\nNat’l Bank I, 988 F.2d at 1276.\\n\\n In 1998, the banking industry successfully challenged as\\ncontrary to the Act the 1981 rule that had eased size limitations\\nfor certain federal credit unions. See First Nat’l Bank III, 522\\nU.S. at 503. Congress swiftly responded. In less than six\\nmonths, legislators amended the Act, superseding the holding\\nin First National Bank III, loosening size limitations on certain\\nfederal credit unions, and adding other reforms. See Credit\\nUnion Membership Access Act, Pub. L. No. 105-219, 112 Stat.\\n913 (1998) (codified as amended in scattered sections of 12\\nU.S.C.). Partly because of the 1998 amendments and related\\nNCUA regulations, credit unions continued to merge and grow\\nin membership. Now, more than 61 million customers perform\\ntheir banking services at about 3,400 federal credit unions. See\\n2018 NAT’L CREDIT UNION ADMIN. ANN. REP. 192.\\n\\f 7\\n B.\\n\\n Federal credit unions pool funds from – and give loans\\nto – their members and other credit-union entities. 12 U.S.C.\\n§ 1757(5), (6). A credit union’s members, whether individual\\nor corporate, must come from the credit union’s membership\\n“field,” id. § 1753(5), which is based on a shared occupation,\\nassociation, or geographical area. Members receive regular\\ndividends. Id. § 1763. Congress has shielded federal credit\\nunions from federal corporate income taxes and most state and\\nlocal taxes, but members must pay taxes on their dividends.\\nSee JAMES M. BICKLEY, CONG. RESEARCH SERV., 97-548 E,\\nSHOULD CREDIT UNIONS BE TAXED? 3-5 (2005).\\n\\n To create a federal credit union, at least seven individuals\\nmust present a proposed charter and pay a fee to the NCUA.\\nSee MICHAEL P. MALLOY, BANKING LAW & REGULATION\\n§ 2.04 (2d ed. 2019). In the application, the organizers must\\npledge to deposit funds for shares in the institution and must\\ndescribe the credit union’s proposed membership field. 12\\nU.S.C. § 1753(3), (5). The NCUA must approve the charter\\nbefore the institution may start. See id. § 1754. The agency\\nwill complete an “appropriate investigation” and determine the\\n“general character and fitness” of the organizers, the\\n“economic advisability of establishing” the credit union, and\\nthe “conform[ity]” of proposal details with the Act. Id.\\n\\n The Act governs two types of federal credit unions:\\n“common-bond” credit unions and “community” credit unions.\\nSee id. § 1759(b). This case deals with the latter category. The\\n1934 version of the Act required a community credit union’s\\nmembership field to reflect a particular geographical area – to\\nwit, “a well-defined neighborhood, community, or rural\\ndistrict.” § 9, 48 Stat. at 1219. As amended in 1998, the Act\\nprovides that membership for a community credit union “shall\\n\\f 8\\nbe limited to . . . [p]ersons or organizations within a well-\\ndefined local community, neighborhood, or rural district.” 12\\nU.S.C. § 1759(b) (emphasis added). The 1998 version calls on\\nthe NCUA to “prescribe, by regulation, a definition for the term\\n‘well-defined local community, neighborhood, or rural\\ndistrict.’” Id. § 1759(g)(1). Thus, under the new regime,\\nindividuals seeking to organize a new community credit union\\n(or alter an existing one) must commit to serving members\\nwithin the NCUA’s contemporaneous definition of “local\\ncommunity, neighborhood, or rural district.” See S. REP. NO.\\n105-193, at 4, 8 (1998); H.R. REP. NO. 105-472, at 21 (1998).\\nAs part of their application to the NCUA, they must provide a\\nproposed description of the precise geographical area that the\\ncredit union would serve.\\n\\n Since 1998, there has been “dramatic growth” in the\\nnumber of community credit unions. U.S. GOV’T\\nACCOUNTABILITY OFF., GAO-07-29, CREDIT UNIONS:\\nGREATER TRANSPARENCY NEEDED ON WHO CREDIT UNIONS\\nSERVE AND ON SENIOR EXECUTIVE COMPENSATION\\nARRANGEMENTS 4 (2006). Despite a 11-percent drop in the\\nnumber of federal credit unions from 2000 to 2005, community\\ncredit unions doubled to 1,115. Id. at 4, 12. Meanwhile, the\\namount of assets in community credit unions quadrupled to\\n$104 billion. Id. at 4.\\n\\n C.\\n\\n On December 7, 2016, the NCUA amended its\\nmembership-field rules for community credit unions. See\\nChartering and Field of Membership Manual, 81 Fed. Reg.\\n88,412 (Dec. 7, 2016). Several changes rely on two terms\\ndevised by the Office of Management and Budget (OMB) and\\nbased on data collected by the Census Bureau (Census): “Core\\nBased Statistical Areas” and “Combined Statistical Areas.”\\n\\f 9\\n\\n The OMB has designated numerous regions around the\\ncountry as Core Based Statistical Areas, which comprise at\\nleast one urban cluster, or core, of 10,000 or more people and\\nadjacent counties with substantial commuting ties to that core.\\nSee U.S. CENSUS BUREAU, GEOGRAPHICAL PROGRAM,\\nGLOSSARY, https://www.census.gov/programs-surveys/geogra\\nphy/about/glossary.html. In layman’s terms, a Core Based\\nStatistical Area is a city or town and its suburbs.\\n\\n Meanwhile, a Combined Statistical Area is a conglomerate\\nof two or more adjoining Core Based Statistical Areas, each of\\nwhich has substantial commuting ties with at least one other\\nCore Based Statistical Area in the group. Id. Essentially, a\\nCombined Statistical Area is a regional hub with urban centers\\nconnected by commuting patterns. Combined Statistical Areas\\nmay “reflect broader social and economic interactions, such as\\nwholesaling, commodity distribution, and weekend recreation\\nactivities.” OFFICE OF MGMT. & BUDGET, EXEC. OFFICE OF THE\\nPRESIDENT, OMB BULL. NO. 15-01, REVISED DELINEATIONS OF\\nMETROPOLITAN STATISTICAL AREAS, MICROPOLITAN\\nSTATISTICAL AREAS, AND COMBINED STATISTICAL AREAS, AND\\nGUIDANCE ON USES OF THE DELINEATIONS OF THESE AREAS\\napp. 2-3 (2015).\\n\\n Relevant here, the 2016 rule made two changes to the\\nNCUA’s definition of the term “local community” under\\n§ 1759(b)(3) and one to that of “rural district.” The changes\\naffect what proposed membership areas satisfy the\\ngeographical limitation imposed by the Act.\\n\\n The first change to the “local community” definition\\ninvolves Combined Statistical Areas. A proposed area\\nqualifies as a local community if it encompasses the whole or\\na portion of a Combined Statistical Area and does not exceed a\\n\\f 10\\ndesignated population limit. See 81 Fed. Reg. at 88,440. The\\nNCUA has set that cap at 2.5 million people.\\n\\n The second change involves Core Based Statistical Areas.\\nThe parties agree that all or part of a Core Based Statistical\\nArea may qualify as a local community so long as it does not\\nexceed the population limit. But since 2010, the NCUA\\nrequired such a membership area to include the urban core.\\nThe new rule no longer requires that the core be included in the\\nlocal community that a credit union proposes to serve. See id.\\nat 88,413, 88,440.\\n\\n As for the “rural district” definition, the new rule increases\\nthe population cap for valid rural districts from 250,000 people\\n(or 3 percent of the population of the state where most eligible\\nresidents are located) to 1 million people. See id. at 88,416,\\n88,440. The new population limit works with two other\\nconstraints set by the rule: (1) an outer geographical limit on\\nhow far a rural district may extend past the borders of the credit\\nunion’s headquarters state; and (2) a requirement either that\\nmost eligible residents reside in Census-designated rural areas,\\nor that the population density of the proposed district equals\\n100 or fewer people per square mile. See id. at 88,440.\\n\\n D.\\n\\n On the day the NCUA published the rule, the Association\\nfiled this injunctive and declaratory action in the District Court.\\nThe Association claimed that the three changes described\\nabove were not only arbitrary and capricious under the\\nAdministrative Procedure Act (APA), Pub. L. No. 79-404, 60\\nStat. 237 (1946) (codified as 5 U.S.C. §§ 701-06), but also\\nunreasonable and entitled to no deference under Chevron. The\\nagency and Association filed cross-motions for summary\\n\\f 11\\njudgment. On March 29, 2018, the District Court granted both\\nmotions in part and denied them in part.\\n\\n The court made three relevant holdings. First, it rejected\\nas unreasonable the qualification of certain Combined\\nStatistical Areas as local communities. See Am. Bankers Ass’n,\\n306 F. Supp. 3d at 61. Second, it sustained as well-reasoned\\nthe elimination of the core requirement from the Core Based\\nStatistical Area a credit union proposes to serve as its local\\ncommunity. Id. at 64-65. Third, it rejected as unreasonable the\\nincreased population cap for rural districts. Id. at 69-70. (It\\nalso sustained a separate portion of the rule, which the\\nAssociation does not challenge here.)\\n\\n The NCUA and Association timely appealed. We have\\nappellate jurisdiction under 28 U.S.C. § 1291.\\n\\n II.\\n\\n At the outset, we must assure ourselves of our subject\\nmatter jurisdiction over the appellate proceeding. See, e.g.,\\nUnited States v. Gooch, 842 F.3d 1274, 1277 (D.C. Cir. 2016).\\n\\n In their original briefing, the parties failed to apprise us of\\na rule that was promulgated while this appeal was pending and\\nthat changed membership-field requirements for community\\ncredit unions. See Chartering and Field of Membership, 83\\nFed. Reg. 30,289 (June 28, 2018). The 2018 rule eliminated\\nthe portion of the 2016 rule allowing Combined Statistical\\nAreas to qualify as local communities. Compare 12 C.F.R. pt.\\n701, app. B, ch. 2 § V.A.2 (2018), with id. (2019). The 2018\\nrule preamble did not specifically discuss the removal but\\nconcluded that “[a]ny modification in th[e] final rule is\\nconsistent with the District Court decision” in this action. 83\\nFed. Reg. at 30,291.\\n\\f 12\\n\\n “Under the mootness doctrine, we cannot decide a case if\\n‘events have so transpired that the decision will neither\\npresently affect the parties’ rights nor have a more-than-\\nspeculative chance of affecting them in the future.’” Reid v.\\nHurwitz, 920 F.3d 828, 832 (D.C. Cir. 2019) (citation omitted).\\nThe same principle applies to individual claims. See Tucson\\nMed. Ctr. v. Sullivan, 947 F.2d 971, 977 (D.C. Cir. 1991).\\nAccordingly, if a rule under review (or a portion of it) is\\nsuperseded or amended during an appeal, the proceeding (or\\nrelevant part) might be moot. See, e.g., Am. Bankers Ass’n v.\\nNCUA, 271 F.3d 262, 274 (D.C. Cir. 2001). We therefore\\nrequested supplemental briefing on the issue of appellate\\njurisdiction.\\n\\n Based on the government’s submission and\\nrepresentations at oral argument, we hold that the portion of the\\nappeal related to Combined Statistical Areas is not moot.\\n“[T]he mere power to [reinstitute] a challenged law is not a\\nsufficient basis on which a court can conclude that” a challenge\\nremains live. Nat’l Black Police Ass’n v. District of Columbia,\\n108 F.3d 346, 349 (D.C. Cir. 1997). The government has stated\\nthat, if we reverse the relevant part of the District Court’s\\ndecision, all three members of the NCUA’s board intend to\\nreinstitute the Combined Statistical Area portion of the 2016\\nrule. See Decl. of Michael McKenna ¶ 3, ECF No. 1781123;\\nOral Arg. Recording 11:33-48. Accordingly, City of Mesquite\\nv. Aladdin’s Castle, Inc., 455 U.S. 283 (1982), governs this\\ncase. In Aladdin’s Castle, the Supreme Court deemed the\\ncontroversy live in part because the government had announced\\n“an intention” to restore the rule under challenge if the lower-\\ncourt decision were vacated. Id. at 289 & n.11; see also Am.\\nBankers Ass’n, 271 F.3d at 274. The NCUA’s submission and\\nrepresentations evince such an intention here, and the\\nAssociation – which bears the “heavy burden” of proving\\n\\f 13\\nmootness, Friends of the Earth, Inc. v. Laidlaw Envtl. Servs.\\n(TOC), Inc., 528 U.S. 167, 189 (2000) (citation\\nomitted) – offers no evidence to the contrary.\\n\\n The Association attempts to distinguish Aladdin’s Castle\\non two grounds, but neither sways us. First, the Association\\nnotes that the city government in Aladdin’s Castle said it would\\nreenact “precisely the same” law, see 455 U.S. at 289, but that\\nin this case any future notice-and-comment proceedings might\\nproduce a different “local community” definition, perhaps not\\neven relying on Combined Statistical Areas. After all, the\\nagency must keep a “flexible and open-minded attitude” during\\nthe process. See Nat’l Tour Brokers Ass’n v. United States, 591\\nF.2d 896, 902 (D.C. Cir. 1978). We grant that commenters\\nmight convince the NCUA to change its mind; mann tracht,\\nund Gott lacht. But that strikes us as speculative as public input\\nconvincing Mesquite legislators to enact a different law. We\\ndo not see Aladdin’s Castle turning on such conjecture.\\nInstead, we see a live dispute because there is “no certainty”\\nthat the NCUA will forego reinstating the same Combined\\nStatistical Area definition. Aladdin’s Castle, 455 U.S. at 289;\\nsee also Trinity Lutheran Church of Columbia, Inc. v. Comer,\\n137 S. Ct. 2012, 2019 n.1 (2017) (holding that the court must\\nfind “absolutely clear that the allegedly wrongful behavior\\ncould not reasonably be expected to recur” (emphasis added)\\n(quoting Friends of the Earth, 528 U.S. at 189)). After all, the\\nNCUA continues to defend the definition here. See Knox v.\\nServ. Emps. Int’l Union, Local 100, 567 U.S. 298, 307 (2012).\\n\\n Second, the Association observes that both sides in\\nAladdin’s Castle urged the Supreme Court to treat their dispute\\nas live. In contrast, only the NCUA seeks to proceed here; the\\nAssociation would prefer to wait until the agency reinstitutes\\nthe rule. But the existence or absence of jurisdiction does not\\nturn on which parties challenge or defend it. Cf. Bender v.\\n\\f 14\\nWilliamsport Area Sch. Dist., 475 U.S. 534, 541 (1986).\\nBecause the NCUA remains bound by the lower-court\\njudgment, the present injury renders irrelevant the\\nAssociation’s preference. “Jurisdiction existing,” our duty to\\ndecide the appeal “is ‘virtually unflagging.’” Sprint\\nCommc’ns, Inc. v. Jacobs, 571 U.S. 69, 77 (2013) (citation\\nomitted).\\n\\n In short, we may review the challenge to the rule change\\ninvolving Combined Statistical Areas. We see no jurisdictional\\nissues with the rest of the appeal. We thus turn to the merits.\\n\\n III.\\n\\n We review de novo the District Court’s rulings on\\nsummary judgment. See Loan Syndications & Trading Ass’n\\nv. SEC, 882 F.3d 220, 222 (D.C. Cir. 2018). We review the\\nadministrative record and give “no particular deference” to the\\nDistrict Court’s views. Oceana, Inc. v. Ross, 920 F.3d 855, 860\\n(D.C. Cir. 2019) (citation omitted).\\n\\n The APA governs this suit. In relevant part, the statute\\nprovides that we “decide all relevant questions of law” and\\n“interpret . . . statutory provisions.” 5 U.S.C. § 706. We\\nordinarily set aside agency actions that are either “arbitrary,\\ncapricious, an abuse of discretion, or otherwise not in\\naccordance with law,” id. § 706(2)(A), or “in excess of\\nstatutory jurisdiction, authority, or limitations, or short of\\nstatutory right,” id. § 706(2)(C).\\n\\n We review the agency rule in accordance with the familiar\\nChevron doctrine, a two-prong test for determining whether an\\nagency “has stayed within the bounds of its statutory authority”\\nwhen issuing its action. City of Arlington v. FCC, 569 U.S.\\n290, 297 (2013) (emphasis omitted). At the first step, we\\n\\f 15\\ndetermine “whether Congress has directly spoken to the precise\\nquestion at issue,” and we “give effect” to any “unambiguously\\nexpressed intent.” Chevron, 467 U.S. at 842-43 & n.9.\\n\\n If we glean no such unambiguous intent, we turn to the\\nsecond step and determine “whether the agency’s answer” to\\nthe question “is based on a permissible construction of the\\nstatute.” Id. at 843. By arriving at the second step, we have\\nconcluded that Congress either explicitly or implicitly\\ndelegated to the agency the lawmaking authority to clarify the\\nstatute. We presume that Congress would not authorize the\\npromulgation of an “[im]permissible construction.” Chevron,\\n467 U.S. at 843. Accordingly, we will set aside agency actions\\nbased on such a construction, because they are either “not in\\naccordance with law” or “in excess of statutory jurisdiction.”\\n5 U.S.C. § 706(2)(A), (C); see also CSX Transp., Inc. v.\\nSurface Transp. Bd., 754 F.3d 1056, 1063 (D.C. Cir. 2014);\\nAss’n of Privacy Sector Colls. & Univs. v. Duncan, 681 F.3d\\n427, 441 (D.C. Cir. 2012).\\n\\n IV.\\n\\n For all three challenges, the first step of the Chevron\\nanalysis proceeds in the same way. “We begin our analysis, as\\nalways, with the statutory text.” Tesoro Alaska Co. v. FERC,\\n778 F.3d 1034, 1038 (D.C. Cir. 2015). Congress having\\nexpressly assigned the NCUA the power to define the\\nchallenged terms, see 12 U.S.C. § 1759(g)(1), we may proceed\\nto Chevron’s second prong without further analysis, see U.S.\\nTelecom Ass’n v. FCC, 825 F.3d 674, 717 (D.C. Cir. 2016);\\nRush Univ. Med. Ctr. v. Burwell, 763 F.3d 754, 760 (7th Cir.\\n2014); Comm’r v. Pepsi-Cola Niagara Bottling Corp., 399\\nF.2d 390, 393 (2d Cir. 1968) (Friendly, J.) (“When Congress\\nhas used a general term and has empowered an administrator\\n\\f 16\\nto define it, the courts must respect his construction if this is\\nwithin the range of reason.”).\\n\\n An express delegation of definitional power “necessarily\\nsuggests that Congress did not intend the [terms] to be applied\\nin [their] plain meaning sense,” Women Involved in Farm\\nEcon. v. U.S. Dep’t of Agric., 876 F.2d 994, 1000 (D.C. Cir.\\n1989), that they are not “self-defining,” id., and that the agency\\n“enjoy[s] broad discretion” in how to define them, Lindeen v.\\nSEC, 825 F.3d 646, 653 (D.C. Cir. 2016). In Chevron terms,\\nCongress through explicit language “has directly spoken to the\\nprecise question” of whether the identified terms must carry\\ncertain meanings. 467 U.S. at 842-43. The answer is no. See\\nBuongiorno v. Sullivan, 912 F.2d 504, 509 (D.C. Cir. 1990)\\n(Thomas, J.) (“When Congress expressly delegates the\\nauthority to fill a gap in a statute, Congress speaks, in effect,\\ndirectly, and says, succinctly, that it wants the agency to\\nannotate its words.”). To hold otherwise at the first Chevron\\nstep would “undermine” the ability of Congress to delegate\\ndefinitional power. Rush Univ. Med. Ctr., 763 F.3d at 760.\\n\\n Consequently, we turn to whether the NCUA’s definitions\\nare “based on a permissible construction of the statute.”\\nChevron, 467 U.S. at 843.\\n\\n V.\\n\\n Agency interpretations promulgated to fill an explicit\\nlegislative gap “are given controlling weight unless they are\\narbitrary, capricious, or manifestly contrary to the statute.”\\nChevron, 467 U.S. at 844; see also Pharm. Research & Mfrs.\\nof Am. v. FTC, 790 F.3d 198, 204 (D.C. Cir. 2015).\\n\\n Under arbitrary and capricious review, “we may not\\nsubstitute our own judgment for that” of the agency. FERC v.\\n\\f 17\\nElec. Power Supply Ass’n, 136 S. Ct. 760, 782 (2016); accord\\nDep’t of Commerce v. New York, 139 S. Ct. 2551, 2569 (2019).\\nStill, we are “not a ‘rubber stamp,’” Oceana, 920 F.3d at 863;\\n“the agency must examine the relevant data and articulate a\\nsatisfactory explanation for its action,” Motor Vehicle Mfrs.\\nAss’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co. (State\\nFarm), 463 U.S. 29, 43 (1983). A rule is arbitrary and\\ncapricious if (1) the agency “has relied on factors which\\nCongress has not intended it to consider”; (2) the agency\\n“entirely failed to consider an important aspect of the\\nproblem”; (3) the agency’s explanation “runs counter to the\\nevidence before the agency”; or (4) the explanation “is so\\nimplausible that it could not be ascribed to a difference in view\\nor the product of agency expertise.” Id.\\n\\n In turn, we assess three definitional changes in the 2016\\nrule: (1) qualifying Combined Statistical Areas as local\\ncommunities; (2) eliminating the core requirement for local\\ncommunities based on Core Based Statistical Areas; and\\n(3) raising the population cap for rural districts. We sustain the\\nfirst and third amendments in full. As for the second, we hold\\nthat it is rationally related to the Act’s text and purposes, but\\nthat it is insufficiently explained.\\n\\n A.\\n\\n The District Court rejected the first change because it\\napproved certain Combined Statistical Areas “no matter how\\ngeographically dispersed and unconnected” the “members may\\nbe.” Am. Bankers Ass’n, 306 F. Supp. 3d at 61. To the court,\\nthe approval did not fit with the term “local community,”\\nwhich, in its view, “encompass[es] an area no larger than a\\ncounty.” Id. at 58, 61. We respectfully disagree.\\n\\f 18\\n The NCUA possesses vast discretion to define terms\\nbecause Congress expressly has given it such power. But the\\nauthority is not boundless. The agency must craft a reasonable\\ndefinition consistent with the Act’s text and purposes; that is\\ncentral to the review we apply at Chevron’s second step. Here,\\nthe NCUA’s definition meets the standard.\\n\\n We first focus on the text. Congress introduced the phrase\\n“local community” in the 1998 amendments. The word\\n“community” had a broad scope at the time. It meant not only\\n“society at large” but also a “body of individuals organized into\\na unit or manifesting usu[ally] with awareness of some\\nunifying trait.” See Community, WEBSTER’S THIRD\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE\\nUNABRIDGED 460 (1993). The group could be “united by\\nhistorical consciousness or . . . common social, economic, and\\npolitical interests.” Id. But the unifying trait could also be\\nsimply “living in a particular place or region.” Id.\\n\\n The NCUA has recognized that the modifier “local”\\n“reflects congressional intent that it takes ‘a more circumspect\\nand restricted approach to chartering community credit\\nunions.’” Am. Bankers Ass’n, 271 F.3d at 273 (citation\\nomitted); see also Oral Arg. Recording 3:16-18. Indeed,\\nCongress made clear its intention to “modif[y]” the “current\\nlaw regarding community credit unions” by adding the word\\n“local” to community. S. REP. NO. 105-193, at 6-7.\\n\\n Insertion of the modifier “local” before “community”\\nimplies that the community “relate[s] to” a “particular limited\\ndistrict” or is “confined to a particular place.” Local,\\nWEBSTER’S THIRD INTERNATIONAL DICTIONARY OF THE\\nENGLISH LANGUAGE UNABRIDGED 1327 (1993). But that place\\nneed not be the size of a county, as the District Court held. The\\nSupreme Court has recognized that the geographical areas\\n\\f 19\\n“need not be small.” First Nat’l Bank III, 522 U.S. at 492. And\\nif Congress wanted a local community to correspond to a\\nparticular geographical unit, such as a county, “‘it easily could\\nhave written’ that limitation explicitly.” NCUA Br. 25-26\\n(quoting Ali v. Fed. Bureau of Prisons, 552 U.S. 214, 227\\n(2008)); see also Credit Union Nat’l Ass’n Amicus Br. 15. The\\nNCUA sensibly reads the term “local” to mean simply that the\\ncommunity, regardless of shape or size, should be neither\\n“broad” nor “general.” Local, SUPRA, at 1327.\\n\\n To be clear, we do not hold today that the NCUA must\\nconsider only bonds and social connections as understood in\\n1998. The parties agree, and thus we assume, that the NCUA,\\ndespite its expressly delegated authority, must adopt a\\ndefinition consistent with what the term “local community”\\nmeant in 1998, the time of its adoption.\\n\\n After consulting state statutes and invoking the canon of\\nnoscitur a sociis, the District Court developed a rather size-\\nrestrictive meaning for the phrase. See Am. Bankers Ass’n, 306\\nF. Supp. 3d at 57-59. But we do not think that defining a “local\\ncommunity” to refer to an area larger than a county is an\\nunreasonable interpretation of the Act’s text.\\n\\n We receive little guidance from the state statutes in effect\\nin 1998. Indeed, some of the statutes considered “local\\ncommunities” to be quite large. See, e.g., ALASKA STAT.\\n§ 18.66.990(7) (1998) (defining “local community entity” as “a\\ncity or borough or other political subdivision of the state, a\\nnonprofit organization, or a combination of these” (emphasis\\nadded)); see also NCUA Br. 25 n.4.\\n\\n We also reject the District Court’s invocation of the\\nnoscitur a sociis canon. When several terms “are associated in\\na context suggesting that [they] have something in common,\\n\\f 20\\nthey should be assigned a permissible meaning that makes\\nthem similar.” ANTONIN SCALIA & BRYAN A. GARNER,\\nREADING LAW: THE INTERPRETATION OF LEGAL TEXTS 195\\n(2012). But the “substantive connection, or fit, between” the\\nterms here – local community, neighborhood, and rural\\ndistrict – is “not so tight or so self-evident.” Graham Cty. Soil\\n& Water Conservation Dist. v. United States ex rel. Wilson, 559\\nU.S. 280, 288 (2010). Only the word “neighborhood” has a\\ndictionary definition clearly suggesting a region of small size;\\nthe phrase “local community” does not, as we have explained\\nabove, and neither does “rural district,” as we explain below.\\nA size restriction would impermissibly “submerge[]” the\\nindependent “character” of the latter two terms. Babbitt v.\\nSweet Home Chapter of Cmtys. for a Great Or. (Sweet Home),\\n515 U.S. 687, 702 (1995) (citation omitted).\\n\\n The Association also points to other textual indicators.\\nBut contrary to what it suggests, see Am. Bankers Ass’n Br.\\n27-28, we see nothing in the record suggesting that “local\\ncommunity” is a term of art. The Association also says the\\nusage of word “local” in two other federal statutes indicates\\nthat rules permitting coverage areas of larger than a county\\nwould be manifestly contrary to the Act. See id. at 32; see also\\n15 U.S.C. § 2203 (1998) (defining “local” as “city, town,\\ncounty, . . . or other political subdivision of a State”); 18 U.S.C.\\n§ 666 (1998) (same). (The Association pointed to other laws,\\nbut they did not exist in 1998. See National Defense\\nAuthorization Act for Fiscal Year 2010, Pub. L. No. 111-84,\\n§ 4703(b), 123 Stat. 2190, 2837 (2009); PRIME Act Grants, 66\\nFed. Reg. 29,010 (May 29, 2001).) True enough. But the two\\nstatutes address materially distinct issues – fire prevention and\\nembezzlement – and thus do not indicate that the term “local”\\nmust imply a size limitation here. See Envtl. Def. v. Duke\\nEnergy Corp., 549 U.S. 561, 574 (2007) (“[M]ost words have\\n\\f 21\\ndifferent shades of meaning and consequently may be variously\\nconstrued . . . in different statutes . . . .” (citation omitted)).\\n\\n In addition to being consistent with the Act’s text, the\\nCombined Statistical Area definition rationally advances the\\nAct’s underlying purposes. In the 1998 amendments, Congress\\nmade two relevant findings about purpose. First, legislators\\nfound “essential” to the credit-union system a “meaningful\\naffinity and bond among members, manifested by a\\ncommonality of routine interaction[;] shared and related work\\nexperiences, interests, or activities[;] or the maintenance of an\\notherwise well-understood sense of cohesion or identity.” § 2,\\n112 Stat. at 914. Second, Congress highlighted the importance\\nof “credit union safety and soundness,” because a credit union\\non firm financial footing “will enhance the public benefit that\\ncitizens receive.” Id. The legislative history also confirms the\\nimportance of common bonds, see S. REP. NO. 73-555, at 2; see\\nalso H.R. REP. NO. 105-472, at 12; H.R. REP. NO. 73-2021, at\\n2; 144 CONG. REC. S9094 (daily ed. July 28, 1998) (statement\\nof Sen. Mikulski); id. at S8971-72 (daily ed. July 24, 1998)\\n(statement of Sen. Moseley-Braun), and economic “integrity,”\\nS. REP. NO. 105-193, at 3; see also 144 CONG. REC. S9094\\n(statement of Sen. Mikulski); id. at S8972 (statement of Sen.\\nMoseley-Braun), to federal credit unions.\\n\\n We recognize that there may be some tension between the\\nAct’s principal purposes: A credit union with exceedingly\\nclose ties among its members is unlikely to have a large enough\\ncustomer base to thrive economically. To the extent that such\\ntension exists, the Act leaves to the NCUA to strike a\\nreasonable balance. Congress was well aware that a viable\\ncredit union might serve a relatively large geographical area.\\nSee, e.g., H.R. REP. NO. 73-2021, at 2 (“[T]here are cases in\\nwhich communities and organizations cross State lines.”).\\n\\f 22\\n The NCUA did just that in promulgating the Combined\\nStatistical Area definition. That definition allows for larger\\ncommunity credit unions; the decision is consistent with\\ndecades of history promoting the economic viability of credit\\nunions in the face of banks and other competing financial\\ninstitutions. Nonetheless, the NCUA struck a balance by\\nensuring that members within the local community maintain\\nsomewhat of a commuter relationship with each other. As the\\nAssociation even acknowledges, commuting patterns “may\\nsometimes serve as a proxy for community interaction.” Am.\\nBankers Ass’n Br. 38 n.25. We see nothing irrational about\\nadopting the factor as a proxy for the common bond\\ncontemplated by Congress. Perhaps we would have made a\\ndifferent call had we been the policymakers. Perhaps we would\\nhave sought a tighter bond. Or perhaps we would not have\\nprioritized credit-union growth. See Iowa Bankers Ass’n\\nAmicus Br. 24-25. But we must “refrain from substituting\\n[our] own interstitial lawmaking for that” of the agency.\\nHousehold Credit Servs., Inc. v. Pfennig, 541 U.S. 232, 244\\n(2004) (citation omitted).\\n\\n The NCUA also reasonably explained its amendment to\\nthe “local community” definition. To begin with, the agency\\nreasonably circumscribed the size of a local community under\\nthe Combined Statistical Area rule by imposing a 2.5 million-\\nperson population limit. Used by the OMB in analogous\\ncircumstances, the cap is a “logical breaking point in terms of\\ncommunity cohesiveness with respect to a multijurisdictional\\narea.” Chartering and Field of Membership for Federal Credit\\nUnions, 75 Fed. Reg. 36,257, 36,259 (June 25, 2010).\\nMoreover, the NCUA reasonably relied on its prior experience\\nwith Core Based Statistical Areas, which no party disputes can\\nserve as local communities under the Act. The agency noted\\nthat the average geographic size of Combined Statistical Areas\\nwith populations of up to 2.5 million people is 4,553 square\\n\\f 23\\nmiles, and that the average size of the Core Based Statistical\\nAreas approved by the NCUA as local communities is 4,572\\nsquare miles. See 81 Fed. Reg. at 88,414-15. Given that\\nsimilarity in size, the NCUA reasonably considered adopted its\\npopulation-limited Combined Statistical Areas standard.\\n\\n Finally, the NCUA’s definition does not readily create\\ngeneral, widely dispersed regions. Cf. First Nat’l Bank III, 522\\nU.S. at 502 (indicating that community credit unions may not\\nbe “composed of members from an unlimited number of\\nunrelated geographical units”). Combined Statistical Areas are\\ngeographical units well-accepted within the government. See\\n81 Fed. Reg. at 88,414. Because they essentially are regional\\nhubs, the Combined Statistical Areas concentrate around\\ncentral locations. The OMB carved out the areas so that their\\nconstituent parts share commuting ties and they reflect\\n“broader social and economic interactions.” OFFICE OF MGMT.\\n& BUDGET, SUPRA, at app. 2-3. The NCUA rationally believed\\nthat such “real-world interconnections” would qualify as the\\ntype of mutual bonds suggested by the term “local\\ncommunity.” Credit Union Nat’l Ass’n Amicus Br. 9. Thus,\\nthe agency reasonably determined that Combined Statistical\\nAreas “simply unif[y], as a single community,” already\\nconnected neighboring regions. See 81 Fed. Reg. at 88,415.\\n\\n The Association raises a potpourri of objections to the\\nNCUA’s decision-making. See Am. Bankers Ass’n Br. 33-48.\\nVirtually all of its gripes are forfeited because it failed first to\\nraise them to the agency, see, e.g., Koretoff v. Vilsack, 707 F.3d\\n394, 397-98 (D.C. Cir. 2013) (per curiam), or lack merit\\nbecause they involve outdated “local community” definitions,\\nwhich either did not allow for the qualification of certain\\ngeographical areas as local communities or lacked a population\\ncap of 2.5 million people.\\n\\f 24\\n But one of the complaints is “deserving of sustained\\nconsideration.” Ortiz v. United States, 138 S. Ct. 2165, 2173\\n(2018). The Association contends that some large Combined\\nStatistical Areas are so sprawling that the NCUA’s definition,\\nwhich treats a 2.5-million-person portion of them as a local\\ncommunity, must be unreasonable. Recall that a Combined\\nStatistical Area is a regional hub with more than one urban\\ncenter. Under the OMB’s technical specifications, each center\\nneed not be connected to every other by commuting ties; rather,\\neach need only have ties to one other center within the hub. As\\nthe Association colorfully puts it, the OMB may designate as a\\nCombined Statistical Area a mere “daisy chain” of urban\\ncenters “that are linked to their neighbors but have nothing to\\nwith those at the other end of the chain.” See Am. Bankers\\nAss’n Br. 36 (citation omitted). Theoretically, the Association\\ncontinues, certain 2.5-million-person communities might bring\\ntogether parts of different urban centers with no connection\\nwith one another. The Association also suggests that the rule\\nmight permit local communities comprising non-contiguous\\nportions of a Combined Statistical Area. See id. at 39.\\n\\n We understand the Association’s argument to be attacking\\nthe Combined Statistical Area definition as unreasonable. To\\nthe Association, the NCUA failed sufficiently to consider the\\npotential for the rule to create unreasonable results. One\\nhypothetical application disturbed the District Court here: the\\nprospect that, within the Combined Statistical Area including\\nthe District of Columbia and counties from three states\\n(Maryland, Pennsylvania, and Virginia), one could create a\\nlocal community bringing together Doylesburg, Pennsylvania,\\nand Partlow, Virginia, towns located 200 miles apart. See Am.\\nBankers Ass’n, 306 F. Supp. 3d at 59-60.\\n\\n We might well agree with the District Court that the\\napproval of such a geographical area would contravene the Act.\\n\\f 25\\nBut even so, the Association would need much more to mount\\nits facial pre-enforcement challenge in this case. As the\\nSupreme Court repeatedly has held, “the fact that petitioner can\\npoint to a hypothetical case in which the rule might lead to an\\narbitrary result does not render the rule” facially invalid. Am.\\nHosp. Ass’n v. NLRB, 499 U.S. 606, 619 (1991); see also EPA\\nv. EME Homer City Generation, L.P. (EME Homer), 572 U.S.\\n489, 524 (2014) (“The possibility that the rule, in uncommon\\nparticular applications, might exceed [the agency]’s statutory\\nauthority does not warrant judicial condemnation of the rule in\\nits entirety.”); INS v. Nat’l Ctr. for Immigrants’ Rights, Inc.,\\n502 U.S. 183, 188 (1991) (“That the regulation may be invalid\\nas applied in s[ome] cases . . . does not mean that the regulation\\nis facially invalid because it is without statutory authority.”);\\ncf. Barnhart v. Thomas, 540 U.S. 20, 29 (2003) (“Virtually\\nevery legal (or other) rule has imperfect applications in\\nparticular circumstances.”).\\n\\n Here, the Association’s complaint and the District Court’s\\naccompanying worry strike us as too conjectural. The NCUA\\nmust assess the “economic advisability of establishing” the\\nproposed credit union before approving it, 12 U.S.C. § 1754,\\nand as part of the assessment, the organizers must propose a\\n“realistic” business plan showing how the institution and its\\nbranches would serve all members in the local community, see\\n12 C.F.R. pt. 701, app. B, ch. 1 § IV.D. The Association has\\nfailed to demonstrate the plausibility of a local community that\\nis defined like the hypothetical narrow, multi-state strip and\\naccompanies a realistic business plan. And if the agency were\\nto receive and approve such an application, a petitioner can\\nmake an as-applied challenge. See, e.g., EME Homer, 572 U.S.\\nat 523-24; Buongiorno, 912 F.2d at 510. The Association has\\nsucceeded in such challenges in the past. See, e.g., Am.\\nBankers Ass’n v. NCUA, No. 1:05-CV-2247, 2008 WL\\n2857678, at *1 (M.D. Pa. July 21, 2008); Am. Bankers Ass’n v.\\n\\f 26\\nNCUA, 347 F. Supp. 2d 1061, 1074 (D. Utah 2004). Thus, we\\nreject the facial attack on the amended “local community”\\ndefinition involving Combined Statistical Areas.\\n\\n B.\\n\\n We turn to the next rule change. The District Court upheld\\nthe eliminated core requirement for a local community based\\non a Core Based Statistical Area. The court acknowledged that\\ndefining a local community without its urban core “does not\\nalter the . . . common bond” shared by the members in the\\nremainder. Am. Bankers Ass’n, 306 F. Supp. 3d at 62-63.\\nMeanwhile, the court accepted the agency’s response to the\\nclaim that the new definition encouraged redlining – a broad\\ncategory of lending practices with negative impact on “areas\\nwhere low-income and minority populations are concentrated.”\\nId. at 65 (quoting 81 Fed. Reg. at 88,413).\\n\\n Like the District Court, we hold that the eliminated core\\nrequirement is consistent with the Act’s text and purposes.\\nStill, we see merit in the Association’s redlining argument and\\nthus hold the definitional change to be arbitrary and capricious.\\n\\n 1.\\n\\n We will sustain the eliminated core requirement if it\\nreflects a reasonable interpretation of “local community” that\\nis rationally related to the dual purposes of promoting credit-\\nunion growth and ensuring some cohesion among members. It\\ndoes. Omission of the urban core from proposed geographical\\narea will permit community credit unions to reach new\\nmembers in the suburban parts of the Core Based Statistical\\nArea and thus to maintain a healthy membership. Because the\\nsuburbs under the OMB’s definition have substantial\\ncommuting ties to the urban cluster, they all will be “within a\\n\\f 27\\nfeasible commuting radius” and thus “share[] at least some\\ngeographic ties.” Am. Bankers Ass’n, 306 F. Supp. 3d at 64.\\nThose bonds do not disappear if the local community lacks the\\ncore. The Association seeks greater “assurance[s]” of a\\nmeaningful bond among members, Am. Bankers Ass’n Br. 66-\\n67; see also Am. Bankers Ass’n Reply 11, but we do not\\nreplace the agency’s policy judgment with ours.\\n\\n We also do not believe that the eliminated core\\nrequirement would create sprawling, boundless geographical\\nregions. No one disputes that, as a general matter, a\\nmembership area comprising an intact Core Based Statistical\\nArea will satisfy any definition of “local community.” If the\\nlocal community with the core poses no problem, we fail to see\\nhow a local community without one would. And even as to\\nCore Based Statistical Areas that do not qualify as local\\ncommunities (because they have populations of more than 2.5\\nmillion), the geographical ties ensure that the proposed\\nmembership area will still be contained within the boundaries\\nof a single, well-recognized metropolitan region. A single\\nregion is not what concerned the Supreme Court in First\\nNational Bank III. See 522 U.S. at 502.\\n\\n The Association objects to the expansive hypothetical\\nmembership fields, highlighting two Core Based Statistical\\nAreas whose populations exceed the NCUA’s cap of 2.5\\nmillion. The Association asserts that the rule change “makes it\\nmuch easier to unite far-distant edges” of those sprawling areas\\nin a single membership area. Am. Bankers Ass’n Reply 9. Fair\\npoint. But economic realities do not make it plausible that\\norganizers would propose such a local community or that the\\nNCUA would approve it. Like the Combined Statistical Area\\ndefinition, the eliminated core requirement does not become\\nfacially infirm because of farfetched hypotheticals. To the\\nextent they occur in the future, troublesome rule applications\\n\\f 28\\nmight be subject to as-applied challenges. See Am. Bankers\\nAss’n, 271 F.3d at 267.\\n\\n 2\\n\\n Despite the eliminated core requirement’s consistency\\nwith the Act, we cannot sustain the definitional change because\\nthe NCUA has not adequately explained it.\\n\\n The Association contends that the rule change “effectively\\nallows credit unions to engage in ‘redlining’ by denying service\\nto urban areas with large numbers of minority and lower-\\nincome residents.” Am. Bankers Ass’n Br. 65. The NCUA\\nattempted to address this concern in the preamble. At first\\nblush, the agency’s statements appear persuasive. Still, the\\nAssociation persuasively argues that the response fails to\\n“consider an important aspect” of the redlining issue or is\\notherwise “so implausible” as to be unreasonable. State Farm,\\n463 U.S. at 43. Thus, the eliminated core requirement is\\narbitrary and capricious.\\n\\n During the notice-and-comment proceedings, the\\nAssociation warned against redlining and objected that\\ncommunity credit unions could now “serv[e] wealthier\\nsuburban counties and exclud[e] markets containing low-\\nincome and minority communities that reside in the core area.”\\nLetter from James Chessen, Exec. Vice President & Chief\\nEconomist, Am. Bankers Ass’n, to Gerard S. Poliquin, Sec’y\\nof the Board, Nat’l Credit Union Admin. (Feb 5, 2016), Am.\\nBankers Ass’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C.\\nfiled Aug. 23, 2017), ECF No. 26-1 at 228. Fairly read, the\\nAssociation’s objection is not to traditional redlining, which\\nencompasses the refusal to make loans in low-income or\\nminority neighborhoods within a service area, see Baradaran,\\nsupra, at 494; see also Cassity, supra, at 348, 355. Federal\\n\\f 29\\ncredit unions “cannot ‘redline’” in the traditional sense because\\n“everyone in a credit union’s ‘community’ is a member who is\\neligible to take advantage of credit union services” and credit\\nunions “by definition cannot reach beyond their member\\ncommunities to offer credit to the general public.” Cassity,\\nsupra, at 355. But a community credit union can engage in\\nmore unconventional redlining practices: “gerrymander[ing] to\\ncreate its own community of exclusively higher-income\\nmembers.” Id. at 359. We think it evident that the Association\\nwas focusing on such gerrymandering.\\n\\n In response, the NCUA acknowledged that it originally\\nkept the core requirement as a benefit to “low-income and\\nunderserved populations.” 81 Fed. Reg. at 88,413. Some\\ncommenters criticized the core requirement as failing to\\nachieve that goal; it caused community credit unions “to\\nsacrifice service to other areas” within the Core Based\\nStatistical Area, id., and such service arguably could benefit\\npoor and minority customers residing outside the core, see id.\\nat 88,414 (remarking on the importance of credit unions\\n“providing financial services to low income and underserved\\npopulations without regard to where they are located within a\\ncommunity, i.e., beyond its ‘core area’”).\\n\\n Still, the agency dismissed the redlining concern on other\\ngrounds, pointing to its “supervisory process to assess [credit-\\nunion] management’s efforts to offer service to the entire\\ncommunity [the credit union] seeks to serve.” Id. The NCUA\\nfocused on two aspects of its process. First, the agency touted\\nits “annual evaluation,” which “encompasses [the credit\\nunion]’s implementation of its business and marketing plans.”\\nId. Citing its “[e]xperience” as support, the agency identified\\nthe evaluation as a “more effective means” than a core\\nrequirement to “ensur[e] that the low-income and underserved\\npopulations are fairly served.” Id. Indeed, prior evaluations\\n\\f 30\\nconfirm that the agency has had “success in providing financial\\nservices to low-income and underserved populations without\\nregard to where they are located within” the membership area.\\nId. at 88,414. Second, the NCUA noted its “mandate to\\nconsider member complaints alleging discriminatory practices\\naffecting low-income and underserved populations, such as\\nredlining, and to respond as necessary when such practices are\\nshown to exist.” Id.\\n\\n Both aspects of the NCUA’s supervisory process fail to\\naddress the redlining issue raised by the Association. The\\nannual evaluation process might be an adequate response to\\ntraditional redlining, because it might ensure that community\\ncredit unions adequately serve poor and minority residents\\nliving in their local communities. But we do not see how it\\nfixes gerrymandering or the potential discriminatory economic\\nimpact on urban residents. See State Farm, 463 U.S. at 43\\n(requiring the agency to consider relevant factors). The annual\\nevaluation process necessarily does not come into effect until\\nthe NCUA already has approved the charter, business plan, and\\nproposed local community. See Iowa Bankers Ass’n Amicus\\nBr. 12. And nothing in the record suggests that business plans\\nmay focus on residents residing outside the finalized\\nmembership area; in fact, the law forbids federal credit unions\\nfrom serving those residents.\\n\\n The complaint process also does not work in the\\ngerrymandering context. As the preamble points out,\\ncomplaints are raised by the membership, which would not\\ninclude the affected urban residents because of the rule change.\\nIt seems quite implausible, absent some contrary evidence the\\nagency failed to detail, that members will file grievances based\\non gerrymandering harms suffered by residents outside the\\ncoverage area. See State Farm, 463 U.S. at 43 (rejecting\\nimplausible explanations).\\n\\f 31\\n\\n The NCUA attempts to defend the rule change by offering\\na buffet of other potential rationales in its briefing and at oral\\nargument. They all fail because the agency did not adopt them\\nwhen promulgating the rule change. See State Farm, 463 U.S.\\nat 43 (noting that courts may not “supply a reasoned basis for\\nthe agency’s action that the agency itself has not given”\\n(quoting SEC v. Chenery Corp., 332 U.S. 194, 196 (1947))).\\nWe pause, though, to caution the NCUA about two proffered\\nreasons. At oral argument, the government counsel suggested\\nthat the agency may reject proposed local communities if it\\nsuspects they discriminate against residents in the urban core.\\nSee Oral Arg. Recording 5:28-49, 13:43-14:09. But current\\nreviewing guidelines do not indicate that the agency looks for\\nsuch discrimination. See 12 C.F.R. pt. 701, app. B, ch. 1\\n§ VII.A. And the NCUA made the opposite representations to\\nthe District Court. See Transcript of Motion Hearing at 48:23-\\n49:7 (“[District Court]: . . . If a credit union comes to the\\nagency and says I want to serve X area, either in a rural district\\nor a combined statistical area, and they meet the definition, the\\nagency has no authority to reject that application, as long as the\\ncredit union can demonstrate that they can serve the\\narea? [NCUA]: . . . I think that’s probably right, your\\nHonor.”), Am. Bankers Ass’n v. NCUA, No. 1:16-cv-02394-\\nDLF (D.D.C. filed Mar. 27, 2018), ECF No. 33. The\\ngovernment counsel also suggested that community credit\\nunions already cover the vast majority of urban cores. See Oral\\nArg. Recording 8:10-22, 12:30-41; see also id. at 7:39-8:01; cf.\\nNCUA Reply 27-28. Perhaps, but the current record does not\\nsupport the assertion.\\n\\f 32\\n C.\\n\\n Finally, the District Court rejected the increased\\npopulation cap for rural districts. The court worried that a rural\\ndistrict satisfying the new, higher limit could be too large or\\ncould contain “numerous urban centers.” Am. Bankers Ass’n,\\n306 F. Supp. 3d at 69. Because the rule qualified certain\\ndistricts “no matter [their] geographic size or the number or\\nsize of metropolitan areas falling within [their] proposed\\nboundaries,” the court held that the NCUA’s new definition is\\nunreasonable. Id. at 69-70. But in our view, the new “rural\\ndistrict” definition is reasonable.\\n\\n As suggested above, we assume that the NCUA must\\nadopt definitions consistent with the statutory terms as\\nunderstood in 1934, not today. The terms “rural” and “district”\\ndo not connote specific population or geographical constraints.\\nSee First Nat’l Bank III, 522 U.S. at 492 (noting that markets\\n“need not be small”). A “district” means a “portion of a state,\\ncounty, country, town, or city . . . made for administrative,\\nelectoral, or other purposes.” District, WEBSTER’S NEW\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 649\\n(2d ed. 1934) (emphasis added). The district may be “of\\nundefined extent.” Id. Meanwhile, “rural” indicates a\\n“country” or “agricultur[al]” lifestyle, as opposed to that of a\\n“city or town.” Rural, WEBSTER’S NEW INTERNATIONAL\\nDICTIONARY OF THE ENGLISH LANGUAGE 1861 (2d ed. 1934).\\n“Rural” lands generally “resembl[e]” the “country[side].” Id.\\n\\n Nothing about the 1-million-person cap prevents the rural\\ndistrict from resembling the countryside. And one of the\\nunchallenged restrictions helps provide a rural character to\\nsuch districts. Either most residents in the proposed district\\nlive on rural land, or the population density is 100 or fewer\\npeople per square mile. See 81 Fed. Reg. at 88,440.\\n\\f 33\\nAccordingly, even if most residents live on urban areas in the\\nrural district, those areas will be surrounded by rural land.\\nThat’s because 100 people per square mile – or one person for\\nevery six or so acres – is a rural population density.\\n\\n As for size, the contemporaneous definition of “district”\\nreassures us that rural districts may cross state lines. And a\\nsecond unchallenged restriction assures that the 1-million-\\nperson cap will not support gigantic or straggly rural districts.\\nAs the rule explains, the boundaries of a community credit\\nunion’s district may not “exceed the outer boundaries of the\\nstates” immediately surrounding the state where the proposed\\ncredit union would have its headquarters. 81 Fed. Reg. at\\n88,440. Thus, even though the population density of the entire\\nUnited States is less than 100 people per square mile, see\\nNCUA Br. 56 n.44; Oral Arg. Recording 39:34-38, the\\ngeographical limitation forces a rural district to be much\\nsmaller. We are confident that such districts will not be so\\nenormous as to amount to federations of “unrelated geographic\\nunits.” See First Nat’l Bank III, 522 U.S. at 502.\\n\\n In arguing that the amended “rural district” definition is\\nunreasonable, the Association relies heavily on the interpretive\\nanalysis performed by the District Court. See Am. Bankers\\nAss’n, 306 F. Supp. 3d at 67-69. The court’s core contention\\nis that we must construe “rural” and “district” together as a\\nspecialized phrase that meant, in 1934, an “area[] much smaller\\nthan a state.” Id. at 68. For support, the District Court not only\\nconsulted two 1934-era dictionaries, results of a Westlaw\\nsearch of contemporaneous opinions, and a database containing\\nhistorical uses of the phrase, but also invoked the noscitur a\\nsociis canon. See id. at 67-69.\\n\\n The proffered definitional evidence is pretty thin. The\\n1934-era dictionaries described what the specialized phrase\\n\\f 34\\n“rural district” meant in Great Britain, not in the much larger\\nand more expansive United States, which by the 1930s\\nencompassed forty-eight continental states. See, e.g., Rural\\nDistrict, WEBSTER’S NEW INTERNATIONAL DICTIONARY OF THE\\nENGLISH LANGUAGE 1861 (1934) (“in England, a subdivision\\nof an administrative county embracing . . . county parishes”).\\n\\n The proffered Westlaw search results included scores of\\nfederal and state judicial and administrative opinions. See\\nPlaintiff’s Supplemental Memorandum app. 6, Am. Bankers\\nAss’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C. filed Mar.\\n16, 2018), ECF No. 32-6. But we find little of use in those\\ndocuments. Many of those opinions do not discuss size or\\npopulation. See, e.g., Nicolai v. Wis. Power & Light Co., 269\\nN.W. 281, 282-83 (Wis. 1936). Those that do involve specific\\nrural districts whose sizes were between a town and a state.\\nSee, e.g., Sarther Grocery Co. v Comm’r, 22 B.T.A. 1273, 1274\\n(1931). But none declares that rural districts by definition are\\nrestricted to any particular size or population.\\n\\n The District Court next turned to the Corpus of Historical\\nAmerican Usage, a free and public online database. See Am.\\nBankers Ass’n, 306 F. Supp. 3d at 68 & n.5. The database notes\\n197 mentions of the phrase in the first half of the twentieth\\ncentury and only 37 in the second. See Search Results for\\nRURAL DISTRICTS and RURAL DISTRICT, CORPUS OF\\nHISTORICAL AMERICAN ENGLISH, https://www.english-\\ncorpora.org/coha/ (follow “List” hyperlink; then search\\nmatching strings field for “RURAL DISTRICT”). The\\nperceived drop-off led the District Court to determine that,\\n“even if rural district does not carry meaning distinct from its\\nindividual words today, it did in 1934.” Am. Bankers Ass’n,\\n306 F. Supp. 3d at 68.\\n\\f 35\\n Although federal courts may use “crude[]” searches on\\ndatabases to learn of ambiguities in a statutory term,\\nMuscarello v. United States, 524 U.S. 125, 129-30 (1998), the\\nsearch here did not suffice to show that the agency’s definition\\nwas unreasonable. Much more is required to cabin the\\nagency’s discretion. A search of a commercial database,\\nProQuest, reveals that the phrase appeared at least 500 times in\\nthe second half of the century. We are not confident that the\\nproffered evidence establishes a particular historical trend.\\n\\n The District Court lastly invoked noscitur a sociis. See\\nAm. Bankers Ass’n, 306 F. Supp. 3d at 68-69. The canon is\\ninapposite. Recall that the term “rural district” was listed with\\n“community,” not “local community,” in 1934. At the time,\\n“community” could refer to “[s]ociety at large” or a\\n“commonwealth or state.” Community, WEBSTER’S NEW\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 452\\n(1934). Indeed, we said that the word is “too indefinite to be\\nused for purposes of exact measurement in terms of acres or\\nsquare miles.” Lukens Steel Co. v. Perkins, 107 F.2d 627, 631\\n(D.C. Cir. 1939), rev’d sub nom., Perkins v. Lukens Steel Co.,\\n310 U.S. 113 (1940). Thus, we do not see size or population\\nas a connection between the terms.\\n\\n The increased population cap is consistent with not only\\nthe Act’s text but also its purposes. For instance, as the\\npreamble noted, the expanded definition allows community\\ncredit unions in rural districts to reach more “persons of modest\\nmeans who may reside” in those areas. 81 Fed. Reg. at 88,416;\\nsee also § 2(1), (2), 112 Stat. at 913.\\n\\n And the change is neither arbitrary nor capricious. The\\nnew rule explains that it provides a “level of operating\\nefficiencies and scale” making rural districts attractive options\\nfor prospective credit unions. 81 Fed. Reg. at 88,416. “[A]\\n\\f 36\\nsufficiently large population base is essential to enable” them\\n“to offer financial services economically.” Id. at 88,417. The\\nNCUA also chose the 1-million figure based on prior\\nexperience. The agency noted that it had approved of eight\\ndistricts with populations of more than 250,000, and that one\\nof them already had exceeded 1 million. See id. “Having set a\\n1 million precedent in one state,” the agency felt justified in\\nhaving a “fixed 1 million population cap for the other 49\\nstates.” Id. (emphasis added). We cannot say this policy\\nchoice lacks rational explanation.\\n\\n The Association says the NCUA failed to consider prior\\nagency decisions. See Am. Bankers Ass’n Br. 61. But we see\\nno discrepancy and thus summarily reject the objection. The\\nAssociation also turns to troubling hypothetical examples of\\nrural districts with unruly shapes and those with dense urban\\nareas such as Denver, Colorado. See NCUA Br. 55-61; Oral\\nArg. Recording 37:55-38:05. Again, such implausible outliers\\ndo not impugn the rule’s general reasonableness.\\n\\n VI.\\n\\n We now consider the remedy. To sum up, we hold that\\ndefining certain Combined Statistical Areas or portions of them\\nas local communities and raising the population cap for rural\\ndistricts are consistent with the Act and reasonably explained.\\nThus, we sustain both aspects of the challenged rule. We also\\nleave undisturbed the portion of the District Court’s opinion\\nthat the parties do not contest on appeal.\\n\\n But we deem the eliminated core requirement to be\\narbitrary and capricious. When a rule is contrary to law, the\\n“ordinary practice is to vacate” it. United Steel v. Mine Safety\\n& Health Admin., 925 F.3d 1279, 1287 (D.C. Cir. 2019); see\\nalso 5 U.S.C. § 706(2) (noting that the “reviewing court\\n\\f 37\\nshall . . . set aside” unlawful agency action). But in “rare\\ncases,” we will opt instead to remand without vacating the rule,\\nso that the agency can “correct its errors.” United Steel, 925\\nF.3d at 1287; see also 5 U.S.C. § 702 (“Nothing\\nherein . . . affects . . . the power or duty of the court to . . . deny\\nrelief on any . . . appropriate . . . equitable ground . . . .”); Oglala\\nSioux Tribe v. U.S. Nuclear Regulatory Comm’n, 896 F.3d 520,\\n536 (D.C. Cir. 2018). In considering whether to adopt the latter\\nequitable remedy, we balance “(1) the seriousness of the\\ndeficiencies of the action, that is, how likely it is the agency\\nwill be able to justify its decision on remand; and (2) the\\ndisruptive consequences of vacatur.” United Steel, 925 F.3d at\\n1287 (citation omitted). A strong showing of one factor may\\nobviate the need to find a similar showing of the other. See,\\ne.g., Fox Television Stations, Inc. v. FCC, 280 F.3d 1027, 1049\\n(D.C. Cir. 2002) (holding that, because the agency likely could\\njustify its action on remand, the potential for disruption was\\n“only barely relevant”); Allied-Signal, Inc. v. U.S. Nuclear\\nRegulatory Comm’n, 988 F.2d 146, 152 (D.C. Cir. 1993)\\n(determining that, because vacatur would give regulated parties\\na “peculiar windfall,” the meager chance of justifying the\\naction was given “little weight” in the remedial analysis).\\n\\n The NCUA has not requested remand without vacatur in\\nthis case. But because we have a “duty” to ensure the propriety\\nof the APA remedy, 5 U.S.C. § 702, we hold that we have the\\ndiscretion to raise the issue sua sponte, cf. Igonia v. Califano,\\n568 F.2d 1383, 1387 (D.C. Cir. 1977); Gaddis v. Dixie Realty\\nCo., 420 F.2d 245, 247 (D.C. Cir. 1969) (per curiam). Remand\\nwithout vacatur is appropriate here.\\n\\n We conclude that the NCUA might be able to offer a\\nsatisfactory reason on remand. And as for disruptive effect, we\\nperceive a substantial likelihood that vacating the rule could\\nmake it more difficult for some poor and minority suburban\\n\\f 38\\nresidents to receive adequate financial services. Even\\ntemporarily depriving these members of the opportunity is\\ninequitable, because it would “set back” the Act’s objective of\\noffering financial services to people of small means. See Nat’l\\nRes. Def. Council v. EPA, 489 F.3d 1364, 1374 (D.C. Cir.\\n2007); see also Advocates for Highway & Auto Safety v. Fed.\\nMotor Carrier Safety Admin., 429 F.3d 1136, 1152 (D.C. Cir.\\n2005) (declining to vacate rule because, in the interim, it “may\\ndo some good, if it does anything at all”).\\n\\n Given the potential for sufficient justification and the\\nsubstantial likelihood of disruptive effect, we have a rare case\\nin which vacatur is inappropriate. See United Steel, 925 F.3d\\nat 1287. Thus, we remand without vacating the relevant part\\nof the 2016 rule. We trust that the agency will act\\nexpeditiously. See EME Homer City Generation, L.P. v. EPA,\\n795 F.3d 118, 132 (D.C. Cir. 2015).\\n\\n * * *\\n\\n In short, we reverse the challenged portions of the District\\nCourt’s summary judgment order. With respect to the\\nqualification of certain Combined Statistical Areas as local\\ncommunities and the increased population cap for rural\\ndistricts, we direct the District Court to issue summary\\njudgment in favor of the NCUA. As to the elimination of the\\nurban-core requirement for local communities based on Core\\nBased Statistical Areas, we direct the District Court to issue\\nsummary judgment in favor of the Association and to remand,\\nwithout vacating, the relevant portion of the 2016 rule for\\nfurther explanation.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4432669/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge WILKINS\"}]}"},"judge1":{"kind":"string","value":"HENDERSON"},"judge2":{"kind":"string","value":"PILLARD"},"judge3":{"kind":"string","value":"WILKINS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"HENDERSON\": \", Circuit\", \"PILLARD\": \", Circuit\", \"WILKINS\": \", Circuit\"}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4655416/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: American Bankers Association v. National Credit Union Administration\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-20\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued April 16, 2019 Decided August 20, 2019\n\n No. 18-5154\n\n AMERICAN BANKERS ASSOCIATION,\n APPELLEE\n\n v.\n\n NATIONAL CREDIT UNION ADMINISTRATION,\n APPELLANT\n\n\n Consolidated with 18-5181\n\n\n Appeals from the United States District Court\n for the District of Columbia\n (No. 1:16-cv-02394)\n\n\n Daniel Aguilar, Attorney, U.S. Department of Justice,\nargued the cause for Appellant-Cross-Appellee. With him on\nthe briefs was Mark B. Stern, Attorney.\n\n Allison Jones Rushing was on the brief for amici curiae\nCredit Union National Association, et al. in support of\nAppellant-Cross-Appellee. Nicholas G. Gamse entered an\nappearance.\n\f 2\n Steven D. Gordon was on the brief for amici curiae State\nBankers Associations in support Appellee-Cross-Appellant\nAmerican Bankers Association.\n\n Robert A. Long Jr. argued the cause for Appellee-Cross-\nAppellant. With him on the briefs were Andrew J. Soukup,\nPhilip Levitz, and Lauren Moxley.\n\n Before: HENDERSON, PILLARD, and WILKINS, Circuit\nJudges.\n\n Opinion for the Court filed by Circuit Judge WILKINS.\n\n WILKINS, Circuit Judge: Longstanding principles of\nadministrative law teach us to give federal agencies breathing\nroom when they make policy and “resolv[e] the struggle\nbetween competing views of the public interest.” Chevron,\nU.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 866\n(1984). And because many policy decisions merge with legal\nones, Chevron requires us frequently to sustain agency\ninterpretations of certain federal statutes. Congress often\nexpects agencies, with their political accountability, “bod[ies]\nof experience[,] and informed judgment,” to make sound\ninterpretive choices “with the force of law.” United States v.\nMead Corp., 533 U.S. 218, 227, 229 (2001) (citation omitted).\n\n Congress expressly tasked the National Credit Union\nAdministration (NCUA) with making such choices in defining\nthe reach of federal credit unions. Since the Great Depression,\nCongress has maintained a “system of federal credit unions\nthat . . . provide credit at reasonable rates” and banking\nservices to “people of ‘small means.’” First Nat’l Bank & Tr.\nCo. v. NCUA (First Nat’l Bank I), 988 F.2d 1272, 1274 (D.C.\nCir. 1993) (citation omitted), aff’d, 522 U.S. 479 (1998).\nAlthough a private bank may solicit and welcome customers\n\f 3\nfrom anywhere, Congress has limited whom these federal\nfinancial institutions may serve. For instance, certain\ninstitutions called “community credit unions” may cover\nindividuals and entities only within a preapproved\ngeographical area. The credit union will not receive a federal\ncharter (and thus cannot start operations) unless it first proffers\na geographical coverage area and the NCUA accepts the\nproposal. Congress explicitly assigns the agency the task of\ncreating vetting standards.\n\n Exercising its expressly delegated power, the NCUA has\npromulgated a final rule that makes it easier for community\ncredit unions to expand their geographical coverage and thus to\nreach more potential members. Representing competitors to\nthe credit unions, the American Bankers Association\n(Association) has challenged the NCUA’s new rule as neither\n“in accordance with law” nor within “statutory jurisdiction.” 5\nU.S.C. § 706(2)(A), (C). The District Court vacated significant\nportions of the rule, deeming them to be based on unreasonable\nagency interpretations of the Federal Credit Union Act (Act),\nPub. L. No. 73-467, 48 Stat. 1216 (1934) (codified as amended\nat 12 U.S.C. §§ 1751 to 1795k). See Am. Bankers Ass’n v.\nNCUA, 306 F. Supp. 3d 44, 61, 69-70 (D.D.C. 2018).\n\n We appreciate the District Court’s conclusions, made after\na thoughtful analysis of the Act. But we ultimately disagree\nwith many of them. In this facial challenge, we review the rule\nnot as armchair bankers or geographers, but rather as lay judges\ncognizant that Congress expressly delegated certain policy\nchoices to the NCUA. After considering the Act’s text,\npurpose, and legislative history, we hold the agency’s policy\nchoices “entirely appropriate” for the most part. Chevron, 467\nU.S. at 865. We therefore sustain the bulk of the rule. Still, we\ndo not rubber-stamp this regulation. We remand, without\nvacating, one portion for further consideration of the\n\f 4\ndiscriminatory impact it might have on poor and minority\nurban residents.\n\n I.\n\n A.\n\n The nation’s credit unions started in the early twentieth\ncentury “as a populist mechanism designed to empower\nfarmers against bad loans.” Mehrsa Baradaran, How the Poor\nGot Cut Out of Banking, 62 EMORY L.J. 483, 500 (2013).\nWalloped by crop failures and the Great Depression, farmers\nseeking credit became not only increasingly suspicious of\ntraditional bankers, who “disregard[ed]” poor individuals and\nstayed in the big cities, but also fearful of loan sharks, “who\nwould extract ‘up to a thousand percent’ in interest rates.” Id.\nat 500-01 (quoting 80 CONG. REC. 6752 (1936) (statement of\nRep. Lundeen)). The farmers thus began to build their own\ncredit networks.\n\n In a national grassroots campaign, farmers created\nlocalized, non-profit “credit groups” collecting funds from and\nloaning small sums to one another at low interest rates. See id.\nat 501-02. The success of any such self-help institution\n“hinge[s] on the interpersonal dynamics of its members:\nLenders must be able to evaluate the ability and willingness of\npotential borrowers to pay back their loans and borrowers must\nfeel obligated to pay back those loans.” Wendy Cassity, Note,\nThe Case for a Credit Union Community Reinvestment Act, 100\nCOLUM. L. REV. 331, 337 (2000); see also First Nat’l Bank &\nTr. Co. v. NCUA (First Nat’l Bank II), 90 F.3d 525, 526 (D.C.\nCir. 1996), aff’d, 522 U.S. 479 (1998).\n\n By 1934, individuals had organized about 3,000 local\ncredit unions, with about 750,000 members. See 80 CONG.\n\f 5\nREC. at 6753. Recognizing the success of credit unions at the\nstate level, Congress created a federal system that year by\npassing the Act. Legislators worried that “usurious money\nlending . . . obviously destroy[ed] vast totals of buying power\n[once held by] . . . the average worker.” H.R. REP. NO. 73-\n2021, at 1-2 (1934); see also S. REP. NO. 73-555, at 1 (1934).\nCongress touted the Act’s ability to “make more available to\npeople of small means credit for provident purposes.” H.R.\nREP. NO. 73-2021, at 1; see also S. REP. NO. 73-555, at 1.\n\n Credit unions multiplied over the ensuing decades. By\n1970, Congress created an independent agency to supervise\nfederal credit unions: the NCUA. See Pub. L. No. 91-468, 84\nStat. 994 (1970) (codified as amended in scattered sections of\n12 U.S.C.); see also Swan v. Clinton, 100 F.3d 973, 974 (D.C.\nCir. 1996) (noting that Congress “entrusted” the agency with\n“the responsibility of overseeing” federal credit unions).\nLegislators thought that the agency would be “more responsive\nto the needs of credit unions” and would “provide more flexible\nand innovative regulation” than prior government agencies,\nwhich did not have federal credit unions as their sole focus. S.\nREP. NO. 91-518, at 3 (1969).\n\n The NCUA faced its first major crisis at the end of the\n1970s. After years of economic decline in several industrial\nsectors, federal credit unions tied to those business sectors\nbegan to suffer. The resulting liquidation of numerous credit\nunions “threaten[ed] ‘the safety and soundness of the federal\ncredit union system.’” Cassity, supra, at 338-39 (footnote\nomitted). Reacting to the emergency, the NCUA in 1981\npromulgated a groundbreaking rule that loosened a major size\nlimitation on certain federal credit unions. Almost\nimmediately, those financial institutions grew in membership.\n\f 6\n Meanwhile, credit unions became “caught up in the\nbroader changes in banking and faced internal as well as\nexternal pressure to compete with [private] banks and seek\nhigher profits.” Baradaran, supra, at 505. Unlike credit\nunions, private, for-profit banks were “owned by equity holders\nwho may not necessarily be customers (depositors or\nborrowers),” and they did “not have similar membership and\ncommercial lending restrictions” as credit unions. DARRYL E.\nGETTER, CONG. RESEARCH SERV., IF11048, INTRODUCTION TO\nBANK REGULATION: CREDIT UNIONS AND COMMUNITY BANKS:\nA COMPARISON 1 (2018). To remain viable, credit unions\n“started to focus on attracting more customers and expanding\nthe industry.” Baradaran, supra, at 505. As part of that\nstrategy, many consolidated through mergers. And private\nbanks soon treated credit unions as serious competitors,\nseeking to curb their growth. See NCUA v. First Nat’l Bank &\nTr. Co. (First Nat’l Bank III), 522 U.S. 479, 485 (1998); First\nNat’l Bank I, 988 F.2d at 1276.\n\n In 1998, the banking industry successfully challenged as\ncontrary to the Act the 1981 rule that had eased size limitations\nfor certain federal credit unions. See First Nat’l Bank III, 522\nU.S. at 503. Congress swiftly responded. In less than six\nmonths, legislators amended the Act, superseding the holding\nin First National Bank III, loosening size limitations on certain\nfederal credit unions, and adding other reforms. See Credit\nUnion Membership Access Act, Pub. L. No. 105-219, 112 Stat.\n913 (1998) (codified as amended in scattered sections of 12\nU.S.C.). Partly because of the 1998 amendments and related\nNCUA regulations, credit unions continued to merge and grow\nin membership. Now, more than 61 million customers perform\ntheir banking services at about 3,400 federal credit unions. See\n2018 NAT’L CREDIT UNION ADMIN. ANN. REP. 192.\n\f 7\n B.\n\n Federal credit unions pool funds from – and give loans\nto – their members and other credit-union entities. 12 U.S.C.\n§ 1757(5), (6). A credit union’s members, whether individual\nor corporate, must come from the credit union’s membership\n“field,” id. § 1753(5), which is based on a shared occupation,\nassociation, or geographical area. Members receive regular\ndividends. Id. § 1763. Congress has shielded federal credit\nunions from federal corporate income taxes and most state and\nlocal taxes, but members must pay taxes on their dividends.\nSee JAMES M. BICKLEY, CONG. RESEARCH SERV., 97-548 E,\nSHOULD CREDIT UNIONS BE TAXED? 3-5 (2005).\n\n To create a federal credit union, at least seven individuals\nmust present a proposed charter and pay a fee to the NCUA.\nSee MICHAEL P. MALLOY, BANKING LAW & REGULATION\n§ 2.04 (2d ed. 2019). In the application, the organizers must\npledge to deposit funds for shares in the institution and must\ndescribe the credit union’s proposed membership field. 12\nU.S.C. § 1753(3), (5). The NCUA must approve the charter\nbefore the institution may start. See id. § 1754. The agency\nwill complete an “appropriate investigation” and determine the\n“general character and fitness” of the organizers, the\n“economic advisability of establishing” the credit union, and\nthe “conform[ity]” of proposal details with the Act. Id.\n\n The Act governs two types of federal credit unions:\n“common-bond” credit unions and “community” credit unions.\nSee id. § 1759(b). This case deals with the latter category. The\n1934 version of the Act required a community credit union’s\nmembership field to reflect a particular geographical area – to\nwit, “a well-defined neighborhood, community, or rural\ndistrict.” § 9, 48 Stat. at 1219. As amended in 1998, the Act\nprovides that membership for a community credit union “shall\n\f 8\nbe limited to . . . [p]ersons or organizations within a well-\ndefined local community, neighborhood, or rural district.” 12\nU.S.C. § 1759(b) (emphasis added). The 1998 version calls on\nthe NCUA to “prescribe, by regulation, a definition for the term\n‘well-defined local community, neighborhood, or rural\ndistrict.’” Id. § 1759(g)(1). Thus, under the new regime,\nindividuals seeking to organize a new community credit union\n(or alter an existing one) must commit to serving members\nwithin the NCUA’s contemporaneous definition of “local\ncommunity, neighborhood, or rural district.” See S. REP. NO.\n105-193, at 4, 8 (1998); H.R. REP. NO. 105-472, at 21 (1998).\nAs part of their application to the NCUA, they must provide a\nproposed description of the precise geographical area that the\ncredit union would serve.\n\n Since 1998, there has been “dramatic growth” in the\nnumber of community credit unions. U.S. GOV’T\nACCOUNTABILITY OFF., GAO-07-29, CREDIT UNIONS:\nGREATER TRANSPARENCY NEEDED ON WHO CREDIT UNIONS\nSERVE AND ON SENIOR EXECUTIVE COMPENSATION\nARRANGEMENTS 4 (2006). Despite a 11-percent drop in the\nnumber of federal credit unions from 2000 to 2005, community\ncredit unions doubled to 1,115. Id. at 4, 12. Meanwhile, the\namount of assets in community credit unions quadrupled to\n$104 billion. Id. at 4.\n\n C.\n\n On December 7, 2016, the NCUA amended its\nmembership-field rules for community credit unions. See\nChartering and Field of Membership Manual, 81 Fed. Reg.\n88,412 (Dec. 7, 2016). Several changes rely on two terms\ndevised by the Office of Management and Budget (OMB) and\nbased on data collected by the Census Bureau (Census): “Core\nBased Statistical Areas” and “Combined Statistical Areas.”\n\f 9\n\n The OMB has designated numerous regions around the\ncountry as Core Based Statistical Areas, which comprise at\nleast one urban cluster, or core, of 10,000 or more people and\nadjacent counties with substantial commuting ties to that core.\nSee U.S. CENSUS BUREAU, GEOGRAPHICAL PROGRAM,\nGLOSSARY, https://www.census.gov/programs-surveys/geogra\nphy/about/glossary.html. In layman’s terms, a Core Based\nStatistical Area is a city or town and its suburbs.\n\n Meanwhile, a Combined Statistical Area is a conglomerate\nof two or more adjoining Core Based Statistical Areas, each of\nwhich has substantial commuting ties with at least one other\nCore Based Statistical Area in the group. Id. Essentially, a\nCombined Statistical Area is a regional hub with urban centers\nconnected by commuting patterns. Combined Statistical Areas\nmay “reflect broader social and economic interactions, such as\nwholesaling, commodity distribution, and weekend recreation\nactivities.” OFFICE OF MGMT. & BUDGET, EXEC. OFFICE OF THE\nPRESIDENT, OMB BULL. NO. 15-01, REVISED DELINEATIONS OF\nMETROPOLITAN STATISTICAL AREAS, MICROPOLITAN\nSTATISTICAL AREAS, AND COMBINED STATISTICAL AREAS, AND\nGUIDANCE ON USES OF THE DELINEATIONS OF THESE AREAS\napp. 2-3 (2015).\n\n Relevant here, the 2016 rule made two changes to the\nNCUA’s definition of the term “local community” under\n§ 1759(b)(3) and one to that of “rural district.” The changes\naffect what proposed membership areas satisfy the\ngeographical limitation imposed by the Act.\n\n The first change to the “local community” definition\ninvolves Combined Statistical Areas. A proposed area\nqualifies as a local community if it encompasses the whole or\na portion of a Combined Statistical Area and does not exceed a\n\f 10\ndesignated population limit. See 81 Fed. Reg. at 88,440. The\nNCUA has set that cap at 2.5 million people.\n\n The second change involves Core Based Statistical Areas.\nThe parties agree that all or part of a Core Based Statistical\nArea may qualify as a local community so long as it does not\nexceed the population limit. But since 2010, the NCUA\nrequired such a membership area to include the urban core.\nThe new rule no longer requires that the core be included in the\nlocal community that a credit union proposes to serve. See id.\nat 88,413, 88,440.\n\n As for the “rural district” definition, the new rule increases\nthe population cap for valid rural districts from 250,000 people\n(or 3 percent of the population of the state where most eligible\nresidents are located) to 1 million people. See id. at 88,416,\n88,440. The new population limit works with two other\nconstraints set by the rule: (1) an outer geographical limit on\nhow far a rural district may extend past the borders of the credit\nunion’s headquarters state; and (2) a requirement either that\nmost eligible residents reside in Census-designated rural areas,\nor that the population density of the proposed district equals\n100 or fewer people per square mile. See id. at 88,440.\n\n D.\n\n On the day the NCUA published the rule, the Association\nfiled this injunctive and declaratory action in the District Court.\nThe Association claimed that the three changes described\nabove were not only arbitrary and capricious under the\nAdministrative Procedure Act (APA), Pub. L. No. 79-404, 60\nStat. 237 (1946) (codified as 5 U.S.C. §§ 701-06), but also\nunreasonable and entitled to no deference under Chevron. The\nagency and Association filed cross-motions for summary\n\f 11\njudgment. On March 29, 2018, the District Court granted both\nmotions in part and denied them in part.\n\n The court made three relevant holdings. First, it rejected\nas unreasonable the qualification of certain Combined\nStatistical Areas as local communities. See Am. Bankers Ass’n,\n306 F. Supp. 3d at 61. Second, it sustained as well-reasoned\nthe elimination of the core requirement from the Core Based\nStatistical Area a credit union proposes to serve as its local\ncommunity. Id. at 64-65. Third, it rejected as unreasonable the\nincreased population cap for rural districts. Id. at 69-70. (It\nalso sustained a separate portion of the rule, which the\nAssociation does not challenge here.)\n\n The NCUA and Association timely appealed. We have\nappellate jurisdiction under 28 U.S.C. § 1291.\n\n II.\n\n At the outset, we must assure ourselves of our subject\nmatter jurisdiction over the appellate proceeding. See, e.g.,\nUnited States v. Gooch, 842 F.3d 1274, 1277 (D.C. Cir. 2016).\n\n In their original briefing, the parties failed to apprise us of\na rule that was promulgated while this appeal was pending and\nthat changed membership-field requirements for community\ncredit unions. See Chartering and Field of Membership, 83\nFed. Reg. 30,289 (June 28, 2018). The 2018 rule eliminated\nthe portion of the 2016 rule allowing Combined Statistical\nAreas to qualify as local communities. Compare 12 C.F.R. pt.\n701, app. B, ch. 2 § V.A.2 (2018), with id. (2019). The 2018\nrule preamble did not specifically discuss the removal but\nconcluded that “[a]ny modification in th[e] final rule is\nconsistent with the District Court decision” in this action. 83\nFed. Reg. at 30,291.\n\f 12\n\n “Under the mootness doctrine, we cannot decide a case if\n‘events have so transpired that the decision will neither\npresently affect the parties’ rights nor have a more-than-\nspeculative chance of affecting them in the future.’” Reid v.\nHurwitz, 920 F.3d 828, 832 (D.C. Cir. 2019) (citation omitted).\nThe same principle applies to individual claims. See Tucson\nMed. Ctr. v. Sullivan, 947 F.2d 971, 977 (D.C. Cir. 1991).\nAccordingly, if a rule under review (or a portion of it) is\nsuperseded or amended during an appeal, the proceeding (or\nrelevant part) might be moot. See, e.g., Am. Bankers Ass’n v.\nNCUA, 271 F.3d 262, 274 (D.C. Cir. 2001). We therefore\nrequested supplemental briefing on the issue of appellate\njurisdiction.\n\n Based on the government’s submission and\nrepresentations at oral argument, we hold that the portion of the\nappeal related to Combined Statistical Areas is not moot.\n“[T]he mere power to [reinstitute] a challenged law is not a\nsufficient basis on which a court can conclude that” a challenge\nremains live. Nat’l Black Police Ass’n v. District of Columbia,\n108 F.3d 346, 349 (D.C. Cir. 1997). The government has stated\nthat, if we reverse the relevant part of the District Court’s\ndecision, all three members of the NCUA’s board intend to\nreinstitute the Combined Statistical Area portion of the 2016\nrule. See Decl. of Michael McKenna ¶ 3, ECF No. 1781123;\nOral Arg. Recording 11:33-48. Accordingly, City of Mesquite\nv. Aladdin’s Castle, Inc., 455 U.S. 283 (1982), governs this\ncase. In Aladdin’s Castle, the Supreme Court deemed the\ncontroversy live in part because the government had announced\n“an intention” to restore the rule under challenge if the lower-\ncourt decision were vacated. Id. at 289 & n.11; see also Am.\nBankers Ass’n, 271 F.3d at 274. The NCUA’s submission and\nrepresentations evince such an intention here, and the\nAssociation – which bears the “heavy burden” of proving\n\f 13\nmootness, Friends of the Earth, Inc. v. Laidlaw Envtl. Servs.\n(TOC), Inc., 528 U.S. 167, 189 (2000) (citation\nomitted) – offers no evidence to the contrary.\n\n The Association attempts to distinguish Aladdin’s Castle\non two grounds, but neither sways us. First, the Association\nnotes that the city government in Aladdin’s Castle said it would\nreenact “precisely the same” law, see 455 U.S. at 289, but that\nin this case any future notice-and-comment proceedings might\nproduce a different “local community” definition, perhaps not\neven relying on Combined Statistical Areas. After all, the\nagency must keep a “flexible and open-minded attitude” during\nthe process. See Nat’l Tour Brokers Ass’n v. United States, 591\nF.2d 896, 902 (D.C. Cir. 1978). We grant that commenters\nmight convince the NCUA to change its mind; mann tracht,\nund Gott lacht. But that strikes us as speculative as public input\nconvincing Mesquite legislators to enact a different law. We\ndo not see Aladdin’s Castle turning on such conjecture.\nInstead, we see a live dispute because there is “no certainty”\nthat the NCUA will forego reinstating the same Combined\nStatistical Area definition. Aladdin’s Castle, 455 U.S. at 289;\nsee also Trinity Lutheran Church of Columbia, Inc. v. Comer,\n137 S. Ct. 2012, 2019 n.1 (2017) (holding that the court must\nfind “absolutely clear that the allegedly wrongful behavior\ncould not reasonably be expected to recur” (emphasis added)\n(quoting Friends of the Earth, 528 U.S. at 189)). After all, the\nNCUA continues to defend the definition here. See Knox v.\nServ. Emps. Int’l Union, Local 100, 567 U.S. 298, 307 (2012).\n\n Second, the Association observes that both sides in\nAladdin’s Castle urged the Supreme Court to treat their dispute\nas live. In contrast, only the NCUA seeks to proceed here; the\nAssociation would prefer to wait until the agency reinstitutes\nthe rule. But the existence or absence of jurisdiction does not\nturn on which parties challenge or defend it. Cf. Bender v.\n\f 14\nWilliamsport Area Sch. Dist., 475 U.S. 534, 541 (1986).\nBecause the NCUA remains bound by the lower-court\njudgment, the present injury renders irrelevant the\nAssociation’s preference. “Jurisdiction existing,” our duty to\ndecide the appeal “is ‘virtually unflagging.’” Sprint\nCommc’ns, Inc. v. Jacobs, 571 U.S. 69, 77 (2013) (citation\nomitted).\n\n In short, we may review the challenge to the rule change\ninvolving Combined Statistical Areas. We see no jurisdictional\nissues with the rest of the appeal. We thus turn to the merits.\n\n III.\n\n We review de novo the District Court’s rulings on\nsummary judgment. See Loan Syndications & Trading Ass’n\nv. SEC, 882 F.3d 220, 222 (D.C. Cir. 2018). We review the\nadministrative record and give “no particular deference” to the\nDistrict Court’s views. Oceana, Inc. v. Ross, 920 F.3d 855, 860\n(D.C. Cir. 2019) (citation omitted).\n\n The APA governs this suit. In relevant part, the statute\nprovides that we “decide all relevant questions of law” and\n“interpret . . . statutory provisions.” 5 U.S.C. § 706. We\nordinarily set aside agency actions that are either “arbitrary,\ncapricious, an abuse of discretion, or otherwise not in\naccordance with law,” id. § 706(2)(A), or “in excess of\nstatutory jurisdiction, authority, or limitations, or short of\nstatutory right,” id. § 706(2)(C).\n\n We review the agency rule in accordance with the familiar\nChevron doctrine, a two-prong test for determining whether an\nagency “has stayed within the bounds of its statutory authority”\nwhen issuing its action. City of Arlington v. FCC, 569 U.S.\n290, 297 (2013) (emphasis omitted). At the first step, we\n\f 15\ndetermine “whether Congress has directly spoken to the precise\nquestion at issue,” and we “give effect” to any “unambiguously\nexpressed intent.” Chevron, 467 U.S. at 842-43 & n.9.\n\n If we glean no such unambiguous intent, we turn to the\nsecond step and determine “whether the agency’s answer” to\nthe question “is based on a permissible construction of the\nstatute.” Id. at 843. By arriving at the second step, we have\nconcluded that Congress either explicitly or implicitly\ndelegated to the agency the lawmaking authority to clarify the\nstatute. We presume that Congress would not authorize the\npromulgation of an “[im]permissible construction.” Chevron,\n467 U.S. at 843. Accordingly, we will set aside agency actions\nbased on such a construction, because they are either “not in\naccordance with law” or “in excess of statutory jurisdiction.”\n5 U.S.C. § 706(2)(A), (C); see also CSX Transp., Inc. v.\nSurface Transp. Bd., 754 F.3d 1056, 1063 (D.C. Cir. 2014);\nAss’n of Privacy Sector Colls. & Univs. v. Duncan, 681 F.3d\n427, 441 (D.C. Cir. 2012).\n\n IV.\n\n For all three challenges, the first step of the Chevron\nanalysis proceeds in the same way. “We begin our analysis, as\nalways, with the statutory text.” Tesoro Alaska Co. v. FERC,\n778 F.3d 1034, 1038 (D.C. Cir. 2015). Congress having\nexpressly assigned the NCUA the power to define the\nchallenged terms, see 12 U.S.C. § 1759(g)(1), we may proceed\nto Chevron’s second prong without further analysis, see U.S.\nTelecom Ass’n v. FCC, 825 F.3d 674, 717 (D.C. Cir. 2016);\nRush Univ. Med. Ctr. v. Burwell, 763 F.3d 754, 760 (7th Cir.\n2014); Comm’r v. Pepsi-Cola Niagara Bottling Corp., 399\nF.2d 390, 393 (2d Cir. 1968) (Friendly, J.) (“When Congress\nhas used a general term and has empowered an administrator\n\f 16\nto define it, the courts must respect his construction if this is\nwithin the range of reason.”).\n\n An express delegation of definitional power “necessarily\nsuggests that Congress did not intend the [terms] to be applied\nin [their] plain meaning sense,” Women Involved in Farm\nEcon. v. U.S. Dep’t of Agric., 876 F.2d 994, 1000 (D.C. Cir.\n1989), that they are not “self-defining,” id., and that the agency\n“enjoy[s] broad discretion” in how to define them, Lindeen v.\nSEC, 825 F.3d 646, 653 (D.C. Cir. 2016). In Chevron terms,\nCongress through explicit language “has directly spoken to the\nprecise question” of whether the identified terms must carry\ncertain meanings. 467 U.S. at 842-43. The answer is no. See\nBuongiorno v. Sullivan, 912 F.2d 504, 509 (D.C. Cir. 1990)\n(Thomas, J.) (“When Congress expressly delegates the\nauthority to fill a gap in a statute, Congress speaks, in effect,\ndirectly, and says, succinctly, that it wants the agency to\nannotate its words.”). To hold otherwise at the first Chevron\nstep would “undermine” the ability of Congress to delegate\ndefinitional power. Rush Univ. Med. Ctr., 763 F.3d at 760.\n\n Consequently, we turn to whether the NCUA’s definitions\nare “based on a permissible construction of the statute.”\nChevron, 467 U.S. at 843.\n\n V.\n\n Agency interpretations promulgated to fill an explicit\nlegislative gap “are given controlling weight unless they are\narbitrary, capricious, or manifestly contrary to the statute.”\nChevron, 467 U.S. at 844; see also Pharm. Research & Mfrs.\nof Am. v. FTC, 790 F.3d 198, 204 (D.C. Cir. 2015).\n\n Under arbitrary and capricious review, “we may not\nsubstitute our own judgment for that” of the agency. FERC v.\n\f 17\nElec. Power Supply Ass’n, 136 S. Ct. 760, 782 (2016); accord\nDep’t of Commerce v. New York, 139 S. Ct. 2551, 2569 (2019).\nStill, we are “not a ‘rubber stamp,’” Oceana, 920 F.3d at 863;\n“the agency must examine the relevant data and articulate a\nsatisfactory explanation for its action,” Motor Vehicle Mfrs.\nAss’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co. (State\nFarm), 463 U.S. 29, 43 (1983). A rule is arbitrary and\ncapricious if (1) the agency “has relied on factors which\nCongress has not intended it to consider”; (2) the agency\n“entirely failed to consider an important aspect of the\nproblem”; (3) the agency’s explanation “runs counter to the\nevidence before the agency”; or (4) the explanation “is so\nimplausible that it could not be ascribed to a difference in view\nor the product of agency expertise.” Id.\n\n In turn, we assess three definitional changes in the 2016\nrule: (1) qualifying Combined Statistical Areas as local\ncommunities; (2) eliminating the core requirement for local\ncommunities based on Core Based Statistical Areas; and\n(3) raising the population cap for rural districts. We sustain the\nfirst and third amendments in full. As for the second, we hold\nthat it is rationally related to the Act’s text and purposes, but\nthat it is insufficiently explained.\n\n A.\n\n The District Court rejected the first change because it\napproved certain Combined Statistical Areas “no matter how\ngeographically dispersed and unconnected” the “members may\nbe.” Am. Bankers Ass’n, 306 F. Supp. 3d at 61. To the court,\nthe approval did not fit with the term “local community,”\nwhich, in its view, “encompass[es] an area no larger than a\ncounty.” Id. at 58, 61. We respectfully disagree.\n\f 18\n The NCUA possesses vast discretion to define terms\nbecause Congress expressly has given it such power. But the\nauthority is not boundless. The agency must craft a reasonable\ndefinition consistent with the Act’s text and purposes; that is\ncentral to the review we apply at Chevron’s second step. Here,\nthe NCUA’s definition meets the standard.\n\n We first focus on the text. Congress introduced the phrase\n“local community” in the 1998 amendments. The word\n“community” had a broad scope at the time. It meant not only\n“society at large” but also a “body of individuals organized into\na unit or manifesting usu[ally] with awareness of some\nunifying trait.” See Community, WEBSTER’S THIRD\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE\nUNABRIDGED 460 (1993). The group could be “united by\nhistorical consciousness or . . . common social, economic, and\npolitical interests.” Id. But the unifying trait could also be\nsimply “living in a particular place or region.” Id.\n\n The NCUA has recognized that the modifier “local”\n“reflects congressional intent that it takes ‘a more circumspect\nand restricted approach to chartering community credit\nunions.’” Am. Bankers Ass’n, 271 F.3d at 273 (citation\nomitted); see also Oral Arg. Recording 3:16-18. Indeed,\nCongress made clear its intention to “modif[y]” the “current\nlaw regarding community credit unions” by adding the word\n“local” to community. S. REP. NO. 105-193, at 6-7.\n\n Insertion of the modifier “local” before “community”\nimplies that the community “relate[s] to” a “particular limited\ndistrict” or is “confined to a particular place.” Local,\nWEBSTER’S THIRD INTERNATIONAL DICTIONARY OF THE\nENGLISH LANGUAGE UNABRIDGED 1327 (1993). But that place\nneed not be the size of a county, as the District Court held. The\nSupreme Court has recognized that the geographical areas\n\f 19\n“need not be small.” First Nat’l Bank III, 522 U.S. at 492. And\nif Congress wanted a local community to correspond to a\nparticular geographical unit, such as a county, “‘it easily could\nhave written’ that limitation explicitly.” NCUA Br. 25-26\n(quoting Ali v. Fed. Bureau of Prisons, 552 U.S. 214, 227\n(2008)); see also Credit Union Nat’l Ass’n Amicus Br. 15. The\nNCUA sensibly reads the term “local” to mean simply that the\ncommunity, regardless of shape or size, should be neither\n“broad” nor “general.” Local, SUPRA, at 1327.\n\n To be clear, we do not hold today that the NCUA must\nconsider only bonds and social connections as understood in\n1998. The parties agree, and thus we assume, that the NCUA,\ndespite its expressly delegated authority, must adopt a\ndefinition consistent with what the term “local community”\nmeant in 1998, the time of its adoption.\n\n After consulting state statutes and invoking the canon of\nnoscitur a sociis, the District Court developed a rather size-\nrestrictive meaning for the phrase. See Am. Bankers Ass’n, 306\nF. Supp. 3d at 57-59. But we do not think that defining a “local\ncommunity” to refer to an area larger than a county is an\nunreasonable interpretation of the Act’s text.\n\n We receive little guidance from the state statutes in effect\nin 1998. Indeed, some of the statutes considered “local\ncommunities” to be quite large. See, e.g., ALASKA STAT.\n§ 18.66.990(7) (1998) (defining “local community entity” as “a\ncity or borough or other political subdivision of the state, a\nnonprofit organization, or a combination of these” (emphasis\nadded)); see also NCUA Br. 25 n.4.\n\n We also reject the District Court’s invocation of the\nnoscitur a sociis canon. When several terms “are associated in\na context suggesting that [they] have something in common,\n\f 20\nthey should be assigned a permissible meaning that makes\nthem similar.” ANTONIN SCALIA & BRYAN A. GARNER,\nREADING LAW: THE INTERPRETATION OF LEGAL TEXTS 195\n(2012). But the “substantive connection, or fit, between” the\nterms here – local community, neighborhood, and rural\ndistrict – is “not so tight or so self-evident.” Graham Cty. Soil\n& Water Conservation Dist. v. United States ex rel. Wilson, 559\nU.S. 280, 288 (2010). Only the word “neighborhood” has a\ndictionary definition clearly suggesting a region of small size;\nthe phrase “local community” does not, as we have explained\nabove, and neither does “rural district,” as we explain below.\nA size restriction would impermissibly “submerge[]” the\nindependent “character” of the latter two terms. Babbitt v.\nSweet Home Chapter of Cmtys. for a Great Or. (Sweet Home),\n515 U.S. 687, 702 (1995) (citation omitted).\n\n The Association also points to other textual indicators.\nBut contrary to what it suggests, see Am. Bankers Ass’n Br.\n27-28, we see nothing in the record suggesting that “local\ncommunity” is a term of art. The Association also says the\nusage of word “local” in two other federal statutes indicates\nthat rules permitting coverage areas of larger than a county\nwould be manifestly contrary to the Act. See id. at 32; see also\n15 U.S.C. § 2203 (1998) (defining “local” as “city, town,\ncounty, . . . or other political subdivision of a State”); 18 U.S.C.\n§ 666 (1998) (same). (The Association pointed to other laws,\nbut they did not exist in 1998. See National Defense\nAuthorization Act for Fiscal Year 2010, Pub. L. No. 111-84,\n§ 4703(b), 123 Stat. 2190, 2837 (2009); PRIME Act Grants, 66\nFed. Reg. 29,010 (May 29, 2001).) True enough. But the two\nstatutes address materially distinct issues – fire prevention and\nembezzlement – and thus do not indicate that the term “local”\nmust imply a size limitation here. See Envtl. Def. v. Duke\nEnergy Corp., 549 U.S. 561, 574 (2007) (“[M]ost words have\n\f 21\ndifferent shades of meaning and consequently may be variously\nconstrued . . . in different statutes . . . .” (citation omitted)).\n\n In addition to being consistent with the Act’s text, the\nCombined Statistical Area definition rationally advances the\nAct’s underlying purposes. In the 1998 amendments, Congress\nmade two relevant findings about purpose. First, legislators\nfound “essential” to the credit-union system a “meaningful\naffinity and bond among members, manifested by a\ncommonality of routine interaction[;] shared and related work\nexperiences, interests, or activities[;] or the maintenance of an\notherwise well-understood sense of cohesion or identity.” § 2,\n112 Stat. at 914. Second, Congress highlighted the importance\nof “credit union safety and soundness,” because a credit union\non firm financial footing “will enhance the public benefit that\ncitizens receive.” Id. The legislative history also confirms the\nimportance of common bonds, see S. REP. NO. 73-555, at 2; see\nalso H.R. REP. NO. 105-472, at 12; H.R. REP. NO. 73-2021, at\n2; 144 CONG. REC. S9094 (daily ed. July 28, 1998) (statement\nof Sen. Mikulski); id. at S8971-72 (daily ed. July 24, 1998)\n(statement of Sen. Moseley-Braun), and economic “integrity,”\nS. REP. NO. 105-193, at 3; see also 144 CONG. REC. S9094\n(statement of Sen. Mikulski); id. at S8972 (statement of Sen.\nMoseley-Braun), to federal credit unions.\n\n We recognize that there may be some tension between the\nAct’s principal purposes: A credit union with exceedingly\nclose ties among its members is unlikely to have a large enough\ncustomer base to thrive economically. To the extent that such\ntension exists, the Act leaves to the NCUA to strike a\nreasonable balance. Congress was well aware that a viable\ncredit union might serve a relatively large geographical area.\nSee, e.g., H.R. REP. NO. 73-2021, at 2 (“[T]here are cases in\nwhich communities and organizations cross State lines.”).\n\f 22\n The NCUA did just that in promulgating the Combined\nStatistical Area definition. That definition allows for larger\ncommunity credit unions; the decision is consistent with\ndecades of history promoting the economic viability of credit\nunions in the face of banks and other competing financial\ninstitutions. Nonetheless, the NCUA struck a balance by\nensuring that members within the local community maintain\nsomewhat of a commuter relationship with each other. As the\nAssociation even acknowledges, commuting patterns “may\nsometimes serve as a proxy for community interaction.” Am.\nBankers Ass’n Br. 38 n.25. We see nothing irrational about\nadopting the factor as a proxy for the common bond\ncontemplated by Congress. Perhaps we would have made a\ndifferent call had we been the policymakers. Perhaps we would\nhave sought a tighter bond. Or perhaps we would not have\nprioritized credit-union growth. See Iowa Bankers Ass’n\nAmicus Br. 24-25. But we must “refrain from substituting\n[our] own interstitial lawmaking for that” of the agency.\nHousehold Credit Servs., Inc. v. Pfennig, 541 U.S. 232, 244\n(2004) (citation omitted).\n\n The NCUA also reasonably explained its amendment to\nthe “local community” definition. To begin with, the agency\nreasonably circumscribed the size of a local community under\nthe Combined Statistical Area rule by imposing a 2.5 million-\nperson population limit. Used by the OMB in analogous\ncircumstances, the cap is a “logical breaking point in terms of\ncommunity cohesiveness with respect to a multijurisdictional\narea.” Chartering and Field of Membership for Federal Credit\nUnions, 75 Fed. Reg. 36,257, 36,259 (June 25, 2010).\nMoreover, the NCUA reasonably relied on its prior experience\nwith Core Based Statistical Areas, which no party disputes can\nserve as local communities under the Act. The agency noted\nthat the average geographic size of Combined Statistical Areas\nwith populations of up to 2.5 million people is 4,553 square\n\f 23\nmiles, and that the average size of the Core Based Statistical\nAreas approved by the NCUA as local communities is 4,572\nsquare miles. See 81 Fed. Reg. at 88,414-15. Given that\nsimilarity in size, the NCUA reasonably considered adopted its\npopulation-limited Combined Statistical Areas standard.\n\n Finally, the NCUA’s definition does not readily create\ngeneral, widely dispersed regions. Cf. First Nat’l Bank III, 522\nU.S. at 502 (indicating that community credit unions may not\nbe “composed of members from an unlimited number of\nunrelated geographical units”). Combined Statistical Areas are\ngeographical units well-accepted within the government. See\n81 Fed. Reg. at 88,414. Because they essentially are regional\nhubs, the Combined Statistical Areas concentrate around\ncentral locations. The OMB carved out the areas so that their\nconstituent parts share commuting ties and they reflect\n“broader social and economic interactions.” OFFICE OF MGMT.\n& BUDGET, SUPRA, at app. 2-3. The NCUA rationally believed\nthat such “real-world interconnections” would qualify as the\ntype of mutual bonds suggested by the term “local\ncommunity.” Credit Union Nat’l Ass’n Amicus Br. 9. Thus,\nthe agency reasonably determined that Combined Statistical\nAreas “simply unif[y], as a single community,” already\nconnected neighboring regions. See 81 Fed. Reg. at 88,415.\n\n The Association raises a potpourri of objections to the\nNCUA’s decision-making. See Am. Bankers Ass’n Br. 33-48.\nVirtually all of its gripes are forfeited because it failed first to\nraise them to the agency, see, e.g., Koretoff v. Vilsack, 707 F.3d\n394, 397-98 (D.C. Cir. 2013) (per curiam), or lack merit\nbecause they involve outdated “local community” definitions,\nwhich either did not allow for the qualification of certain\ngeographical areas as local communities or lacked a population\ncap of 2.5 million people.\n\f 24\n But one of the complaints is “deserving of sustained\nconsideration.” Ortiz v. United States, 138 S. Ct. 2165, 2173\n(2018). The Association contends that some large Combined\nStatistical Areas are so sprawling that the NCUA’s definition,\nwhich treats a 2.5-million-person portion of them as a local\ncommunity, must be unreasonable. Recall that a Combined\nStatistical Area is a regional hub with more than one urban\ncenter. Under the OMB’s technical specifications, each center\nneed not be connected to every other by commuting ties; rather,\neach need only have ties to one other center within the hub. As\nthe Association colorfully puts it, the OMB may designate as a\nCombined Statistical Area a mere “daisy chain” of urban\ncenters “that are linked to their neighbors but have nothing to\nwith those at the other end of the chain.” See Am. Bankers\nAss’n Br. 36 (citation omitted). Theoretically, the Association\ncontinues, certain 2.5-million-person communities might bring\ntogether parts of different urban centers with no connection\nwith one another. The Association also suggests that the rule\nmight permit local communities comprising non-contiguous\nportions of a Combined Statistical Area. See id. at 39.\n\n We understand the Association’s argument to be attacking\nthe Combined Statistical Area definition as unreasonable. To\nthe Association, the NCUA failed sufficiently to consider the\npotential for the rule to create unreasonable results. One\nhypothetical application disturbed the District Court here: the\nprospect that, within the Combined Statistical Area including\nthe District of Columbia and counties from three states\n(Maryland, Pennsylvania, and Virginia), one could create a\nlocal community bringing together Doylesburg, Pennsylvania,\nand Partlow, Virginia, towns located 200 miles apart. See Am.\nBankers Ass’n, 306 F. Supp. 3d at 59-60.\n\n We might well agree with the District Court that the\napproval of such a geographical area would contravene the Act.\n\f 25\nBut even so, the Association would need much more to mount\nits facial pre-enforcement challenge in this case. As the\nSupreme Court repeatedly has held, “the fact that petitioner can\npoint to a hypothetical case in which the rule might lead to an\narbitrary result does not render the rule” facially invalid. Am.\nHosp. Ass’n v. NLRB, 499 U.S. 606, 619 (1991); see also EPA\nv. EME Homer City Generation, L.P. (EME Homer), 572 U.S.\n489, 524 (2014) (“The possibility that the rule, in uncommon\nparticular applications, might exceed [the agency]’s statutory\nauthority does not warrant judicial condemnation of the rule in\nits entirety.”); INS v. Nat’l Ctr. for Immigrants’ Rights, Inc.,\n502 U.S. 183, 188 (1991) (“That the regulation may be invalid\nas applied in s[ome] cases . . . does not mean that the regulation\nis facially invalid because it is without statutory authority.”);\ncf. Barnhart v. Thomas, 540 U.S. 20, 29 (2003) (“Virtually\nevery legal (or other) rule has imperfect applications in\nparticular circumstances.”).\n\n Here, the Association’s complaint and the District Court’s\naccompanying worry strike us as too conjectural. The NCUA\nmust assess the “economic advisability of establishing” the\nproposed credit union before approving it, 12 U.S.C. § 1754,\nand as part of the assessment, the organizers must propose a\n“realistic” business plan showing how the institution and its\nbranches would serve all members in the local community, see\n12 C.F.R. pt. 701, app. B, ch. 1 § IV.D. The Association has\nfailed to demonstrate the plausibility of a local community that\nis defined like the hypothetical narrow, multi-state strip and\naccompanies a realistic business plan. And if the agency were\nto receive and approve such an application, a petitioner can\nmake an as-applied challenge. See, e.g., EME Homer, 572 U.S.\nat 523-24; Buongiorno, 912 F.2d at 510. The Association has\nsucceeded in such challenges in the past. See, e.g., Am.\nBankers Ass’n v. NCUA, No. 1:05-CV-2247, 2008 WL\n2857678, at *1 (M.D. Pa. July 21, 2008); Am. Bankers Ass’n v.\n\f 26\nNCUA, 347 F. Supp. 2d 1061, 1074 (D. Utah 2004). Thus, we\nreject the facial attack on the amended “local community”\ndefinition involving Combined Statistical Areas.\n\n B.\n\n We turn to the next rule change. The District Court upheld\nthe eliminated core requirement for a local community based\non a Core Based Statistical Area. The court acknowledged that\ndefining a local community without its urban core “does not\nalter the . . . common bond” shared by the members in the\nremainder. Am. Bankers Ass’n, 306 F. Supp. 3d at 62-63.\nMeanwhile, the court accepted the agency’s response to the\nclaim that the new definition encouraged redlining – a broad\ncategory of lending practices with negative impact on “areas\nwhere low-income and minority populations are concentrated.”\nId. at 65 (quoting 81 Fed. Reg. at 88,413).\n\n Like the District Court, we hold that the eliminated core\nrequirement is consistent with the Act’s text and purposes.\nStill, we see merit in the Association’s redlining argument and\nthus hold the definitional change to be arbitrary and capricious.\n\n 1.\n\n We will sustain the eliminated core requirement if it\nreflects a reasonable interpretation of “local community” that\nis rationally related to the dual purposes of promoting credit-\nunion growth and ensuring some cohesion among members. It\ndoes. Omission of the urban core from proposed geographical\narea will permit community credit unions to reach new\nmembers in the suburban parts of the Core Based Statistical\nArea and thus to maintain a healthy membership. Because the\nsuburbs under the OMB’s definition have substantial\ncommuting ties to the urban cluster, they all will be “within a\n\f 27\nfeasible commuting radius” and thus “share[] at least some\ngeographic ties.” Am. Bankers Ass’n, 306 F. Supp. 3d at 64.\nThose bonds do not disappear if the local community lacks the\ncore. The Association seeks greater “assurance[s]” of a\nmeaningful bond among members, Am. Bankers Ass’n Br. 66-\n67; see also Am. Bankers Ass’n Reply 11, but we do not\nreplace the agency’s policy judgment with ours.\n\n We also do not believe that the eliminated core\nrequirement would create sprawling, boundless geographical\nregions. No one disputes that, as a general matter, a\nmembership area comprising an intact Core Based Statistical\nArea will satisfy any definition of “local community.” If the\nlocal community with the core poses no problem, we fail to see\nhow a local community without one would. And even as to\nCore Based Statistical Areas that do not qualify as local\ncommunities (because they have populations of more than 2.5\nmillion), the geographical ties ensure that the proposed\nmembership area will still be contained within the boundaries\nof a single, well-recognized metropolitan region. A single\nregion is not what concerned the Supreme Court in First\nNational Bank III. See 522 U.S. at 502.\n\n The Association objects to the expansive hypothetical\nmembership fields, highlighting two Core Based Statistical\nAreas whose populations exceed the NCUA’s cap of 2.5\nmillion. The Association asserts that the rule change “makes it\nmuch easier to unite far-distant edges” of those sprawling areas\nin a single membership area. Am. Bankers Ass’n Reply 9. Fair\npoint. But economic realities do not make it plausible that\norganizers would propose such a local community or that the\nNCUA would approve it. Like the Combined Statistical Area\ndefinition, the eliminated core requirement does not become\nfacially infirm because of farfetched hypotheticals. To the\nextent they occur in the future, troublesome rule applications\n\f 28\nmight be subject to as-applied challenges. See Am. Bankers\nAss’n, 271 F.3d at 267.\n\n 2\n\n Despite the eliminated core requirement’s consistency\nwith the Act, we cannot sustain the definitional change because\nthe NCUA has not adequately explained it.\n\n The Association contends that the rule change “effectively\nallows credit unions to engage in ‘redlining’ by denying service\nto urban areas with large numbers of minority and lower-\nincome residents.” Am. Bankers Ass’n Br. 65. The NCUA\nattempted to address this concern in the preamble. At first\nblush, the agency’s statements appear persuasive. Still, the\nAssociation persuasively argues that the response fails to\n“consider an important aspect” of the redlining issue or is\notherwise “so implausible” as to be unreasonable. State Farm,\n463 U.S. at 43. Thus, the eliminated core requirement is\narbitrary and capricious.\n\n During the notice-and-comment proceedings, the\nAssociation warned against redlining and objected that\ncommunity credit unions could now “serv[e] wealthier\nsuburban counties and exclud[e] markets containing low-\nincome and minority communities that reside in the core area.”\nLetter from James Chessen, Exec. Vice President & Chief\nEconomist, Am. Bankers Ass’n, to Gerard S. Poliquin, Sec’y\nof the Board, Nat’l Credit Union Admin. (Feb 5, 2016), Am.\nBankers Ass’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C.\nfiled Aug. 23, 2017), ECF No. 26-1 at 228. Fairly read, the\nAssociation’s objection is not to traditional redlining, which\nencompasses the refusal to make loans in low-income or\nminority neighborhoods within a service area, see Baradaran,\nsupra, at 494; see also Cassity, supra, at 348, 355. Federal\n\f 29\ncredit unions “cannot ‘redline’” in the traditional sense because\n“everyone in a credit union’s ‘community’ is a member who is\neligible to take advantage of credit union services” and credit\nunions “by definition cannot reach beyond their member\ncommunities to offer credit to the general public.” Cassity,\nsupra, at 355. But a community credit union can engage in\nmore unconventional redlining practices: “gerrymander[ing] to\ncreate its own community of exclusively higher-income\nmembers.” Id. at 359. We think it evident that the Association\nwas focusing on such gerrymandering.\n\n In response, the NCUA acknowledged that it originally\nkept the core requirement as a benefit to “low-income and\nunderserved populations.” 81 Fed. Reg. at 88,413. Some\ncommenters criticized the core requirement as failing to\nachieve that goal; it caused community credit unions “to\nsacrifice service to other areas” within the Core Based\nStatistical Area, id., and such service arguably could benefit\npoor and minority customers residing outside the core, see id.\nat 88,414 (remarking on the importance of credit unions\n“providing financial services to low income and underserved\npopulations without regard to where they are located within a\ncommunity, i.e., beyond its ‘core area’”).\n\n Still, the agency dismissed the redlining concern on other\ngrounds, pointing to its “supervisory process to assess [credit-\nunion] management’s efforts to offer service to the entire\ncommunity [the credit union] seeks to serve.” Id. The NCUA\nfocused on two aspects of its process. First, the agency touted\nits “annual evaluation,” which “encompasses [the credit\nunion]’s implementation of its business and marketing plans.”\nId. Citing its “[e]xperience” as support, the agency identified\nthe evaluation as a “more effective means” than a core\nrequirement to “ensur[e] that the low-income and underserved\npopulations are fairly served.” Id. Indeed, prior evaluations\n\f 30\nconfirm that the agency has had “success in providing financial\nservices to low-income and underserved populations without\nregard to where they are located within” the membership area.\nId. at 88,414. Second, the NCUA noted its “mandate to\nconsider member complaints alleging discriminatory practices\naffecting low-income and underserved populations, such as\nredlining, and to respond as necessary when such practices are\nshown to exist.” Id.\n\n Both aspects of the NCUA’s supervisory process fail to\naddress the redlining issue raised by the Association. The\nannual evaluation process might be an adequate response to\ntraditional redlining, because it might ensure that community\ncredit unions adequately serve poor and minority residents\nliving in their local communities. But we do not see how it\nfixes gerrymandering or the potential discriminatory economic\nimpact on urban residents. See State Farm, 463 U.S. at 43\n(requiring the agency to consider relevant factors). The annual\nevaluation process necessarily does not come into effect until\nthe NCUA already has approved the charter, business plan, and\nproposed local community. See Iowa Bankers Ass’n Amicus\nBr. 12. And nothing in the record suggests that business plans\nmay focus on residents residing outside the finalized\nmembership area; in fact, the law forbids federal credit unions\nfrom serving those residents.\n\n The complaint process also does not work in the\ngerrymandering context. As the preamble points out,\ncomplaints are raised by the membership, which would not\ninclude the affected urban residents because of the rule change.\nIt seems quite implausible, absent some contrary evidence the\nagency failed to detail, that members will file grievances based\non gerrymandering harms suffered by residents outside the\ncoverage area. See State Farm, 463 U.S. at 43 (rejecting\nimplausible explanations).\n\f 31\n\n The NCUA attempts to defend the rule change by offering\na buffet of other potential rationales in its briefing and at oral\nargument. They all fail because the agency did not adopt them\nwhen promulgating the rule change. See State Farm, 463 U.S.\nat 43 (noting that courts may not “supply a reasoned basis for\nthe agency’s action that the agency itself has not given”\n(quoting SEC v. Chenery Corp., 332 U.S. 194, 196 (1947))).\nWe pause, though, to caution the NCUA about two proffered\nreasons. At oral argument, the government counsel suggested\nthat the agency may reject proposed local communities if it\nsuspects they discriminate against residents in the urban core.\nSee Oral Arg. Recording 5:28-49, 13:43-14:09. But current\nreviewing guidelines do not indicate that the agency looks for\nsuch discrimination. See 12 C.F.R. pt. 701, app. B, ch. 1\n§ VII.A. And the NCUA made the opposite representations to\nthe District Court. See Transcript of Motion Hearing at 48:23-\n49:7 (“[District Court]: . . . If a credit union comes to the\nagency and says I want to serve X area, either in a rural district\nor a combined statistical area, and they meet the definition, the\nagency has no authority to reject that application, as long as the\ncredit union can demonstrate that they can serve the\narea? [NCUA]: . . . I think that’s probably right, your\nHonor.”), Am. Bankers Ass’n v. NCUA, No. 1:16-cv-02394-\nDLF (D.D.C. filed Mar. 27, 2018), ECF No. 33. The\ngovernment counsel also suggested that community credit\nunions already cover the vast majority of urban cores. See Oral\nArg. Recording 8:10-22, 12:30-41; see also id. at 7:39-8:01; cf.\nNCUA Reply 27-28. Perhaps, but the current record does not\nsupport the assertion.\n\f 32\n C.\n\n Finally, the District Court rejected the increased\npopulation cap for rural districts. The court worried that a rural\ndistrict satisfying the new, higher limit could be too large or\ncould contain “numerous urban centers.” Am. Bankers Ass’n,\n306 F. Supp. 3d at 69. Because the rule qualified certain\ndistricts “no matter [their] geographic size or the number or\nsize of metropolitan areas falling within [their] proposed\nboundaries,” the court held that the NCUA’s new definition is\nunreasonable. Id. at 69-70. But in our view, the new “rural\ndistrict” definition is reasonable.\n\n As suggested above, we assume that the NCUA must\nadopt definitions consistent with the statutory terms as\nunderstood in 1934, not today. The terms “rural” and “district”\ndo not connote specific population or geographical constraints.\nSee First Nat’l Bank III, 522 U.S. at 492 (noting that markets\n“need not be small”). A “district” means a “portion of a state,\ncounty, country, town, or city . . . made for administrative,\nelectoral, or other purposes.” District, WEBSTER’S NEW\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 649\n(2d ed. 1934) (emphasis added). The district may be “of\nundefined extent.” Id. Meanwhile, “rural” indicates a\n“country” or “agricultur[al]” lifestyle, as opposed to that of a\n“city or town.” Rural, WEBSTER’S NEW INTERNATIONAL\nDICTIONARY OF THE ENGLISH LANGUAGE 1861 (2d ed. 1934).\n“Rural” lands generally “resembl[e]” the “country[side].” Id.\n\n Nothing about the 1-million-person cap prevents the rural\ndistrict from resembling the countryside. And one of the\nunchallenged restrictions helps provide a rural character to\nsuch districts. Either most residents in the proposed district\nlive on rural land, or the population density is 100 or fewer\npeople per square mile. See 81 Fed. Reg. at 88,440.\n\f 33\nAccordingly, even if most residents live on urban areas in the\nrural district, those areas will be surrounded by rural land.\nThat’s because 100 people per square mile – or one person for\nevery six or so acres – is a rural population density.\n\n As for size, the contemporaneous definition of “district”\nreassures us that rural districts may cross state lines. And a\nsecond unchallenged restriction assures that the 1-million-\nperson cap will not support gigantic or straggly rural districts.\nAs the rule explains, the boundaries of a community credit\nunion’s district may not “exceed the outer boundaries of the\nstates” immediately surrounding the state where the proposed\ncredit union would have its headquarters. 81 Fed. Reg. at\n88,440. Thus, even though the population density of the entire\nUnited States is less than 100 people per square mile, see\nNCUA Br. 56 n.44; Oral Arg. Recording 39:34-38, the\ngeographical limitation forces a rural district to be much\nsmaller. We are confident that such districts will not be so\nenormous as to amount to federations of “unrelated geographic\nunits.” See First Nat’l Bank III, 522 U.S. at 502.\n\n In arguing that the amended “rural district” definition is\nunreasonable, the Association relies heavily on the interpretive\nanalysis performed by the District Court. See Am. Bankers\nAss’n, 306 F. Supp. 3d at 67-69. The court’s core contention\nis that we must construe “rural” and “district” together as a\nspecialized phrase that meant, in 1934, an “area[] much smaller\nthan a state.” Id. at 68. For support, the District Court not only\nconsulted two 1934-era dictionaries, results of a Westlaw\nsearch of contemporaneous opinions, and a database containing\nhistorical uses of the phrase, but also invoked the noscitur a\nsociis canon. See id. at 67-69.\n\n The proffered definitional evidence is pretty thin. The\n1934-era dictionaries described what the specialized phrase\n\f 34\n“rural district” meant in Great Britain, not in the much larger\nand more expansive United States, which by the 1930s\nencompassed forty-eight continental states. See, e.g., Rural\nDistrict, WEBSTER’S NEW INTERNATIONAL DICTIONARY OF THE\nENGLISH LANGUAGE 1861 (1934) (“in England, a subdivision\nof an administrative county embracing . . . county parishes”).\n\n The proffered Westlaw search results included scores of\nfederal and state judicial and administrative opinions. See\nPlaintiff’s Supplemental Memorandum app. 6, Am. Bankers\nAss’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C. filed Mar.\n16, 2018), ECF No. 32-6. But we find little of use in those\ndocuments. Many of those opinions do not discuss size or\npopulation. See, e.g., Nicolai v. Wis. Power & Light Co., 269\nN.W. 281, 282-83 (Wis. 1936). Those that do involve specific\nrural districts whose sizes were between a town and a state.\nSee, e.g., Sarther Grocery Co. v Comm’r, 22 B.T.A. 1273, 1274\n(1931). But none declares that rural districts by definition are\nrestricted to any particular size or population.\n\n The District Court next turned to the Corpus of Historical\nAmerican Usage, a free and public online database. See Am.\nBankers Ass’n, 306 F. Supp. 3d at 68 & n.5. The database notes\n197 mentions of the phrase in the first half of the twentieth\ncentury and only 37 in the second. See Search Results for\nRURAL DISTRICTS and RURAL DISTRICT, CORPUS OF\nHISTORICAL AMERICAN ENGLISH, https://www.english-\ncorpora.org/coha/ (follow “List” hyperlink; then search\nmatching strings field for “RURAL DISTRICT”). The\nperceived drop-off led the District Court to determine that,\n“even if rural district does not carry meaning distinct from its\nindividual words today, it did in 1934.” Am. Bankers Ass’n,\n306 F. Supp. 3d at 68.\n\f 35\n Although federal courts may use “crude[]” searches on\ndatabases to learn of ambiguities in a statutory term,\nMuscarello v. United States, 524 U.S. 125, 129-30 (1998), the\nsearch here did not suffice to show that the agency’s definition\nwas unreasonable. Much more is required to cabin the\nagency’s discretion. A search of a commercial database,\nProQuest, reveals that the phrase appeared at least 500 times in\nthe second half of the century. We are not confident that the\nproffered evidence establishes a particular historical trend.\n\n The District Court lastly invoked noscitur a sociis. See\nAm. Bankers Ass’n, 306 F. Supp. 3d at 68-69. The canon is\ninapposite. Recall that the term “rural district” was listed with\n“community,” not “local community,” in 1934. At the time,\n“community” could refer to “[s]ociety at large” or a\n“commonwealth or state.” Community, WEBSTER’S NEW\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 452\n(1934). Indeed, we said that the word is “too indefinite to be\nused for purposes of exact measurement in terms of acres or\nsquare miles.” Lukens Steel Co. v. Perkins, 107 F.2d 627, 631\n(D.C. Cir. 1939), rev’d sub nom., Perkins v. Lukens Steel Co.,\n310 U.S. 113 (1940). Thus, we do not see size or population\nas a connection between the terms.\n\n The increased population cap is consistent with not only\nthe Act’s text but also its purposes. For instance, as the\npreamble noted, the expanded definition allows community\ncredit unions in rural districts to reach more “persons of modest\nmeans who may reside” in those areas. 81 Fed. Reg. at 88,416;\nsee also § 2(1), (2), 112 Stat. at 913.\n\n And the change is neither arbitrary nor capricious. The\nnew rule explains that it provides a “level of operating\nefficiencies and scale” making rural districts attractive options\nfor prospective credit unions. 81 Fed. Reg. at 88,416. “[A]\n\f 36\nsufficiently large population base is essential to enable” them\n“to offer financial services economically.” Id. at 88,417. The\nNCUA also chose the 1-million figure based on prior\nexperience. The agency noted that it had approved of eight\ndistricts with populations of more than 250,000, and that one\nof them already had exceeded 1 million. See id. “Having set a\n1 million precedent in one state,” the agency felt justified in\nhaving a “fixed 1 million population cap for the other 49\nstates.” Id. (emphasis added). We cannot say this policy\nchoice lacks rational explanation.\n\n The Association says the NCUA failed to consider prior\nagency decisions. See Am. Bankers Ass’n Br. 61. But we see\nno discrepancy and thus summarily reject the objection. The\nAssociation also turns to troubling hypothetical examples of\nrural districts with unruly shapes and those with dense urban\nareas such as Denver, Colorado. See NCUA Br. 55-61; Oral\nArg. Recording 37:55-38:05. Again, such implausible outliers\ndo not impugn the rule’s general reasonableness.\n\n VI.\n\n We now consider the remedy. To sum up, we hold that\ndefining certain Combined Statistical Areas or portions of them\nas local communities and raising the population cap for rural\ndistricts are consistent with the Act and reasonably explained.\nThus, we sustain both aspects of the challenged rule. We also\nleave undisturbed the portion of the District Court’s opinion\nthat the parties do not contest on appeal.\n\n But we deem the eliminated core requirement to be\narbitrary and capricious. When a rule is contrary to law, the\n“ordinary practice is to vacate” it. United Steel v. Mine Safety\n& Health Admin., 925 F.3d 1279, 1287 (D.C. Cir. 2019); see\nalso 5 U.S.C. § 706(2) (noting that the “reviewing court\n\f 37\nshall . . . set aside” unlawful agency action). But in “rare\ncases,” we will opt instead to remand without vacating the rule,\nso that the agency can “correct its errors.” United Steel, 925\nF.3d at 1287; see also 5 U.S.C. § 702 (“Nothing\nherein . . . affects . . . the power or duty of the court to . . . deny\nrelief on any . . . appropriate . . . equitable ground . . . .”); Oglala\nSioux Tribe v. U.S. Nuclear Regulatory Comm’n, 896 F.3d 520,\n536 (D.C. Cir. 2018). In considering whether to adopt the latter\nequitable remedy, we balance “(1) the seriousness of the\ndeficiencies of the action, that is, how likely it is the agency\nwill be able to justify its decision on remand; and (2) the\ndisruptive consequences of vacatur.” United Steel, 925 F.3d at\n1287 (citation omitted). A strong showing of one factor may\nobviate the need to find a similar showing of the other. See,\ne.g., Fox Television Stations, Inc. v. FCC, 280 F.3d 1027, 1049\n(D.C. Cir. 2002) (holding that, because the agency likely could\njustify its action on remand, the potential for disruption was\n“only barely relevant”); Allied-Signal, Inc. v. U.S. Nuclear\nRegulatory Comm’n, 988 F.2d 146, 152 (D.C. Cir. 1993)\n(determining that, because vacatur would give regulated parties\na “peculiar windfall,” the meager chance of justifying the\naction was given “little weight” in the remedial analysis).\n\n The NCUA has not requested remand without vacatur in\nthis case. But because we have a “duty” to ensure the propriety\nof the APA remedy, 5 U.S.C. § 702, we hold that we have the\ndiscretion to raise the issue sua sponte, cf. Igonia v. Califano,\n568 F.2d 1383, 1387 (D.C. Cir. 1977); Gaddis v. Dixie Realty\nCo., 420 F.2d 245, 247 (D.C. Cir. 1969) (per curiam). Remand\nwithout vacatur is appropriate here.\n\n We conclude that the NCUA might be able to offer a\nsatisfactory reason on remand. And as for disruptive effect, we\nperceive a substantial likelihood that vacating the rule could\nmake it more difficult for some poor and minority suburban\n\f 38\nresidents to receive adequate financial services. Even\ntemporarily depriving these members of the opportunity is\ninequitable, because it would “set back” the Act’s objective of\noffering financial services to people of small means. See Nat’l\nRes. Def. Council v. EPA, 489 F.3d 1364, 1374 (D.C. Cir.\n2007); see also Advocates for Highway & Auto Safety v. Fed.\nMotor Carrier Safety Admin., 429 F.3d 1136, 1152 (D.C. Cir.\n2005) (declining to vacate rule because, in the interim, it “may\ndo some good, if it does anything at all”).\n\n Given the potential for sufficient justification and the\nsubstantial likelihood of disruptive effect, we have a rare case\nin which vacatur is inappropriate. See United Steel, 925 F.3d\nat 1287. Thus, we remand without vacating the relevant part\nof the 2016 rule. We trust that the agency will act\nexpeditiously. See EME Homer City Generation, L.P. v. EPA,\n795 F.3d 118, 132 (D.C. Cir. 2015).\n\n * * *\n\n In short, we reverse the challenged portions of the District\nCourt’s summary judgment order. With respect to the\nqualification of certain Combined Statistical Areas as local\ncommunities and the increased population cap for rural\ndistricts, we direct the District Court to issue summary\njudgment in favor of the NCUA. As to the elimination of the\nurban-core requirement for local communities based on Core\nBased Statistical Areas, we direct the District Court to issue\nsummary judgment in favor of the Association and to remand,\nwithout vacating, the relevant portion of the 2016 rule for\nfurther explanation.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: American Bankers Association v. National Credit Union Administration\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-20\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued April 16, 2019 Decided August 20, 2019\\n\\n No. 18-5154\\n\\n AMERICAN BANKERS ASSOCIATION,\\n APPELLEE\\n\\n v.\\n\\n NATIONAL CREDIT UNION ADMINISTRATION,\\n APPELLANT\\n\\n\\n Consolidated with 18-5181\\n\\n\\n Appeals from the United States District Court\\n for the District of Columbia\\n (No. 1:16-cv-02394)\\n\\n\\n Daniel Aguilar, Attorney, U.S. Department of Justice,\\nargued the cause for Appellant-Cross-Appellee. With him on\\nthe briefs was Mark B. Stern, Attorney.\\n\\n Allison Jones Rushing was on the brief for amici curiae\\nCredit Union National Association, et al. in support of\\nAppellant-Cross-Appellee. Nicholas G. Gamse entered an\\nappearance.\\n\\f 2\\n Steven D. Gordon was on the brief for amici curiae State\\nBankers Associations in support Appellee-Cross-Appellant\\nAmerican Bankers Association.\\n\\n Robert A. Long Jr. argued the cause for Appellee-Cross-\\nAppellant. With him on the briefs were Andrew J. Soukup,\\nPhilip Levitz, and Lauren Moxley.\\n\\n Before: HENDERSON, PILLARD, and WILKINS, Circuit\\nJudges.\\n\\n Opinion for the Court filed by Circuit Judge WILKINS.\\n\\n WILKINS, Circuit Judge: Longstanding principles of\\nadministrative law teach us to give federal agencies breathing\\nroom when they make policy and “resolv[e] the struggle\\nbetween competing views of the public interest.” Chevron,\\nU.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 866\\n(1984). And because many policy decisions merge with legal\\nones, Chevron requires us frequently to sustain agency\\ninterpretations of certain federal statutes. Congress often\\nexpects agencies, with their political accountability, “bod[ies]\\nof experience[,] and informed judgment,” to make sound\\ninterpretive choices “with the force of law.” United States v.\\nMead Corp., 533 U.S. 218, 227, 229 (2001) (citation omitted).\\n\\n Congress expressly tasked the National Credit Union\\nAdministration (NCUA) with making such choices in defining\\nthe reach of federal credit unions. Since the Great Depression,\\nCongress has maintained a “system of federal credit unions\\nthat . . . provide credit at reasonable rates” and banking\\nservices to “people of ‘small means.’” First Nat’l Bank & Tr.\\nCo. v. NCUA (First Nat’l Bank I), 988 F.2d 1272, 1274 (D.C.\\nCir. 1993) (citation omitted), aff’d, 522 U.S. 479 (1998).\\nAlthough a private bank may solicit and welcome customers\\n\\f 3\\nfrom anywhere, Congress has limited whom these federal\\nfinancial institutions may serve. For instance, certain\\ninstitutions called “community credit unions” may cover\\nindividuals and entities only within a preapproved\\ngeographical area. The credit union will not receive a federal\\ncharter (and thus cannot start operations) unless it first proffers\\na geographical coverage area and the NCUA accepts the\\nproposal. Congress explicitly assigns the agency the task of\\ncreating vetting standards.\\n\\n Exercising its expressly delegated power, the NCUA has\\npromulgated a final rule that makes it easier for community\\ncredit unions to expand their geographical coverage and thus to\\nreach more potential members. Representing competitors to\\nthe credit unions, the American Bankers Association\\n(Association) has challenged the NCUA’s new rule as neither\\n“in accordance with law” nor within “statutory jurisdiction.” 5\\nU.S.C. § 706(2)(A), (C). The District Court vacated significant\\nportions of the rule, deeming them to be based on unreasonable\\nagency interpretations of the Federal Credit Union Act (Act),\\nPub. L. No. 73-467, 48 Stat. 1216 (1934) (codified as amended\\nat 12 U.S.C. §§ 1751 to 1795k). See Am. Bankers Ass’n v.\\nNCUA, 306 F. Supp. 3d 44, 61, 69-70 (D.D.C. 2018).\\n\\n We appreciate the District Court’s conclusions, made after\\na thoughtful analysis of the Act. But we ultimately disagree\\nwith many of them. In this facial challenge, we review the rule\\nnot as armchair bankers or geographers, but rather as lay judges\\ncognizant that Congress expressly delegated certain policy\\nchoices to the NCUA. After considering the Act’s text,\\npurpose, and legislative history, we hold the agency’s policy\\nchoices “entirely appropriate” for the most part. Chevron, 467\\nU.S. at 865. We therefore sustain the bulk of the rule. Still, we\\ndo not rubber-stamp this regulation. We remand, without\\nvacating, one portion for further consideration of the\\n\\f 4\\ndiscriminatory impact it might have on poor and minority\\nurban residents.\\n\\n I.\\n\\n A.\\n\\n The nation’s credit unions started in the early twentieth\\ncentury “as a populist mechanism designed to empower\\nfarmers against bad loans.” Mehrsa Baradaran, How the Poor\\nGot Cut Out of Banking, 62 EMORY L.J. 483, 500 (2013).\\nWalloped by crop failures and the Great Depression, farmers\\nseeking credit became not only increasingly suspicious of\\ntraditional bankers, who “disregard[ed]” poor individuals and\\nstayed in the big cities, but also fearful of loan sharks, “who\\nwould extract ‘up to a thousand percent’ in interest rates.” Id.\\nat 500-01 (quoting 80 CONG. REC. 6752 (1936) (statement of\\nRep. Lundeen)). The farmers thus began to build their own\\ncredit networks.\\n\\n In a national grassroots campaign, farmers created\\nlocalized, non-profit “credit groups” collecting funds from and\\nloaning small sums to one another at low interest rates. See id.\\nat 501-02. The success of any such self-help institution\\n“hinge[s] on the interpersonal dynamics of its members:\\nLenders must be able to evaluate the ability and willingness of\\npotential borrowers to pay back their loans and borrowers must\\nfeel obligated to pay back those loans.” Wendy Cassity, Note,\\nThe Case for a Credit Union Community Reinvestment Act, 100\\nCOLUM. L. REV. 331, 337 (2000); see also First Nat’l Bank &\\nTr. Co. v. NCUA (First Nat’l Bank II), 90 F.3d 525, 526 (D.C.\\nCir. 1996), aff’d, 522 U.S. 479 (1998).\\n\\n By 1934, individuals had organized about 3,000 local\\ncredit unions, with about 750,000 members. See 80 CONG.\\n\\f 5\\nREC. at 6753. Recognizing the success of credit unions at the\\nstate level, Congress created a federal system that year by\\npassing the Act. Legislators worried that “usurious money\\nlending . . . obviously destroy[ed] vast totals of buying power\\n[once held by] . . . the average worker.” H.R. REP. NO. 73-\\n2021, at 1-2 (1934); see also S. REP. NO. 73-555, at 1 (1934).\\nCongress touted the Act’s ability to “make more available to\\npeople of small means credit for provident purposes.” H.R.\\nREP. NO. 73-2021, at 1; see also S. REP. NO. 73-555, at 1.\\n\\n Credit unions multiplied over the ensuing decades. By\\n1970, Congress created an independent agency to supervise\\nfederal credit unions: the NCUA. See Pub. L. No. 91-468, 84\\nStat. 994 (1970) (codified as amended in scattered sections of\\n12 U.S.C.); see also Swan v. Clinton, 100 F.3d 973, 974 (D.C.\\nCir. 1996) (noting that Congress “entrusted” the agency with\\n“the responsibility of overseeing” federal credit unions).\\nLegislators thought that the agency would be “more responsive\\nto the needs of credit unions” and would “provide more flexible\\nand innovative regulation” than prior government agencies,\\nwhich did not have federal credit unions as their sole focus. S.\\nREP. NO. 91-518, at 3 (1969).\\n\\n The NCUA faced its first major crisis at the end of the\\n1970s. After years of economic decline in several industrial\\nsectors, federal credit unions tied to those business sectors\\nbegan to suffer. The resulting liquidation of numerous credit\\nunions “threaten[ed] ‘the safety and soundness of the federal\\ncredit union system.’” Cassity, supra, at 338-39 (footnote\\nomitted). Reacting to the emergency, the NCUA in 1981\\npromulgated a groundbreaking rule that loosened a major size\\nlimitation on certain federal credit unions. Almost\\nimmediately, those financial institutions grew in membership.\\n\\f 6\\n Meanwhile, credit unions became “caught up in the\\nbroader changes in banking and faced internal as well as\\nexternal pressure to compete with [private] banks and seek\\nhigher profits.” Baradaran, supra, at 505. Unlike credit\\nunions, private, for-profit banks were “owned by equity holders\\nwho may not necessarily be customers (depositors or\\nborrowers),” and they did “not have similar membership and\\ncommercial lending restrictions” as credit unions. DARRYL E.\\nGETTER, CONG. RESEARCH SERV., IF11048, INTRODUCTION TO\\nBANK REGULATION: CREDIT UNIONS AND COMMUNITY BANKS:\\nA COMPARISON 1 (2018). To remain viable, credit unions\\n“started to focus on attracting more customers and expanding\\nthe industry.” Baradaran, supra, at 505. As part of that\\nstrategy, many consolidated through mergers. And private\\nbanks soon treated credit unions as serious competitors,\\nseeking to curb their growth. See NCUA v. First Nat’l Bank &\\nTr. Co. (First Nat’l Bank III), 522 U.S. 479, 485 (1998); First\\nNat’l Bank I, 988 F.2d at 1276.\\n\\n In 1998, the banking industry successfully challenged as\\ncontrary to the Act the 1981 rule that had eased size limitations\\nfor certain federal credit unions. See First Nat’l Bank III, 522\\nU.S. at 503. Congress swiftly responded. In less than six\\nmonths, legislators amended the Act, superseding the holding\\nin First National Bank III, loosening size limitations on certain\\nfederal credit unions, and adding other reforms. See Credit\\nUnion Membership Access Act, Pub. L. No. 105-219, 112 Stat.\\n913 (1998) (codified as amended in scattered sections of 12\\nU.S.C.). Partly because of the 1998 amendments and related\\nNCUA regulations, credit unions continued to merge and grow\\nin membership. Now, more than 61 million customers perform\\ntheir banking services at about 3,400 federal credit unions. See\\n2018 NAT’L CREDIT UNION ADMIN. ANN. REP. 192.\\n\\f 7\\n B.\\n\\n Federal credit unions pool funds from – and give loans\\nto – their members and other credit-union entities. 12 U.S.C.\\n§ 1757(5), (6). A credit union’s members, whether individual\\nor corporate, must come from the credit union’s membership\\n“field,” id. § 1753(5), which is based on a shared occupation,\\nassociation, or geographical area. Members receive regular\\ndividends. Id. § 1763. Congress has shielded federal credit\\nunions from federal corporate income taxes and most state and\\nlocal taxes, but members must pay taxes on their dividends.\\nSee JAMES M. BICKLEY, CONG. RESEARCH SERV., 97-548 E,\\nSHOULD CREDIT UNIONS BE TAXED? 3-5 (2005).\\n\\n To create a federal credit union, at least seven individuals\\nmust present a proposed charter and pay a fee to the NCUA.\\nSee MICHAEL P. MALLOY, BANKING LAW & REGULATION\\n§ 2.04 (2d ed. 2019). In the application, the organizers must\\npledge to deposit funds for shares in the institution and must\\ndescribe the credit union’s proposed membership field. 12\\nU.S.C. § 1753(3), (5). The NCUA must approve the charter\\nbefore the institution may start. See id. § 1754. The agency\\nwill complete an “appropriate investigation” and determine the\\n“general character and fitness” of the organizers, the\\n“economic advisability of establishing” the credit union, and\\nthe “conform[ity]” of proposal details with the Act. Id.\\n\\n The Act governs two types of federal credit unions:\\n“common-bond” credit unions and “community” credit unions.\\nSee id. § 1759(b). This case deals with the latter category. The\\n1934 version of the Act required a community credit union’s\\nmembership field to reflect a particular geographical area – to\\nwit, “a well-defined neighborhood, community, or rural\\ndistrict.” § 9, 48 Stat. at 1219. As amended in 1998, the Act\\nprovides that membership for a community credit union “shall\\n\\f 8\\nbe limited to . . . [p]ersons or organizations within a well-\\ndefined local community, neighborhood, or rural district.” 12\\nU.S.C. § 1759(b) (emphasis added). The 1998 version calls on\\nthe NCUA to “prescribe, by regulation, a definition for the term\\n‘well-defined local community, neighborhood, or rural\\ndistrict.’” Id. § 1759(g)(1). Thus, under the new regime,\\nindividuals seeking to organize a new community credit union\\n(or alter an existing one) must commit to serving members\\nwithin the NCUA’s contemporaneous definition of “local\\ncommunity, neighborhood, or rural district.” See S. REP. NO.\\n105-193, at 4, 8 (1998); H.R. REP. NO. 105-472, at 21 (1998).\\nAs part of their application to the NCUA, they must provide a\\nproposed description of the precise geographical area that the\\ncredit union would serve.\\n\\n Since 1998, there has been “dramatic growth” in the\\nnumber of community credit unions. U.S. GOV’T\\nACCOUNTABILITY OFF., GAO-07-29, CREDIT UNIONS:\\nGREATER TRANSPARENCY NEEDED ON WHO CREDIT UNIONS\\nSERVE AND ON SENIOR EXECUTIVE COMPENSATION\\nARRANGEMENTS 4 (2006). Despite a 11-percent drop in the\\nnumber of federal credit unions from 2000 to 2005, community\\ncredit unions doubled to 1,115. Id. at 4, 12. Meanwhile, the\\namount of assets in community credit unions quadrupled to\\n$104 billion. Id. at 4.\\n\\n C.\\n\\n On December 7, 2016, the NCUA amended its\\nmembership-field rules for community credit unions. See\\nChartering and Field of Membership Manual, 81 Fed. Reg.\\n88,412 (Dec. 7, 2016). Several changes rely on two terms\\ndevised by the Office of Management and Budget (OMB) and\\nbased on data collected by the Census Bureau (Census): “Core\\nBased Statistical Areas” and “Combined Statistical Areas.”\\n\\f 9\\n\\n The OMB has designated numerous regions around the\\ncountry as Core Based Statistical Areas, which comprise at\\nleast one urban cluster, or core, of 10,000 or more people and\\nadjacent counties with substantial commuting ties to that core.\\nSee U.S. CENSUS BUREAU, GEOGRAPHICAL PROGRAM,\\nGLOSSARY, https://www.census.gov/programs-surveys/geogra\\nphy/about/glossary.html. In layman’s terms, a Core Based\\nStatistical Area is a city or town and its suburbs.\\n\\n Meanwhile, a Combined Statistical Area is a conglomerate\\nof two or more adjoining Core Based Statistical Areas, each of\\nwhich has substantial commuting ties with at least one other\\nCore Based Statistical Area in the group. Id. Essentially, a\\nCombined Statistical Area is a regional hub with urban centers\\nconnected by commuting patterns. Combined Statistical Areas\\nmay “reflect broader social and economic interactions, such as\\nwholesaling, commodity distribution, and weekend recreation\\nactivities.” OFFICE OF MGMT. & BUDGET, EXEC. OFFICE OF THE\\nPRESIDENT, OMB BULL. NO. 15-01, REVISED DELINEATIONS OF\\nMETROPOLITAN STATISTICAL AREAS, MICROPOLITAN\\nSTATISTICAL AREAS, AND COMBINED STATISTICAL AREAS, AND\\nGUIDANCE ON USES OF THE DELINEATIONS OF THESE AREAS\\napp. 2-3 (2015).\\n\\n Relevant here, the 2016 rule made two changes to the\\nNCUA’s definition of the term “local community” under\\n§ 1759(b)(3) and one to that of “rural district.” The changes\\naffect what proposed membership areas satisfy the\\ngeographical limitation imposed by the Act.\\n\\n The first change to the “local community” definition\\ninvolves Combined Statistical Areas. A proposed area\\nqualifies as a local community if it encompasses the whole or\\na portion of a Combined Statistical Area and does not exceed a\\n\\f 10\\ndesignated population limit. See 81 Fed. Reg. at 88,440. The\\nNCUA has set that cap at 2.5 million people.\\n\\n The second change involves Core Based Statistical Areas.\\nThe parties agree that all or part of a Core Based Statistical\\nArea may qualify as a local community so long as it does not\\nexceed the population limit. But since 2010, the NCUA\\nrequired such a membership area to include the urban core.\\nThe new rule no longer requires that the core be included in the\\nlocal community that a credit union proposes to serve. See id.\\nat 88,413, 88,440.\\n\\n As for the “rural district” definition, the new rule increases\\nthe population cap for valid rural districts from 250,000 people\\n(or 3 percent of the population of the state where most eligible\\nresidents are located) to 1 million people. See id. at 88,416,\\n88,440. The new population limit works with two other\\nconstraints set by the rule: (1) an outer geographical limit on\\nhow far a rural district may extend past the borders of the credit\\nunion’s headquarters state; and (2) a requirement either that\\nmost eligible residents reside in Census-designated rural areas,\\nor that the population density of the proposed district equals\\n100 or fewer people per square mile. See id. at 88,440.\\n\\n D.\\n\\n On the day the NCUA published the rule, the Association\\nfiled this injunctive and declaratory action in the District Court.\\nThe Association claimed that the three changes described\\nabove were not only arbitrary and capricious under the\\nAdministrative Procedure Act (APA), Pub. L. No. 79-404, 60\\nStat. 237 (1946) (codified as 5 U.S.C. §§ 701-06), but also\\nunreasonable and entitled to no deference under Chevron. The\\nagency and Association filed cross-motions for summary\\n\\f 11\\njudgment. On March 29, 2018, the District Court granted both\\nmotions in part and denied them in part.\\n\\n The court made three relevant holdings. First, it rejected\\nas unreasonable the qualification of certain Combined\\nStatistical Areas as local communities. See Am. Bankers Ass’n,\\n306 F. Supp. 3d at 61. Second, it sustained as well-reasoned\\nthe elimination of the core requirement from the Core Based\\nStatistical Area a credit union proposes to serve as its local\\ncommunity. Id. at 64-65. Third, it rejected as unreasonable the\\nincreased population cap for rural districts. Id. at 69-70. (It\\nalso sustained a separate portion of the rule, which the\\nAssociation does not challenge here.)\\n\\n The NCUA and Association timely appealed. We have\\nappellate jurisdiction under 28 U.S.C. § 1291.\\n\\n II.\\n\\n At the outset, we must assure ourselves of our subject\\nmatter jurisdiction over the appellate proceeding. See, e.g.,\\nUnited States v. Gooch, 842 F.3d 1274, 1277 (D.C. Cir. 2016).\\n\\n In their original briefing, the parties failed to apprise us of\\na rule that was promulgated while this appeal was pending and\\nthat changed membership-field requirements for community\\ncredit unions. See Chartering and Field of Membership, 83\\nFed. Reg. 30,289 (June 28, 2018). The 2018 rule eliminated\\nthe portion of the 2016 rule allowing Combined Statistical\\nAreas to qualify as local communities. Compare 12 C.F.R. pt.\\n701, app. B, ch. 2 § V.A.2 (2018), with id. (2019). The 2018\\nrule preamble did not specifically discuss the removal but\\nconcluded that “[a]ny modification in th[e] final rule is\\nconsistent with the District Court decision” in this action. 83\\nFed. Reg. at 30,291.\\n\\f 12\\n\\n “Under the mootness doctrine, we cannot decide a case if\\n‘events have so transpired that the decision will neither\\npresently affect the parties’ rights nor have a more-than-\\nspeculative chance of affecting them in the future.’” Reid v.\\nHurwitz, 920 F.3d 828, 832 (D.C. Cir. 2019) (citation omitted).\\nThe same principle applies to individual claims. See Tucson\\nMed. Ctr. v. Sullivan, 947 F.2d 971, 977 (D.C. Cir. 1991).\\nAccordingly, if a rule under review (or a portion of it) is\\nsuperseded or amended during an appeal, the proceeding (or\\nrelevant part) might be moot. See, e.g., Am. Bankers Ass’n v.\\nNCUA, 271 F.3d 262, 274 (D.C. Cir. 2001). We therefore\\nrequested supplemental briefing on the issue of appellate\\njurisdiction.\\n\\n Based on the government’s submission and\\nrepresentations at oral argument, we hold that the portion of the\\nappeal related to Combined Statistical Areas is not moot.\\n“[T]he mere power to [reinstitute] a challenged law is not a\\nsufficient basis on which a court can conclude that” a challenge\\nremains live. Nat’l Black Police Ass’n v. District of Columbia,\\n108 F.3d 346, 349 (D.C. Cir. 1997). The government has stated\\nthat, if we reverse the relevant part of the District Court’s\\ndecision, all three members of the NCUA’s board intend to\\nreinstitute the Combined Statistical Area portion of the 2016\\nrule. See Decl. of Michael McKenna ¶ 3, ECF No. 1781123;\\nOral Arg. Recording 11:33-48. Accordingly, City of Mesquite\\nv. Aladdin’s Castle, Inc., 455 U.S. 283 (1982), governs this\\ncase. In Aladdin’s Castle, the Supreme Court deemed the\\ncontroversy live in part because the government had announced\\n“an intention” to restore the rule under challenge if the lower-\\ncourt decision were vacated. Id. at 289 & n.11; see also Am.\\nBankers Ass’n, 271 F.3d at 274. The NCUA’s submission and\\nrepresentations evince such an intention here, and the\\nAssociation – which bears the “heavy burden” of proving\\n\\f 13\\nmootness, Friends of the Earth, Inc. v. Laidlaw Envtl. Servs.\\n(TOC), Inc., 528 U.S. 167, 189 (2000) (citation\\nomitted) – offers no evidence to the contrary.\\n\\n The Association attempts to distinguish Aladdin’s Castle\\non two grounds, but neither sways us. First, the Association\\nnotes that the city government in Aladdin’s Castle said it would\\nreenact “precisely the same” law, see 455 U.S. at 289, but that\\nin this case any future notice-and-comment proceedings might\\nproduce a different “local community” definition, perhaps not\\neven relying on Combined Statistical Areas. After all, the\\nagency must keep a “flexible and open-minded attitude” during\\nthe process. See Nat’l Tour Brokers Ass’n v. United States, 591\\nF.2d 896, 902 (D.C. Cir. 1978). We grant that commenters\\nmight convince the NCUA to change its mind; mann tracht,\\nund Gott lacht. But that strikes us as speculative as public input\\nconvincing Mesquite legislators to enact a different law. We\\ndo not see Aladdin’s Castle turning on such conjecture.\\nInstead, we see a live dispute because there is “no certainty”\\nthat the NCUA will forego reinstating the same Combined\\nStatistical Area definition. Aladdin’s Castle, 455 U.S. at 289;\\nsee also Trinity Lutheran Church of Columbia, Inc. v. Comer,\\n137 S. Ct. 2012, 2019 n.1 (2017) (holding that the court must\\nfind “absolutely clear that the allegedly wrongful behavior\\ncould not reasonably be expected to recur” (emphasis added)\\n(quoting Friends of the Earth, 528 U.S. at 189)). After all, the\\nNCUA continues to defend the definition here. See Knox v.\\nServ. Emps. Int’l Union, Local 100, 567 U.S. 298, 307 (2012).\\n\\n Second, the Association observes that both sides in\\nAladdin’s Castle urged the Supreme Court to treat their dispute\\nas live. In contrast, only the NCUA seeks to proceed here; the\\nAssociation would prefer to wait until the agency reinstitutes\\nthe rule. But the existence or absence of jurisdiction does not\\nturn on which parties challenge or defend it. Cf. Bender v.\\n\\f 14\\nWilliamsport Area Sch. Dist., 475 U.S. 534, 541 (1986).\\nBecause the NCUA remains bound by the lower-court\\njudgment, the present injury renders irrelevant the\\nAssociation’s preference. “Jurisdiction existing,” our duty to\\ndecide the appeal “is ‘virtually unflagging.’” Sprint\\nCommc’ns, Inc. v. Jacobs, 571 U.S. 69, 77 (2013) (citation\\nomitted).\\n\\n In short, we may review the challenge to the rule change\\ninvolving Combined Statistical Areas. We see no jurisdictional\\nissues with the rest of the appeal. We thus turn to the merits.\\n\\n III.\\n\\n We review de novo the District Court’s rulings on\\nsummary judgment. See Loan Syndications & Trading Ass’n\\nv. SEC, 882 F.3d 220, 222 (D.C. Cir. 2018). We review the\\nadministrative record and give “no particular deference” to the\\nDistrict Court’s views. Oceana, Inc. v. Ross, 920 F.3d 855, 860\\n(D.C. Cir. 2019) (citation omitted).\\n\\n The APA governs this suit. In relevant part, the statute\\nprovides that we “decide all relevant questions of law” and\\n“interpret . . . statutory provisions.” 5 U.S.C. § 706. We\\nordinarily set aside agency actions that are either “arbitrary,\\ncapricious, an abuse of discretion, or otherwise not in\\naccordance with law,” id. § 706(2)(A), or “in excess of\\nstatutory jurisdiction, authority, or limitations, or short of\\nstatutory right,” id. § 706(2)(C).\\n\\n We review the agency rule in accordance with the familiar\\nChevron doctrine, a two-prong test for determining whether an\\nagency “has stayed within the bounds of its statutory authority”\\nwhen issuing its action. City of Arlington v. FCC, 569 U.S.\\n290, 297 (2013) (emphasis omitted). At the first step, we\\n\\f 15\\ndetermine “whether Congress has directly spoken to the precise\\nquestion at issue,” and we “give effect” to any “unambiguously\\nexpressed intent.” Chevron, 467 U.S. at 842-43 & n.9.\\n\\n If we glean no such unambiguous intent, we turn to the\\nsecond step and determine “whether the agency’s answer” to\\nthe question “is based on a permissible construction of the\\nstatute.” Id. at 843. By arriving at the second step, we have\\nconcluded that Congress either explicitly or implicitly\\ndelegated to the agency the lawmaking authority to clarify the\\nstatute. We presume that Congress would not authorize the\\npromulgation of an “[im]permissible construction.” Chevron,\\n467 U.S. at 843. Accordingly, we will set aside agency actions\\nbased on such a construction, because they are either “not in\\naccordance with law” or “in excess of statutory jurisdiction.”\\n5 U.S.C. § 706(2)(A), (C); see also CSX Transp., Inc. v.\\nSurface Transp. Bd., 754 F.3d 1056, 1063 (D.C. Cir. 2014);\\nAss’n of Privacy Sector Colls. & Univs. v. Duncan, 681 F.3d\\n427, 441 (D.C. Cir. 2012).\\n\\n IV.\\n\\n For all three challenges, the first step of the Chevron\\nanalysis proceeds in the same way. “We begin our analysis, as\\nalways, with the statutory text.” Tesoro Alaska Co. v. FERC,\\n778 F.3d 1034, 1038 (D.C. Cir. 2015). Congress having\\nexpressly assigned the NCUA the power to define the\\nchallenged terms, see 12 U.S.C. § 1759(g)(1), we may proceed\\nto Chevron’s second prong without further analysis, see U.S.\\nTelecom Ass’n v. FCC, 825 F.3d 674, 717 (D.C. Cir. 2016);\\nRush Univ. Med. Ctr. v. Burwell, 763 F.3d 754, 760 (7th Cir.\\n2014); Comm’r v. Pepsi-Cola Niagara Bottling Corp., 399\\nF.2d 390, 393 (2d Cir. 1968) (Friendly, J.) (“When Congress\\nhas used a general term and has empowered an administrator\\n\\f 16\\nto define it, the courts must respect his construction if this is\\nwithin the range of reason.”).\\n\\n An express delegation of definitional power “necessarily\\nsuggests that Congress did not intend the [terms] to be applied\\nin [their] plain meaning sense,” Women Involved in Farm\\nEcon. v. U.S. Dep’t of Agric., 876 F.2d 994, 1000 (D.C. Cir.\\n1989), that they are not “self-defining,” id., and that the agency\\n“enjoy[s] broad discretion” in how to define them, Lindeen v.\\nSEC, 825 F.3d 646, 653 (D.C. Cir. 2016). In Chevron terms,\\nCongress through explicit language “has directly spoken to the\\nprecise question” of whether the identified terms must carry\\ncertain meanings. 467 U.S. at 842-43. The answer is no. See\\nBuongiorno v. Sullivan, 912 F.2d 504, 509 (D.C. Cir. 1990)\\n(Thomas, J.) (“When Congress expressly delegates the\\nauthority to fill a gap in a statute, Congress speaks, in effect,\\ndirectly, and says, succinctly, that it wants the agency to\\nannotate its words.”). To hold otherwise at the first Chevron\\nstep would “undermine” the ability of Congress to delegate\\ndefinitional power. Rush Univ. Med. Ctr., 763 F.3d at 760.\\n\\n Consequently, we turn to whether the NCUA’s definitions\\nare “based on a permissible construction of the statute.”\\nChevron, 467 U.S. at 843.\\n\\n V.\\n\\n Agency interpretations promulgated to fill an explicit\\nlegislative gap “are given controlling weight unless they are\\narbitrary, capricious, or manifestly contrary to the statute.”\\nChevron, 467 U.S. at 844; see also Pharm. Research & Mfrs.\\nof Am. v. FTC, 790 F.3d 198, 204 (D.C. Cir. 2015).\\n\\n Under arbitrary and capricious review, “we may not\\nsubstitute our own judgment for that” of the agency. FERC v.\\n\\f 17\\nElec. Power Supply Ass’n, 136 S. Ct. 760, 782 (2016); accord\\nDep’t of Commerce v. New York, 139 S. Ct. 2551, 2569 (2019).\\nStill, we are “not a ‘rubber stamp,’” Oceana, 920 F.3d at 863;\\n“the agency must examine the relevant data and articulate a\\nsatisfactory explanation for its action,” Motor Vehicle Mfrs.\\nAss’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co. (State\\nFarm), 463 U.S. 29, 43 (1983). A rule is arbitrary and\\ncapricious if (1) the agency “has relied on factors which\\nCongress has not intended it to consider”; (2) the agency\\n“entirely failed to consider an important aspect of the\\nproblem”; (3) the agency’s explanation “runs counter to the\\nevidence before the agency”; or (4) the explanation “is so\\nimplausible that it could not be ascribed to a difference in view\\nor the product of agency expertise.” Id.\\n\\n In turn, we assess three definitional changes in the 2016\\nrule: (1) qualifying Combined Statistical Areas as local\\ncommunities; (2) eliminating the core requirement for local\\ncommunities based on Core Based Statistical Areas; and\\n(3) raising the population cap for rural districts. We sustain the\\nfirst and third amendments in full. As for the second, we hold\\nthat it is rationally related to the Act’s text and purposes, but\\nthat it is insufficiently explained.\\n\\n A.\\n\\n The District Court rejected the first change because it\\napproved certain Combined Statistical Areas “no matter how\\ngeographically dispersed and unconnected” the “members may\\nbe.” Am. Bankers Ass’n, 306 F. Supp. 3d at 61. To the court,\\nthe approval did not fit with the term “local community,”\\nwhich, in its view, “encompass[es] an area no larger than a\\ncounty.” Id. at 58, 61. We respectfully disagree.\\n\\f 18\\n The NCUA possesses vast discretion to define terms\\nbecause Congress expressly has given it such power. But the\\nauthority is not boundless. The agency must craft a reasonable\\ndefinition consistent with the Act’s text and purposes; that is\\ncentral to the review we apply at Chevron’s second step. Here,\\nthe NCUA’s definition meets the standard.\\n\\n We first focus on the text. Congress introduced the phrase\\n“local community” in the 1998 amendments. The word\\n“community” had a broad scope at the time. It meant not only\\n“society at large” but also a “body of individuals organized into\\na unit or manifesting usu[ally] with awareness of some\\nunifying trait.” See Community, WEBSTER’S THIRD\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE\\nUNABRIDGED 460 (1993). The group could be “united by\\nhistorical consciousness or . . . common social, economic, and\\npolitical interests.” Id. But the unifying trait could also be\\nsimply “living in a particular place or region.” Id.\\n\\n The NCUA has recognized that the modifier “local”\\n“reflects congressional intent that it takes ‘a more circumspect\\nand restricted approach to chartering community credit\\nunions.’” Am. Bankers Ass’n, 271 F.3d at 273 (citation\\nomitted); see also Oral Arg. Recording 3:16-18. Indeed,\\nCongress made clear its intention to “modif[y]” the “current\\nlaw regarding community credit unions” by adding the word\\n“local” to community. S. REP. NO. 105-193, at 6-7.\\n\\n Insertion of the modifier “local” before “community”\\nimplies that the community “relate[s] to” a “particular limited\\ndistrict” or is “confined to a particular place.” Local,\\nWEBSTER’S THIRD INTERNATIONAL DICTIONARY OF THE\\nENGLISH LANGUAGE UNABRIDGED 1327 (1993). But that place\\nneed not be the size of a county, as the District Court held. The\\nSupreme Court has recognized that the geographical areas\\n\\f 19\\n“need not be small.” First Nat’l Bank III, 522 U.S. at 492. And\\nif Congress wanted a local community to correspond to a\\nparticular geographical unit, such as a county, “‘it easily could\\nhave written’ that limitation explicitly.” NCUA Br. 25-26\\n(quoting Ali v. Fed. Bureau of Prisons, 552 U.S. 214, 227\\n(2008)); see also Credit Union Nat’l Ass’n Amicus Br. 15. The\\nNCUA sensibly reads the term “local” to mean simply that the\\ncommunity, regardless of shape or size, should be neither\\n“broad” nor “general.” Local, SUPRA, at 1327.\\n\\n To be clear, we do not hold today that the NCUA must\\nconsider only bonds and social connections as understood in\\n1998. The parties agree, and thus we assume, that the NCUA,\\ndespite its expressly delegated authority, must adopt a\\ndefinition consistent with what the term “local community”\\nmeant in 1998, the time of its adoption.\\n\\n After consulting state statutes and invoking the canon of\\nnoscitur a sociis, the District Court developed a rather size-\\nrestrictive meaning for the phrase. See Am. Bankers Ass’n, 306\\nF. Supp. 3d at 57-59. But we do not think that defining a “local\\ncommunity” to refer to an area larger than a county is an\\nunreasonable interpretation of the Act’s text.\\n\\n We receive little guidance from the state statutes in effect\\nin 1998. Indeed, some of the statutes considered “local\\ncommunities” to be quite large. See, e.g., ALASKA STAT.\\n§ 18.66.990(7) (1998) (defining “local community entity” as “a\\ncity or borough or other political subdivision of the state, a\\nnonprofit organization, or a combination of these” (emphasis\\nadded)); see also NCUA Br. 25 n.4.\\n\\n We also reject the District Court’s invocation of the\\nnoscitur a sociis canon. When several terms “are associated in\\na context suggesting that [they] have something in common,\\n\\f 20\\nthey should be assigned a permissible meaning that makes\\nthem similar.” ANTONIN SCALIA & BRYAN A. GARNER,\\nREADING LAW: THE INTERPRETATION OF LEGAL TEXTS 195\\n(2012). But the “substantive connection, or fit, between” the\\nterms here – local community, neighborhood, and rural\\ndistrict – is “not so tight or so self-evident.” Graham Cty. Soil\\n& Water Conservation Dist. v. United States ex rel. Wilson, 559\\nU.S. 280, 288 (2010). Only the word “neighborhood” has a\\ndictionary definition clearly suggesting a region of small size;\\nthe phrase “local community” does not, as we have explained\\nabove, and neither does “rural district,” as we explain below.\\nA size restriction would impermissibly “submerge[]” the\\nindependent “character” of the latter two terms. Babbitt v.\\nSweet Home Chapter of Cmtys. for a Great Or. (Sweet Home),\\n515 U.S. 687, 702 (1995) (citation omitted).\\n\\n The Association also points to other textual indicators.\\nBut contrary to what it suggests, see Am. Bankers Ass’n Br.\\n27-28, we see nothing in the record suggesting that “local\\ncommunity” is a term of art. The Association also says the\\nusage of word “local” in two other federal statutes indicates\\nthat rules permitting coverage areas of larger than a county\\nwould be manifestly contrary to the Act. See id. at 32; see also\\n15 U.S.C. § 2203 (1998) (defining “local” as “city, town,\\ncounty, . . . or other political subdivision of a State”); 18 U.S.C.\\n§ 666 (1998) (same). (The Association pointed to other laws,\\nbut they did not exist in 1998. See National Defense\\nAuthorization Act for Fiscal Year 2010, Pub. L. No. 111-84,\\n§ 4703(b), 123 Stat. 2190, 2837 (2009); PRIME Act Grants, 66\\nFed. Reg. 29,010 (May 29, 2001).) True enough. But the two\\nstatutes address materially distinct issues – fire prevention and\\nembezzlement – and thus do not indicate that the term “local”\\nmust imply a size limitation here. See Envtl. Def. v. Duke\\nEnergy Corp., 549 U.S. 561, 574 (2007) (“[M]ost words have\\n\\f 21\\ndifferent shades of meaning and consequently may be variously\\nconstrued . . . in different statutes . . . .” (citation omitted)).\\n\\n In addition to being consistent with the Act’s text, the\\nCombined Statistical Area definition rationally advances the\\nAct’s underlying purposes. In the 1998 amendments, Congress\\nmade two relevant findings about purpose. First, legislators\\nfound “essential” to the credit-union system a “meaningful\\naffinity and bond among members, manifested by a\\ncommonality of routine interaction[;] shared and related work\\nexperiences, interests, or activities[;] or the maintenance of an\\notherwise well-understood sense of cohesion or identity.” § 2,\\n112 Stat. at 914. Second, Congress highlighted the importance\\nof “credit union safety and soundness,” because a credit union\\non firm financial footing “will enhance the public benefit that\\ncitizens receive.” Id. The legislative history also confirms the\\nimportance of common bonds, see S. REP. NO. 73-555, at 2; see\\nalso H.R. REP. NO. 105-472, at 12; H.R. REP. NO. 73-2021, at\\n2; 144 CONG. REC. S9094 (daily ed. July 28, 1998) (statement\\nof Sen. Mikulski); id. at S8971-72 (daily ed. July 24, 1998)\\n(statement of Sen. Moseley-Braun), and economic “integrity,”\\nS. REP. NO. 105-193, at 3; see also 144 CONG. REC. S9094\\n(statement of Sen. Mikulski); id. at S8972 (statement of Sen.\\nMoseley-Braun), to federal credit unions.\\n\\n We recognize that there may be some tension between the\\nAct’s principal purposes: A credit union with exceedingly\\nclose ties among its members is unlikely to have a large enough\\ncustomer base to thrive economically. To the extent that such\\ntension exists, the Act leaves to the NCUA to strike a\\nreasonable balance. Congress was well aware that a viable\\ncredit union might serve a relatively large geographical area.\\nSee, e.g., H.R. REP. NO. 73-2021, at 2 (“[T]here are cases in\\nwhich communities and organizations cross State lines.”).\\n\\f 22\\n The NCUA did just that in promulgating the Combined\\nStatistical Area definition. That definition allows for larger\\ncommunity credit unions; the decision is consistent with\\ndecades of history promoting the economic viability of credit\\nunions in the face of banks and other competing financial\\ninstitutions. Nonetheless, the NCUA struck a balance by\\nensuring that members within the local community maintain\\nsomewhat of a commuter relationship with each other. As the\\nAssociation even acknowledges, commuting patterns “may\\nsometimes serve as a proxy for community interaction.” Am.\\nBankers Ass’n Br. 38 n.25. We see nothing irrational about\\nadopting the factor as a proxy for the common bond\\ncontemplated by Congress. Perhaps we would have made a\\ndifferent call had we been the policymakers. Perhaps we would\\nhave sought a tighter bond. Or perhaps we would not have\\nprioritized credit-union growth. See Iowa Bankers Ass’n\\nAmicus Br. 24-25. But we must “refrain from substituting\\n[our] own interstitial lawmaking for that” of the agency.\\nHousehold Credit Servs., Inc. v. Pfennig, 541 U.S. 232, 244\\n(2004) (citation omitted).\\n\\n The NCUA also reasonably explained its amendment to\\nthe “local community” definition. To begin with, the agency\\nreasonably circumscribed the size of a local community under\\nthe Combined Statistical Area rule by imposing a 2.5 million-\\nperson population limit. Used by the OMB in analogous\\ncircumstances, the cap is a “logical breaking point in terms of\\ncommunity cohesiveness with respect to a multijurisdictional\\narea.” Chartering and Field of Membership for Federal Credit\\nUnions, 75 Fed. Reg. 36,257, 36,259 (June 25, 2010).\\nMoreover, the NCUA reasonably relied on its prior experience\\nwith Core Based Statistical Areas, which no party disputes can\\nserve as local communities under the Act. The agency noted\\nthat the average geographic size of Combined Statistical Areas\\nwith populations of up to 2.5 million people is 4,553 square\\n\\f 23\\nmiles, and that the average size of the Core Based Statistical\\nAreas approved by the NCUA as local communities is 4,572\\nsquare miles. See 81 Fed. Reg. at 88,414-15. Given that\\nsimilarity in size, the NCUA reasonably considered adopted its\\npopulation-limited Combined Statistical Areas standard.\\n\\n Finally, the NCUA’s definition does not readily create\\ngeneral, widely dispersed regions. Cf. First Nat’l Bank III, 522\\nU.S. at 502 (indicating that community credit unions may not\\nbe “composed of members from an unlimited number of\\nunrelated geographical units”). Combined Statistical Areas are\\ngeographical units well-accepted within the government. See\\n81 Fed. Reg. at 88,414. Because they essentially are regional\\nhubs, the Combined Statistical Areas concentrate around\\ncentral locations. The OMB carved out the areas so that their\\nconstituent parts share commuting ties and they reflect\\n“broader social and economic interactions.” OFFICE OF MGMT.\\n& BUDGET, SUPRA, at app. 2-3. The NCUA rationally believed\\nthat such “real-world interconnections” would qualify as the\\ntype of mutual bonds suggested by the term “local\\ncommunity.” Credit Union Nat’l Ass’n Amicus Br. 9. Thus,\\nthe agency reasonably determined that Combined Statistical\\nAreas “simply unif[y], as a single community,” already\\nconnected neighboring regions. See 81 Fed. Reg. at 88,415.\\n\\n The Association raises a potpourri of objections to the\\nNCUA’s decision-making. See Am. Bankers Ass’n Br. 33-48.\\nVirtually all of its gripes are forfeited because it failed first to\\nraise them to the agency, see, e.g., Koretoff v. Vilsack, 707 F.3d\\n394, 397-98 (D.C. Cir. 2013) (per curiam), or lack merit\\nbecause they involve outdated “local community” definitions,\\nwhich either did not allow for the qualification of certain\\ngeographical areas as local communities or lacked a population\\ncap of 2.5 million people.\\n\\f 24\\n But one of the complaints is “deserving of sustained\\nconsideration.” Ortiz v. United States, 138 S. Ct. 2165, 2173\\n(2018). The Association contends that some large Combined\\nStatistical Areas are so sprawling that the NCUA’s definition,\\nwhich treats a 2.5-million-person portion of them as a local\\ncommunity, must be unreasonable. Recall that a Combined\\nStatistical Area is a regional hub with more than one urban\\ncenter. Under the OMB’s technical specifications, each center\\nneed not be connected to every other by commuting ties; rather,\\neach need only have ties to one other center within the hub. As\\nthe Association colorfully puts it, the OMB may designate as a\\nCombined Statistical Area a mere “daisy chain” of urban\\ncenters “that are linked to their neighbors but have nothing to\\nwith those at the other end of the chain.” See Am. Bankers\\nAss’n Br. 36 (citation omitted). Theoretically, the Association\\ncontinues, certain 2.5-million-person communities might bring\\ntogether parts of different urban centers with no connection\\nwith one another. The Association also suggests that the rule\\nmight permit local communities comprising non-contiguous\\nportions of a Combined Statistical Area. See id. at 39.\\n\\n We understand the Association’s argument to be attacking\\nthe Combined Statistical Area definition as unreasonable. To\\nthe Association, the NCUA failed sufficiently to consider the\\npotential for the rule to create unreasonable results. One\\nhypothetical application disturbed the District Court here: the\\nprospect that, within the Combined Statistical Area including\\nthe District of Columbia and counties from three states\\n(Maryland, Pennsylvania, and Virginia), one could create a\\nlocal community bringing together Doylesburg, Pennsylvania,\\nand Partlow, Virginia, towns located 200 miles apart. See Am.\\nBankers Ass’n, 306 F. Supp. 3d at 59-60.\\n\\n We might well agree with the District Court that the\\napproval of such a geographical area would contravene the Act.\\n\\f 25\\nBut even so, the Association would need much more to mount\\nits facial pre-enforcement challenge in this case. As the\\nSupreme Court repeatedly has held, “the fact that petitioner can\\npoint to a hypothetical case in which the rule might lead to an\\narbitrary result does not render the rule” facially invalid. Am.\\nHosp. Ass’n v. NLRB, 499 U.S. 606, 619 (1991); see also EPA\\nv. EME Homer City Generation, L.P. (EME Homer), 572 U.S.\\n489, 524 (2014) (“The possibility that the rule, in uncommon\\nparticular applications, might exceed [the agency]’s statutory\\nauthority does not warrant judicial condemnation of the rule in\\nits entirety.”); INS v. Nat’l Ctr. for Immigrants’ Rights, Inc.,\\n502 U.S. 183, 188 (1991) (“That the regulation may be invalid\\nas applied in s[ome] cases . . . does not mean that the regulation\\nis facially invalid because it is without statutory authority.”);\\ncf. Barnhart v. Thomas, 540 U.S. 20, 29 (2003) (“Virtually\\nevery legal (or other) rule has imperfect applications in\\nparticular circumstances.”).\\n\\n Here, the Association’s complaint and the District Court’s\\naccompanying worry strike us as too conjectural. The NCUA\\nmust assess the “economic advisability of establishing” the\\nproposed credit union before approving it, 12 U.S.C. § 1754,\\nand as part of the assessment, the organizers must propose a\\n“realistic” business plan showing how the institution and its\\nbranches would serve all members in the local community, see\\n12 C.F.R. pt. 701, app. B, ch. 1 § IV.D. The Association has\\nfailed to demonstrate the plausibility of a local community that\\nis defined like the hypothetical narrow, multi-state strip and\\naccompanies a realistic business plan. And if the agency were\\nto receive and approve such an application, a petitioner can\\nmake an as-applied challenge. See, e.g., EME Homer, 572 U.S.\\nat 523-24; Buongiorno, 912 F.2d at 510. The Association has\\nsucceeded in such challenges in the past. See, e.g., Am.\\nBankers Ass’n v. NCUA, No. 1:05-CV-2247, 2008 WL\\n2857678, at *1 (M.D. Pa. July 21, 2008); Am. Bankers Ass’n v.\\n\\f 26\\nNCUA, 347 F. Supp. 2d 1061, 1074 (D. Utah 2004). Thus, we\\nreject the facial attack on the amended “local community”\\ndefinition involving Combined Statistical Areas.\\n\\n B.\\n\\n We turn to the next rule change. The District Court upheld\\nthe eliminated core requirement for a local community based\\non a Core Based Statistical Area. The court acknowledged that\\ndefining a local community without its urban core “does not\\nalter the . . . common bond” shared by the members in the\\nremainder. Am. Bankers Ass’n, 306 F. Supp. 3d at 62-63.\\nMeanwhile, the court accepted the agency’s response to the\\nclaim that the new definition encouraged redlining – a broad\\ncategory of lending practices with negative impact on “areas\\nwhere low-income and minority populations are concentrated.”\\nId. at 65 (quoting 81 Fed. Reg. at 88,413).\\n\\n Like the District Court, we hold that the eliminated core\\nrequirement is consistent with the Act’s text and purposes.\\nStill, we see merit in the Association’s redlining argument and\\nthus hold the definitional change to be arbitrary and capricious.\\n\\n 1.\\n\\n We will sustain the eliminated core requirement if it\\nreflects a reasonable interpretation of “local community” that\\nis rationally related to the dual purposes of promoting credit-\\nunion growth and ensuring some cohesion among members. It\\ndoes. Omission of the urban core from proposed geographical\\narea will permit community credit unions to reach new\\nmembers in the suburban parts of the Core Based Statistical\\nArea and thus to maintain a healthy membership. Because the\\nsuburbs under the OMB’s definition have substantial\\ncommuting ties to the urban cluster, they all will be “within a\\n\\f 27\\nfeasible commuting radius” and thus “share[] at least some\\ngeographic ties.” Am. Bankers Ass’n, 306 F. Supp. 3d at 64.\\nThose bonds do not disappear if the local community lacks the\\ncore. The Association seeks greater “assurance[s]” of a\\nmeaningful bond among members, Am. Bankers Ass’n Br. 66-\\n67; see also Am. Bankers Ass’n Reply 11, but we do not\\nreplace the agency’s policy judgment with ours.\\n\\n We also do not believe that the eliminated core\\nrequirement would create sprawling, boundless geographical\\nregions. No one disputes that, as a general matter, a\\nmembership area comprising an intact Core Based Statistical\\nArea will satisfy any definition of “local community.” If the\\nlocal community with the core poses no problem, we fail to see\\nhow a local community without one would. And even as to\\nCore Based Statistical Areas that do not qualify as local\\ncommunities (because they have populations of more than 2.5\\nmillion), the geographical ties ensure that the proposed\\nmembership area will still be contained within the boundaries\\nof a single, well-recognized metropolitan region. A single\\nregion is not what concerned the Supreme Court in First\\nNational Bank III. See 522 U.S. at 502.\\n\\n The Association objects to the expansive hypothetical\\nmembership fields, highlighting two Core Based Statistical\\nAreas whose populations exceed the NCUA’s cap of 2.5\\nmillion. The Association asserts that the rule change “makes it\\nmuch easier to unite far-distant edges” of those sprawling areas\\nin a single membership area. Am. Bankers Ass’n Reply 9. Fair\\npoint. But economic realities do not make it plausible that\\norganizers would propose such a local community or that the\\nNCUA would approve it. Like the Combined Statistical Area\\ndefinition, the eliminated core requirement does not become\\nfacially infirm because of farfetched hypotheticals. To the\\nextent they occur in the future, troublesome rule applications\\n\\f 28\\nmight be subject to as-applied challenges. See Am. Bankers\\nAss’n, 271 F.3d at 267.\\n\\n 2\\n\\n Despite the eliminated core requirement’s consistency\\nwith the Act, we cannot sustain the definitional change because\\nthe NCUA has not adequately explained it.\\n\\n The Association contends that the rule change “effectively\\nallows credit unions to engage in ‘redlining’ by denying service\\nto urban areas with large numbers of minority and lower-\\nincome residents.” Am. Bankers Ass’n Br. 65. The NCUA\\nattempted to address this concern in the preamble. At first\\nblush, the agency’s statements appear persuasive. Still, the\\nAssociation persuasively argues that the response fails to\\n“consider an important aspect” of the redlining issue or is\\notherwise “so implausible” as to be unreasonable. State Farm,\\n463 U.S. at 43. Thus, the eliminated core requirement is\\narbitrary and capricious.\\n\\n During the notice-and-comment proceedings, the\\nAssociation warned against redlining and objected that\\ncommunity credit unions could now “serv[e] wealthier\\nsuburban counties and exclud[e] markets containing low-\\nincome and minority communities that reside in the core area.”\\nLetter from James Chessen, Exec. Vice President & Chief\\nEconomist, Am. Bankers Ass’n, to Gerard S. Poliquin, Sec’y\\nof the Board, Nat’l Credit Union Admin. (Feb 5, 2016), Am.\\nBankers Ass’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C.\\nfiled Aug. 23, 2017), ECF No. 26-1 at 228. Fairly read, the\\nAssociation’s objection is not to traditional redlining, which\\nencompasses the refusal to make loans in low-income or\\nminority neighborhoods within a service area, see Baradaran,\\nsupra, at 494; see also Cassity, supra, at 348, 355. Federal\\n\\f 29\\ncredit unions “cannot ‘redline’” in the traditional sense because\\n“everyone in a credit union’s ‘community’ is a member who is\\neligible to take advantage of credit union services” and credit\\nunions “by definition cannot reach beyond their member\\ncommunities to offer credit to the general public.” Cassity,\\nsupra, at 355. But a community credit union can engage in\\nmore unconventional redlining practices: “gerrymander[ing] to\\ncreate its own community of exclusively higher-income\\nmembers.” Id. at 359. We think it evident that the Association\\nwas focusing on such gerrymandering.\\n\\n In response, the NCUA acknowledged that it originally\\nkept the core requirement as a benefit to “low-income and\\nunderserved populations.” 81 Fed. Reg. at 88,413. Some\\ncommenters criticized the core requirement as failing to\\nachieve that goal; it caused community credit unions “to\\nsacrifice service to other areas” within the Core Based\\nStatistical Area, id., and such service arguably could benefit\\npoor and minority customers residing outside the core, see id.\\nat 88,414 (remarking on the importance of credit unions\\n“providing financial services to low income and underserved\\npopulations without regard to where they are located within a\\ncommunity, i.e., beyond its ‘core area’”).\\n\\n Still, the agency dismissed the redlining concern on other\\ngrounds, pointing to its “supervisory process to assess [credit-\\nunion] management’s efforts to offer service to the entire\\ncommunity [the credit union] seeks to serve.” Id. The NCUA\\nfocused on two aspects of its process. First, the agency touted\\nits “annual evaluation,” which “encompasses [the credit\\nunion]’s implementation of its business and marketing plans.”\\nId. Citing its “[e]xperience” as support, the agency identified\\nthe evaluation as a “more effective means” than a core\\nrequirement to “ensur[e] that the low-income and underserved\\npopulations are fairly served.” Id. Indeed, prior evaluations\\n\\f 30\\nconfirm that the agency has had “success in providing financial\\nservices to low-income and underserved populations without\\nregard to where they are located within” the membership area.\\nId. at 88,414. Second, the NCUA noted its “mandate to\\nconsider member complaints alleging discriminatory practices\\naffecting low-income and underserved populations, such as\\nredlining, and to respond as necessary when such practices are\\nshown to exist.” Id.\\n\\n Both aspects of the NCUA’s supervisory process fail to\\naddress the redlining issue raised by the Association. The\\nannual evaluation process might be an adequate response to\\ntraditional redlining, because it might ensure that community\\ncredit unions adequately serve poor and minority residents\\nliving in their local communities. But we do not see how it\\nfixes gerrymandering or the potential discriminatory economic\\nimpact on urban residents. See State Farm, 463 U.S. at 43\\n(requiring the agency to consider relevant factors). The annual\\nevaluation process necessarily does not come into effect until\\nthe NCUA already has approved the charter, business plan, and\\nproposed local community. See Iowa Bankers Ass’n Amicus\\nBr. 12. And nothing in the record suggests that business plans\\nmay focus on residents residing outside the finalized\\nmembership area; in fact, the law forbids federal credit unions\\nfrom serving those residents.\\n\\n The complaint process also does not work in the\\ngerrymandering context. As the preamble points out,\\ncomplaints are raised by the membership, which would not\\ninclude the affected urban residents because of the rule change.\\nIt seems quite implausible, absent some contrary evidence the\\nagency failed to detail, that members will file grievances based\\non gerrymandering harms suffered by residents outside the\\ncoverage area. See State Farm, 463 U.S. at 43 (rejecting\\nimplausible explanations).\\n\\f 31\\n\\n The NCUA attempts to defend the rule change by offering\\na buffet of other potential rationales in its briefing and at oral\\nargument. They all fail because the agency did not adopt them\\nwhen promulgating the rule change. See State Farm, 463 U.S.\\nat 43 (noting that courts may not “supply a reasoned basis for\\nthe agency’s action that the agency itself has not given”\\n(quoting SEC v. Chenery Corp., 332 U.S. 194, 196 (1947))).\\nWe pause, though, to caution the NCUA about two proffered\\nreasons. At oral argument, the government counsel suggested\\nthat the agency may reject proposed local communities if it\\nsuspects they discriminate against residents in the urban core.\\nSee Oral Arg. Recording 5:28-49, 13:43-14:09. But current\\nreviewing guidelines do not indicate that the agency looks for\\nsuch discrimination. See 12 C.F.R. pt. 701, app. B, ch. 1\\n§ VII.A. And the NCUA made the opposite representations to\\nthe District Court. See Transcript of Motion Hearing at 48:23-\\n49:7 (“[District Court]: . . . If a credit union comes to the\\nagency and says I want to serve X area, either in a rural district\\nor a combined statistical area, and they meet the definition, the\\nagency has no authority to reject that application, as long as the\\ncredit union can demonstrate that they can serve the\\narea? [NCUA]: . . . I think that’s probably right, your\\nHonor.”), Am. Bankers Ass’n v. NCUA, No. 1:16-cv-02394-\\nDLF (D.D.C. filed Mar. 27, 2018), ECF No. 33. The\\ngovernment counsel also suggested that community credit\\nunions already cover the vast majority of urban cores. See Oral\\nArg. Recording 8:10-22, 12:30-41; see also id. at 7:39-8:01; cf.\\nNCUA Reply 27-28. Perhaps, but the current record does not\\nsupport the assertion.\\n\\f 32\\n C.\\n\\n Finally, the District Court rejected the increased\\npopulation cap for rural districts. The court worried that a rural\\ndistrict satisfying the new, higher limit could be too large or\\ncould contain “numerous urban centers.” Am. Bankers Ass’n,\\n306 F. Supp. 3d at 69. Because the rule qualified certain\\ndistricts “no matter [their] geographic size or the number or\\nsize of metropolitan areas falling within [their] proposed\\nboundaries,” the court held that the NCUA’s new definition is\\nunreasonable. Id. at 69-70. But in our view, the new “rural\\ndistrict” definition is reasonable.\\n\\n As suggested above, we assume that the NCUA must\\nadopt definitions consistent with the statutory terms as\\nunderstood in 1934, not today. The terms “rural” and “district”\\ndo not connote specific population or geographical constraints.\\nSee First Nat’l Bank III, 522 U.S. at 492 (noting that markets\\n“need not be small”). A “district” means a “portion of a state,\\ncounty, country, town, or city . . . made for administrative,\\nelectoral, or other purposes.” District, WEBSTER’S NEW\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 649\\n(2d ed. 1934) (emphasis added). The district may be “of\\nundefined extent.” Id. Meanwhile, “rural” indicates a\\n“country” or “agricultur[al]” lifestyle, as opposed to that of a\\n“city or town.” Rural, WEBSTER’S NEW INTERNATIONAL\\nDICTIONARY OF THE ENGLISH LANGUAGE 1861 (2d ed. 1934).\\n“Rural” lands generally “resembl[e]” the “country[side].” Id.\\n\\n Nothing about the 1-million-person cap prevents the rural\\ndistrict from resembling the countryside. And one of the\\nunchallenged restrictions helps provide a rural character to\\nsuch districts. Either most residents in the proposed district\\nlive on rural land, or the population density is 100 or fewer\\npeople per square mile. See 81 Fed. Reg. at 88,440.\\n\\f 33\\nAccordingly, even if most residents live on urban areas in the\\nrural district, those areas will be surrounded by rural land.\\nThat’s because 100 people per square mile – or one person for\\nevery six or so acres – is a rural population density.\\n\\n As for size, the contemporaneous definition of “district”\\nreassures us that rural districts may cross state lines. And a\\nsecond unchallenged restriction assures that the 1-million-\\nperson cap will not support gigantic or straggly rural districts.\\nAs the rule explains, the boundaries of a community credit\\nunion’s district may not “exceed the outer boundaries of the\\nstates” immediately surrounding the state where the proposed\\ncredit union would have its headquarters. 81 Fed. Reg. at\\n88,440. Thus, even though the population density of the entire\\nUnited States is less than 100 people per square mile, see\\nNCUA Br. 56 n.44; Oral Arg. Recording 39:34-38, the\\ngeographical limitation forces a rural district to be much\\nsmaller. We are confident that such districts will not be so\\nenormous as to amount to federations of “unrelated geographic\\nunits.” See First Nat’l Bank III, 522 U.S. at 502.\\n\\n In arguing that the amended “rural district” definition is\\nunreasonable, the Association relies heavily on the interpretive\\nanalysis performed by the District Court. See Am. Bankers\\nAss’n, 306 F. Supp. 3d at 67-69. The court’s core contention\\nis that we must construe “rural” and “district” together as a\\nspecialized phrase that meant, in 1934, an “area[] much smaller\\nthan a state.” Id. at 68. For support, the District Court not only\\nconsulted two 1934-era dictionaries, results of a Westlaw\\nsearch of contemporaneous opinions, and a database containing\\nhistorical uses of the phrase, but also invoked the noscitur a\\nsociis canon. See id. at 67-69.\\n\\n The proffered definitional evidence is pretty thin. The\\n1934-era dictionaries described what the specialized phrase\\n\\f 34\\n“rural district” meant in Great Britain, not in the much larger\\nand more expansive United States, which by the 1930s\\nencompassed forty-eight continental states. See, e.g., Rural\\nDistrict, WEBSTER’S NEW INTERNATIONAL DICTIONARY OF THE\\nENGLISH LANGUAGE 1861 (1934) (“in England, a subdivision\\nof an administrative county embracing . . . county parishes”).\\n\\n The proffered Westlaw search results included scores of\\nfederal and state judicial and administrative opinions. See\\nPlaintiff’s Supplemental Memorandum app. 6, Am. Bankers\\nAss’n v. NCUA, No. 1:16-cv-02394-DLF (D.D.C. filed Mar.\\n16, 2018), ECF No. 32-6. But we find little of use in those\\ndocuments. Many of those opinions do not discuss size or\\npopulation. See, e.g., Nicolai v. Wis. Power & Light Co., 269\\nN.W. 281, 282-83 (Wis. 1936). Those that do involve specific\\nrural districts whose sizes were between a town and a state.\\nSee, e.g., Sarther Grocery Co. v Comm’r, 22 B.T.A. 1273, 1274\\n(1931). But none declares that rural districts by definition are\\nrestricted to any particular size or population.\\n\\n The District Court next turned to the Corpus of Historical\\nAmerican Usage, a free and public online database. See Am.\\nBankers Ass’n, 306 F. Supp. 3d at 68 & n.5. The database notes\\n197 mentions of the phrase in the first half of the twentieth\\ncentury and only 37 in the second. See Search Results for\\nRURAL DISTRICTS and RURAL DISTRICT, CORPUS OF\\nHISTORICAL AMERICAN ENGLISH, https://www.english-\\ncorpora.org/coha/ (follow “List” hyperlink; then search\\nmatching strings field for “RURAL DISTRICT”). The\\nperceived drop-off led the District Court to determine that,\\n“even if rural district does not carry meaning distinct from its\\nindividual words today, it did in 1934.” Am. Bankers Ass’n,\\n306 F. Supp. 3d at 68.\\n\\f 35\\n Although federal courts may use “crude[]” searches on\\ndatabases to learn of ambiguities in a statutory term,\\nMuscarello v. United States, 524 U.S. 125, 129-30 (1998), the\\nsearch here did not suffice to show that the agency’s definition\\nwas unreasonable. Much more is required to cabin the\\nagency’s discretion. A search of a commercial database,\\nProQuest, reveals that the phrase appeared at least 500 times in\\nthe second half of the century. We are not confident that the\\nproffered evidence establishes a particular historical trend.\\n\\n The District Court lastly invoked noscitur a sociis. See\\nAm. Bankers Ass’n, 306 F. Supp. 3d at 68-69. The canon is\\ninapposite. Recall that the term “rural district” was listed with\\n“community,” not “local community,” in 1934. At the time,\\n“community” could refer to “[s]ociety at large” or a\\n“commonwealth or state.” Community, WEBSTER’S NEW\\nINTERNATIONAL DICTIONARY OF THE ENGLISH LANGUAGE 452\\n(1934). Indeed, we said that the word is “too indefinite to be\\nused for purposes of exact measurement in terms of acres or\\nsquare miles.” Lukens Steel Co. v. Perkins, 107 F.2d 627, 631\\n(D.C. Cir. 1939), rev’d sub nom., Perkins v. Lukens Steel Co.,\\n310 U.S. 113 (1940). Thus, we do not see size or population\\nas a connection between the terms.\\n\\n The increased population cap is consistent with not only\\nthe Act’s text but also its purposes. For instance, as the\\npreamble noted, the expanded definition allows community\\ncredit unions in rural districts to reach more “persons of modest\\nmeans who may reside” in those areas. 81 Fed. Reg. at 88,416;\\nsee also § 2(1), (2), 112 Stat. at 913.\\n\\n And the change is neither arbitrary nor capricious. The\\nnew rule explains that it provides a “level of operating\\nefficiencies and scale” making rural districts attractive options\\nfor prospective credit unions. 81 Fed. Reg. at 88,416. “[A]\\n\\f 36\\nsufficiently large population base is essential to enable” them\\n“to offer financial services economically.” Id. at 88,417. The\\nNCUA also chose the 1-million figure based on prior\\nexperience. The agency noted that it had approved of eight\\ndistricts with populations of more than 250,000, and that one\\nof them already had exceeded 1 million. See id. “Having set a\\n1 million precedent in one state,” the agency felt justified in\\nhaving a “fixed 1 million population cap for the other 49\\nstates.” Id. (emphasis added). We cannot say this policy\\nchoice lacks rational explanation.\\n\\n The Association says the NCUA failed to consider prior\\nagency decisions. See Am. Bankers Ass’n Br. 61. But we see\\nno discrepancy and thus summarily reject the objection. The\\nAssociation also turns to troubling hypothetical examples of\\nrural districts with unruly shapes and those with dense urban\\nareas such as Denver, Colorado. See NCUA Br. 55-61; Oral\\nArg. Recording 37:55-38:05. Again, such implausible outliers\\ndo not impugn the rule’s general reasonableness.\\n\\n VI.\\n\\n We now consider the remedy. To sum up, we hold that\\ndefining certain Combined Statistical Areas or portions of them\\nas local communities and raising the population cap for rural\\ndistricts are consistent with the Act and reasonably explained.\\nThus, we sustain both aspects of the challenged rule. We also\\nleave undisturbed the portion of the District Court’s opinion\\nthat the parties do not contest on appeal.\\n\\n But we deem the eliminated core requirement to be\\narbitrary and capricious. When a rule is contrary to law, the\\n“ordinary practice is to vacate” it. United Steel v. Mine Safety\\n& Health Admin., 925 F.3d 1279, 1287 (D.C. Cir. 2019); see\\nalso 5 U.S.C. § 706(2) (noting that the “reviewing court\\n\\f 37\\nshall . . . set aside” unlawful agency action). But in “rare\\ncases,” we will opt instead to remand without vacating the rule,\\nso that the agency can “correct its errors.” United Steel, 925\\nF.3d at 1287; see also 5 U.S.C. § 702 (“Nothing\\nherein . . . affects . . . the power or duty of the court to . . . deny\\nrelief on any . . . appropriate . . . equitable ground . . . .”); Oglala\\nSioux Tribe v. U.S. Nuclear Regulatory Comm’n, 896 F.3d 520,\\n536 (D.C. Cir. 2018). In considering whether to adopt the latter\\nequitable remedy, we balance “(1) the seriousness of the\\ndeficiencies of the action, that is, how likely it is the agency\\nwill be able to justify its decision on remand; and (2) the\\ndisruptive consequences of vacatur.” United Steel, 925 F.3d at\\n1287 (citation omitted). A strong showing of one factor may\\nobviate the need to find a similar showing of the other. See,\\ne.g., Fox Television Stations, Inc. v. FCC, 280 F.3d 1027, 1049\\n(D.C. Cir. 2002) (holding that, because the agency likely could\\njustify its action on remand, the potential for disruption was\\n“only barely relevant”); Allied-Signal, Inc. v. U.S. Nuclear\\nRegulatory Comm’n, 988 F.2d 146, 152 (D.C. Cir. 1993)\\n(determining that, because vacatur would give regulated parties\\na “peculiar windfall,” the meager chance of justifying the\\naction was given “little weight” in the remedial analysis).\\n\\n The NCUA has not requested remand without vacatur in\\nthis case. But because we have a “duty” to ensure the propriety\\nof the APA remedy, 5 U.S.C. § 702, we hold that we have the\\ndiscretion to raise the issue sua sponte, cf. Igonia v. Califano,\\n568 F.2d 1383, 1387 (D.C. Cir. 1977); Gaddis v. Dixie Realty\\nCo., 420 F.2d 245, 247 (D.C. Cir. 1969) (per curiam). Remand\\nwithout vacatur is appropriate here.\\n\\n We conclude that the NCUA might be able to offer a\\nsatisfactory reason on remand. And as for disruptive effect, we\\nperceive a substantial likelihood that vacating the rule could\\nmake it more difficult for some poor and minority suburban\\n\\f 38\\nresidents to receive adequate financial services. Even\\ntemporarily depriving these members of the opportunity is\\ninequitable, because it would “set back” the Act’s objective of\\noffering financial services to people of small means. See Nat’l\\nRes. Def. Council v. EPA, 489 F.3d 1364, 1374 (D.C. Cir.\\n2007); see also Advocates for Highway & Auto Safety v. Fed.\\nMotor Carrier Safety Admin., 429 F.3d 1136, 1152 (D.C. Cir.\\n2005) (declining to vacate rule because, in the interim, it “may\\ndo some good, if it does anything at all”).\\n\\n Given the potential for sufficient justification and the\\nsubstantial likelihood of disruptive effect, we have a rare case\\nin which vacatur is inappropriate. See United Steel, 925 F.3d\\nat 1287. Thus, we remand without vacating the relevant part\\nof the 2016 rule. We trust that the agency will act\\nexpeditiously. See EME Homer City Generation, L.P. v. EPA,\\n795 F.3d 118, 132 (D.C. Cir. 2015).\\n\\n * * *\\n\\n In short, we reverse the challenged portions of the District\\nCourt’s summary judgment order. With respect to the\\nqualification of certain Combined Statistical Areas as local\\ncommunities and the increased population cap for rural\\ndistricts, we direct the District Court to issue summary\\njudgment in favor of the NCUA. As to the elimination of the\\nurban-core requirement for local communities based on Core\\nBased Statistical Areas, we direct the District Court to issue\\nsummary judgment in favor of the Association and to remand,\\nwithout vacating, the relevant portion of the 2016 rule for\\nfurther explanation.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4184,"cells":{"id":{"kind":"number","value":7336843,"string":"7,336,843"},"name":{"kind":"string","value":"SELECT SPECIALTY HOSPITAL-DENVER, INC. v. Alex M. AZAR II, Secretary, U.S. Department of Health and Human Services"},"name_abbreviation":{"kind":"string","value":"Select Specialty Hospital-Denver, Inc. v. Azar"},"decision_date":{"kind":"string","value":"2019-08-22"},"docket_number":{"kind":"string","value":"Civil Action No. 10-cv-1356 (BAH)"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"\", \"parties\": \"\", \"opinions\": [{\"author\": \"\", \"type\": \"020lead\", \"text\": \"\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/7254763/\"}]}"},"judge1":{"kind":"string","value":""},"judge2":{"kind":"string","value":""},"judge3":{"kind":"string","value":""},"exactly_three_caps":{"kind":"number","value":0,"string":"0"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/7336843/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":0,"string":"0"},"020lead_count":{"kind":"number","value":1,"string":"1"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: SELECT SPECIALTY HOSPITAL-DENVER, INC. v. Alex M. AZAR II, Secretary, U.S. Department of Health and Human Services\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-22\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\n\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: SELECT SPECIALTY HOSPITAL-DENVER, INC. v. Alex M. AZAR II, Secretary, U.S. Department of Health and Human Services\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-22\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\n\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4185,"cells":{"id":{"kind":"number","value":4655414,"string":"4,655,414"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"Murray Energy Corporation v. EPA"},"decision_date":{"kind":"string","value":"2019-08-23"},"docket_number":{"kind":"string","value":"15-1385"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before: GRIFFITH, PILLARD and WILKINS, Circuit Judges.\", \"parties\": \"\", \"opinions\": [{\"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued December 18, 2018 Decided August 23, 2019\\n\\n No. 15-1385\\n\\n MURRAY ENERGY CORPORATION,\\n PETITIONER\\n\\n v.\\n\\n ENVIRONMENTAL PROTECTION AGENCY,\\n RESPONDENT\\n\\n AMERICAN LUNG ASSOCIATION, ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 15-1392, 15-1490, 15-1491, 15-1494\\n\\n\\n On Petitions for Review of Final Agency Action of\\n the United States Environmental Protection Agency\\n\\n\\n Seth L. Johnson argued the cause for Public Health and\\nEnvironmental Petitioners. With him on the briefs were Joshua\\nStebbins, Joshua Berman, David S. Baron, and Paul Cort.\\n\\f 2\\n Dominic E. Draye, Deputy Solicitor General, Office of the\\nAttorney General for the State of Arizona, argued the cause for\\nState Petitioners. With him on the briefs were Mark Brnovich,\\nAttorney General, John R. Lopez, IV, Solicitor General, Keith\\nMiller, Associate Solicitor, Joshua L. Kaul, Attorney General,\\nOffice of the Attorney General for the State of Wisconsin,\\nMisha Tseytlin, Solicitor General, Daniel P. Lennington,\\nAssistant Attorney General, Leslie C. Rutledge, Attorney\\nGeneral, Office of the Attorney General for the State of\\nArkansas, Lee Rudofsky, Solicitor General, Jamie L. Ewing,\\nAssistant Attorney General, Jeff Landry, Attorney General,\\nOffice of the Attorney General for the State of Louisiana,\\nSteven B. ABeaux@ Jones, Assistant Attorney General, Wayne\\nStenehjem, Attorney General, Office of the Attorney General\\nfor the State of North Dakota, Margaret I. Olson, Assistant\\nAttorney General, Peter S. Glaser, Carroll W. AMack@\\nMcGuffey, Andy Beshear, Attorney General, Office of the\\nAttorney General for the Commonwealth of Kentucky, Joseph\\nA. Newberg, II, Assistant Attorney General, Lara Katz,\\nAssistant Attorney General, Office of the Attorney General for\\nthe State of New Mexico, Ken Paxton, Attorney General,\\nOffice of the Attorney General for the State of Texas, Priscilla\\nM. Hubenak, Chief, Craig J. Pritzlaff, Assistant Attorney\\nGeneral, E. Scott Pruitt, Attorney General at the time the brief\\nwas filed, Office of the Attorney General for the State of\\nOklahoma, P. Clayton Eubanks, Deputy Solicitor General, and\\nSean Reyes, Attorney General, Office of the State of Utah.\\nMithun Mansinghani, Attorney, Office of the Attorney General\\nfor the State of Oklahoma, Lisa M. Mitchell, Assistant Attorney\\nGeneral, Office of the Attorney General for the State of Texas,\\nOramel H. Skinner, III, Solicitor, Office of the Attorney\\nGeneral for the State of Arizona, Sarah Adkins and Samuel R.\\nFlynn, Assistant Attorneys General, and Gregory T. Dutton,\\nCounsel, Office of the Attorney General for the\\nCommonwealth of Kentucky, Steven C. Kilpatrick, Assistant\\n\\f 3\\nAttorney General, Office of the Attorney General for the State\\nof Wisconsin, Elizabeth B. Murrill and Harry J. Vorhoff,\\nAssistant Attorneys General, Office of the Attorney General\\nfor the State of Louisiana, entered appearances.\\n\\n James R. Bieke argued the cause for Industry Petitioners.\\nWith him on the briefs were Roger R. Martella, Joel F. Visser,\\nScott C. Oostdyk, E. Duncan Getchell, Jr., Michael H. Brady,\\nThomas A. Lorenzen, Robert J. Meyers, Linda E. Kelly,\\nQuentin Riegel, Leland P. Frost, Michael B. Schon, Elizabeth\\nL. Horner, Lucinda Minton Langworthy, Aaron M. Flynn,\\nSteven P. Lehotsky, Sheldon B. Gilbert, Stacy Linden, and\\nRichard S. Moskowitz. Peter C. Tolsdorf entered an\\nappearance.\\n\\n Hope M. Babcock and Sarah J. Fox were on the brief for\\namici curiae American Thoracic Society, et al. in support of\\npetitioners Sierra Club, et al.\\n\\n Thomas J. Ward was on the brief for amicus curiae The\\nNational Association of Home Builders in support of Industry\\nand State Petitioners.\\n\\n Justin Heminger and Simi Bhat, Trial Attorneys, U.S.\\nDepartment of Justice, argued the causes for respondent. With\\nthem on the brief were John C. Cruden, Assistant Attorney\\nGeneral at the time the brief was filed, and David Orlin, Steven\\nSilverman, and Brian Doster, Attorneys, U.S. Environmental\\nProtection Agency. Jon M. Lipshultz, Attorney, U.S.\\nDepartment of Justice, entered an appearance.\\n\\n James R. Bieke argued the cause for Industry Respondent-\\nIntervenors. With him on the brief were Roger R. Martella,\\nJoel F. Visser, Lucinda Minton Langworthy, Aaron M. Flynn,\\nThomas A. Lorenzen, Robert J. Meyers, Stacy Linden, Richard\\n\\f 4\\nS. Moskowitz, Steven P. Lehotsky, Sheldon B. Gilbert, Linda E.\\nKelly, Quentin Riegel, Leland P. Frost, Michael B. Schon,\\nElizabeth Horner, and Leslie A. Hulse. Peter C. Tolsdorf\\nentered an appearance.\\n\\n Seth L. Johnson argued the cause for Health and\\nEnvironmental Respondent-Intervenors. With him on the brief\\nwas David S. Baron. Joshua A. Berman and Joshua R. Stebbins\\nentered appearances.\\n\\n Jonathan Weiner, Deputy Attorney General, Office of the\\nAttorney General for the State of California, argued the cause\\nfor State Amici in support of respondent. With him on the brief\\nwere Kamala D. Harris, Attorney General at the time the brief\\nwas filed, Robert W. Byrne, Senior Assistant Attorney General,\\nand Gavin G. McCabe, Supervising Deputy Attorney General.\\nMelinda Pilling, Attorney, entered an appearance.\\n\\n Richard L. Revesz, Denise A. Grab, Jack Lienke, Michael\\nA. Livermore, and Jason A. Schwartz were on the brief for\\namicus curiae The Institute for Policy Integrity at New York\\nUniversity School of Law in support of respondent.\\n\\n Before: GRIFFITH, PILLARD and WILKINS, Circuit Judges.\\n\\n Opinion for the Court filed PER CURIAM.\\n\\n PER CURIAM: In this opinion, we consider various\\nchallenges to the Environmental Protection Agency’s 2015\\nrevisions to the primary and secondary national ambient air\\nquality standards for ozone. For the reasons given below, we\\ndeny the petitions, except with respect to the secondary ozone\\nstandard, which we remand for reconsideration, and\\ngrandfathering provision, which we vacate.\\n\\f 5\\n I. Background\\n\\nA. Statutory and Procedural Background\\n\\n Congress enacted the modern version of the Clean Air Act\\n(the “Act”), codified at 42 U.S.C. § 7401 et seq., in 1970 to\\ncontrol and reduce contaminants responsible for air pollution\\nwith the overarching goal to protect human health and the\\nenvironment. Pursuant to Title I, EPA must establish, publish,\\nand periodically review primary and secondary national\\nambient air quality standards (“NAAQS”) for air pollutants that\\n“may reasonably be anticipated to endanger public health or\\nwelfare.” 42 U.S.C. § 7408(a)(1)(A). The primary NAAQS\\nare to be set at levels “the attainment and maintenance of which\\nin the judgment of the Administrator, . . . allowing an adequate\\nmargin of safety, are requisite to protect the public health.” Id.\\n§ 7409(b)(1). The secondary NAAQS “shall specify a level of\\nair quality the attainment and maintenance of which in the\\njudgment of the Administrator, . . . is requisite to protect the\\npublic welfare from any known or anticipated adverse effects.”\\nId. § 7409(b)(2). Thus, primary NAAQS protect the public\\nhealth, while the secondary NAAQS protect the public welfare.\\n“Public health” includes adverse health effects for both the\\npopulation at large and sensitive populations such as children,\\nolder adults, and people with asthma or other lung diseases.\\nThe term “public welfare” encompasses a wide variety of\\neffects on soil, plants, wildlife and biota, property damage,\\naesthetic concerns, and other non-health-related impacts such\\nas hazards to economic values and personal comfort. Id.\\n§ 7602(h).\\n\\n Pursuant to section 109(d) of the Clean Air Act, EPA must\\ncomplete a “thorough” review of the NAAQS every five years.\\nId. § 7409(d)(1). During this review, EPA must revise the\\ncriteria and standards or promulgate new standards as\\n\\f 6\\nappropriate. Id. To assist in this process, the Clean Air\\nScientific Advisory Committee (“CASAC”) undertakes an\\nexamination of the current air quality criteria, primary\\nNAAQS, and secondary NAAQS, and submits\\nrecommendations to EPA for “any new [NAAQS] and\\nrevisions of existing criteria and standards as may be\\nappropriate.” Id. § 7409(d)(2)(A)-(B). Congress required EPA\\nto take CASAC’s recommendations into account when\\npromulgating revised NAAQS and to fully explain its reasons\\nwhen it departs from CASAC’s advice. However, the ultimate\\ndecision to revise the NAAQS—and the determination of the\\nnew level—rests with the Administrator. Id. § 7407(d)(3).\\n\\n These petitions concern EPA’s promulgation of revised\\nNAAQS related to ozone. Ozone is a colorless gas that occurs\\nboth in the Earth’s upper atmosphere and at ground level.\\nAlthough ozone is an “essential presence in the atmosphere’s\\nstratospheric layer,” ground-level ozone is an air pollutant that\\nis harmful to breathe and damages crops, trees, and other\\nvegetation. S. Coast Air Quality Mgmt. Dist. v. EPA, 472 F.3d\\n882, 887 (D.C. Cir. 2006); see Allied Local & Reg’l Mfrs.\\nCaucus v. EPA, 215 F.3d 61, 66 (D.C. Cir. 2000). Ground-\\nlevel ozone is not a direct product of human activity, but\\ninstead forms when atmospheric pollutants (including nitrogen\\noxides and volatile organic compounds) react in the presence\\nof sunlight. See Mississippi v. EPA, 744 F.3d 1334, 1340 (D.C.\\nCir. 2013). These precursor atmospheric pollutants are created\\nprimarily from emissions produced by cars, power plants, and\\nchemical solvents. NRDC v. EPA, 777 F.3d 456, 459 (D.C.\\nCir. 2014).\\n\\n In 1979, EPA issued primary and secondary NAAQS for\\nozone with a limit of 0.12 parts per millions (ppm), and a one-\\nhour averaging time. See id. This “one-hour standard”\\nmeasured average ozone levels over one-hour periods, and\\n\\f 7\\nEPA would deem an area in compliance with this standard if it\\ndid not exceed the level for more than one day per calendar\\nyear. Id. EPA next revised the ozone NAAQS in 1997, having\\ndetermined that no revisions to the standards were necessary in\\n1993. National Ambient Air Quality Standards for Ozone, 58\\nFed. Reg. 13,008 (Mar. 9, 1993). The agency replaced the one-\\nhour, 0.12 ppm standard with a 0.08 ppm standard measured\\nover eight hours. National Ambient Air Quality Standards for\\nOzone, 62 Fed. Reg. 38,856 (July 18, 1997). EPA also altered\\nthe form of compliance to the annual fourth-highest daily\\nmaximum 8-hour concentration, averaged over three years. Id.\\n\\n In 2008, EPA lowered the primary and secondary\\nstandards to 0.075 ppm but kept the same eight-hour averaging\\ntime and form as in 1997. NRDC, 777 F.3d at 462-63; National\\nAmbient Air Quality Standards for Ozone, 73 Fed. Reg. 16,436\\n(Mar. 27, 2008). In developing the 2008 standards, EPA relied\\non scientific evidence showing that ozone causes health effects\\nat and above 0.08 ppm and examined two new clinical studies\\nthat found negative health effects from ozone at lower levels.\\nSee Mississippi, 744 F.3d at 1340. While CASAC reviewed\\nthis same evidence and recommended a primary level between\\n0.06 ppm and 0.07 ppm, EPA departed from this advice and\\nexplained that the scientific data regarding negative health\\neffects at 0.06 ppm was too limited and inconclusive to support\\na standard below 0.075 ppm. See id. at 1340-41. We upheld\\nEPA’s primary standard on this basis but found that EPA had\\nnot adequately explained its revision of the secondary standard.\\nId. at 1359-62. We noted that EPA had not properly\\ndetermined what level of air quality was requisite to protect the\\npublic welfare. Id. Accordingly, we remanded for further\\nexplanation and reconsideration of the secondary level.\\n\\n Following the promulgation of the 2008 ozone NAAQS,\\nEPA undertook another comprehensive review of the ozone\\n\\f 8\\nstandards and scientific data. EPA consolidated its review on\\nremand of the 2008 secondary standard with its ongoing review\\nfor the 2015 ozone NAAQS. After conducting public hearings\\nand receiving approximately 430,000 written comments on its\\nproposed revision of the primary and secondary ozone\\nNAAQS, EPA published its final 2015 ozone NAAQS on\\nOctober 26, 2015. National Ambient Air Quality Standards for\\nOzone (“2015 Rule”), 80 Fed. Reg. 65,292 (Oct. 26, 2015). For\\nboth the primary and secondary standards, EPA lowered the\\nlevel from 0.075 ppm to 0.07 ppm, while retaining the 2008\\nindicator (ozone), averaging time (8 hours), and form (three-\\nyear average of the fourth-highest daily maximum 8-hour\\nconcentration). See id. at 65,294, 65,301, 65,347, 65,349-50,\\n65,352.\\n\\n Based on scientific data and CASAC’s recommendation,\\nthe Administrator concluded that the 2008 primary ozone\\nstandard (0.075 ppm) was not at a level requisite to protect\\npublic health while allowing an adequate margin of safety. Id.\\nat 65,326, 65,344, 65,346. The Administrator placed\\nsignificant weight on new clinical studies linking short-term\\nozone exposure with respiratory effects, including lung\\ninflammation. See id. at 65,302-03, 65,341, 65,352, 65,359.\\nEPA conducted its own exposure study and found that a revised\\nprimary NAAQS standard of 0.07 ppm would “eliminate the\\noccurrence of two or more exposures of concern to [ozone]\\nconcentrations at and above [0.08 ppm]” and would “virtually\\neliminate the occurrence of two or more exposures of concern\\nto [ozone] concentrations at and above [0.07 ppm], even in the\\nworst-case urban study area and year evaluated.” Id. at 65,353.\\nThe Administrator additionally found that a level of 0.07 ppm\\nwould “protect the large majority of children in the urban study\\nareas (i.e., about 96% to more than 99% of children in\\nindividual urban study areas) from experiencing two or more\\nexposures of concern at or above the [0.06 ppm] benchmark.”\\n\\f 9\\nId. at 65,353, 65,360-64. EPA also evaluated hundreds of\\nepidemiologic studies that provided statistically relevant\\ninformation about a broader population of individuals who are\\nexposed to uncontrolled air pollution. See id. at 65,304,\\n65,364.\\n\\n Further, the Administrator considered CASAC’s advice on\\nthe new primary ozone standard. See id. at 65,346, 65,361. In\\nadvance of the 2015 Rule, CASAC stated that “there is clear\\nscientific support for the need to revise the standard” in place\\nsince 2008 and recommended setting the standard within a\\nrange of 0.06 ppm to 0.07 ppm, while leaving the form,\\naveraging time, and indicator unchanged. Id. at 65,322,\\n65,361.\\n\\n EPA also assessed the secondary standard and concluded\\nthat the 2008 secondary standard (0.075 ppm) was not requisite\\nto protect public welfare. Id. at 65,382, 65,389-90. Rather,\\nEPA set the secondary standard at 0.07 ppm and kept the\\nindicator, averaging time, and form the same. Id. at 65,369,\\n65,403, 65,409-10. EPA noted that more than four hundred\\nnew studies examining the interplay between ozone and public\\nwelfare had been developed since the promulgation of the 2008\\nozone NAAQS. Id. at 65,369. These studies strengthened and\\nexpanded the scientific understanding of ozone’s effects on\\nplants. The Administrator analyzed the causal relationship\\nbetween ozone exposure and vegetation effects, examining tree\\ngrowth impacts, crop yield loss, and visible leaf injury. Id. at\\n65,370; see id. at 65,294, 65,369-70. The Administrator gave\\n“primary consideration” to tree growth impacts, judging that it\\nwas more difficult to assess the welfare significance of crop\\nyield loss (given that crops are heavily managed) and visible\\nleaf injury (citing the “lack of established criteria or\\nobjectives”). Id. at 65,407.\\n\\f 10\\n In setting the secondary standard, EPA departed from\\nCASAC’s recommendations as to the form and standard of the\\nsecondary ozone NAAQS. CASAC recommended using a\\nsingle-year averaging approach, but EPA opted for a three-year\\naverage because “the public welfare significance of effects\\nassociated with multiple years of critical exposures are\\npotentially greater than those associated with a single year of\\nsuch exposure.” Id. at 65,404. The Administrator also rejected\\nCASAC’s recommendation to use the W126 exposure index, a\\ncumulative, seasonal ozone exposure metric, as the form of the\\nsecondary standard, rather than the same form as the primary\\nstandard (three-year average of the fourth-highest daily level),\\nfinding that the latter form was adequate to restrict cumulative\\nozone exposures that are detrimental to vegetation. Id. at\\n65,408.\\n\\n In addition to revising the ozone standards, the 2015\\nNAAQS also updated the regulations for the Prevention of\\nSignificant Deterioration (“PSD”) permitting program. See id.\\nat 65,431-34. Under the PSD program, no construction on a\\nmajor emitting facility may be commenced in an area that has\\nattained the air quality standards for any criteria pollutant\\nunless “the owner or operator of such facility demonstrates . . .\\nthat emissions from construction or operation of such facility\\nwill not cause, or contribute to, air pollution in excess of any\\n. . . [NAAQS].” 42 U.S.C. § 7475(a)(3)(B). The permitting\\nauthority (either a state or EPA) must grant or deny a completed\\npreconstruction permit application under the PSD program\\nwithin one year of filing. Id. § 7475(c).\\n\\n EPA claimed that sections 7475(a)(3)(B) and 7475(c) have\\nthe “potential for conflict,” which it was authorized under\\nChevron to resolve. EPA Br. 132; see also 80 Fed. Reg. at\\n65,433-34. EPA worried that, for a limited subset of pending\\npermit applications, complying with section 7475(a)(3)(B)’s\\n\\f 11\\ndemonstration requirement for the 2015 NAAQS “could hinder\\ncompliance with the requirement under section [7475](c) to\\nissue a permit within one year of the completeness\\ndetermination.” 80 Fed. Reg. at 65,434. In other words, EPA\\nwas unsure how to handle completed permit applications where\\nsources would have complied with the NAAQS in existence at\\nthe time the application was filed (the 2008 ozone standards),\\nbut where the NAAQS were revised before the permit was\\napproved. See id. at 65,433. EPA resolved this apparent\\nconflict by implementing a grandfathering provision for\\npending permit applications that satisfy one of two permitting\\ncriteria. See id. at 65,431-34. These two applicable categories\\ninclude: (1) permit applications that are deemed complete on\\nor before the signature date of the final rule revising the ozone\\nNAAQS; and (2) permit applications “for which the reviewing\\nauthority has first published a public notice of the draft permit\\nor preliminary determination before the effective date of the\\nrevised NAAQS.” Id. at 65,431, 65,433. If a permit\\napplication satisfies either requirement, the owner or operator\\nmust show compliance with the 2008 rather than the 2015\\nNAAQS before initiating construction.\\n\\n Petitioners from the oil and gas industry (“Industry\\nPetitioners”) and from several states (“State Petitioners”) claim\\nthat the primary and secondary NAAQS are too protective.\\nConversely, petitioners from public health and environmental\\ngroups (“Environmental Petitioners”) claim that these NAAQS\\nare not protective enough. Environmental Petitioners also\\ncontend that EPA’s decision to allow limited grandfathering of\\ncertain permit applications was inconsistent with the Clean Air\\nAct.\\n\\n This opinion proceeds in five parts. Part II considers the\\nchallenges to the primary standard, and Part III the challenges\\nto the secondary standard. Part IV reviews the cross-cutting\\n\\f 12\\nchallenges to EPA’s promulgation of the NAAQS, and Part V\\naddresses the challenge to the grandfathering provision.\\n\\nB. Jurisdiction and Standard of Review\\n\\n This court has jurisdiction to review EPA’s 2015 Rule\\npursuant to 42 U.S.C. § 7607(b)(1). Under section 7607,\\nagency action may be reversed if it is arbitrary, capricious, an\\nabuse of discretion, not in accordance with law, or in excess of\\nstatutory authority. Id. § 7607(d)(9)(A), (C); see Am.\\nPetroleum Inst. v. EPA, 684 F.3d 1342, 1347 (D.C. Cir. 2012).\\nTo withstand review, an agency must have examined all\\nrelevant facts and data, and articulated a rational explanation\\nfor its decision, including a reasonable connection between the\\nfacts and ultimate outcome. See Allied Local, 215 F.3d at 68.\\nWe cannot look at EPA’s decision as would a scientist, but\\ninstead must exercise our “narrowly defined duty of holding\\nagencies to certain minimal standards of rationality.”\\nMississippi, 744 F.3d at 1342 (quoting Nat’l Envtl. Dev.\\nAss’n’s Clean Air Project v. EPA, 686 F.3d 803, 810 (D.C. Cir.\\n2012)); see id. at 1348 (“We repeat: it is not our job to referee\\nbattles among experts; ours is only to evaluate the rationality\\nof EPA’s decision . . . .”).\\n\\n On questions of statutory interpretation, the court must\\nreview EPA’s actions in accordance with the standard set forth\\nin Chevron, U.S.A., Inc. v. Natural Resources Defense Council,\\nInc., 467 U.S. 837, 842-44 (1984). See Michigan v. EPA, 135\\nS. Ct. 2699, 2706-07 (2015); NRDC, 777 F.3d at 463. Chevron\\ndeference involves a two-step analysis. First, if a statute is\\nclear, the court must give effect to Congress’s unambiguous\\nlanguage and intent. Chevron, 467 U.S. at 842-43. Where a\\nstatute that an agency administers is ambiguous, the court must\\nturn to the second Chevron principle and give deference to the\\n\\f 13\\nagency’s reasonable interpretation of the statute. Id. at 843; see\\nalso S. Coast Air Quality Mgmt. Dist., 472 F.3d at 891.\\n\\n II. Primary Standard Challenges\\n\\n Industry and State Petitioners contend that EPA’s\\npromulgation of the 2015 Rule’s primary standard was\\narbitrary and capricious because, they say, EPA failed to\\nprovide a reasoned explanation for departing from the 2008\\nNAAQS. Environmental Petitioners argue that the primary\\nozone standard is too lenient because it occasionally permits\\nozone levels to exceed 0.07 ppm and will allegedly tolerate\\nadverse health effects. For the reasons below, we hold that\\nthese arguments lack merit.\\n\\nA. Industry and State Petitioners’ Challenge\\n\\n The Clean Air Act requires EPA to set primary NAAQS\\nthat are “requisite” to protect public health with an adequate\\nmargin of safety. 42 U.S.C. § 7409(b)(1). The term “requisite”\\nmeans “sufficient, but not more than necessary.” See Whitman\\nv. Am. Trucking Ass’ns, 531 U.S. 457, 473 (2001). While the\\ndetermination of what is “requisite” may require a contextual\\nassessment of acceptable risk, our precedent is clear that prior\\nNAAQS are not sacrosanct and are not granted presumptive\\nvalidity. See Mississippi, 744 F.3d at 1343. Prior NAAQS\\nneed not remain the governing standard until every aspect of\\nthose prior NAAQS is undermined. Id. To hold otherwise\\n“would bind EPA to potential deficiencies in past reviews\\nbecause discrepancies between past and current judgments as\\neasily reflect problems in the past as in the present.” Id. Thus,\\nwhen EPA reviews and revises the NAAQS, it does so against\\ncurrent policy considerations and existing scientific\\nknowledge. Id. Accordingly, “[t]he statutory framework\\nrequires us to ask only whether EPA’s proposed NAAQS is\\n\\f 14\\n‘requisite’; we need not ask why the prior NAAQS once was\\n‘requisite’ but is no longer up to the task.” Id.\\n\\n Given our decision in Mississippi, we must reject Industry\\nand State Petitioners’ argument that EPA departed from the\\n2008 NAAQS without adequate explanation. EPA has\\nadequately explained why on the record before it the revised\\nstandard is requisite to protect the public health. As the record\\nmakes clear, the Administrator considered the entire body of\\nscientific evidence available, including clinical studies,\\nepidemiologic evidence, human exposure and health risk\\nassessments, CASAC’s recommendations, and over 430,000\\npublic comments. See 80 Fed. Reg. at 65,293, 65,299, 65,301,\\n65,323, 65,326. And consistent with CASAC’s advice, the\\nAdministrator placed the most weight on clinical studies. See\\nid. at 65,302-03, 65,341, 65,352, 65,359. State Petitioners take\\nparticular issue with EPA’s reliance on the 2009 Schelegle\\nstudy, which, they argue, does not support EPA’s finding of\\nadverse effects. But the 2009 Schelegle study is merely one of\\nmany clinical studies that EPA relied on. State Petitioners\\nignore the fact that additional clinical studies and\\nepidemiologic evidence show substantial health effects at\\nozone levels as low as 0.06 ppm. See id. at 65,331, 65,334-35\\n& n.96, 65,344, 65,364.\\n\\n In addition to evaluating clinical studies, the Administrator\\nalso examined epidemiologic evidence and EPA’s exposure\\nand risk assessments. See 80 Fed. Reg. at 65,304, 65,314-17,\\n65,364. While the Administrator placed less weight on the\\nepidemiologic results than on clinical studies, see id. at 65,320,\\n65,324, 65,335, she nonetheless found that recent\\nepidemiologic studies suggested adverse health effects would\\noccur below the 2008 NAAQS standard. While Industry\\nPetitioners challenge the use of epidemiologic evidence given\\nthe uncertainties presented in these studies, even CASAC\\n\\f 15\\nconcluded that the epidemiologic evidence would have alone\\nbeen strong enough to justify revision of the 2008 NAAQS.\\n\\nB. Environmental Petitioners’ Challenges\\n\\n In contrast to Industry and State Petitioners,\\nEnvironmental Petitioners argue that the primary ozone\\nstandard is too lenient. They offer two reasons. First,\\nEnvironmental Petitioners contend that the form of the revised\\nstandard is not health-protective because it permits ozone\\nlevels to exceed 0.07 ppm on some days. Second, they argue\\nthat ozone exposures of 0.07 ppm will cause adverse health\\neffects, particularly in sensitive populations. For the reasons\\nbelow, we hold that these arguments lack merit.\\n\\n EPA reasonably explained its decision to retain\\n the form of the primary standard.\\n\\n Environmental Petitioners first take issue with the form of\\nthe revised standard—that the average, over three years, of the\\nannual fourth-highest maximum daily 8-hour average ozone\\nlevel must not exceed 0.07 ppm. By using this form, EPA\\n“allows ozone levels to exceed—multiple times in any year—\\nlevels that EPA itself agrees cause adverse health effects.”\\nEnvtl. Pet’rs Br. 20. Environmental Petitioners contend that\\nbecause areas with ozone levels above the 0.07 ppm threshold\\ncan still comply with EPA’s standards, the standard is unlawful\\nand arbitrary. We disagree.\\n\\n The Administrator’s decision to retain the same form used\\nin the 1997 and 2008 primary standards was in part based on\\nhow many people are estimated to experience unhealthy ozone\\nexposures. To this end, the Administrator utilized EPA’s\\nexposure assessment to gauge how often various population\\nsubgroups will be exposed to potentially health-impairing\\nozone concentrations while experiencing elevated breathing\\n\\f 16\\nrates. See 80 Fed. Reg. at 65,356. Rather than considering the\\nozone level of an area in isolation, the Administrator also\\n“consider[ed] activity patterns in the exposed population”\\nbecause adverse health responses to ozone exposure are\\ncritically dependent on breathing rates. Id. at 65,363; see also\\nid. (“Not considering activity patterns, and corresponding\\nventilation rates, can result in a standard that provides more\\nprotection than is requisite.”). This use of the exposure\\nassessment was rational.\\n\\n Further, the Administrator adequately interpreted the\\nexposure assessment. The Administrator correctly observed\\nthat the exposure assessment only chronicled how many people\\nwith an elevated breathing rate will be exposed to a specific\\nozone level, not how many will have an adverse effect. Given\\nthis limitation, the Administrator focused on estimates of two\\nor more exposures of concern to assess the potential for adverse\\neffects at and above an ozone concentration of 0.07 ppm. See\\nid. at 65,310, 65,325, 65,330, 65,343, 65,345-46, 65,354,\\n65,358, 65,361, 65,363. Based on the data, the Administrator\\nconcluded that a level of 0.07 ppm ozone would eliminate the\\noccasions on which school-age children experience two or\\nmore exposures of concern at ozone concentrations at and\\nabove 0.08 ppm and, even in the worst-case years and\\nlocations, would “virtually eliminate” the occasions on which\\nsuch children experience two or more exposures of concern at\\n0.07 ppm. Id. at 65,353. In particular, the Administrator noted\\nthat a primary level of 0.07 ppm would protect approximately\\n96% to 99% of children in the urban study areas from\\nexperiencing two or more exposures of concern above the 0.06\\nppm benchmark. Id. at 65,353; see also id. at 65,360-64. That\\nEnvironmental Petitioners cite data suggesting that 18,000\\nchildren would experience multiple exposures of concern at or\\nabove 0.07 ppm during the worst year and location is not\\ndeterminative because the primary standard for a non-threshold\\n\\f 17\\npollutant like ozone is not required to produce zero risk, and\\n“[t]he task of determining what standard is ‘requisite’ to protect\\nthe qualitative value of public health or what margin of safety\\nis ‘adequate’ to protect sensitive subpopulations necessarily\\nrequires the exercise of policy judgment.” Mississippi, 744\\nF.3d at 1351, 1358. EPA has reasonably exercised its\\njudgment.\\n\\n Given that the Administrator appropriately examined and\\nconsidered EPA’s exposure assessment, her decision to retain\\nthe form of the primary standard was also rational. The\\nAdministrator determined that, when combined with an ozone\\nlevel of 0.07 ppm, the form of the standard (three-year average\\nof the fourth-highest daily level) was requisite. In particular,\\nthe Administrator found that most areas that meet the revised\\nstandard will have an 8-hour ozone concentration below 0.07\\nppm. See 80 Fed. Reg. at 65,363. Environmental Petitioners,\\nhowever, criticize this form for allegedly failing to account for\\nindividual ozone days above 0.07 ppm. But Environmental\\nPetitioners elide a crucial detail of EPA’s methodology: the\\nexposure assessment study that “informed the Administrator’s\\nconsideration of the degree of public health protection provided\\nby various standard levels” considered the air quality at\\n“various standards with the current 8-hour averaging time and\\nfourth-high, 3-year average form.” Id. at 65,351-52 (emphasis\\nadded). As such, the Administrator reasonably accounted for\\ndays when ozone levels may exceed 0.07 ppm. Taken together,\\nwe conclude that the Administrator’s decision to retain the\\nform of the standard was appropriate and neither arbitrary nor\\ncapricious.\\n\\f 18\\n EPA reasonably set the primary standard at 0.07\\n ppm.\\n\\n Environmental Petitioners next argue that the\\nAdministrator’s decision to lower the primary standard from\\nfrom 0.075 to 0.07 ppm is an unexplained departure from\\nCASAC’s recommendation and from EPA’s prior position\\nregarding the adversity of certain lung function decrements.\\nWe reject both contentions.\\n\\n Environmental Petitioners argue that EPA failed to fulfill\\nits duty under the Act to provide “an explanation of the\\nreasons” for departing from CASAC’s scientific\\nrecommendations. 42 U.S.C. § 7607(d)(3); see also id.\\n§ 7607(d)(6). They contend CASAC told EPA that “[a]t [0.07\\nppm] there is substantial scientific certainty of a variety of\\nadverse effects, including decrease in lung function, increase\\nin respiratory symptoms, and increase in airway\\ninflammation,” Envtl. Pet’rs Br. 31 (quoting J.A. 531), and\\nEPA failed to rationally dispute or refute this finding.\\n\\n EPA adequately considered CASAC’s advice when setting\\nthe primary NAAQS to 0.07 ppm. While CASAC advised EPA\\nto “set the level of the standard lower than [0.07 ppm],” J.A.\\n531, that recommendation constituted policy—not scientific—\\nadvice. This distinction is crucial because we defer to EPA’s\\njudgment on issues of policy but require EPA to “fully explain\\nits reasons for any departure from” “CASAC’s expert scientific\\nrecommendations.” Mississippi, 744 F.3d at 1358, 1354-55\\n(emphasis added). CASAC’s letter to EPA makes clear that\\n“based on the scientific evidence” it “recommend[ed] a range\\nof levels for a revised primary ozone standard from [0.07 ppm]\\nto [0.06 ppm].” J.A. 531. CASAC then “acknowledge[d] that\\nthe choice of a level within the range recommended based on\\nscientific evidence is a policy judgment.” Id.\\n\\f 19\\n In an effort to influence EPA’s policy judgment, CASAC\\nnoted that given the “substantial scientific certainty of a variety\\nof adverse effects” at 0.07 ppm, that level may not be adequate\\nto protect public health. Id. So, CASAC’s “policy advice\\n[was] to set the level of the standard lower than [0.07 ppm]\\nwithin a range down to [0.06 ppm], taking into account [EPA’s]\\njudgment regarding the desired margin of safety to protect\\npublic health.” Id. EPA did not take CASAC’s policy advice\\nbecause the Administrator found that 0.07 ppm would still\\n“provide substantial protection against the broader range of\\n[ozone] exposure concentrations that have been shown in\\ncontrolled human exposure studies to result in respiratory\\neffects, including exposure concentrations below [0.07 ppm].”\\n80 Fed. Reg. at 65,363. EPA nonetheless chose a level for the\\nprimary standard that was within CASAC’s scientifically\\nrecommended range. In so doing, EPA did not abrogate its\\nduty under the Clean Air Act to consider CASAC’s scientific\\nrecommendations but instead made a valid policy decision.\\n\\n Environmental Petitioners also argue that EPA should\\nhave set the primary standard lower than 0.07 ppm, given the\\ncontrolled human exposure studies that had been published\\nsince the agency’s 2008 NAAQS review. Specifically,\\nEnvironmental Petitioners highlight two new studies\\nevaluating exposures to 0.06-0.063 ppm ozone that concluded\\nthat exposures at 0.06 ppm caused lung function decrements of\\n10% or more. Taken with EPA’s statement in the 2008 ozone\\nNAAQS rule that a lung function decrement of 10% or more\\n“should be considered adverse for sensitive populations,” 73\\nFed. Reg. at 16,454-55, Environmental Petitioners contend that\\nEPA acted arbitrarily by giving insufficient weight to those\\nnew studies. We are unpersuaded by this argument, which fails\\nto appreciate the dynamic nature of adversity determinations at\\neach NAAQS review.\\n\\f 20\\n In determining the appropriate level for the 2015 primary\\nstandard, the Administrator was not bound by adversity\\njudgments in the 2008 NAAQS review. Indeed, “as the\\ncontours and texture of scientific knowledge change, the\\nepistemological posture of EPA’s NAAQS review necessarily\\nchanges as well.” Mississippi, 744 F.3d at 1344. Thus, we\\nconsider only whether EPA’s 2015 Rule offers a rational\\nexplanation of why EPA chose a new adversity definition and\\nwhether the Administrator reasonably evaluated the evidence\\nof adversity. We hold that EPA’s actions were reasonable and\\nreasonably explained. Rather than applying a rigid test for\\ndetermining what level of decrement is adverse, the\\nAdministrator took a more comprehensive approach provided\\nby American Thoracic Society (ATS) guidelines. ATS\\nguidelines provide that “reversible loss of lung function in\\ncombination with the presence of [respiratory] symptoms\\nshould be considered adverse.” 80 Fed. Reg. at 65,357\\n(internal quotation marks omitted). EPA reasonably explained\\nthat it chose to adopt the ATS definition of adversity because\\n“the available evidence does not provide information on the\\nextent to which a short-term, transient decrease in lung function\\nin a population,” without more, could “change the risk profile\\nof the population,” and that CASAC was “conditional” about\\nwhether “the lung function decrements observed in some\\npeople at [0.06 ppm] . . . are adverse.” Id. at 65,358. The\\nclinical studies that Environmental Petitioners contend EPA\\ndismissed concluded that lung function decrements (such as\\n10%) occurred in some individuals at lower ozone\\nconcentrations, including 0.06 and 0.063 ppm, but not in\\n“combination [with] statistically significant increases in\\nrespiratory symptoms.” See id. at 65,357. Under ATS\\nguidance, the Administrator declined to designate such loss of\\nlung function as adverse. This decision was rational, and\\nEnvironmental Petitioners cannot show that the evidence\\nrequired EPA to decide differently.\\n\\f 21\\n III. Secondary Standard Challenges\\n\\n Next, we turn to the challenges to EPA’s secondary\\nstandard. The Environmental Petitioners contend that: (1) in\\nconsidering tree growth loss, EPA acted arbitrarily in setting\\nthe target level of air quality and therefore fell short of the\\nstatutory requirement to set a standard “requisite” to protect\\nagainst such harm; (2) EPA arbitrarily used a three-year\\naverage rather than a single-year, cumulative measurement of\\nozone exposure as a benchmark to gauge the protectiveness of\\nits standard, and arbitrarily declined to adopt the single-year\\ncumulative exposure index as the form and averaging time for\\nthe secondary standard; and (3) EPA unreasonably failed to\\nidentify a level of air quality requisite to protect the public\\nwelfare against adverse effects from visible leaf injury.\\nIndustry Petitioners argue only that EPA failed to justify its\\ndecision to lower the secondary standard from its 2008 level.\\nWe hold that EPA has not explained its decision to set a target\\nlevel of protection against tree growth loss based on a three-\\nyear average of cumulative, single-year ozone exposures, nor\\nhas it justified its decision not to specify any level of air quality\\nrequisite to protect against visible leaf injury. We reserve\\njudgment on whether EPA reasonably declined to adopt the\\ncumulative exposure index as the form and averaging time, and\\ndeny the remainder of the challenges.\\n\\n We begin by reviewing EPA’s secondary standard-setting\\nprocess. EPA concentrated its review on the association\\nbetween ozone exposure and “vegetation effects,” focusing on\\ntree growth loss (also referred to as “relative biomass loss”),\\ncrop yield loss, and visible leaf (or “foliar”) injury. 80 Fed.\\nReg. at 65,370-71. CASAC agreed that those three effects\\nwere “appropriate surrogates [for] a wide range of damage that\\nis adverse to public welfare” because ozone damage to trees,\\nleaves, and crops can directly affect numerous resources and\\n\\f 22\\necosystem services that are important to the public and\\nindirectly affect a wide array of ecosystem components and\\nfunctions. J.A. 532-33.\\n\\n In performing its analysis, EPA used an “exposure metric”\\ncalled the “W126 index” to gauge how differing levels of ozone\\nexposure correspond to effects on tree growth, crop yields, and\\nvisible leaf injury. 80 Fed. Reg. at 65,373. The W126 index,\\nwhich EPA deemed to be the “most biologically relevant\\nmetric[] for consideration of [ozone] exposures eliciting\\nvegetation-related effects,” measures the cumulative amount of\\nozone to which a plant is exposed over a single three-month\\ngrowing season. Id. at 65,373-74. W126 levels are expressed\\nas parts-per-million hours (ppm-hrs). Id.\\n\\n CASAC recommended that EPA use the single-year W126\\nexposure index as the form and averaging time for the\\nsecondary standard, J.A. 518, meaning that compliance with\\nthe standard would be measured based on a single growing\\nseason’s worth of cumulative ozone exposure. CASAC\\nadvised EPA to set the level for the secondary standard within\\na range of 7 ppm-hrs and 15 ppm-hrs. Based on the data\\nCASAC used, that corresponded to median annual tree growth\\nloss between 2% and 5.2%. J.A. 518, 534-36, 631. CASAC\\ncautioned that “at 17 ppm-hrs, the median tree species has 6%\\nrelative biomass loss,” which would be “unacceptably high.”\\nJ.A. 518. Regarding the other two surrogates, CASAC\\ncounseled that, to protect against loss of crop yield, “a level of\\n15 ppm-hrs for the highest 3-month sum in a single year is\\nrequisite,” and to reduce foliar injury, a “level below 10 ppm-\\nhrs is required.” Id. All of those levels were based on single-\\nyear measuring periods, so CASAC advised that if EPA were\\nto base its secondary standard on a three-year average of the\\nrelevant measurements, it should lower the level of the standard\\n\\f 23\\n“to protect against single unusually damaging years that will be\\nobscured in the average.” J.A. 536.\\n\\n EPA agreed with CASAC’s recommendation that 6% tree\\ngrowth loss would be “unacceptably high.” 80 Fed. Reg. at\\n65,406. However, as EPA explained, id. at 65,384, 65,392\\nn.197, CASAC’s advice on that was based in part on a study of\\ncottonwood trees, which are uniquely ozone-sensitive, and\\nCASAC itself advised EPA that the “cottonwood data . . .\\nreceive[d] too much emphasis” in EPA’s analysis, J.A. 533.\\nAfter excluding the cottonwood data, EPA concluded that\\nozone exposure of 19 ppm-hrs was associated with 6% tree\\ngrowth loss, and that 17 ppm-hrs of ozone exposure brought it\\ndown to 5.3% loss. 80 Fed. Reg. at 65,396, 65,407. EPA\\naccordingly chose to focus on a standard “somewhat below”\\nthe 19 ppm-hrs level associated with 6% median tree growth\\nloss. Id. at 65,406-07.\\n\\n EPA then departed from CASAC’s advice in several ways.\\nFirst, it chose not to use the W126 cumulative seasonal\\nexposure index as the form and averaging time of the standard,\\nopting instead to retain the averaging time (8 hours) and form\\nit had used in the 2008 rule (the three-year average of the\\nfourth-highest daily level). EPA purported to use as a\\n“benchmark,” or target level of protectiveness, the high end of\\nthe range of W126 levels CASAC had recommended. Recall\\nthat CASAC had recommended that the maximum seasonal\\nozone exposure not exceed (the cottonwood-adjusted) 17 ppm-\\nhrs in any single growing season. In deciding what exposure\\nlevel correlated to CASAC’s 17 ppm-hrs, however, EPA used\\na three-year average of anticipated seasonal exposure levels.\\nThus, EPA chose to set the standard at a level that it projected\\n(based on a statistical analysis of past ozone exposure data)\\nwould “in nearly all instances” going forward restrict the\\naverage of three growing seasons’ ozone exposures to the\\n\\f 24\\nequivalent of 17 ppm-hrs. Id. at 65,407. But by using that\\nmethod to set the level, EPA arrived at a standard that\\nstatistically tolerates cumulative ozone exposures in a single\\ngrowing season that are higher than CASAC’s maximum\\nrecommended level. EPA also did not specify any level of\\nozone to protect against visible leaf injury or crop loss. The\\nagency reasoned that data for those public-welfare harms were\\ntoo uncertain to permit it to discern a level that would provide\\nthe requisite protections, and that the standard it set to protect\\nagainst adverse effects from tree growth loss would at least\\nprovide “some increased protection” against visible leaf injury\\nand crop damage. Id. at 65,407-08.\\n\\n In sum, EPA ultimately chose to set the level of the\\nsecondary standard at 0.07 ppm while retaining the form and\\naveraging time it had previously used. Id. at 65,410.\\n\\nA. Environmental Petitioners’ Challenges\\n\\n The essence of the Environmental Petitioners’ petition is\\nthat EPA did not adequately explain its deviations from\\nCASAC’s advice. “In order to enable judicial review and to\\nsatisfy its statutory obligation to explain its reasons for\\ndeparting from CASAC, EPA must be precise in describing the\\nbasis for its disagreement with CASAC.” Mississippi, 744 F.3d\\nat 1355. Where EPA diverges from CASAC’s scientific\\nadvice, there must be “substantial evidence in the record when\\nconsidered as a whole which supports the Administrator’s\\ndeterminations.” Id. (quoting Lead Indus. Ass’n v. EPA, 647\\nF.2d 1130, 1146 (D.C. Cir. 1980)).\\n\\f 25\\n EPA reasonably used 17 ppm-hrs as the\\n benchmark level to protect against adverse effects\\n associated with tree growth loss.\\n\\n The Environmental Petitioners argue that, by establishing\\n17 ppm-hrs as the target level of protection against adverse\\nwelfare effects associated with tree growth loss, EPA\\nimpermissibly departed upward from CASAC’s advice in two\\nways—first, by excluding the data from the cottonwood tree\\nstudy from its tree growth loss analysis and, second, by\\nfocusing on limiting tree growth loss to under 6%, rather than\\nthe 2% target they say CASAC’s analysis required. Setting\\naside for the moment EPA’s decision to average the benchmark\\nover three years, we conclude that EPA adequately explained\\nits decision to exclude the cottonwood data and acted\\nconsistently with CASAC’s advice in choosing to limit tree\\ngrowth loss to under 6%.\\n\\n Cottonwood Data. EPA reviewed seedling studies of\\ntwelve different tree species to determine the median percentile\\nof tree growth loss at varying ozone levels. See 80 Fed. Reg.\\nat 65,372. One of those studies—of cottonwood seedlings—\\nsignificantly lowered the ozone level at which the twelve tree\\nspecies experience 6% median growth loss. See id. at 65,384,\\n65,391-92. CASAC had itself relied on that study, see J.A.\\n537, but advised EPA that the cottonwood data had received\\n“too much emphasis” because the study “did not control for\\nozone and climatic conditions,” and the results “show[ed]\\nextreme sensitivity to ozone compared to other studies,” J.A.\\n533. EPA accordingly excluded the cottonwood data from its\\ntree growth analysis, which increased the ozone level\\nassociated with 6% median growth tree loss from 17 ppm-hrs\\nup to 19 ppm-hrs. 80 Fed. Reg. at 65,384, 65,391-92.\\n\\f 26\\n The Environmental Petitioners object that EPA arbitrarily\\ndisregarded CASAC’s advice because, they say, CASAC never\\nexpressly advised EPA to disregard the cottonwood data\\naltogether, and CASAC relied on it in setting its recommended\\nexposure range. But we can only discern an unreasonable\\ndeviation where CASAC itself has been “precise about the\\nbasis for its recommendations.” Mississippi, 744 F.3d at 1358.\\nIt was far from precise on the disputed point. It warned EPA\\nto treat the cottonwood data with caution, but recommended a\\nrange of ozone levels in reliance on the median tree growth loss\\nestimates (outlined in Table 6-1 of EPA’s staff’s second draft\\nPolicy Assessment), which weigh the cottonwood data equally\\nwith other seedling studies. See J.A. 537 (citing EPA, Policy\\nAssessment for the Review of the Ozone NAAQS, Second\\nExternal Review Draft 6-19, tbl. 6-1 (2014), J.A. 631). The\\nEnvironmental Petitioners suggest that CASAC was only\\nwarning EPA to take care in describing the cottonwood data in\\nthe narrative portion of its Policy Assessment, not suggesting\\nanything about how the data was actually used, but that makes\\nlittle sense in the context of an exchange about quantitative\\ngrowth loss estimates. In light of CASAC’s mixed messages,\\nwe cannot say that EPA’s decision to exclude the cottonwood\\ndata was arbitrary.\\n\\n The 6% Target. Emphasizing CASAC’s admonition that\\n6% tree growth loss would be “unacceptably high,” EPA\\ndecided “to focus on a standard that would generally limit\\ncumulative exposures to those for which the median [tree\\ngrowth loss] estimate would be somewhat lower than 6%.” 80\\nFed. Reg. at 65,406-07. The Environmental Petitioners\\ncontend that CASAC “plainly specified” that 2% median\\ngrowth loss was the requisite level to avoid adverse welfare\\neffects related to tree growth loss, Envtl. Pet’rs Br. 42-44, but\\nthey misconstrue CASAC’s advice.\\n\\f 27\\n CASAC did not identify 2% growth loss as the only\\nsufficiently protective level, but as “an appropriate\\nscientifically based value to consider as a benchmark of\\nadverse impact,” elaborating that it would be “appropriate to\\nidentify a range of levels of alternative W126-based standards\\nthat includes levels that aim for not greater than 2% [growth\\nloss] for the median tree species.” J.A. 537 (emphases added).\\nCASAC in fact ultimately recommended that EPA set the\\nstandard between 7 ppm-hrs and 15 ppm-hrs, J.A. 518, which,\\nbased on the seedling data upon which CASAC relied\\n(including the cottonwood data), corresponds to median annual\\ntree growth loss between 2% and 5.2%, see J.A. 631. In other\\nwords, CASAC recommended that EPA consider limiting tree\\ngrowth loss to 2% as the lower end of a range of permissible\\ntarget levels. EPA followed CASAC’s advice by considering\\na 2% growth loss target as part of a range of growth loss targets\\nand determined that the studies underlying CASAC’s low-end\\nrecommendation were insufficiently reliable to base the\\nsecondary standard on that low level of tree growth loss. See\\n80 Fed. Reg. at 65,394.\\n\\n The Environmental Petitioners passingly disparage EPA’s\\nassessment of the studies associated with CASAC’s 2% growth\\nloss target, but that argument is equally unavailing. EPA\\nreasonably explained why it thought those studies were\\nunreliable, and its “evaluation of ‘scientific data within its\\ntechnical expertise’” is entitled significant deference. See\\nMiss. Comm’n on Envtl. Quality v. EPA, 790 F.3d 138, 150,\\n155-56 (D.C. Cir. 2015) (quoting City of Waukesha v. EPA,\\n320 F.3d 228, 247 (D.C. Cir. 2003)). For instance, in support\\nof its advice that EPA consider limiting median annual tree\\ngrowth loss to 2%, CASAC highlighted a study purporting to\\nshow that stands of aspen that experienced ozone-induced\\nbiomass loss of 2% annually had cumulative biomass loss of\\n21% over seven years, suggesting that even relatively minor\\n\\f 28\\nannual growth loss can lead to significant loss over longer\\nperiods of time. J.A. 537. But CASAC had misread the study.\\nAs EPA explained, the aspen in that study in fact experienced\\nmore than 20% biomass loss annually, suggesting that their\\ncumulative biomass loss was not much worse than their loss in\\nany single year. 80 Fed. Reg. at 65,394.\\n\\n EPA also clarified that the only other report CASAC relied\\non for its 2% target recommendation provided no explicit\\nrationale for selecting that level and did not identify any new\\ndata in support of its recommendations. See id. at 65,394-95 &\\nn.200. While it is true that EPA has cited that same report\\nfavorably in the past, see, e.g., National Ambient Air Quality\\nStandards for Ozone, 75 Fed. Reg. 2,938, 3,025 (Jan. 19, 2010),\\nthat alone does not make EPA’s current assessment arbitrary.\\nEPA must simply provide a “reasoned explanation” for its\\ndeparture from its past position—which it has done. See\\nEncino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125-26\\n(2016); see also Mississippi, 744 F.3d at 1351.\\n\\n EPA did not adequately explain its decision to use\\n a three-year average of the W126 index as the\\n benchmark.\\n\\n The Environmental Petitioners raise two challenges\\npertaining to the W126 index, both as a benchmark and as a\\nform and averaging time. First, they argue that EPA\\nimpermissibly departed from CASAC’s advice by setting the\\nsecondary standard level using a three-year average W126\\nbenchmark without lowering the level to protect against single-\\nyear exposures associated with median annual tree growth loss\\nof 6%, which CASAC had advised was “unacceptably high.”\\nJ.A. 518. They also contend that EPA arbitrarily disregarded\\nCASAC’s advice to adopt the W126 index as the form and\\n\\f 29\\naveraging time for the secondary standard. We remand to EPA\\non the first issue and do not reach the second.\\n\\n The Three-Year Average Benchmark. CASAC advised\\nEPA that basing the secondary standard on a “single-year\\nperiod” would provide “more protection for annual crops and\\nfor the anticipated cumulative effects on perennial species”\\nthan a three-year average. J.A. 518. It explained that EPA’s\\nproposal to use a three-year averaging period was “not\\nsupported by the available data,” J.A. 536, and that if EPA\\nchose to “base the secondary standard on a three-year\\naveraging period,” then “the level of the standard should be\\nrevised downward such that the level for the highest three-\\nmonth summation in any given year of the three-year period\\nwould not exceed [its] scientifically recommended” range of\\nsingle-year, W126 exposure levels, J.A. 518. This was\\nnecessary, CASAC explained, to “protect against single\\nunusually damaging years that will be obscured in the\\naverage.” J.A. 536.\\n\\n EPA argues it gave effect to CASAC’s recommendation\\nby using a three-year average benchmark of 17 ppm-hrs,\\nwhich, after adjusting for the cottonwood data, was “somewhat\\nbelow” the 19 ppm-hrs associated with 6% annual growth loss\\nthat CASAC had advised was “unacceptably high.” 80 Fed.\\nReg. at 65,406-07 (quoting J.A. 518). But CASAC had advised\\na maximum level associated with 5.2% annual biomass loss,\\nsee J.A. 631, and it expressly cautioned that 6% median growth\\nloss in a single year was unacceptable, see J.A. 518. EPA’s use\\nof a benchmark that averages out to less than 6% biomass loss\\nover three years does not accord with CASAC’s advice.\\nIndeed, as commenters informed EPA during the rulemaking,\\nsee J.A. 1836-40, EPA’s own air quality data suggests that\\nmany large national parks and wilderness areas that have met\\nEPA’s chosen three-year average 17 ppm-hrs benchmark—\\n\\f 30\\nareas that Congress considers significant to the public welfare,\\nsee 42 U.S.C. §§ 7470(2), 7472(a)—have meanwhile recorded\\nsingle-year W126 values at and above 19 ppm-hrs, which is\\nassociated with “unacceptably high” annual biomass loss of\\n6% and higher. See J.A. 1061-64; 80 Fed. Reg. at 65,391. EPA\\ncritiques that data as marred by outdated handling procedures,\\nbut the agency acknowledged that other data derived through\\nupdated procedures produced results “similar to” those\\nshowing harmful exposure spikes. J.A. 1213. Critically, EPA\\npoints us to no data or analysis (based on new or old\\nprocedures) suggesting that the chosen benchmark prevents\\nsingle seasonal exposures of 19 ppm-hrs or higher. In short,\\nEPA has not demonstrated how its chosen benchmark protects\\nagainst “unusually damaging years that will be obscured in the\\naverage.” J.A. 536.\\n\\n EPA alternatively defends its decision to use the three-\\nyear-average benchmark as providing a focus on public welfare\\neffects of “potentially greater” significance than effects\\n“associated with a single year” of exposure. 80 Fed. Reg. at\\n65,404. This position, however, is inconsistent with EPA’s\\nother actions. In establishing the secondary standard, for\\nexample, EPA heavily relied on data showing annual 6%\\nmedian tree growth loss at 19 ppm-hrs, see id. at 65,406, and\\nacknowledged the potential for a single season of high ozone\\nexposure to “alter biomass allocation and plant reproduction in\\nseasons subsequent to [that season’s] exposure,” thereby\\nleading to “a negative impact on species regeneration in\\nsubsequent years,” id. at 65,371-72; see also J.A. 740-41. EPA\\nadditionally recognized that “ozone effects in plants are\\ncumulative,” id. at 65,373 (quoting EPA, Integrated Science\\nAssessment 2-44 (2013)), meaning that the adverse vegetative\\neffects from single, high-ozone years are not offset by\\nsubsequent low-ozone years. EPA has identified no contrary\\n\\f 31\\nevidence in the record demonstrating why these single-year\\neffects matter less than a three-year average.\\n\\n We therefore remand this issue for EPA to either lower the\\nstandard to protect against unusually damaging cumulative\\nseasonal exposures that will be obscured in its three-year\\naverage, or explain its conclusion that the unadjusted average\\nis an appropriate benchmark notwithstanding CASAC’s\\ncontrary advice. Alternatively, EPA could adopt the single-\\nyear W126 exposure index as the form and averaging time,\\nwhich would presumably moot any problems with the way it\\ntranslated that index to use as a benchmark.\\n\\n The Form and Averaging Time. CASAC also\\nrecommended that EPA use the single-year W126 index as the\\nform and averaging time for the secondary standard. J.A. 518.\\nEPA instead chose to retain the existing form and averaging\\ntime—the three-year average of the fourth-highest daily\\nmaximum eight-hour concentration. Adopting the W126 index\\nas the form and averaging time was unnecessary, EPA\\nexplained, because the ozone exposure levels associated with\\nthe existing form and averaging time and a three-year average\\nof the W126 index are “highly correlated,” especially at lower\\nlevels, and “future control programs designed to help meet a\\nprimary [ozone] standard based on the” current form and\\naveraging time should “provide similar improvements in terms\\nof the 3-year average of the annual W126 metric.” J.A. 1253;\\nsee also 80 Fed Reg. at 65,400-01, 408-09.\\n\\n The Environmental Petitioners argue that EPA did not\\njustify its decision not to adopt the W126 index as the form and\\naveraging time. We lack any basis to assess the reasonableness\\nof EPA’s actions, however, because a critical piece of the\\npuzzle is missing. To review: EPA chose not to use the W126\\nindex as the form and averaging time because it found that\\n\\f 32\\nozone exposure levels associated with the existing form and\\naveraging time are “highly correlated” to a three-year average\\nof the W126 index. But, as discussed, EPA never explained\\nwhy it is reasonable to focus on a three-year average of the\\nW126 index in the first place. Therefore, we cannot assess the\\nrelevance of the claim that the two are “highly correlated.”\\nEPA’s reconsideration on remand of the three-year averaging\\nissue should supply us with the information necessary to\\nresolve this question, or, if EPA chooses to follow CASAC’s\\nadvice to lower the standard to control for unusually high single\\nyears, potentially moot the Environmental Petitioners’ concern\\nthat the current form tolerates even three-year average W126\\nlevels higher than 17 ppm-hrs during periods when a 0.07 ppm,\\n8-hour level is met. Accordingly, we decline to reach this\\nquestion.\\n\\n EPA arbitrarily failed to identify a level of air\\n quality requisite to protect against adverse effects\\n from visible leaf injury.\\n\\n The Environmental Petitioners also fault EPA for deciding\\nnot to specify a level of air quality to protect against adverse\\nwelfare effects from ozone-induced visible leaf injury.\\nBecause EPA failed to offer a reasoned explanation, we remand\\nfor it to reconsider.\\n\\n EPA has found that the “strongest evidence for effects\\nfrom [ozone] exposure on vegetation is from controlled\\nexposure studies, which ‘have clearly shown that exposure to\\n[ozone] is causally linked to visible foliar injury.’” 80 Fed.\\nReg. at 65,370 (quoting EPA, Integrated Science Assessment\\n1-15). Relying on that evidence, CASAC advised EPA of its\\n“scientific judgment” that “a level of 10 ppm-hrs is required to\\nreduce foliar injury.” J.A. 538. EPA nonetheless concluded\\nthat there were too many “uncertainties and complexities” in\\n\\f 33\\nthe evidence to specify a level of air quality to protect against\\nleaf injury. See 80 Fed. Reg. at 65,407-08. But the mere\\ninvocation of “substantial uncertainty” is not a justification for\\nthe agency’s failure to fulfill its statutory mandate. Motor\\nVehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins.\\nCo., 463 U.S. 29, 52 (1983) (internal quotation marks omitted).\\nEPA “must explain the evidence which is available, and must\\noffer a ‘rational connection between the facts found and the\\nchoice made.’” Id. (quoting Burlington Truck Lines, Inc. v.\\nUnited States, 371 U.S. 156, 168 (1962)). Where CASAC has\\n“reached a scientific conclusion that adverse [welfare] effects\\n[are] likely to occur,” EPA must “explain why the evidence on\\nwhich CASAC relied cannot support the degree of confidence\\nCASAC placed in it.” Mississippi, 744 F.3d at 1357; see also\\nAm. Farm Bureau Fed’n v. EPA, 559 F.3d 512, 529-30 (D.C.\\nCir. 2009). We defer to EPA’s judgment that the available\\nevidence is too uncertain only when the agency reasonably\\nexplains its decision, see State Farm, 463 U.S. at 51-53, and\\nEPA has failed to carry that burden here.\\n\\n None of the identified uncertainties justifies EPA’s\\ndecision not to set a standard to protect against visible leaf\\ninjury. For instance, EPA argues that it lacked criteria for\\nassessing the welfare effects of visible leaf injury, but that does\\nnot square with its conclusion in the Integrated Science\\nAssessment (compiling all available scientific criteria) that\\nvisible leaf injury “has been well characterized and\\ndocumented over several decades.” J.A. 985-86. EPA had at\\nits fingertips a wealth of new data regarding visible leaf injury,\\nincluding new controlled exposure studies, multiyear field\\nsurveys, and biomonitoring data of ozone-induced visible leaf\\ninjury on public lands. J.A. 749. Using that data, EPA\\nanalyzed how visible leaf injury affects “ecosystem services”\\nlike aesthetic value and recreation on “Class I” lands—lands\\nthat “have particular public welfare significance.” J.A. 749-50,\\n\\f 34\\n1036-44; see also 42 U.S.C. §§ 7470(2), 7472(a). CASAC\\npointed out that those same ecosystem services are relevant in\\nidentifying “damage that is adverse to public welfare.” J.A.\\n532-33.\\n\\n CASAC concluded, based on the same data available to\\nEPA, that “a level of 10 ppm-hrs is required to reduce foliar\\ninjury.” J.A. 538. EPA counters that “decreases in leaf injury\\n[also occur] with decreasing ozone exposures across a range of\\nvalues well above 10 ppm-hrs,” and that CASAC\\nrecommended a range of ozone levels that included values\\nabove 10 ppm-hrs. See EPA Br. 89; 80 Fed. Reg. at 65,395-\\n96. Be that as it may, EPA has not explained why, unlike\\nCASAC, it was unable to choose a level at all. Indeed,\\n“[b]ecause the EPA failed to identify any target level, we need\\nnot decide whether it was reasonable for the agency to reject\\nthe target recommended by . . . CASAC.” Am. Farm Bureau,\\n559 F.3d at 530.\\n\\n EPA objects that it only found that visible leaf injury has\\nthe “potential to be adverse to the public welfare,” and that the\\nClean Air Act does not require EPA “to identify a precise,\\nquantified level of public welfare protection for every\\npotentially adverse public welfare impact” it considers when\\nrevising the secondary standard. EPA Br. 87-88 (first quoting\\n80 Fed. Reg. at 65,388). But EPA has failed to justify its\\ninaction where CASAC unequivocally found that “damage to\\nresource use from foliar injury” was an “adverse welfare\\neffect.” J.A. 518. In light of CASAC’s “scientific conclusion\\nthat adverse [welfare] effects were likely to occur” from visible\\nleaf injury, EPA’s inaction in the face of such effects is\\n“unacceptable.” Mississippi, 744 F.3d at 1357. Nor was it\\nsufficient for EPA simply to conclude that the standard set to\\nprotect against tree growth loss sufficed because it provided\\n“additional” protection against leaf injury. 80 Fed. Reg. at\\n\\f 35\\n65,407-08. Nothing in the record suggests that tree growth loss\\nis a surrogate for leaf injury. EPA’s assurance that the standard\\nset to protect against tree growth loss will provide\\n“incidental[]” protection against visible leaf injury “cannot\\nsave its decision.” Am. Farm Bureau, 559 F.3d at 529-30.\\n\\n Other purported uncertainties do not support EPA’s\\ninaction. For instance, EPA asserts that the Administrator\\n“lacked evidence that would allow her to measure the\\nrelationship between leaf injury and other vegetation effects\\nthat she might find adverse,” EPA Br. 87, but EPA and CASAC\\nboth determined that visible leaf injury itself can “impact the\\npublic welfare” by harming “aesthetic value and outdoor\\nrecreation” in public lands, 80 Fed. Reg. at 65,379; J.A. 533.\\nVisible leaf injury impairs a variety of outdoor activities,\\nincluding “scenic viewing, wildlife watching, hiking, and\\ncamping, that are of significance to the public welfare” and\\ngenerate “millions of dollars in economic value” annually. 80\\nFed. Reg. at 65,381; J.A. 749. Much of the documented visible\\nleaf injury is in Class I areas, see 80 Fed. Reg. at 65,378, the\\npreservation and enjoyment of which is a “clear public\\ninterest,” id. at 65,377 (quoting 73 Fed. Reg. at 16,496). Those\\ninterests are manifest whether or not the relationship between\\nleaf injury and other vegetative effects has been well\\ncharacterized.\\n\\n This case is a far cry from Center for Biological Diversity\\nv. EPA, where we upheld EPA’s determination that the\\navailable data was too uncertain to support setting a standard\\nfor acid rain precursors. 749 F.3d 1079 (D.C. Cir. 2014). In\\nthe administrative record in that case, EPA “explained in great\\ndetail” why the scientific uncertainties were so “unusually\\nprofound” that EPA “could not form” a reasoned judgment as\\nto a requisite level of protection. Id. at 1088, 1090-91. CASAC\\nconcurred with that assessment. Id. at 1086 n.11. Here, in\\n\\f 36\\ncontrast, CASAC expressed its scientific judgment that a target\\nlevel of 10 ppm-hrs was “required to reduce” leaf injury. J.A.\\n538. EPA’s failure to explain why it could not accept data that\\nCASAC deemed informative falls short of reasoned\\ndecisionmaking.\\n\\nB. Industry Petitioners’ Challenge\\n\\n The Industry Petitioners glancingly claim that EPA failed\\nto explain why, when faced with the same scientific evidence\\nthat was available in 2008, it revised the secondary standard\\ndownward. As we have explained in reference to the primary\\nstandard, it is a “conceptual error” to assume that “EPA is\\nsomehow bound by” a prior standard in a subsequent review.\\nMississippi, 844 F.3d at 1344. What is more, EPA did in fact\\nrely on new scientific evidence, including more than 400 new\\nstudies and various new analyses of existing data, which\\n“strengthen[ed] [its] confidence” that the 2008 secondary\\nstandard was no longer as demanding as was “requisite” to\\nprotect the public welfare. See 80 Fed. Reg. at 65,369-70,\\n65,384. “[A]dditional certainty . . . that the line marked by the\\nterm ‘requisite’ has shifted” is a reasonable—if not\\nparadigmatic—basis for EPA to revise its prior standards.\\nMississippi, 844 F.3d at 1344.\\n\\n IV. Cross-Cutting NAAQS Challenges\\n\\n We now turn to the arguments of State and Industry\\nPetitioners that EPA failed to take into account all the factors\\nrequired by law when setting the primary and secondary\\nstandards. First, they fault EPA for not considering the “overall\\nadverse economic, social, and energy impacts” of the revised\\nNAAQS. Second, they argue that the Clean Air Act requires\\nEPA to consider the impact of background ozone on the ability\\nof states to attain the revised NAAQS. And, lastly, they argue\\nthat EPA’s interpretation of the Act, which excluded\\n\\f 37\\nconsideration of background ozone, leaves no “intelligible\\nprinciple” by which to set NAAQS, creating a constitutional\\nnondelegation issue. None of these arguments has merit.\\n\\nA. Adverse Impacts\\n\\n According to section 109(b) of the Clean Air Act, EPA\\nmust set the NAAQS at a level “requisite to protect the public\\nhealth” and “the public welfare.” 42 U.S.C. § 7409(b). The\\nNAAQS should be revised periodically “as may be\\nappropriate.” Id. § 7409(d)(1). Industry Petitioners invoke the\\nterm “appropriate” and claim that “an evaluation of\\n‘appropriateness’ must take into account the adverse\\nsocioeconomic and energy impacts” of the revised NAAQS.\\nIndus. Pet’rs Br. 33. To bolster this argument, they point to\\nMichigan v. EPA, which held that the phrase “appropriate and\\nnecessary” in section 112(n) of the Act—governing power\\nplants’ hazardous air pollution—“requires at least some\\nattention to cost” because “‘appropriate’ is ‘the classic broad\\nand all-encompassing term that naturally and traditionally\\nincludes consideration of all the relevant factors.’” 135 S. Ct.\\nat 2707 (quoting White Stallion Energy Ctr., LLC v. EPA, 748\\nF.3d 1222, 1266 (D.C. Cir. 2014) (Kavanaugh, J., concurring\\nin part and dissenting in part)).\\n\\n In Whitman, the Supreme Court considered a similar\\nargument regarding NAAQS-setting, with challengers\\nasserting that the words “requisite” and “adequate” in section\\n109(b)(1) meant that EPA must consider the economic costs of\\nimplementing revised NAAQS. 531 U.S. at 468. The Whitman\\nCourt rejected this reading, finding that the plain text of the Act\\n“unambiguously bars cost considerations from the NAAQS-\\nsetting process.” Id. at 471. The text of the Act refers only to\\nthe “public health,” and though it uses the terms “requisite” and\\n“adequate,” the Whitman Court found it “implausible that\\n\\f 38\\nCongress would give to the EPA through these modest words\\nthe power to determine whether implementation costs should\\nmoderate national air quality standards,” especially given that\\nother Clean Air Act provisions “explicitly permit[] or require[]\\neconomic costs to be taken into account in implementing\\n[other] air quality standards.” Id. at 467-68.\\n\\n According to Whitman, the Act commands EPA to follow\\na specific process when setting primary NAAQS: “[I]dentify\\nthe maximum airborne concentration of a pollutant that the\\npublic health can tolerate, decrease the concentration to\\nprovide an ‘adequate’ margin of safety, and set the standard at\\nthat level.” Id. at 465. Industry Petitioners argue that this\\naccount of the NAAQS-setting process is incomplete, and that\\nconsideration of other “adverse economic, social, and energy\\nimpacts” is required. But at bottom this is the same argument\\nrejected in Whitman, with the “costs” of the revised and more\\nstringent NAAQS merely reframed as “impacts.” In listing\\nthose impacts, Industry Petitioners emphasize that stricter\\nstandards could reduce gross domestic product, drive up energy\\ncosts, “stymie economic growth by forcing the early retirement\\nof facilities unable to implement controls, contributing to job\\nlosses[,] discourag[e] existing businesses from expanding in\\nnonattainment regions[,] and driv[e] away potential new\\ninvestments,” Indus. Pet’rs Br. 35, but these “impacts” are no\\ndifferent than the “economic costs” that the petitioners in\\nWhitman worried “might produce health losses sufficient to\\noffset the health gains achieved in cleaning the air—for\\nexample, by closing down whole industries and thereby\\nimpoverishing the workers and consumers dependent upon\\nthose industries,” 531 U.S. at 466. Whitman forbids EPA from\\ntaking these considerations into account, however\\ndenominated.\\n\\f 39\\n As for Petitioners’ reliance on Michigan and the Act’s use\\nof the word “appropriate,” that argument fails twice over. We\\nhave already rejected the idea that “appropriate” in section\\n109(d) requires consideration of economic costs. Am. Trucking\\nAss’ns, Inc. v. EPA, 175 F.3d 1027, 1040-41 (D.C. Cir. 1999).\\nFurther, Michigan involved a different provision of the Clean\\nAir Act, and the Court was careful to emphasize that its reading\\nof “appropriate” was dependent on the statutory context,\\nexplaining that “[t]here are undoubtedly settings in which the\\nphrase ‘appropriate and necessary’ does not encompass cost.”\\n135 S. Ct. at 2707. Indeed, Michigan explicitly distinguished\\nsection 112(n) from section 109(b)(1), explaining yet again that\\nthe criteria for setting NAAQS “do[] not encompass cost.” Id.\\nat 2709.\\n\\n Industry Petitioners also point to section 109(d)(2)(C) of\\nthe Act, which requires CASAC to advise EPA “of any adverse\\npublic health, welfare, social, economic, or energy effects\\nwhich may result from various strategies for attainment and\\nmaintenance” of revised NAAQS. 42 U.S.C. § 7409(d)(2)(C).\\nAccording to Petitioners, the fact that CASAC is required to\\nsupply information to EPA about the “social, economic, or\\nenergy effects” of the revised NAAQS implies that EPA is\\nobliged to consider that information in setting the NAAQS.\\nThat argument, however, was also raised and rejected in\\nWhitman, where the Supreme Court concluded that this\\nprovision was intended to “enable the [EPA] to assist the States\\nin carrying out their statutory role as primary implementers of\\nthe NAAQS,” but had “no bearing upon whether cost\\nconsiderations are to be taken into account in formulating the\\n[NAAQS].” 531 U.S. at 470-71 (second emphasis added).\\n\\n Petitioners’ remaining arguments rely heavily on Justice\\nBreyer’s concurrence in Whitman, in which he urged that the\\nAct “does not require the EPA to eliminate every health risk,\\n\\f 40\\nhowever slight, at any economic cost, however great, to the\\npoint of hurtling industry over the brink of ruin, or even forcing\\ndeindustrialization.” Id. at 494 (Breyer, J., concurring in part\", \"author\": \"PER CURIAM\"}, {\"type\": \"concurrence\", \"author\": \"and\", \"text\": \"and concurring in the judgment) (internal quotation marks\\nomitted). But the concurrence does not govern our decision,\\nand in any event, Justice Breyer agreed with the majority that\\neconomic costs could not be considered in setting NAAQS. Id.\\nat 490.\\n\\nB. Background Ozone\\n\\n Under the Clean Air Act, “background ozone” is ozone\\n“that would exist in the absence of any man-made emissions\\ninside the U.S.” 80 Fed. Reg. at 65,436. This includes ozone\\ngenerated by both natural sources anywhere (e.g., a wildfire)\\nand foreign man-made sources (e.g., a factory in Mexico). See\\nid. According to the Industry and State Petitioners, EPA failed\\nto take into account background ozone when setting the new\\nNAAQS. As a result, certain areas of the country with high\\nlevels of background ozone fail right off the bat to meet the\\nnew standards, which is unlawful, Petitioners argue, because\\nthe Act requires EPA to set NAAQS that are attainable.\\nTherefore, they contend EPA’s failure to consider background\\nozone when revising the NAAQS was arbitrary and capricious.\\n\\n The text of the Act forecloses this argument. Section\\n109(b) directs EPA to set NAAQS “requisite to protect the\\npublic health” and “the public welfare.” 42 U.S.C. § 7409(b).\\nAccepting Petitioners’ argument would mean that, if the level\\nof background ozone in any part of the country exceeds the\\nlevel of ozone that is “requisite to protect the public health,”\\nEPA must set the NAAQS at the higher, unhealthy level. The\\nstatutory text leaves no room for this hidden caveat: “[W]hen\\nCongress directs an agency to consider only certain factors in\\nreaching an administrative decision, the agency is not free to\\n\\f 41\\ntrespass beyond the bounds of its statutory authority by taking\\nother factors into account.” Lead Indus., 647 F.2d at 1150.\\n\\n Petitioners argue that states cannot be required to do the\\nimpossible, and the presence of background ozone will make it\\nimpossible to achieve attainment. But Congress recognized the\\npossibility that some states could not achieve attainment\\nbecause of the presence of background ozone and, rather than\\nwatering down the nationally applicable standards, allowed\\nEPA to relax enforcement on a case-by-case basis. When\\nozone exceedances are caused by events beyond a state’s\\ncontrol, three enforcement exceptions in the Act allow the state\\nsome leeway as a practical matter: A showing of an\\n“Exceptional Event” may allow a state to avoid a\\nnonattainment designation. See 42 U.S.C. § 7619(b) (defining\\n“exceptional event” as one caused by “human activity that is\\nunlikely to recur at a particular location” or “a natural event,”\\nand that “affects air quality” but is “not reasonably controllable\\nor preventable”). And a state that meets the requirements of\\neither the “International Transport” or “Rural Transport”\\nprovision may avoid having its state implementation plan (SIP)\\nrejected due to the exceedance. See id. §§ 7509a(a)-(b)\\n(directing EPA to approve a SIP if attainment is not met due to\\n“emissions emanating from outside of the United States”),\\n7511a(h) (excluding nonattainment regions that themselves\\n“do not make a significant contribution” to ozone levels, so\\nlong as they are not within or next to a “Metropolitan Statistical\\nArea”). These provisions make little sense under Petitioners’\\nreading of the Act. And even if, as the states claim, it is more\\ndifficult to meet the terms of these exceptions than EPA asserts,\\nState Pet’rs Br. 34-44, the fact remains that Congress decided\\nthat EPA should account for background ozone during\\nenforcement, not when setting standards.\\n\\f 42\\n We rejected a version of Petitioners’ argument in\\nAmerican Petroleum Institute v. Costle, 665 F.2d 1176 (D.C.\\nCir. 1981). There, the City of Houston urged us to find that\\nEPA’s primary ozone standard was arbitrary because “natural\\nfactors ma[de] attainment impossible” for that City. Id. at\\n1185. We emphasized that “Congress [was] aware that some\\nregions [were] having difficulty in meeting the national\\nstandard,” and had responded not by requiring EPA to ease the\\nNAAQS but by developing a distinct program for\\nnonattainment areas that gives states resources to bring those\\nareas into compliance while also protecting the public health\\nand welfare. Id. at 1185-86. Industry Petitioners argue that\\nCostle, a case about one city, is not binding when “numerous\\nareas of the country cannot attain the NAAQS due to\\nbackground levels” of ozone. Indus. Pet’rs Br. 26-27\\n(emphasis added). But neither our precedent nor the Act allows\\nfor such a distinction. Indeed, in Costle, we rejected the\\nargument that EPA had ignored comments related to the issue\\nof whether “attainment of the proposed standards would be\\nprecluded in most areas of the nation by natural background\\nlevels of ozone.” 665 F.2d at 1190 (emphasis added) (internal\\nquotation marks omitted). Simply put, “the question of\\nattainability is not relevant to the setting of ambient air quality\\nstandards under the Clean Air Act.” Id.\\n\\n State Petitioners argue that Costle is distinguishable\\nbecause the petitioner there was not a state, and several parts of\\nthe Act place the burden to attain NAAQS on states rather than\\ncities. Indus. Pet’rs Br. 25-26 (citing 42 U.S.C. §§ 7407(a),\\n7410(a)(2)(C)); State Pet’rs Br. 32-33 (citing 42 U.S.C.\\n§ 7407(a)). True, if Houston cannot attain NAAQS due to\\nbackground ozone, the Act places the burden on Texas, not the\\nCity, to identify a plan to bring the area into attainment. But\\nCostle was not limited to its particular facts—it relied instead\\non the premise that “[a]ttainability and technological feasibility\\n\\f 43\\nare not relevant considerations in the promulgation of\\n[NAAQS],” 665 F.2d at 1185, a premise that does not change\\nwhether it is a state or city arguing that the NAAQS are\\nunattainable. Congress intended NAAQS to be national\\nambient air quality standards, and EPA is not required to “tailor\\nnational regulations to fit each region or locale.” Id. Because\\nthe Clean Air Act prohibits EPA from adjusting for background\\nozone in setting the NAAQS, EPA did not act unlawfully or\\narbitrarily and capriciously in setting the NAAQS without\\nregard for background ozone.\\n\\nC. Nondelegation\\n\\n Finally, State Petitioners argue that EPA’s interpretation\\nof the Clean Air Act runs afoul of the Constitution’s limit on\\nthe authority of Congress to delegate lawmaking power to an\\nagency. “Congress generally cannot delegate its legislative\\npower to another Branch” unless it provides an “intelligible\\nprinciple” for the agency to follow in exercising that delegated\\npower. Mistretta v. United States, 488 U.S. 361, 372 (1989)\\n(second quoting J.W. Hampton, Jr., & Co. v. United States, 276\\nU.S. 394, 409 (1928)). We refer to this as the nondelegation\\ndoctrine. Id.\\n\\n State Petitioners do not argue that the Act lacks an\\nintelligible principle. Nor could they, as the Supreme Court\\nheld in Whitman that Congress provided one when it directed\\nEPA to set NAAQS “requisite to protect public health”—\\nmeaning “sufficient, but not more than necessary.” 531 U.S. at\\n473-74 (internal quotation marks omitted). Instead, Petitioners\\nclaim that EPA’s interpretation of the Act has created a\\nnondelegation problem by “ignor[ing] the ‘intelligible’\\nprinciple[] that Congress provided.” State Pet’rs Br. 45. But\\nin a nondelegation challenge, “the constitutional question is\\nwhether the statute has delegated legislative power to the\\n\\f 44\\nagency.” Whitman, 531 U.S. at 472 (emphasis added). EPA\\ncannot alter the text of the Clean Air Act, so EPA’s\\ninterpretation of the Act cannot alter whether the Act included\\nan intelligible principle. Cf. id. at 473 (“Whether the statute\\ndelegates legislative power is a question for the courts, and an\\nagency’s [interpretation] has no bearing upon the answer.”).\\nThere is no nondelegation issue here.\\n\\n V. Grandfathering Challenge\\n\\n Finally, we address EPA’s decision to allow those who\\ncompleted applications for preconstruction permits before the\\n2015 Rule was adopted to demonstrate compliance with the\\nprevious NAAQS rather than the new, more stringent\\nstandards. See 80 Fed. Reg. at 65,431-35; 40 C.F.R.\\n§§ 51.166(i)(11), 52.21(i)(12). The Environmental Petitioners\\nargue that this “grandfathering” provision violates the Clean\\nAir Act, which requires permit applicants to demonstrate\\ncompliance with “any” NAAQS regardless of when their\\napplication was completed. See Envtl. Pet’rs Br. 57-62. EPA\\nargues that the Act is ambiguous with respect to the treatment\\nof permit applications pending at the time that new NAAQS\\nare adopted, and that the grandfathering provision is lawful as\\na reasonable interpretation of the Act. We find no such\\nambiguity.\\n\\n Under section 165(a) of the Act, before beginning\\nconstruction of any “major emitting facility,” the owner or\\noperator must demonstrate “that emissions from construction\\nor operation of such facility will not cause, or contribute to, air\\npollution in excess of any . . . national ambient air quality\\nstandard.” 42 U.S.C. § 7475(a). This “demonstration\\nrequirement” is part of EPA’s PSD program. Generally, EPA\\nrequires permit applicants to demonstrate compliance with the\\nNAAQS in effect at the time a permit is issued. 80 Fed. Reg.\\n\\f 45\\nat 65,433; EPA Br. 130. The revised ozone NAAQS were\\neffective on December 28, 2015, so under the usual rule no\\npermit could be granted for the construction of a major emitting\\nfacility after that date unless the operator demonstrated that the\\nfacility would not cause or contribute to ozone levels in excess\\nof the revised NAAQS. But when EPA issued the 2015 Rule,\\nit waived that requirement for projects that either (1) had\\ncomplete permit applications as of October 1, 2015, or (2) had\\na draft permit or preliminary determination publicly noticed\\nbefore December 28, 2015. 1 See 80 Fed. Reg. at 65,460\\n(codified at 40 C.F.R. §§ 51.166(i)(11), 52.21(i)(12)).\\n\\n In the 2015 Rule, EPA asserted the need for this\\ngrandfathering provision because section 165(c) of the Act\\ninstructs EPA and state permitting authorities that “[a]ny\\ncompleted permit application . . . shall be granted or denied not\\nlater than one year after the date of filing such completed\\napplication.” 42 U.S.C. § 7475(c). As EPA sees it, requiring\\napplicants with already-completed permit applications to\\ndemonstrate compliance with the newly revised NAAQS\\n“could hinder compliance with the requirement under\\nsection 165(c) to issue a permit within one year.” 80 Fed. Reg.\\nat 65,433-34. According to EPA, the Act does not “clearly\\naddress” how to handle such a situation, therefore it was\\n“permissible under the discretion provided by the [Act] for the\\nEPA to craft a reasonable implementation regulation that\\nbalances competing objectives of the statutory PSD program.”\\nId. at 65,433.\\n\\n We evaluate EPA’s interpretation of the Act under\\nChevron. “If the Act unambiguously authorizes or forecloses\\nEPA’s . . . rule, step one of the Chevron analysis requires that\\n\\n 1\\n The second type of waiver was necessary because not all\\nagencies that issue preconstruction permits issue formal\\ncompleteness determinations. See 80 Fed. Reg. at 65,432-33.\\n\\f 46\\nwe follow Congress’s express policy choice. If the Act is\\nunclear on the matter, step two of Chevron requires that we\\ndefer to EPA’s reasonable interpretation.” Sierra Club v. EPA,\\n536 F.3d 673, 677 (D.C. Cir. 2008) (citing Chevron, 467 U.S.\\nat 842-43). We hold, under step one of Chevron, that the Act\\nunambiguously precludes EPA’s interpretation of\\nsection 165(a) and vacate the grandfathering provision of the\\nFinal Rule.\\n\\n The revised NAAQS were effective on December 28,\\n2015. As a result of the grandfathering provision, a major\\nemitting facility that demonstrates compliance with the\\nprevious NAAQS of 0.075 ppm, but not the revised NAAQS\\nof 0.07 ppm, can still be built even after the new NAAQS take\\neffect, provided the project had completed its application\\nwithin certain time limits. That is exactly what the plain text\\nof the Act forbids: the “construction” of a “major emitting\\nfacility” with emissions that will “cause, or contribute to, air\\npollution in excess of any . . . [NAAQS].” 42 U.S.C. § 7475(a)\\n(emphasis added). “Read naturally, the word ‘any’ has an\\nexpansive meaning, that is ‘one or some indiscriminately of\\nwhatever kind.’” United States v. Gonzalez, 520 U.S. 1, 5\\n(1997) (quoting WEBSTER’S THIRD NEW INTERNATIONAL\\nDICTIONARY 97 (1976)). EPA gives no reason why, in the\\ncontext of section 165(a), the word “any” should not carry its\\nnatural meaning and hence refer to the newly revised NAAQS\\nonce they have become effective.\\n\\n In New York v. EPA, we observed that the meaning of\\n“any” can differ if, for instance, the “Supreme Court has\\nrequired heightened standards of clarity to avoid upsetting\\nfundamental policies,” or if there would be “‘strange and\\nindeterminate results’ that would emerge from adopting the\\nnatural meaning of ‘any’” in a statute. 443 F.3d 880, 885-86\\n(D.C. Cir. 2006) (quoting Nixon v. Mo. Mun. League, 541 U.S.\\n\\f 47\\n125, 132-33 (2004)). But none of those factors is present here.\\nIndeed, NAAQS are set based on the “maximum airborne\\nconcentration of a pollutant that the public health can tolerate,”\\nWhitman, 531 U.S. at 465, so the “strange” result would be if\\nthe Act, focused as it is on the protection of public health, left\\nthe door open for construction projects that will cause or\\ncontribute to ozone levels higher than the “public health can\\ntolerate.” Thus, we read section 165(a) to forbid construction\\nof any facility that cannot demonstrate compliance with “any”\\neffective NAAQS.\\n\\n There is no ambiguity created by reading section 165(a)\\nalongside the timeliness requirement of section 165(c). EPA\\nthought the grandfathering provision necessary to allow\\npermitting authorities to “issue a permit within one year.” 80\\nFed. Reg. at 65,433-34. But the Act does not require that a\\npermit be “issued” within one year. It requires only that the\\npermitting authority “grant[] or den[y]” completed permit\\napplications within one year. 42 U.S.C. § 7475(c) (emphasis\\nadded). “[N]othing in the [Act] provides for issuance of a []\\npermit as a matter of right.” Am. Corn Growers Ass’n v. EPA,\\n291 F.3d 1, 12 (D.C. Cir. 2002) (internal quotation marks\\nomitted). If a permit applicant has not shown that it can meet\\nthe new NAAQS, EPA or a state permitting authority can\\ncomply with the timeliness requirement of section 165(c) by\\ndenying the application.\\n\\n This reading is confirmed by the structure of the Act.\\nWhen amending the Act to add section 165(a)’s demonstration\\nrequirement, Congress expressly exempted projects that had\\nbegun construction prior to passage of that amendment. See 42\\nU.S.C. § 7475(a); Pub. L. No. 95-95, § 127(a), 91 Stat. 685,\\n735-39 (Aug. 7, 1977). The inclusion of this grandfather clause\\nimplies that Congress did not intend EPA to have some\\ninherent grandfathering authority, and that, in the future,\\n\\f 48\\nNAAQS would be enforced as enacted. See Andrus v. Glover\\nConstr. Co., 446 U.S. 608, 616-17 (1980) (“Where Congress\\nexplicitly enumerates exceptions to a general prohibition,\\nadditional exceptions are not to be implied, in the absence of\\nevidence of a contrary legislative intent.”). Congress has\\nspoken on this question, and EPA cannot displace the statutory\\ndetermination simply because the agency’s “preferred\\napproach [might] be better policy.” See Engine Mfrs. Ass’n v.\\nEPA, 88 F.3d 1075, 1089 (D.C. Cir. 1996). Nor does it matter\\nthat EPA has evidently relied on this grandfathering authority\\nin the past. See, e.g., National Ambient Air Quality Standards\\nfor Particulate Matter, 78 Fed. Reg. 3,086, 3,258-59 (Jan. 15,\\n2013) (adopting grandfathering provision for revised\\nNAAQS); Regulations for Implementing Revised Particulate\\nMatter Standards, 52 Fed. Reg. 24,672, 24,683 (July 1, 1987)\\n(same). “[P]revious statutory violations cannot excuse the one\\nnow before the court.” New Jersey v. EPA, 517 F.3d 574, 583\\n(D.C. Cir. 2008).\\n\\n The Ninth Circuit addressed a related issue in Sierra Club\\nv. EPA, 762 F.3d 971 (9th Cir. 2014). EPA had failed to grant\\nor deny a completed permit application within one year as\\nrequired by section 165(c), and in the meantime the agency had\\nlowered the relevant NAAQS. See id. at 974-75. After the\\napplicant filed suit to compel issuance of the permit under the\\nold standards, EPA initially took the position—contrary to its\\nstance here, but consistent with our holding—that the Act\\nprohibited it “from granting the Permit unless [the applicant]\\ncomplies with the superseding standards.” Id. at 975. Then,\\nafter a “policy review,” EPA changed its mind and asserted\\n“inherent grandfathering authority,” id., relying on the alleged\\ntension that it advances here between sections 165(a) and (c),\\nsee id. at 978. The Ninth Circuit discerned no such tension and\\nheld that, under Chevron step one, EPA must “apply the\\n\\f 49\\nregulations in effect at the time of the permitting decision.” Id.\\nat 979.\\n\\n After concluding that EPA had violated the plain terms of\\nthe Act, the Sierra Club court added in dicta that the agency\\ncould create a grandfathering exception through “formal notice\\nand comment rulemaking.” Id. at 982. The Ninth Circuit cited\\n42 U.S.C. § 7601(a), which includes a general grant of\\nauthority to the EPA Administrator to “prescribe such\\nregulations as are necessary” to carry out the Act. EPA argues\\nthat we should follow the Ninth Circuit’s dicta because EPA\\nused formal rulemaking to grandfather these applications. But\\nwe decline to do so. Dicta is never binding on any court, Glus\\nv. Brooklyn E. Dist. Terminal, 359 U.S. 231, 235 (1959), nor is\\nit persuasive here, because it is fundamentally incorrect. A\\ngeneral grant of authority cannot displace the clear, specific\\ntext of the Act. Air All. Hous. v. EPA, 906 F.3d 1049, 1061\\n(D.C. Cir. 2018) (“[A]n agency may not circumvent specific\\nstatutory limits on its actions by relying on separate, general\\nrulemaking authority.”). As the Sierra Club court recognized,\\nthe Act requires compliance with the NAAQS that are in effect\\n“at the time of the permitting decision.” 762 F.3d at 979. EPA\\nhas no authority to change that provision of the Act, whether\\nby ad hoc waiver or rulemaking. See also Citizens to Save\\nSpencer Cty. v. EPA, 600 F.2d 844, 873 (D.C. Cir. 1979)\\n(explaining that a general power cannot “provide the [EPA]\\nAdministrator with Carte blanche authority to promulgate any\\nrules, on any matter relating to the [Act], in any matter that the\\nAdministrator wishes”).\\n\\n The grandfathering provision of the 2015 Rule, as codified\\nat 40 C.F.R. §§ 51.166(i)(11) and 52.21(i)(12), contradicts\\nCongress’s “express policy choice” not to allow construction\\nwhich will “cause or contribute to” nonattainment of “any”\\neffective NAAQS, regardless of when they are adopted or\\n\\f 50\\nwhen a permit was completed. Accordingly, we grant the\\npetition for review with respect to the grandfathering provision,\\nand vacate that portion of the Rule.\\n\\n * * *\\n\\n In setting the secondary standard, EPA failed to justify its\\ndecision to use a three-year average benchmark without\\nlowering the level to account for single-year spikes in ozone\\nexposures, and it arbitrarily declined to set a level to protect\\nagainst adverse welfare effects associated with visible leaf\\ninjury. EPA also impermissibly allowed sources that had\\ncompleted applications for preconstruction permits before the\\n2015 Rule was adopted to demonstrate compliance with the\\nprevious NAAQS rather than the new, more stringent primary\\nand secondary standards. Accordingly, we grant those portions\\nof the Environmental Petitioners’ petition, vacate the\\ngrandfathering provision, and remand to EPA for\\nreconsideration of the secondary standard.\\n\\n So ordered.\"}]}"},"judge1":{"kind":"string","value":"GRIFFITH"},"judge2":{"kind":"string","value":"PILLARD"},"judge3":{"kind":"string","value":"WILKINS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"GRIFFITH\": \", Circuit\", \"PILLARD\": \", Circuit\", \"WILKINS\": \", Circuit\"}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":1,"string":"1"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4655414/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Murray Energy Corporation v. EPA\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-23\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: Yes\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued December 18, 2018 Decided August 23, 2019\n\n No. 15-1385\n\n MURRAY ENERGY CORPORATION,\n PETITIONER\n\n v.\n\n ENVIRONMENTAL PROTECTION AGENCY,\n RESPONDENT\n\n AMERICAN LUNG ASSOCIATION, ET AL.,\n INTERVENORS\n\n\n Consolidated with 15-1392, 15-1490, 15-1491, 15-1494\n\n\n On Petitions for Review of Final Agency Action of\n the United States Environmental Protection Agency\n\n\n Seth L. Johnson argued the cause for Public Health and\nEnvironmental Petitioners. With him on the briefs were Joshua\nStebbins, Joshua Berman, David S. Baron, and Paul Cort.\n\f 2\n Dominic E. Draye, Deputy Solicitor General, Office of the\nAttorney General for the State of Arizona, argued the cause for\nState Petitioners. With him on the briefs were Mark Brnovich,\nAttorney General, John R. Lopez, IV, Solicitor General, Keith\nMiller, Associate Solicitor, Joshua L. Kaul, Attorney General,\nOffice of the Attorney General for the State of Wisconsin,\nMisha Tseytlin, Solicitor General, Daniel P. Lennington,\nAssistant Attorney General, Leslie C. Rutledge, Attorney\nGeneral, Office of the Attorney General for the State of\nArkansas, Lee Rudofsky, Solicitor General, Jamie L. Ewing,\nAssistant Attorney General, Jeff Landry, Attorney General,\nOffice of the Attorney General for the State of Louisiana,\nSteven B. ABeaux@ Jones, Assistant Attorney General, Wayne\nStenehjem, Attorney General, Office of the Attorney General\nfor the State of North Dakota, Margaret I. Olson, Assistant\nAttorney General, Peter S. Glaser, Carroll W. AMack@\nMcGuffey, Andy Beshear, Attorney General, Office of the\nAttorney General for the Commonwealth of Kentucky, Joseph\nA. Newberg, II, Assistant Attorney General, Lara Katz,\nAssistant Attorney General, Office of the Attorney General for\nthe State of New Mexico, Ken Paxton, Attorney General,\nOffice of the Attorney General for the State of Texas, Priscilla\nM. Hubenak, Chief, Craig J. Pritzlaff, Assistant Attorney\nGeneral, E. Scott Pruitt, Attorney General at the time the brief\nwas filed, Office of the Attorney General for the State of\nOklahoma, P. Clayton Eubanks, Deputy Solicitor General, and\nSean Reyes, Attorney General, Office of the State of Utah.\nMithun Mansinghani, Attorney, Office of the Attorney General\nfor the State of Oklahoma, Lisa M. Mitchell, Assistant Attorney\nGeneral, Office of the Attorney General for the State of Texas,\nOramel H. Skinner, III, Solicitor, Office of the Attorney\nGeneral for the State of Arizona, Sarah Adkins and Samuel R.\nFlynn, Assistant Attorneys General, and Gregory T. Dutton,\nCounsel, Office of the Attorney General for the\nCommonwealth of Kentucky, Steven C. Kilpatrick, Assistant\n\f 3\nAttorney General, Office of the Attorney General for the State\nof Wisconsin, Elizabeth B. Murrill and Harry J. Vorhoff,\nAssistant Attorneys General, Office of the Attorney General\nfor the State of Louisiana, entered appearances.\n\n James R. Bieke argued the cause for Industry Petitioners.\nWith him on the briefs were Roger R. Martella, Joel F. Visser,\nScott C. Oostdyk, E. Duncan Getchell, Jr., Michael H. Brady,\nThomas A. Lorenzen, Robert J. Meyers, Linda E. Kelly,\nQuentin Riegel, Leland P. Frost, Michael B. Schon, Elizabeth\nL. Horner, Lucinda Minton Langworthy, Aaron M. Flynn,\nSteven P. Lehotsky, Sheldon B. Gilbert, Stacy Linden, and\nRichard S. Moskowitz. Peter C. Tolsdorf entered an\nappearance.\n\n Hope M. Babcock and Sarah J. Fox were on the brief for\namici curiae American Thoracic Society, et al. in support of\npetitioners Sierra Club, et al.\n\n Thomas J. Ward was on the brief for amicus curiae The\nNational Association of Home Builders in support of Industry\nand State Petitioners.\n\n Justin Heminger and Simi Bhat, Trial Attorneys, U.S.\nDepartment of Justice, argued the causes for respondent. With\nthem on the brief were John C. Cruden, Assistant Attorney\nGeneral at the time the brief was filed, and David Orlin, Steven\nSilverman, and Brian Doster, Attorneys, U.S. Environmental\nProtection Agency. Jon M. Lipshultz, Attorney, U.S.\nDepartment of Justice, entered an appearance.\n\n James R. Bieke argued the cause for Industry Respondent-\nIntervenors. With him on the brief were Roger R. Martella,\nJoel F. Visser, Lucinda Minton Langworthy, Aaron M. Flynn,\nThomas A. Lorenzen, Robert J. Meyers, Stacy Linden, Richard\n\f 4\nS. Moskowitz, Steven P. Lehotsky, Sheldon B. Gilbert, Linda E.\nKelly, Quentin Riegel, Leland P. Frost, Michael B. Schon,\nElizabeth Horner, and Leslie A. Hulse. Peter C. Tolsdorf\nentered an appearance.\n\n Seth L. Johnson argued the cause for Health and\nEnvironmental Respondent-Intervenors. With him on the brief\nwas David S. Baron. Joshua A. Berman and Joshua R. Stebbins\nentered appearances.\n\n Jonathan Weiner, Deputy Attorney General, Office of the\nAttorney General for the State of California, argued the cause\nfor State Amici in support of respondent. With him on the brief\nwere Kamala D. Harris, Attorney General at the time the brief\nwas filed, Robert W. Byrne, Senior Assistant Attorney General,\nand Gavin G. McCabe, Supervising Deputy Attorney General.\nMelinda Pilling, Attorney, entered an appearance.\n\n Richard L. Revesz, Denise A. Grab, Jack Lienke, Michael\nA. Livermore, and Jason A. Schwartz were on the brief for\namicus curiae The Institute for Policy Integrity at New York\nUniversity School of Law in support of respondent.\n\n Before: GRIFFITH, PILLARD and WILKINS, Circuit Judges.\n\n Opinion for the Court filed PER CURIAM.\n\n PER CURIAM: In this opinion, we consider various\nchallenges to the Environmental Protection Agency’s 2015\nrevisions to the primary and secondary national ambient air\nquality standards for ozone. For the reasons given below, we\ndeny the petitions, except with respect to the secondary ozone\nstandard, which we remand for reconsideration, and\ngrandfathering provision, which we vacate.\n\f 5\n I. Background\n\nA. Statutory and Procedural Background\n\n Congress enacted the modern version of the Clean Air Act\n(the “Act”), codified at 42 U.S.C. § 7401 et seq., in 1970 to\ncontrol and reduce contaminants responsible for air pollution\nwith the overarching goal to protect human health and the\nenvironment. Pursuant to Title I, EPA must establish, publish,\nand periodically review primary and secondary national\nambient air quality standards (“NAAQS”) for air pollutants that\n“may reasonably be anticipated to endanger public health or\nwelfare.” 42 U.S.C. § 7408(a)(1)(A). The primary NAAQS\nare to be set at levels “the attainment and maintenance of which\nin the judgment of the Administrator, . . . allowing an adequate\nmargin of safety, are requisite to protect the public health.” Id.\n§ 7409(b)(1). The secondary NAAQS “shall specify a level of\nair quality the attainment and maintenance of which in the\njudgment of the Administrator, . . . is requisite to protect the\npublic welfare from any known or anticipated adverse effects.”\nId. § 7409(b)(2). Thus, primary NAAQS protect the public\nhealth, while the secondary NAAQS protect the public welfare.\n“Public health” includes adverse health effects for both the\npopulation at large and sensitive populations such as children,\nolder adults, and people with asthma or other lung diseases.\nThe term “public welfare” encompasses a wide variety of\neffects on soil, plants, wildlife and biota, property damage,\naesthetic concerns, and other non-health-related impacts such\nas hazards to economic values and personal comfort. Id.\n§ 7602(h).\n\n Pursuant to section 109(d) of the Clean Air Act, EPA must\ncomplete a “thorough” review of the NAAQS every five years.\nId. § 7409(d)(1). During this review, EPA must revise the\ncriteria and standards or promulgate new standards as\n\f 6\nappropriate. Id. To assist in this process, the Clean Air\nScientific Advisory Committee (“CASAC”) undertakes an\nexamination of the current air quality criteria, primary\nNAAQS, and secondary NAAQS, and submits\nrecommendations to EPA for “any new [NAAQS] and\nrevisions of existing criteria and standards as may be\nappropriate.” Id. § 7409(d)(2)(A)-(B). Congress required EPA\nto take CASAC’s recommendations into account when\npromulgating revised NAAQS and to fully explain its reasons\nwhen it departs from CASAC’s advice. However, the ultimate\ndecision to revise the NAAQS—and the determination of the\nnew level—rests with the Administrator. Id. § 7407(d)(3).\n\n These petitions concern EPA’s promulgation of revised\nNAAQS related to ozone. Ozone is a colorless gas that occurs\nboth in the Earth’s upper atmosphere and at ground level.\nAlthough ozone is an “essential presence in the atmosphere’s\nstratospheric layer,” ground-level ozone is an air pollutant that\nis harmful to breathe and damages crops, trees, and other\nvegetation. S. Coast Air Quality Mgmt. Dist. v. EPA, 472 F.3d\n882, 887 (D.C. Cir. 2006); see Allied Local & Reg’l Mfrs.\nCaucus v. EPA, 215 F.3d 61, 66 (D.C. Cir. 2000). Ground-\nlevel ozone is not a direct product of human activity, but\ninstead forms when atmospheric pollutants (including nitrogen\noxides and volatile organic compounds) react in the presence\nof sunlight. See Mississippi v. EPA, 744 F.3d 1334, 1340 (D.C.\nCir. 2013). These precursor atmospheric pollutants are created\nprimarily from emissions produced by cars, power plants, and\nchemical solvents. NRDC v. EPA, 777 F.3d 456, 459 (D.C.\nCir. 2014).\n\n In 1979, EPA issued primary and secondary NAAQS for\nozone with a limit of 0.12 parts per millions (ppm), and a one-\nhour averaging time. See id. This “one-hour standard”\nmeasured average ozone levels over one-hour periods, and\n\f 7\nEPA would deem an area in compliance with this standard if it\ndid not exceed the level for more than one day per calendar\nyear. Id. EPA next revised the ozone NAAQS in 1997, having\ndetermined that no revisions to the standards were necessary in\n1993. National Ambient Air Quality Standards for Ozone, 58\nFed. Reg. 13,008 (Mar. 9, 1993). The agency replaced the one-\nhour, 0.12 ppm standard with a 0.08 ppm standard measured\nover eight hours. National Ambient Air Quality Standards for\nOzone, 62 Fed. Reg. 38,856 (July 18, 1997). EPA also altered\nthe form of compliance to the annual fourth-highest daily\nmaximum 8-hour concentration, averaged over three years. Id.\n\n In 2008, EPA lowered the primary and secondary\nstandards to 0.075 ppm but kept the same eight-hour averaging\ntime and form as in 1997. NRDC, 777 F.3d at 462-63; National\nAmbient Air Quality Standards for Ozone, 73 Fed. Reg. 16,436\n(Mar. 27, 2008). In developing the 2008 standards, EPA relied\non scientific evidence showing that ozone causes health effects\nat and above 0.08 ppm and examined two new clinical studies\nthat found negative health effects from ozone at lower levels.\nSee Mississippi, 744 F.3d at 1340. While CASAC reviewed\nthis same evidence and recommended a primary level between\n0.06 ppm and 0.07 ppm, EPA departed from this advice and\nexplained that the scientific data regarding negative health\neffects at 0.06 ppm was too limited and inconclusive to support\na standard below 0.075 ppm. See id. at 1340-41. We upheld\nEPA’s primary standard on this basis but found that EPA had\nnot adequately explained its revision of the secondary standard.\nId. at 1359-62. We noted that EPA had not properly\ndetermined what level of air quality was requisite to protect the\npublic welfare. Id. Accordingly, we remanded for further\nexplanation and reconsideration of the secondary level.\n\n Following the promulgation of the 2008 ozone NAAQS,\nEPA undertook another comprehensive review of the ozone\n\f 8\nstandards and scientific data. EPA consolidated its review on\nremand of the 2008 secondary standard with its ongoing review\nfor the 2015 ozone NAAQS. After conducting public hearings\nand receiving approximately 430,000 written comments on its\nproposed revision of the primary and secondary ozone\nNAAQS, EPA published its final 2015 ozone NAAQS on\nOctober 26, 2015. National Ambient Air Quality Standards for\nOzone (“2015 Rule”), 80 Fed. Reg. 65,292 (Oct. 26, 2015). For\nboth the primary and secondary standards, EPA lowered the\nlevel from 0.075 ppm to 0.07 ppm, while retaining the 2008\nindicator (ozone), averaging time (8 hours), and form (three-\nyear average of the fourth-highest daily maximum 8-hour\nconcentration). See id. at 65,294, 65,301, 65,347, 65,349-50,\n65,352.\n\n Based on scientific data and CASAC’s recommendation,\nthe Administrator concluded that the 2008 primary ozone\nstandard (0.075 ppm) was not at a level requisite to protect\npublic health while allowing an adequate margin of safety. Id.\nat 65,326, 65,344, 65,346. The Administrator placed\nsignificant weight on new clinical studies linking short-term\nozone exposure with respiratory effects, including lung\ninflammation. See id. at 65,302-03, 65,341, 65,352, 65,359.\nEPA conducted its own exposure study and found that a revised\nprimary NAAQS standard of 0.07 ppm would “eliminate the\noccurrence of two or more exposures of concern to [ozone]\nconcentrations at and above [0.08 ppm]” and would “virtually\neliminate the occurrence of two or more exposures of concern\nto [ozone] concentrations at and above [0.07 ppm], even in the\nworst-case urban study area and year evaluated.” Id. at 65,353.\nThe Administrator additionally found that a level of 0.07 ppm\nwould “protect the large majority of children in the urban study\nareas (i.e., about 96% to more than 99% of children in\nindividual urban study areas) from experiencing two or more\nexposures of concern at or above the [0.06 ppm] benchmark.”\n\f 9\nId. at 65,353, 65,360-64. EPA also evaluated hundreds of\nepidemiologic studies that provided statistically relevant\ninformation about a broader population of individuals who are\nexposed to uncontrolled air pollution. See id. at 65,304,\n65,364.\n\n Further, the Administrator considered CASAC’s advice on\nthe new primary ozone standard. See id. at 65,346, 65,361. In\nadvance of the 2015 Rule, CASAC stated that “there is clear\nscientific support for the need to revise the standard” in place\nsince 2008 and recommended setting the standard within a\nrange of 0.06 ppm to 0.07 ppm, while leaving the form,\naveraging time, and indicator unchanged. Id. at 65,322,\n65,361.\n\n EPA also assessed the secondary standard and concluded\nthat the 2008 secondary standard (0.075 ppm) was not requisite\nto protect public welfare. Id. at 65,382, 65,389-90. Rather,\nEPA set the secondary standard at 0.07 ppm and kept the\nindicator, averaging time, and form the same. Id. at 65,369,\n65,403, 65,409-10. EPA noted that more than four hundred\nnew studies examining the interplay between ozone and public\nwelfare had been developed since the promulgation of the 2008\nozone NAAQS. Id. at 65,369. These studies strengthened and\nexpanded the scientific understanding of ozone’s effects on\nplants. The Administrator analyzed the causal relationship\nbetween ozone exposure and vegetation effects, examining tree\ngrowth impacts, crop yield loss, and visible leaf injury. Id. at\n65,370; see id. at 65,294, 65,369-70. The Administrator gave\n“primary consideration” to tree growth impacts, judging that it\nwas more difficult to assess the welfare significance of crop\nyield loss (given that crops are heavily managed) and visible\nleaf injury (citing the “lack of established criteria or\nobjectives”). Id. at 65,407.\n\f 10\n In setting the secondary standard, EPA departed from\nCASAC’s recommendations as to the form and standard of the\nsecondary ozone NAAQS. CASAC recommended using a\nsingle-year averaging approach, but EPA opted for a three-year\naverage because “the public welfare significance of effects\nassociated with multiple years of critical exposures are\npotentially greater than those associated with a single year of\nsuch exposure.” Id. at 65,404. The Administrator also rejected\nCASAC’s recommendation to use the W126 exposure index, a\ncumulative, seasonal ozone exposure metric, as the form of the\nsecondary standard, rather than the same form as the primary\nstandard (three-year average of the fourth-highest daily level),\nfinding that the latter form was adequate to restrict cumulative\nozone exposures that are detrimental to vegetation. Id. at\n65,408.\n\n In addition to revising the ozone standards, the 2015\nNAAQS also updated the regulations for the Prevention of\nSignificant Deterioration (“PSD”) permitting program. See id.\nat 65,431-34. Under the PSD program, no construction on a\nmajor emitting facility may be commenced in an area that has\nattained the air quality standards for any criteria pollutant\nunless “the owner or operator of such facility demonstrates . . .\nthat emissions from construction or operation of such facility\nwill not cause, or contribute to, air pollution in excess of any\n. . . [NAAQS].” 42 U.S.C. § 7475(a)(3)(B). The permitting\nauthority (either a state or EPA) must grant or deny a completed\npreconstruction permit application under the PSD program\nwithin one year of filing. Id. § 7475(c).\n\n EPA claimed that sections 7475(a)(3)(B) and 7475(c) have\nthe “potential for conflict,” which it was authorized under\nChevron to resolve. EPA Br. 132; see also 80 Fed. Reg. at\n65,433-34. EPA worried that, for a limited subset of pending\npermit applications, complying with section 7475(a)(3)(B)’s\n\f 11\ndemonstration requirement for the 2015 NAAQS “could hinder\ncompliance with the requirement under section [7475](c) to\nissue a permit within one year of the completeness\ndetermination.” 80 Fed. Reg. at 65,434. In other words, EPA\nwas unsure how to handle completed permit applications where\nsources would have complied with the NAAQS in existence at\nthe time the application was filed (the 2008 ozone standards),\nbut where the NAAQS were revised before the permit was\napproved. See id. at 65,433. EPA resolved this apparent\nconflict by implementing a grandfathering provision for\npending permit applications that satisfy one of two permitting\ncriteria. See id. at 65,431-34. These two applicable categories\ninclude: (1) permit applications that are deemed complete on\nor before the signature date of the final rule revising the ozone\nNAAQS; and (2) permit applications “for which the reviewing\nauthority has first published a public notice of the draft permit\nor preliminary determination before the effective date of the\nrevised NAAQS.” Id. at 65,431, 65,433. If a permit\napplication satisfies either requirement, the owner or operator\nmust show compliance with the 2008 rather than the 2015\nNAAQS before initiating construction.\n\n Petitioners from the oil and gas industry (“Industry\nPetitioners”) and from several states (“State Petitioners”) claim\nthat the primary and secondary NAAQS are too protective.\nConversely, petitioners from public health and environmental\ngroups (“Environmental Petitioners”) claim that these NAAQS\nare not protective enough. Environmental Petitioners also\ncontend that EPA’s decision to allow limited grandfathering of\ncertain permit applications was inconsistent with the Clean Air\nAct.\n\n This opinion proceeds in five parts. Part II considers the\nchallenges to the primary standard, and Part III the challenges\nto the secondary standard. Part IV reviews the cross-cutting\n\f 12\nchallenges to EPA’s promulgation of the NAAQS, and Part V\naddresses the challenge to the grandfathering provision.\n\nB. Jurisdiction and Standard of Review\n\n This court has jurisdiction to review EPA’s 2015 Rule\npursuant to 42 U.S.C. § 7607(b)(1). Under section 7607,\nagency action may be reversed if it is arbitrary, capricious, an\nabuse of discretion, not in accordance with law, or in excess of\nstatutory authority. Id. § 7607(d)(9)(A), (C); see Am.\nPetroleum Inst. v. EPA, 684 F.3d 1342, 1347 (D.C. Cir. 2012).\nTo withstand review, an agency must have examined all\nrelevant facts and data, and articulated a rational explanation\nfor its decision, including a reasonable connection between the\nfacts and ultimate outcome. See Allied Local, 215 F.3d at 68.\nWe cannot look at EPA’s decision as would a scientist, but\ninstead must exercise our “narrowly defined duty of holding\nagencies to certain minimal standards of rationality.”\nMississippi, 744 F.3d at 1342 (quoting Nat’l Envtl. Dev.\nAss’n’s Clean Air Project v. EPA, 686 F.3d 803, 810 (D.C. Cir.\n2012)); see id. at 1348 (“We repeat: it is not our job to referee\nbattles among experts; ours is only to evaluate the rationality\nof EPA’s decision . . . .”).\n\n On questions of statutory interpretation, the court must\nreview EPA’s actions in accordance with the standard set forth\nin Chevron, U.S.A., Inc. v. Natural Resources Defense Council,\nInc., 467 U.S. 837, 842-44 (1984). See Michigan v. EPA, 135\nS. Ct. 2699, 2706-07 (2015); NRDC, 777 F.3d at 463. Chevron\ndeference involves a two-step analysis. First, if a statute is\nclear, the court must give effect to Congress’s unambiguous\nlanguage and intent. Chevron, 467 U.S. at 842-43. Where a\nstatute that an agency administers is ambiguous, the court must\nturn to the second Chevron principle and give deference to the\n\f 13\nagency’s reasonable interpretation of the statute. Id. at 843; see\nalso S. Coast Air Quality Mgmt. Dist., 472 F.3d at 891.\n\n II. Primary Standard Challenges\n\n Industry and State Petitioners contend that EPA’s\npromulgation of the 2015 Rule’s primary standard was\narbitrary and capricious because, they say, EPA failed to\nprovide a reasoned explanation for departing from the 2008\nNAAQS. Environmental Petitioners argue that the primary\nozone standard is too lenient because it occasionally permits\nozone levels to exceed 0.07 ppm and will allegedly tolerate\nadverse health effects. For the reasons below, we hold that\nthese arguments lack merit.\n\nA. Industry and State Petitioners’ Challenge\n\n The Clean Air Act requires EPA to set primary NAAQS\nthat are “requisite” to protect public health with an adequate\nmargin of safety. 42 U.S.C. § 7409(b)(1). The term “requisite”\nmeans “sufficient, but not more than necessary.” See Whitman\nv. Am. Trucking Ass’ns, 531 U.S. 457, 473 (2001). While the\ndetermination of what is “requisite” may require a contextual\nassessment of acceptable risk, our precedent is clear that prior\nNAAQS are not sacrosanct and are not granted presumptive\nvalidity. See Mississippi, 744 F.3d at 1343. Prior NAAQS\nneed not remain the governing standard until every aspect of\nthose prior NAAQS is undermined. Id. To hold otherwise\n“would bind EPA to potential deficiencies in past reviews\nbecause discrepancies between past and current judgments as\neasily reflect problems in the past as in the present.” Id. Thus,\nwhen EPA reviews and revises the NAAQS, it does so against\ncurrent policy considerations and existing scientific\nknowledge. Id. Accordingly, “[t]he statutory framework\nrequires us to ask only whether EPA’s proposed NAAQS is\n\f 14\n‘requisite’; we need not ask why the prior NAAQS once was\n‘requisite’ but is no longer up to the task.” Id.\n\n Given our decision in Mississippi, we must reject Industry\nand State Petitioners’ argument that EPA departed from the\n2008 NAAQS without adequate explanation. EPA has\nadequately explained why on the record before it the revised\nstandard is requisite to protect the public health. As the record\nmakes clear, the Administrator considered the entire body of\nscientific evidence available, including clinical studies,\nepidemiologic evidence, human exposure and health risk\nassessments, CASAC’s recommendations, and over 430,000\npublic comments. See 80 Fed. Reg. at 65,293, 65,299, 65,301,\n65,323, 65,326. And consistent with CASAC’s advice, the\nAdministrator placed the most weight on clinical studies. See\nid. at 65,302-03, 65,341, 65,352, 65,359. State Petitioners take\nparticular issue with EPA’s reliance on the 2009 Schelegle\nstudy, which, they argue, does not support EPA’s finding of\nadverse effects. But the 2009 Schelegle study is merely one of\nmany clinical studies that EPA relied on. State Petitioners\nignore the fact that additional clinical studies and\nepidemiologic evidence show substantial health effects at\nozone levels as low as 0.06 ppm. See id. at 65,331, 65,334-35\n& n.96, 65,344, 65,364.\n\n In addition to evaluating clinical studies, the Administrator\nalso examined epidemiologic evidence and EPA’s exposure\nand risk assessments. See 80 Fed. Reg. at 65,304, 65,314-17,\n65,364. While the Administrator placed less weight on the\nepidemiologic results than on clinical studies, see id. at 65,320,\n65,324, 65,335, she nonetheless found that recent\nepidemiologic studies suggested adverse health effects would\noccur below the 2008 NAAQS standard. While Industry\nPetitioners challenge the use of epidemiologic evidence given\nthe uncertainties presented in these studies, even CASAC\n\f 15\nconcluded that the epidemiologic evidence would have alone\nbeen strong enough to justify revision of the 2008 NAAQS.\n\nB. Environmental Petitioners’ Challenges\n\n In contrast to Industry and State Petitioners,\nEnvironmental Petitioners argue that the primary ozone\nstandard is too lenient. They offer two reasons. First,\nEnvironmental Petitioners contend that the form of the revised\nstandard is not health-protective because it permits ozone\nlevels to exceed 0.07 ppm on some days. Second, they argue\nthat ozone exposures of 0.07 ppm will cause adverse health\neffects, particularly in sensitive populations. For the reasons\nbelow, we hold that these arguments lack merit.\n\n EPA reasonably explained its decision to retain\n the form of the primary standard.\n\n Environmental Petitioners first take issue with the form of\nthe revised standard—that the average, over three years, of the\nannual fourth-highest maximum daily 8-hour average ozone\nlevel must not exceed 0.07 ppm. By using this form, EPA\n“allows ozone levels to exceed—multiple times in any year—\nlevels that EPA itself agrees cause adverse health effects.”\nEnvtl. Pet’rs Br. 20. Environmental Petitioners contend that\nbecause areas with ozone levels above the 0.07 ppm threshold\ncan still comply with EPA’s standards, the standard is unlawful\nand arbitrary. We disagree.\n\n The Administrator’s decision to retain the same form used\nin the 1997 and 2008 primary standards was in part based on\nhow many people are estimated to experience unhealthy ozone\nexposures. To this end, the Administrator utilized EPA’s\nexposure assessment to gauge how often various population\nsubgroups will be exposed to potentially health-impairing\nozone concentrations while experiencing elevated breathing\n\f 16\nrates. See 80 Fed. Reg. at 65,356. Rather than considering the\nozone level of an area in isolation, the Administrator also\n“consider[ed] activity patterns in the exposed population”\nbecause adverse health responses to ozone exposure are\ncritically dependent on breathing rates. Id. at 65,363; see also\nid. (“Not considering activity patterns, and corresponding\nventilation rates, can result in a standard that provides more\nprotection than is requisite.”). This use of the exposure\nassessment was rational.\n\n Further, the Administrator adequately interpreted the\nexposure assessment. The Administrator correctly observed\nthat the exposure assessment only chronicled how many people\nwith an elevated breathing rate will be exposed to a specific\nozone level, not how many will have an adverse effect. Given\nthis limitation, the Administrator focused on estimates of two\nor more exposures of concern to assess the potential for adverse\neffects at and above an ozone concentration of 0.07 ppm. See\nid. at 65,310, 65,325, 65,330, 65,343, 65,345-46, 65,354,\n65,358, 65,361, 65,363. Based on the data, the Administrator\nconcluded that a level of 0.07 ppm ozone would eliminate the\noccasions on which school-age children experience two or\nmore exposures of concern at ozone concentrations at and\nabove 0.08 ppm and, even in the worst-case years and\nlocations, would “virtually eliminate” the occasions on which\nsuch children experience two or more exposures of concern at\n0.07 ppm. Id. at 65,353. In particular, the Administrator noted\nthat a primary level of 0.07 ppm would protect approximately\n96% to 99% of children in the urban study areas from\nexperiencing two or more exposures of concern above the 0.06\nppm benchmark. Id. at 65,353; see also id. at 65,360-64. That\nEnvironmental Petitioners cite data suggesting that 18,000\nchildren would experience multiple exposures of concern at or\nabove 0.07 ppm during the worst year and location is not\ndeterminative because the primary standard for a non-threshold\n\f 17\npollutant like ozone is not required to produce zero risk, and\n“[t]he task of determining what standard is ‘requisite’ to protect\nthe qualitative value of public health or what margin of safety\nis ‘adequate’ to protect sensitive subpopulations necessarily\nrequires the exercise of policy judgment.” Mississippi, 744\nF.3d at 1351, 1358. EPA has reasonably exercised its\njudgment.\n\n Given that the Administrator appropriately examined and\nconsidered EPA’s exposure assessment, her decision to retain\nthe form of the primary standard was also rational. The\nAdministrator determined that, when combined with an ozone\nlevel of 0.07 ppm, the form of the standard (three-year average\nof the fourth-highest daily level) was requisite. In particular,\nthe Administrator found that most areas that meet the revised\nstandard will have an 8-hour ozone concentration below 0.07\nppm. See 80 Fed. Reg. at 65,363. Environmental Petitioners,\nhowever, criticize this form for allegedly failing to account for\nindividual ozone days above 0.07 ppm. But Environmental\nPetitioners elide a crucial detail of EPA’s methodology: the\nexposure assessment study that “informed the Administrator’s\nconsideration of the degree of public health protection provided\nby various standard levels” considered the air quality at\n“various standards with the current 8-hour averaging time and\nfourth-high, 3-year average form.” Id. at 65,351-52 (emphasis\nadded). As such, the Administrator reasonably accounted for\ndays when ozone levels may exceed 0.07 ppm. Taken together,\nwe conclude that the Administrator’s decision to retain the\nform of the standard was appropriate and neither arbitrary nor\ncapricious.\n\f 18\n EPA reasonably set the primary standard at 0.07\n ppm.\n\n Environmental Petitioners next argue that the\nAdministrator’s decision to lower the primary standard from\nfrom 0.075 to 0.07 ppm is an unexplained departure from\nCASAC’s recommendation and from EPA’s prior position\nregarding the adversity of certain lung function decrements.\nWe reject both contentions.\n\n Environmental Petitioners argue that EPA failed to fulfill\nits duty under the Act to provide “an explanation of the\nreasons” for departing from CASAC’s scientific\nrecommendations. 42 U.S.C. § 7607(d)(3); see also id.\n§ 7607(d)(6). They contend CASAC told EPA that “[a]t [0.07\nppm] there is substantial scientific certainty of a variety of\nadverse effects, including decrease in lung function, increase\nin respiratory symptoms, and increase in airway\ninflammation,” Envtl. Pet’rs Br. 31 (quoting J.A. 531), and\nEPA failed to rationally dispute or refute this finding.\n\n EPA adequately considered CASAC’s advice when setting\nthe primary NAAQS to 0.07 ppm. While CASAC advised EPA\nto “set the level of the standard lower than [0.07 ppm],” J.A.\n531, that recommendation constituted policy—not scientific—\nadvice. This distinction is crucial because we defer to EPA’s\njudgment on issues of policy but require EPA to “fully explain\nits reasons for any departure from” “CASAC’s expert scientific\nrecommendations.” Mississippi, 744 F.3d at 1358, 1354-55\n(emphasis added). CASAC’s letter to EPA makes clear that\n“based on the scientific evidence” it “recommend[ed] a range\nof levels for a revised primary ozone standard from [0.07 ppm]\nto [0.06 ppm].” J.A. 531. CASAC then “acknowledge[d] that\nthe choice of a level within the range recommended based on\nscientific evidence is a policy judgment.” Id.\n\f 19\n In an effort to influence EPA’s policy judgment, CASAC\nnoted that given the “substantial scientific certainty of a variety\nof adverse effects” at 0.07 ppm, that level may not be adequate\nto protect public health. Id. So, CASAC’s “policy advice\n[was] to set the level of the standard lower than [0.07 ppm]\nwithin a range down to [0.06 ppm], taking into account [EPA’s]\njudgment regarding the desired margin of safety to protect\npublic health.” Id. EPA did not take CASAC’s policy advice\nbecause the Administrator found that 0.07 ppm would still\n“provide substantial protection against the broader range of\n[ozone] exposure concentrations that have been shown in\ncontrolled human exposure studies to result in respiratory\neffects, including exposure concentrations below [0.07 ppm].”\n80 Fed. Reg. at 65,363. EPA nonetheless chose a level for the\nprimary standard that was within CASAC’s scientifically\nrecommended range. In so doing, EPA did not abrogate its\nduty under the Clean Air Act to consider CASAC’s scientific\nrecommendations but instead made a valid policy decision.\n\n Environmental Petitioners also argue that EPA should\nhave set the primary standard lower than 0.07 ppm, given the\ncontrolled human exposure studies that had been published\nsince the agency’s 2008 NAAQS review. Specifically,\nEnvironmental Petitioners highlight two new studies\nevaluating exposures to 0.06-0.063 ppm ozone that concluded\nthat exposures at 0.06 ppm caused lung function decrements of\n10% or more. Taken with EPA’s statement in the 2008 ozone\nNAAQS rule that a lung function decrement of 10% or more\n“should be considered adverse for sensitive populations,” 73\nFed. Reg. at 16,454-55, Environmental Petitioners contend that\nEPA acted arbitrarily by giving insufficient weight to those\nnew studies. We are unpersuaded by this argument, which fails\nto appreciate the dynamic nature of adversity determinations at\neach NAAQS review.\n\f 20\n In determining the appropriate level for the 2015 primary\nstandard, the Administrator was not bound by adversity\njudgments in the 2008 NAAQS review. Indeed, “as the\ncontours and texture of scientific knowledge change, the\nepistemological posture of EPA’s NAAQS review necessarily\nchanges as well.” Mississippi, 744 F.3d at 1344. Thus, we\nconsider only whether EPA’s 2015 Rule offers a rational\nexplanation of why EPA chose a new adversity definition and\nwhether the Administrator reasonably evaluated the evidence\nof adversity. We hold that EPA’s actions were reasonable and\nreasonably explained. Rather than applying a rigid test for\ndetermining what level of decrement is adverse, the\nAdministrator took a more comprehensive approach provided\nby American Thoracic Society (ATS) guidelines. ATS\nguidelines provide that “reversible loss of lung function in\ncombination with the presence of [respiratory] symptoms\nshould be considered adverse.” 80 Fed. Reg. at 65,357\n(internal quotation marks omitted). EPA reasonably explained\nthat it chose to adopt the ATS definition of adversity because\n“the available evidence does not provide information on the\nextent to which a short-term, transient decrease in lung function\nin a population,” without more, could “change the risk profile\nof the population,” and that CASAC was “conditional” about\nwhether “the lung function decrements observed in some\npeople at [0.06 ppm] . . . are adverse.” Id. at 65,358. The\nclinical studies that Environmental Petitioners contend EPA\ndismissed concluded that lung function decrements (such as\n10%) occurred in some individuals at lower ozone\nconcentrations, including 0.06 and 0.063 ppm, but not in\n“combination [with] statistically significant increases in\nrespiratory symptoms.” See id. at 65,357. Under ATS\nguidance, the Administrator declined to designate such loss of\nlung function as adverse. This decision was rational, and\nEnvironmental Petitioners cannot show that the evidence\nrequired EPA to decide differently.\n\f 21\n III. Secondary Standard Challenges\n\n Next, we turn to the challenges to EPA’s secondary\nstandard. The Environmental Petitioners contend that: (1) in\nconsidering tree growth loss, EPA acted arbitrarily in setting\nthe target level of air quality and therefore fell short of the\nstatutory requirement to set a standard “requisite” to protect\nagainst such harm; (2) EPA arbitrarily used a three-year\naverage rather than a single-year, cumulative measurement of\nozone exposure as a benchmark to gauge the protectiveness of\nits standard, and arbitrarily declined to adopt the single-year\ncumulative exposure index as the form and averaging time for\nthe secondary standard; and (3) EPA unreasonably failed to\nidentify a level of air quality requisite to protect the public\nwelfare against adverse effects from visible leaf injury.\nIndustry Petitioners argue only that EPA failed to justify its\ndecision to lower the secondary standard from its 2008 level.\nWe hold that EPA has not explained its decision to set a target\nlevel of protection against tree growth loss based on a three-\nyear average of cumulative, single-year ozone exposures, nor\nhas it justified its decision not to specify any level of air quality\nrequisite to protect against visible leaf injury. We reserve\njudgment on whether EPA reasonably declined to adopt the\ncumulative exposure index as the form and averaging time, and\ndeny the remainder of the challenges.\n\n We begin by reviewing EPA’s secondary standard-setting\nprocess. EPA concentrated its review on the association\nbetween ozone exposure and “vegetation effects,” focusing on\ntree growth loss (also referred to as “relative biomass loss”),\ncrop yield loss, and visible leaf (or “foliar”) injury. 80 Fed.\nReg. at 65,370-71. CASAC agreed that those three effects\nwere “appropriate surrogates [for] a wide range of damage that\nis adverse to public welfare” because ozone damage to trees,\nleaves, and crops can directly affect numerous resources and\n\f 22\necosystem services that are important to the public and\nindirectly affect a wide array of ecosystem components and\nfunctions. J.A. 532-33.\n\n In performing its analysis, EPA used an “exposure metric”\ncalled the “W126 index” to gauge how differing levels of ozone\nexposure correspond to effects on tree growth, crop yields, and\nvisible leaf injury. 80 Fed. Reg. at 65,373. The W126 index,\nwhich EPA deemed to be the “most biologically relevant\nmetric[] for consideration of [ozone] exposures eliciting\nvegetation-related effects,” measures the cumulative amount of\nozone to which a plant is exposed over a single three-month\ngrowing season. Id. at 65,373-74. W126 levels are expressed\nas parts-per-million hours (ppm-hrs). Id.\n\n CASAC recommended that EPA use the single-year W126\nexposure index as the form and averaging time for the\nsecondary standard, J.A. 518, meaning that compliance with\nthe standard would be measured based on a single growing\nseason’s worth of cumulative ozone exposure. CASAC\nadvised EPA to set the level for the secondary standard within\na range of 7 ppm-hrs and 15 ppm-hrs. Based on the data\nCASAC used, that corresponded to median annual tree growth\nloss between 2% and 5.2%. J.A. 518, 534-36, 631. CASAC\ncautioned that “at 17 ppm-hrs, the median tree species has 6%\nrelative biomass loss,” which would be “unacceptably high.”\nJ.A. 518. Regarding the other two surrogates, CASAC\ncounseled that, to protect against loss of crop yield, “a level of\n15 ppm-hrs for the highest 3-month sum in a single year is\nrequisite,” and to reduce foliar injury, a “level below 10 ppm-\nhrs is required.” Id. All of those levels were based on single-\nyear measuring periods, so CASAC advised that if EPA were\nto base its secondary standard on a three-year average of the\nrelevant measurements, it should lower the level of the standard\n\f 23\n“to protect against single unusually damaging years that will be\nobscured in the average.” J.A. 536.\n\n EPA agreed with CASAC’s recommendation that 6% tree\ngrowth loss would be “unacceptably high.” 80 Fed. Reg. at\n65,406. However, as EPA explained, id. at 65,384, 65,392\nn.197, CASAC’s advice on that was based in part on a study of\ncottonwood trees, which are uniquely ozone-sensitive, and\nCASAC itself advised EPA that the “cottonwood data . . .\nreceive[d] too much emphasis” in EPA’s analysis, J.A. 533.\nAfter excluding the cottonwood data, EPA concluded that\nozone exposure of 19 ppm-hrs was associated with 6% tree\ngrowth loss, and that 17 ppm-hrs of ozone exposure brought it\ndown to 5.3% loss. 80 Fed. Reg. at 65,396, 65,407. EPA\naccordingly chose to focus on a standard “somewhat below”\nthe 19 ppm-hrs level associated with 6% median tree growth\nloss. Id. at 65,406-07.\n\n EPA then departed from CASAC’s advice in several ways.\nFirst, it chose not to use the W126 cumulative seasonal\nexposure index as the form and averaging time of the standard,\nopting instead to retain the averaging time (8 hours) and form\nit had used in the 2008 rule (the three-year average of the\nfourth-highest daily level). EPA purported to use as a\n“benchmark,” or target level of protectiveness, the high end of\nthe range of W126 levels CASAC had recommended. Recall\nthat CASAC had recommended that the maximum seasonal\nozone exposure not exceed (the cottonwood-adjusted) 17 ppm-\nhrs in any single growing season. In deciding what exposure\nlevel correlated to CASAC’s 17 ppm-hrs, however, EPA used\na three-year average of anticipated seasonal exposure levels.\nThus, EPA chose to set the standard at a level that it projected\n(based on a statistical analysis of past ozone exposure data)\nwould “in nearly all instances” going forward restrict the\naverage of three growing seasons’ ozone exposures to the\n\f 24\nequivalent of 17 ppm-hrs. Id. at 65,407. But by using that\nmethod to set the level, EPA arrived at a standard that\nstatistically tolerates cumulative ozone exposures in a single\ngrowing season that are higher than CASAC’s maximum\nrecommended level. EPA also did not specify any level of\nozone to protect against visible leaf injury or crop loss. The\nagency reasoned that data for those public-welfare harms were\ntoo uncertain to permit it to discern a level that would provide\nthe requisite protections, and that the standard it set to protect\nagainst adverse effects from tree growth loss would at least\nprovide “some increased protection” against visible leaf injury\nand crop damage. Id. at 65,407-08.\n\n In sum, EPA ultimately chose to set the level of the\nsecondary standard at 0.07 ppm while retaining the form and\naveraging time it had previously used. Id. at 65,410.\n\nA. Environmental Petitioners’ Challenges\n\n The essence of the Environmental Petitioners’ petition is\nthat EPA did not adequately explain its deviations from\nCASAC’s advice. “In order to enable judicial review and to\nsatisfy its statutory obligation to explain its reasons for\ndeparting from CASAC, EPA must be precise in describing the\nbasis for its disagreement with CASAC.” Mississippi, 744 F.3d\nat 1355. Where EPA diverges from CASAC’s scientific\nadvice, there must be “substantial evidence in the record when\nconsidered as a whole which supports the Administrator’s\ndeterminations.” Id. (quoting Lead Indus. Ass’n v. EPA, 647\nF.2d 1130, 1146 (D.C. Cir. 1980)).\n\f 25\n EPA reasonably used 17 ppm-hrs as the\n benchmark level to protect against adverse effects\n associated with tree growth loss.\n\n The Environmental Petitioners argue that, by establishing\n17 ppm-hrs as the target level of protection against adverse\nwelfare effects associated with tree growth loss, EPA\nimpermissibly departed upward from CASAC’s advice in two\nways—first, by excluding the data from the cottonwood tree\nstudy from its tree growth loss analysis and, second, by\nfocusing on limiting tree growth loss to under 6%, rather than\nthe 2% target they say CASAC’s analysis required. Setting\naside for the moment EPA’s decision to average the benchmark\nover three years, we conclude that EPA adequately explained\nits decision to exclude the cottonwood data and acted\nconsistently with CASAC’s advice in choosing to limit tree\ngrowth loss to under 6%.\n\n Cottonwood Data. EPA reviewed seedling studies of\ntwelve different tree species to determine the median percentile\nof tree growth loss at varying ozone levels. See 80 Fed. Reg.\nat 65,372. One of those studies—of cottonwood seedlings—\nsignificantly lowered the ozone level at which the twelve tree\nspecies experience 6% median growth loss. See id. at 65,384,\n65,391-92. CASAC had itself relied on that study, see J.A.\n537, but advised EPA that the cottonwood data had received\n“too much emphasis” because the study “did not control for\nozone and climatic conditions,” and the results “show[ed]\nextreme sensitivity to ozone compared to other studies,” J.A.\n533. EPA accordingly excluded the cottonwood data from its\ntree growth analysis, which increased the ozone level\nassociated with 6% median growth tree loss from 17 ppm-hrs\nup to 19 ppm-hrs. 80 Fed. Reg. at 65,384, 65,391-92.\n\f 26\n The Environmental Petitioners object that EPA arbitrarily\ndisregarded CASAC’s advice because, they say, CASAC never\nexpressly advised EPA to disregard the cottonwood data\naltogether, and CASAC relied on it in setting its recommended\nexposure range. But we can only discern an unreasonable\ndeviation where CASAC itself has been “precise about the\nbasis for its recommendations.” Mississippi, 744 F.3d at 1358.\nIt was far from precise on the disputed point. It warned EPA\nto treat the cottonwood data with caution, but recommended a\nrange of ozone levels in reliance on the median tree growth loss\nestimates (outlined in Table 6-1 of EPA’s staff’s second draft\nPolicy Assessment), which weigh the cottonwood data equally\nwith other seedling studies. See J.A. 537 (citing EPA, Policy\nAssessment for the Review of the Ozone NAAQS, Second\nExternal Review Draft 6-19, tbl. 6-1 (2014), J.A. 631). The\nEnvironmental Petitioners suggest that CASAC was only\nwarning EPA to take care in describing the cottonwood data in\nthe narrative portion of its Policy Assessment, not suggesting\nanything about how the data was actually used, but that makes\nlittle sense in the context of an exchange about quantitative\ngrowth loss estimates. In light of CASAC’s mixed messages,\nwe cannot say that EPA’s decision to exclude the cottonwood\ndata was arbitrary.\n\n The 6% Target. Emphasizing CASAC’s admonition that\n6% tree growth loss would be “unacceptably high,” EPA\ndecided “to focus on a standard that would generally limit\ncumulative exposures to those for which the median [tree\ngrowth loss] estimate would be somewhat lower than 6%.” 80\nFed. Reg. at 65,406-07. The Environmental Petitioners\ncontend that CASAC “plainly specified” that 2% median\ngrowth loss was the requisite level to avoid adverse welfare\neffects related to tree growth loss, Envtl. Pet’rs Br. 42-44, but\nthey misconstrue CASAC’s advice.\n\f 27\n CASAC did not identify 2% growth loss as the only\nsufficiently protective level, but as “an appropriate\nscientifically based value to consider as a benchmark of\nadverse impact,” elaborating that it would be “appropriate to\nidentify a range of levels of alternative W126-based standards\nthat includes levels that aim for not greater than 2% [growth\nloss] for the median tree species.” J.A. 537 (emphases added).\nCASAC in fact ultimately recommended that EPA set the\nstandard between 7 ppm-hrs and 15 ppm-hrs, J.A. 518, which,\nbased on the seedling data upon which CASAC relied\n(including the cottonwood data), corresponds to median annual\ntree growth loss between 2% and 5.2%, see J.A. 631. In other\nwords, CASAC recommended that EPA consider limiting tree\ngrowth loss to 2% as the lower end of a range of permissible\ntarget levels. EPA followed CASAC’s advice by considering\na 2% growth loss target as part of a range of growth loss targets\nand determined that the studies underlying CASAC’s low-end\nrecommendation were insufficiently reliable to base the\nsecondary standard on that low level of tree growth loss. See\n80 Fed. Reg. at 65,394.\n\n The Environmental Petitioners passingly disparage EPA’s\nassessment of the studies associated with CASAC’s 2% growth\nloss target, but that argument is equally unavailing. EPA\nreasonably explained why it thought those studies were\nunreliable, and its “evaluation of ‘scientific data within its\ntechnical expertise’” is entitled significant deference. See\nMiss. Comm’n on Envtl. Quality v. EPA, 790 F.3d 138, 150,\n155-56 (D.C. Cir. 2015) (quoting City of Waukesha v. EPA,\n320 F.3d 228, 247 (D.C. Cir. 2003)). For instance, in support\nof its advice that EPA consider limiting median annual tree\ngrowth loss to 2%, CASAC highlighted a study purporting to\nshow that stands of aspen that experienced ozone-induced\nbiomass loss of 2% annually had cumulative biomass loss of\n21% over seven years, suggesting that even relatively minor\n\f 28\nannual growth loss can lead to significant loss over longer\nperiods of time. J.A. 537. But CASAC had misread the study.\nAs EPA explained, the aspen in that study in fact experienced\nmore than 20% biomass loss annually, suggesting that their\ncumulative biomass loss was not much worse than their loss in\nany single year. 80 Fed. Reg. at 65,394.\n\n EPA also clarified that the only other report CASAC relied\non for its 2% target recommendation provided no explicit\nrationale for selecting that level and did not identify any new\ndata in support of its recommendations. See id. at 65,394-95 &\nn.200. While it is true that EPA has cited that same report\nfavorably in the past, see, e.g., National Ambient Air Quality\nStandards for Ozone, 75 Fed. Reg. 2,938, 3,025 (Jan. 19, 2010),\nthat alone does not make EPA’s current assessment arbitrary.\nEPA must simply provide a “reasoned explanation” for its\ndeparture from its past position—which it has done. See\nEncino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125-26\n(2016); see also Mississippi, 744 F.3d at 1351.\n\n EPA did not adequately explain its decision to use\n a three-year average of the W126 index as the\n benchmark.\n\n The Environmental Petitioners raise two challenges\npertaining to the W126 index, both as a benchmark and as a\nform and averaging time. First, they argue that EPA\nimpermissibly departed from CASAC’s advice by setting the\nsecondary standard level using a three-year average W126\nbenchmark without lowering the level to protect against single-\nyear exposures associated with median annual tree growth loss\nof 6%, which CASAC had advised was “unacceptably high.”\nJ.A. 518. They also contend that EPA arbitrarily disregarded\nCASAC’s advice to adopt the W126 index as the form and\n\f 29\naveraging time for the secondary standard. We remand to EPA\non the first issue and do not reach the second.\n\n The Three-Year Average Benchmark. CASAC advised\nEPA that basing the secondary standard on a “single-year\nperiod” would provide “more protection for annual crops and\nfor the anticipated cumulative effects on perennial species”\nthan a three-year average. J.A. 518. It explained that EPA’s\nproposal to use a three-year averaging period was “not\nsupported by the available data,” J.A. 536, and that if EPA\nchose to “base the secondary standard on a three-year\naveraging period,” then “the level of the standard should be\nrevised downward such that the level for the highest three-\nmonth summation in any given year of the three-year period\nwould not exceed [its] scientifically recommended” range of\nsingle-year, W126 exposure levels, J.A. 518. This was\nnecessary, CASAC explained, to “protect against single\nunusually damaging years that will be obscured in the\naverage.” J.A. 536.\n\n EPA argues it gave effect to CASAC’s recommendation\nby using a three-year average benchmark of 17 ppm-hrs,\nwhich, after adjusting for the cottonwood data, was “somewhat\nbelow” the 19 ppm-hrs associated with 6% annual growth loss\nthat CASAC had advised was “unacceptably high.” 80 Fed.\nReg. at 65,406-07 (quoting J.A. 518). But CASAC had advised\na maximum level associated with 5.2% annual biomass loss,\nsee J.A. 631, and it expressly cautioned that 6% median growth\nloss in a single year was unacceptable, see J.A. 518. EPA’s use\nof a benchmark that averages out to less than 6% biomass loss\nover three years does not accord with CASAC’s advice.\nIndeed, as commenters informed EPA during the rulemaking,\nsee J.A. 1836-40, EPA’s own air quality data suggests that\nmany large national parks and wilderness areas that have met\nEPA’s chosen three-year average 17 ppm-hrs benchmark—\n\f 30\nareas that Congress considers significant to the public welfare,\nsee 42 U.S.C. §§ 7470(2), 7472(a)—have meanwhile recorded\nsingle-year W126 values at and above 19 ppm-hrs, which is\nassociated with “unacceptably high” annual biomass loss of\n6% and higher. See J.A. 1061-64; 80 Fed. Reg. at 65,391. EPA\ncritiques that data as marred by outdated handling procedures,\nbut the agency acknowledged that other data derived through\nupdated procedures produced results “similar to” those\nshowing harmful exposure spikes. J.A. 1213. Critically, EPA\npoints us to no data or analysis (based on new or old\nprocedures) suggesting that the chosen benchmark prevents\nsingle seasonal exposures of 19 ppm-hrs or higher. In short,\nEPA has not demonstrated how its chosen benchmark protects\nagainst “unusually damaging years that will be obscured in the\naverage.” J.A. 536.\n\n EPA alternatively defends its decision to use the three-\nyear-average benchmark as providing a focus on public welfare\neffects of “potentially greater” significance than effects\n“associated with a single year” of exposure. 80 Fed. Reg. at\n65,404. This position, however, is inconsistent with EPA’s\nother actions. In establishing the secondary standard, for\nexample, EPA heavily relied on data showing annual 6%\nmedian tree growth loss at 19 ppm-hrs, see id. at 65,406, and\nacknowledged the potential for a single season of high ozone\nexposure to “alter biomass allocation and plant reproduction in\nseasons subsequent to [that season’s] exposure,” thereby\nleading to “a negative impact on species regeneration in\nsubsequent years,” id. at 65,371-72; see also J.A. 740-41. EPA\nadditionally recognized that “ozone effects in plants are\ncumulative,” id. at 65,373 (quoting EPA, Integrated Science\nAssessment 2-44 (2013)), meaning that the adverse vegetative\neffects from single, high-ozone years are not offset by\nsubsequent low-ozone years. EPA has identified no contrary\n\f 31\nevidence in the record demonstrating why these single-year\neffects matter less than a three-year average.\n\n We therefore remand this issue for EPA to either lower the\nstandard to protect against unusually damaging cumulative\nseasonal exposures that will be obscured in its three-year\naverage, or explain its conclusion that the unadjusted average\nis an appropriate benchmark notwithstanding CASAC’s\ncontrary advice. Alternatively, EPA could adopt the single-\nyear W126 exposure index as the form and averaging time,\nwhich would presumably moot any problems with the way it\ntranslated that index to use as a benchmark.\n\n The Form and Averaging Time. CASAC also\nrecommended that EPA use the single-year W126 index as the\nform and averaging time for the secondary standard. J.A. 518.\nEPA instead chose to retain the existing form and averaging\ntime—the three-year average of the fourth-highest daily\nmaximum eight-hour concentration. Adopting the W126 index\nas the form and averaging time was unnecessary, EPA\nexplained, because the ozone exposure levels associated with\nthe existing form and averaging time and a three-year average\nof the W126 index are “highly correlated,” especially at lower\nlevels, and “future control programs designed to help meet a\nprimary [ozone] standard based on the” current form and\naveraging time should “provide similar improvements in terms\nof the 3-year average of the annual W126 metric.” J.A. 1253;\nsee also 80 Fed Reg. at 65,400-01, 408-09.\n\n The Environmental Petitioners argue that EPA did not\njustify its decision not to adopt the W126 index as the form and\naveraging time. We lack any basis to assess the reasonableness\nof EPA’s actions, however, because a critical piece of the\npuzzle is missing. To review: EPA chose not to use the W126\nindex as the form and averaging time because it found that\n\f 32\nozone exposure levels associated with the existing form and\naveraging time are “highly correlated” to a three-year average\nof the W126 index. But, as discussed, EPA never explained\nwhy it is reasonable to focus on a three-year average of the\nW126 index in the first place. Therefore, we cannot assess the\nrelevance of the claim that the two are “highly correlated.”\nEPA’s reconsideration on remand of the three-year averaging\nissue should supply us with the information necessary to\nresolve this question, or, if EPA chooses to follow CASAC’s\nadvice to lower the standard to control for unusually high single\nyears, potentially moot the Environmental Petitioners’ concern\nthat the current form tolerates even three-year average W126\nlevels higher than 17 ppm-hrs during periods when a 0.07 ppm,\n8-hour level is met. Accordingly, we decline to reach this\nquestion.\n\n EPA arbitrarily failed to identify a level of air\n quality requisite to protect against adverse effects\n from visible leaf injury.\n\n The Environmental Petitioners also fault EPA for deciding\nnot to specify a level of air quality to protect against adverse\nwelfare effects from ozone-induced visible leaf injury.\nBecause EPA failed to offer a reasoned explanation, we remand\nfor it to reconsider.\n\n EPA has found that the “strongest evidence for effects\nfrom [ozone] exposure on vegetation is from controlled\nexposure studies, which ‘have clearly shown that exposure to\n[ozone] is causally linked to visible foliar injury.’” 80 Fed.\nReg. at 65,370 (quoting EPA, Integrated Science Assessment\n1-15). Relying on that evidence, CASAC advised EPA of its\n“scientific judgment” that “a level of 10 ppm-hrs is required to\nreduce foliar injury.” J.A. 538. EPA nonetheless concluded\nthat there were too many “uncertainties and complexities” in\n\f 33\nthe evidence to specify a level of air quality to protect against\nleaf injury. See 80 Fed. Reg. at 65,407-08. But the mere\ninvocation of “substantial uncertainty” is not a justification for\nthe agency’s failure to fulfill its statutory mandate. Motor\nVehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins.\nCo., 463 U.S. 29, 52 (1983) (internal quotation marks omitted).\nEPA “must explain the evidence which is available, and must\noffer a ‘rational connection between the facts found and the\nchoice made.’” Id. (quoting Burlington Truck Lines, Inc. v.\nUnited States, 371 U.S. 156, 168 (1962)). Where CASAC has\n“reached a scientific conclusion that adverse [welfare] effects\n[are] likely to occur,” EPA must “explain why the evidence on\nwhich CASAC relied cannot support the degree of confidence\nCASAC placed in it.” Mississippi, 744 F.3d at 1357; see also\nAm. Farm Bureau Fed’n v. EPA, 559 F.3d 512, 529-30 (D.C.\nCir. 2009). We defer to EPA’s judgment that the available\nevidence is too uncertain only when the agency reasonably\nexplains its decision, see State Farm, 463 U.S. at 51-53, and\nEPA has failed to carry that burden here.\n\n None of the identified uncertainties justifies EPA’s\ndecision not to set a standard to protect against visible leaf\ninjury. For instance, EPA argues that it lacked criteria for\nassessing the welfare effects of visible leaf injury, but that does\nnot square with its conclusion in the Integrated Science\nAssessment (compiling all available scientific criteria) that\nvisible leaf injury “has been well characterized and\ndocumented over several decades.” J.A. 985-86. EPA had at\nits fingertips a wealth of new data regarding visible leaf injury,\nincluding new controlled exposure studies, multiyear field\nsurveys, and biomonitoring data of ozone-induced visible leaf\ninjury on public lands. J.A. 749. Using that data, EPA\nanalyzed how visible leaf injury affects “ecosystem services”\nlike aesthetic value and recreation on “Class I” lands—lands\nthat “have particular public welfare significance.” J.A. 749-50,\n\f 34\n1036-44; see also 42 U.S.C. §§ 7470(2), 7472(a). CASAC\npointed out that those same ecosystem services are relevant in\nidentifying “damage that is adverse to public welfare.” J.A.\n532-33.\n\n CASAC concluded, based on the same data available to\nEPA, that “a level of 10 ppm-hrs is required to reduce foliar\ninjury.” J.A. 538. EPA counters that “decreases in leaf injury\n[also occur] with decreasing ozone exposures across a range of\nvalues well above 10 ppm-hrs,” and that CASAC\nrecommended a range of ozone levels that included values\nabove 10 ppm-hrs. See EPA Br. 89; 80 Fed. Reg. at 65,395-\n96. Be that as it may, EPA has not explained why, unlike\nCASAC, it was unable to choose a level at all. Indeed,\n“[b]ecause the EPA failed to identify any target level, we need\nnot decide whether it was reasonable for the agency to reject\nthe target recommended by . . . CASAC.” Am. Farm Bureau,\n559 F.3d at 530.\n\n EPA objects that it only found that visible leaf injury has\nthe “potential to be adverse to the public welfare,” and that the\nClean Air Act does not require EPA “to identify a precise,\nquantified level of public welfare protection for every\npotentially adverse public welfare impact” it considers when\nrevising the secondary standard. EPA Br. 87-88 (first quoting\n80 Fed. Reg. at 65,388). But EPA has failed to justify its\ninaction where CASAC unequivocally found that “damage to\nresource use from foliar injury” was an “adverse welfare\neffect.” J.A. 518. In light of CASAC’s “scientific conclusion\nthat adverse [welfare] effects were likely to occur” from visible\nleaf injury, EPA’s inaction in the face of such effects is\n“unacceptable.” Mississippi, 744 F.3d at 1357. Nor was it\nsufficient for EPA simply to conclude that the standard set to\nprotect against tree growth loss sufficed because it provided\n“additional” protection against leaf injury. 80 Fed. Reg. at\n\f 35\n65,407-08. Nothing in the record suggests that tree growth loss\nis a surrogate for leaf injury. EPA’s assurance that the standard\nset to protect against tree growth loss will provide\n“incidental[]” protection against visible leaf injury “cannot\nsave its decision.” Am. Farm Bureau, 559 F.3d at 529-30.\n\n Other purported uncertainties do not support EPA’s\ninaction. For instance, EPA asserts that the Administrator\n“lacked evidence that would allow her to measure the\nrelationship between leaf injury and other vegetation effects\nthat she might find adverse,” EPA Br. 87, but EPA and CASAC\nboth determined that visible leaf injury itself can “impact the\npublic welfare” by harming “aesthetic value and outdoor\nrecreation” in public lands, 80 Fed. Reg. at 65,379; J.A. 533.\nVisible leaf injury impairs a variety of outdoor activities,\nincluding “scenic viewing, wildlife watching, hiking, and\ncamping, that are of significance to the public welfare” and\ngenerate “millions of dollars in economic value” annually. 80\nFed. Reg. at 65,381; J.A. 749. Much of the documented visible\nleaf injury is in Class I areas, see 80 Fed. Reg. at 65,378, the\npreservation and enjoyment of which is a “clear public\ninterest,” id. at 65,377 (quoting 73 Fed. Reg. at 16,496). Those\ninterests are manifest whether or not the relationship between\nleaf injury and other vegetative effects has been well\ncharacterized.\n\n This case is a far cry from Center for Biological Diversity\nv. EPA, where we upheld EPA’s determination that the\navailable data was too uncertain to support setting a standard\nfor acid rain precursors. 749 F.3d 1079 (D.C. Cir. 2014). In\nthe administrative record in that case, EPA “explained in great\ndetail” why the scientific uncertainties were so “unusually\nprofound” that EPA “could not form” a reasoned judgment as\nto a requisite level of protection. Id. at 1088, 1090-91. CASAC\nconcurred with that assessment. Id. at 1086 n.11. Here, in\n\f 36\ncontrast, CASAC expressed its scientific judgment that a target\nlevel of 10 ppm-hrs was “required to reduce” leaf injury. J.A.\n538. EPA’s failure to explain why it could not accept data that\nCASAC deemed informative falls short of reasoned\ndecisionmaking.\n\nB. Industry Petitioners’ Challenge\n\n The Industry Petitioners glancingly claim that EPA failed\nto explain why, when faced with the same scientific evidence\nthat was available in 2008, it revised the secondary standard\ndownward. As we have explained in reference to the primary\nstandard, it is a “conceptual error” to assume that “EPA is\nsomehow bound by” a prior standard in a subsequent review.\nMississippi, 844 F.3d at 1344. What is more, EPA did in fact\nrely on new scientific evidence, including more than 400 new\nstudies and various new analyses of existing data, which\n“strengthen[ed] [its] confidence” that the 2008 secondary\nstandard was no longer as demanding as was “requisite” to\nprotect the public welfare. See 80 Fed. Reg. at 65,369-70,\n65,384. “[A]dditional certainty . . . that the line marked by the\nterm ‘requisite’ has shifted” is a reasonable—if not\nparadigmatic—basis for EPA to revise its prior standards.\nMississippi, 844 F.3d at 1344.\n\n IV. Cross-Cutting NAAQS Challenges\n\n We now turn to the arguments of State and Industry\nPetitioners that EPA failed to take into account all the factors\nrequired by law when setting the primary and secondary\nstandards. First, they fault EPA for not considering the “overall\nadverse economic, social, and energy impacts” of the revised\nNAAQS. Second, they argue that the Clean Air Act requires\nEPA to consider the impact of background ozone on the ability\nof states to attain the revised NAAQS. And, lastly, they argue\nthat EPA’s interpretation of the Act, which excluded\n\f 37\nconsideration of background ozone, leaves no “intelligible\nprinciple” by which to set NAAQS, creating a constitutional\nnondelegation issue. None of these arguments has merit.\n\nA. Adverse Impacts\n\n According to section 109(b) of the Clean Air Act, EPA\nmust set the NAAQS at a level “requisite to protect the public\nhealth” and “the public welfare.” 42 U.S.C. § 7409(b). The\nNAAQS should be revised periodically “as may be\nappropriate.” Id. § 7409(d)(1). Industry Petitioners invoke the\nterm “appropriate” and claim that “an evaluation of\n‘appropriateness’ must take into account the adverse\nsocioeconomic and energy impacts” of the revised NAAQS.\nIndus. Pet’rs Br. 33. To bolster this argument, they point to\nMichigan v. EPA, which held that the phrase “appropriate and\nnecessary” in section 112(n) of the Act—governing power\nplants’ hazardous air pollution—“requires at least some\nattention to cost” because “‘appropriate’ is ‘the classic broad\nand all-encompassing term that naturally and traditionally\nincludes consideration of all the relevant factors.’” 135 S. Ct.\nat 2707 (quoting White Stallion Energy Ctr., LLC v. EPA, 748\nF.3d 1222, 1266 (D.C. Cir. 2014) (Kavanaugh, J., concurring\nin part and dissenting in part)).\n\n In Whitman, the Supreme Court considered a similar\nargument regarding NAAQS-setting, with challengers\nasserting that the words “requisite” and “adequate” in section\n109(b)(1) meant that EPA must consider the economic costs of\nimplementing revised NAAQS. 531 U.S. at 468. The Whitman\nCourt rejected this reading, finding that the plain text of the Act\n“unambiguously bars cost considerations from the NAAQS-\nsetting process.” Id. at 471. The text of the Act refers only to\nthe “public health,” and though it uses the terms “requisite” and\n“adequate,” the Whitman Court found it “implausible that\n\f 38\nCongress would give to the EPA through these modest words\nthe power to determine whether implementation costs should\nmoderate national air quality standards,” especially given that\nother Clean Air Act provisions “explicitly permit[] or require[]\neconomic costs to be taken into account in implementing\n[other] air quality standards.” Id. at 467-68.\n\n According to Whitman, the Act commands EPA to follow\na specific process when setting primary NAAQS: “[I]dentify\nthe maximum airborne concentration of a pollutant that the\npublic health can tolerate, decrease the concentration to\nprovide an ‘adequate’ margin of safety, and set the standard at\nthat level.” Id. at 465. Industry Petitioners argue that this\naccount of the NAAQS-setting process is incomplete, and that\nconsideration of other “adverse economic, social, and energy\nimpacts” is required. But at bottom this is the same argument\nrejected in Whitman, with the “costs” of the revised and more\nstringent NAAQS merely reframed as “impacts.” In listing\nthose impacts, Industry Petitioners emphasize that stricter\nstandards could reduce gross domestic product, drive up energy\ncosts, “stymie economic growth by forcing the early retirement\nof facilities unable to implement controls, contributing to job\nlosses[,] discourag[e] existing businesses from expanding in\nnonattainment regions[,] and driv[e] away potential new\ninvestments,” Indus. Pet’rs Br. 35, but these “impacts” are no\ndifferent than the “economic costs” that the petitioners in\nWhitman worried “might produce health losses sufficient to\noffset the health gains achieved in cleaning the air—for\nexample, by closing down whole industries and thereby\nimpoverishing the workers and consumers dependent upon\nthose industries,” 531 U.S. at 466. Whitman forbids EPA from\ntaking these considerations into account, however\ndenominated.\n\f 39\n As for Petitioners’ reliance on Michigan and the Act’s use\nof the word “appropriate,” that argument fails twice over. We\nhave already rejected the idea that “appropriate” in section\n109(d) requires consideration of economic costs. Am. Trucking\nAss’ns, Inc. v. EPA, 175 F.3d 1027, 1040-41 (D.C. Cir. 1999).\nFurther, Michigan involved a different provision of the Clean\nAir Act, and the Court was careful to emphasize that its reading\nof “appropriate” was dependent on the statutory context,\nexplaining that “[t]here are undoubtedly settings in which the\nphrase ‘appropriate and necessary’ does not encompass cost.”\n135 S. Ct. at 2707. Indeed, Michigan explicitly distinguished\nsection 112(n) from section 109(b)(1), explaining yet again that\nthe criteria for setting NAAQS “do[] not encompass cost.” Id.\nat 2709.\n\n Industry Petitioners also point to section 109(d)(2)(C) of\nthe Act, which requires CASAC to advise EPA “of any adverse\npublic health, welfare, social, economic, or energy effects\nwhich may result from various strategies for attainment and\nmaintenance” of revised NAAQS. 42 U.S.C. § 7409(d)(2)(C).\nAccording to Petitioners, the fact that CASAC is required to\nsupply information to EPA about the “social, economic, or\nenergy effects” of the revised NAAQS implies that EPA is\nobliged to consider that information in setting the NAAQS.\nThat argument, however, was also raised and rejected in\nWhitman, where the Supreme Court concluded that this\nprovision was intended to “enable the [EPA] to assist the States\nin carrying out their statutory role as primary implementers of\nthe NAAQS,” but had “no bearing upon whether cost\nconsiderations are to be taken into account in formulating the\n[NAAQS].” 531 U.S. at 470-71 (second emphasis added).\n\n Petitioners’ remaining arguments rely heavily on Justice\nBreyer’s concurrence in Whitman, in which he urged that the\nAct “does not require the EPA to eliminate every health risk,\n\f 40\nhowever slight, at any economic cost, however great, to the\npoint of hurtling industry over the brink of ruin, or even forcing\ndeindustrialization.” Id. at 494 (Breyer, J., concurring in part\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Murray Energy Corporation v. EPA\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-23\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: Yes\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued December 18, 2018 Decided August 23, 2019\\n\\n No. 15-1385\\n\\n MURRAY ENERGY CORPORATION,\\n PETITIONER\\n\\n v.\\n\\n ENVIRONMENTAL PROTECTION AGENCY,\\n RESPONDENT\\n\\n AMERICAN LUNG ASSOCIATION, ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 15-1392, 15-1490, 15-1491, 15-1494\\n\\n\\n On Petitions for Review of Final Agency Action of\\n the United States Environmental Protection Agency\\n\\n\\n Seth L. Johnson argued the cause for Public Health and\\nEnvironmental Petitioners. With him on the briefs were Joshua\\nStebbins, Joshua Berman, David S. Baron, and Paul Cort.\\n\\f 2\\n Dominic E. Draye, Deputy Solicitor General, Office of the\\nAttorney General for the State of Arizona, argued the cause for\\nState Petitioners. With him on the briefs were Mark Brnovich,\\nAttorney General, John R. Lopez, IV, Solicitor General, Keith\\nMiller, Associate Solicitor, Joshua L. Kaul, Attorney General,\\nOffice of the Attorney General for the State of Wisconsin,\\nMisha Tseytlin, Solicitor General, Daniel P. Lennington,\\nAssistant Attorney General, Leslie C. Rutledge, Attorney\\nGeneral, Office of the Attorney General for the State of\\nArkansas, Lee Rudofsky, Solicitor General, Jamie L. Ewing,\\nAssistant Attorney General, Jeff Landry, Attorney General,\\nOffice of the Attorney General for the State of Louisiana,\\nSteven B. ABeaux@ Jones, Assistant Attorney General, Wayne\\nStenehjem, Attorney General, Office of the Attorney General\\nfor the State of North Dakota, Margaret I. Olson, Assistant\\nAttorney General, Peter S. Glaser, Carroll W. AMack@\\nMcGuffey, Andy Beshear, Attorney General, Office of the\\nAttorney General for the Commonwealth of Kentucky, Joseph\\nA. Newberg, II, Assistant Attorney General, Lara Katz,\\nAssistant Attorney General, Office of the Attorney General for\\nthe State of New Mexico, Ken Paxton, Attorney General,\\nOffice of the Attorney General for the State of Texas, Priscilla\\nM. Hubenak, Chief, Craig J. Pritzlaff, Assistant Attorney\\nGeneral, E. Scott Pruitt, Attorney General at the time the brief\\nwas filed, Office of the Attorney General for the State of\\nOklahoma, P. Clayton Eubanks, Deputy Solicitor General, and\\nSean Reyes, Attorney General, Office of the State of Utah.\\nMithun Mansinghani, Attorney, Office of the Attorney General\\nfor the State of Oklahoma, Lisa M. Mitchell, Assistant Attorney\\nGeneral, Office of the Attorney General for the State of Texas,\\nOramel H. Skinner, III, Solicitor, Office of the Attorney\\nGeneral for the State of Arizona, Sarah Adkins and Samuel R.\\nFlynn, Assistant Attorneys General, and Gregory T. Dutton,\\nCounsel, Office of the Attorney General for the\\nCommonwealth of Kentucky, Steven C. Kilpatrick, Assistant\\n\\f 3\\nAttorney General, Office of the Attorney General for the State\\nof Wisconsin, Elizabeth B. Murrill and Harry J. Vorhoff,\\nAssistant Attorneys General, Office of the Attorney General\\nfor the State of Louisiana, entered appearances.\\n\\n James R. Bieke argued the cause for Industry Petitioners.\\nWith him on the briefs were Roger R. Martella, Joel F. Visser,\\nScott C. Oostdyk, E. Duncan Getchell, Jr., Michael H. Brady,\\nThomas A. Lorenzen, Robert J. Meyers, Linda E. Kelly,\\nQuentin Riegel, Leland P. Frost, Michael B. Schon, Elizabeth\\nL. Horner, Lucinda Minton Langworthy, Aaron M. Flynn,\\nSteven P. Lehotsky, Sheldon B. Gilbert, Stacy Linden, and\\nRichard S. Moskowitz. Peter C. Tolsdorf entered an\\nappearance.\\n\\n Hope M. Babcock and Sarah J. Fox were on the brief for\\namici curiae American Thoracic Society, et al. in support of\\npetitioners Sierra Club, et al.\\n\\n Thomas J. Ward was on the brief for amicus curiae The\\nNational Association of Home Builders in support of Industry\\nand State Petitioners.\\n\\n Justin Heminger and Simi Bhat, Trial Attorneys, U.S.\\nDepartment of Justice, argued the causes for respondent. With\\nthem on the brief were John C. Cruden, Assistant Attorney\\nGeneral at the time the brief was filed, and David Orlin, Steven\\nSilverman, and Brian Doster, Attorneys, U.S. Environmental\\nProtection Agency. Jon M. Lipshultz, Attorney, U.S.\\nDepartment of Justice, entered an appearance.\\n\\n James R. Bieke argued the cause for Industry Respondent-\\nIntervenors. With him on the brief were Roger R. Martella,\\nJoel F. Visser, Lucinda Minton Langworthy, Aaron M. Flynn,\\nThomas A. Lorenzen, Robert J. Meyers, Stacy Linden, Richard\\n\\f 4\\nS. Moskowitz, Steven P. Lehotsky, Sheldon B. Gilbert, Linda E.\\nKelly, Quentin Riegel, Leland P. Frost, Michael B. Schon,\\nElizabeth Horner, and Leslie A. Hulse. Peter C. Tolsdorf\\nentered an appearance.\\n\\n Seth L. Johnson argued the cause for Health and\\nEnvironmental Respondent-Intervenors. With him on the brief\\nwas David S. Baron. Joshua A. Berman and Joshua R. Stebbins\\nentered appearances.\\n\\n Jonathan Weiner, Deputy Attorney General, Office of the\\nAttorney General for the State of California, argued the cause\\nfor State Amici in support of respondent. With him on the brief\\nwere Kamala D. Harris, Attorney General at the time the brief\\nwas filed, Robert W. Byrne, Senior Assistant Attorney General,\\nand Gavin G. McCabe, Supervising Deputy Attorney General.\\nMelinda Pilling, Attorney, entered an appearance.\\n\\n Richard L. Revesz, Denise A. Grab, Jack Lienke, Michael\\nA. Livermore, and Jason A. Schwartz were on the brief for\\namicus curiae The Institute for Policy Integrity at New York\\nUniversity School of Law in support of respondent.\\n\\n Before: GRIFFITH, PILLARD and WILKINS, Circuit Judges.\\n\\n Opinion for the Court filed PER CURIAM.\\n\\n PER CURIAM: In this opinion, we consider various\\nchallenges to the Environmental Protection Agency’s 2015\\nrevisions to the primary and secondary national ambient air\\nquality standards for ozone. For the reasons given below, we\\ndeny the petitions, except with respect to the secondary ozone\\nstandard, which we remand for reconsideration, and\\ngrandfathering provision, which we vacate.\\n\\f 5\\n I. Background\\n\\nA. Statutory and Procedural Background\\n\\n Congress enacted the modern version of the Clean Air Act\\n(the “Act”), codified at 42 U.S.C. § 7401 et seq., in 1970 to\\ncontrol and reduce contaminants responsible for air pollution\\nwith the overarching goal to protect human health and the\\nenvironment. Pursuant to Title I, EPA must establish, publish,\\nand periodically review primary and secondary national\\nambient air quality standards (“NAAQS”) for air pollutants that\\n“may reasonably be anticipated to endanger public health or\\nwelfare.” 42 U.S.C. § 7408(a)(1)(A). The primary NAAQS\\nare to be set at levels “the attainment and maintenance of which\\nin the judgment of the Administrator, . . . allowing an adequate\\nmargin of safety, are requisite to protect the public health.” Id.\\n§ 7409(b)(1). The secondary NAAQS “shall specify a level of\\nair quality the attainment and maintenance of which in the\\njudgment of the Administrator, . . . is requisite to protect the\\npublic welfare from any known or anticipated adverse effects.”\\nId. § 7409(b)(2). Thus, primary NAAQS protect the public\\nhealth, while the secondary NAAQS protect the public welfare.\\n“Public health” includes adverse health effects for both the\\npopulation at large and sensitive populations such as children,\\nolder adults, and people with asthma or other lung diseases.\\nThe term “public welfare” encompasses a wide variety of\\neffects on soil, plants, wildlife and biota, property damage,\\naesthetic concerns, and other non-health-related impacts such\\nas hazards to economic values and personal comfort. Id.\\n§ 7602(h).\\n\\n Pursuant to section 109(d) of the Clean Air Act, EPA must\\ncomplete a “thorough” review of the NAAQS every five years.\\nId. § 7409(d)(1). During this review, EPA must revise the\\ncriteria and standards or promulgate new standards as\\n\\f 6\\nappropriate. Id. To assist in this process, the Clean Air\\nScientific Advisory Committee (“CASAC”) undertakes an\\nexamination of the current air quality criteria, primary\\nNAAQS, and secondary NAAQS, and submits\\nrecommendations to EPA for “any new [NAAQS] and\\nrevisions of existing criteria and standards as may be\\nappropriate.” Id. § 7409(d)(2)(A)-(B). Congress required EPA\\nto take CASAC’s recommendations into account when\\npromulgating revised NAAQS and to fully explain its reasons\\nwhen it departs from CASAC’s advice. However, the ultimate\\ndecision to revise the NAAQS—and the determination of the\\nnew level—rests with the Administrator. Id. § 7407(d)(3).\\n\\n These petitions concern EPA’s promulgation of revised\\nNAAQS related to ozone. Ozone is a colorless gas that occurs\\nboth in the Earth’s upper atmosphere and at ground level.\\nAlthough ozone is an “essential presence in the atmosphere’s\\nstratospheric layer,” ground-level ozone is an air pollutant that\\nis harmful to breathe and damages crops, trees, and other\\nvegetation. S. Coast Air Quality Mgmt. Dist. v. EPA, 472 F.3d\\n882, 887 (D.C. Cir. 2006); see Allied Local & Reg’l Mfrs.\\nCaucus v. EPA, 215 F.3d 61, 66 (D.C. Cir. 2000). Ground-\\nlevel ozone is not a direct product of human activity, but\\ninstead forms when atmospheric pollutants (including nitrogen\\noxides and volatile organic compounds) react in the presence\\nof sunlight. See Mississippi v. EPA, 744 F.3d 1334, 1340 (D.C.\\nCir. 2013). These precursor atmospheric pollutants are created\\nprimarily from emissions produced by cars, power plants, and\\nchemical solvents. NRDC v. EPA, 777 F.3d 456, 459 (D.C.\\nCir. 2014).\\n\\n In 1979, EPA issued primary and secondary NAAQS for\\nozone with a limit of 0.12 parts per millions (ppm), and a one-\\nhour averaging time. See id. This “one-hour standard”\\nmeasured average ozone levels over one-hour periods, and\\n\\f 7\\nEPA would deem an area in compliance with this standard if it\\ndid not exceed the level for more than one day per calendar\\nyear. Id. EPA next revised the ozone NAAQS in 1997, having\\ndetermined that no revisions to the standards were necessary in\\n1993. National Ambient Air Quality Standards for Ozone, 58\\nFed. Reg. 13,008 (Mar. 9, 1993). The agency replaced the one-\\nhour, 0.12 ppm standard with a 0.08 ppm standard measured\\nover eight hours. National Ambient Air Quality Standards for\\nOzone, 62 Fed. Reg. 38,856 (July 18, 1997). EPA also altered\\nthe form of compliance to the annual fourth-highest daily\\nmaximum 8-hour concentration, averaged over three years. Id.\\n\\n In 2008, EPA lowered the primary and secondary\\nstandards to 0.075 ppm but kept the same eight-hour averaging\\ntime and form as in 1997. NRDC, 777 F.3d at 462-63; National\\nAmbient Air Quality Standards for Ozone, 73 Fed. Reg. 16,436\\n(Mar. 27, 2008). In developing the 2008 standards, EPA relied\\non scientific evidence showing that ozone causes health effects\\nat and above 0.08 ppm and examined two new clinical studies\\nthat found negative health effects from ozone at lower levels.\\nSee Mississippi, 744 F.3d at 1340. While CASAC reviewed\\nthis same evidence and recommended a primary level between\\n0.06 ppm and 0.07 ppm, EPA departed from this advice and\\nexplained that the scientific data regarding negative health\\neffects at 0.06 ppm was too limited and inconclusive to support\\na standard below 0.075 ppm. See id. at 1340-41. We upheld\\nEPA’s primary standard on this basis but found that EPA had\\nnot adequately explained its revision of the secondary standard.\\nId. at 1359-62. We noted that EPA had not properly\\ndetermined what level of air quality was requisite to protect the\\npublic welfare. Id. Accordingly, we remanded for further\\nexplanation and reconsideration of the secondary level.\\n\\n Following the promulgation of the 2008 ozone NAAQS,\\nEPA undertook another comprehensive review of the ozone\\n\\f 8\\nstandards and scientific data. EPA consolidated its review on\\nremand of the 2008 secondary standard with its ongoing review\\nfor the 2015 ozone NAAQS. After conducting public hearings\\nand receiving approximately 430,000 written comments on its\\nproposed revision of the primary and secondary ozone\\nNAAQS, EPA published its final 2015 ozone NAAQS on\\nOctober 26, 2015. National Ambient Air Quality Standards for\\nOzone (“2015 Rule”), 80 Fed. Reg. 65,292 (Oct. 26, 2015). For\\nboth the primary and secondary standards, EPA lowered the\\nlevel from 0.075 ppm to 0.07 ppm, while retaining the 2008\\nindicator (ozone), averaging time (8 hours), and form (three-\\nyear average of the fourth-highest daily maximum 8-hour\\nconcentration). See id. at 65,294, 65,301, 65,347, 65,349-50,\\n65,352.\\n\\n Based on scientific data and CASAC’s recommendation,\\nthe Administrator concluded that the 2008 primary ozone\\nstandard (0.075 ppm) was not at a level requisite to protect\\npublic health while allowing an adequate margin of safety. Id.\\nat 65,326, 65,344, 65,346. The Administrator placed\\nsignificant weight on new clinical studies linking short-term\\nozone exposure with respiratory effects, including lung\\ninflammation. See id. at 65,302-03, 65,341, 65,352, 65,359.\\nEPA conducted its own exposure study and found that a revised\\nprimary NAAQS standard of 0.07 ppm would “eliminate the\\noccurrence of two or more exposures of concern to [ozone]\\nconcentrations at and above [0.08 ppm]” and would “virtually\\neliminate the occurrence of two or more exposures of concern\\nto [ozone] concentrations at and above [0.07 ppm], even in the\\nworst-case urban study area and year evaluated.” Id. at 65,353.\\nThe Administrator additionally found that a level of 0.07 ppm\\nwould “protect the large majority of children in the urban study\\nareas (i.e., about 96% to more than 99% of children in\\nindividual urban study areas) from experiencing two or more\\nexposures of concern at or above the [0.06 ppm] benchmark.”\\n\\f 9\\nId. at 65,353, 65,360-64. EPA also evaluated hundreds of\\nepidemiologic studies that provided statistically relevant\\ninformation about a broader population of individuals who are\\nexposed to uncontrolled air pollution. See id. at 65,304,\\n65,364.\\n\\n Further, the Administrator considered CASAC’s advice on\\nthe new primary ozone standard. See id. at 65,346, 65,361. In\\nadvance of the 2015 Rule, CASAC stated that “there is clear\\nscientific support for the need to revise the standard” in place\\nsince 2008 and recommended setting the standard within a\\nrange of 0.06 ppm to 0.07 ppm, while leaving the form,\\naveraging time, and indicator unchanged. Id. at 65,322,\\n65,361.\\n\\n EPA also assessed the secondary standard and concluded\\nthat the 2008 secondary standard (0.075 ppm) was not requisite\\nto protect public welfare. Id. at 65,382, 65,389-90. Rather,\\nEPA set the secondary standard at 0.07 ppm and kept the\\nindicator, averaging time, and form the same. Id. at 65,369,\\n65,403, 65,409-10. EPA noted that more than four hundred\\nnew studies examining the interplay between ozone and public\\nwelfare had been developed since the promulgation of the 2008\\nozone NAAQS. Id. at 65,369. These studies strengthened and\\nexpanded the scientific understanding of ozone’s effects on\\nplants. The Administrator analyzed the causal relationship\\nbetween ozone exposure and vegetation effects, examining tree\\ngrowth impacts, crop yield loss, and visible leaf injury. Id. at\\n65,370; see id. at 65,294, 65,369-70. The Administrator gave\\n“primary consideration” to tree growth impacts, judging that it\\nwas more difficult to assess the welfare significance of crop\\nyield loss (given that crops are heavily managed) and visible\\nleaf injury (citing the “lack of established criteria or\\nobjectives”). Id. at 65,407.\\n\\f 10\\n In setting the secondary standard, EPA departed from\\nCASAC’s recommendations as to the form and standard of the\\nsecondary ozone NAAQS. CASAC recommended using a\\nsingle-year averaging approach, but EPA opted for a three-year\\naverage because “the public welfare significance of effects\\nassociated with multiple years of critical exposures are\\npotentially greater than those associated with a single year of\\nsuch exposure.” Id. at 65,404. The Administrator also rejected\\nCASAC’s recommendation to use the W126 exposure index, a\\ncumulative, seasonal ozone exposure metric, as the form of the\\nsecondary standard, rather than the same form as the primary\\nstandard (three-year average of the fourth-highest daily level),\\nfinding that the latter form was adequate to restrict cumulative\\nozone exposures that are detrimental to vegetation. Id. at\\n65,408.\\n\\n In addition to revising the ozone standards, the 2015\\nNAAQS also updated the regulations for the Prevention of\\nSignificant Deterioration (“PSD”) permitting program. See id.\\nat 65,431-34. Under the PSD program, no construction on a\\nmajor emitting facility may be commenced in an area that has\\nattained the air quality standards for any criteria pollutant\\nunless “the owner or operator of such facility demonstrates . . .\\nthat emissions from construction or operation of such facility\\nwill not cause, or contribute to, air pollution in excess of any\\n. . . [NAAQS].” 42 U.S.C. § 7475(a)(3)(B). The permitting\\nauthority (either a state or EPA) must grant or deny a completed\\npreconstruction permit application under the PSD program\\nwithin one year of filing. Id. § 7475(c).\\n\\n EPA claimed that sections 7475(a)(3)(B) and 7475(c) have\\nthe “potential for conflict,” which it was authorized under\\nChevron to resolve. EPA Br. 132; see also 80 Fed. Reg. at\\n65,433-34. EPA worried that, for a limited subset of pending\\npermit applications, complying with section 7475(a)(3)(B)’s\\n\\f 11\\ndemonstration requirement for the 2015 NAAQS “could hinder\\ncompliance with the requirement under section [7475](c) to\\nissue a permit within one year of the completeness\\ndetermination.” 80 Fed. Reg. at 65,434. In other words, EPA\\nwas unsure how to handle completed permit applications where\\nsources would have complied with the NAAQS in existence at\\nthe time the application was filed (the 2008 ozone standards),\\nbut where the NAAQS were revised before the permit was\\napproved. See id. at 65,433. EPA resolved this apparent\\nconflict by implementing a grandfathering provision for\\npending permit applications that satisfy one of two permitting\\ncriteria. See id. at 65,431-34. These two applicable categories\\ninclude: (1) permit applications that are deemed complete on\\nor before the signature date of the final rule revising the ozone\\nNAAQS; and (2) permit applications “for which the reviewing\\nauthority has first published a public notice of the draft permit\\nor preliminary determination before the effective date of the\\nrevised NAAQS.” Id. at 65,431, 65,433. If a permit\\napplication satisfies either requirement, the owner or operator\\nmust show compliance with the 2008 rather than the 2015\\nNAAQS before initiating construction.\\n\\n Petitioners from the oil and gas industry (“Industry\\nPetitioners”) and from several states (“State Petitioners”) claim\\nthat the primary and secondary NAAQS are too protective.\\nConversely, petitioners from public health and environmental\\ngroups (“Environmental Petitioners”) claim that these NAAQS\\nare not protective enough. Environmental Petitioners also\\ncontend that EPA’s decision to allow limited grandfathering of\\ncertain permit applications was inconsistent with the Clean Air\\nAct.\\n\\n This opinion proceeds in five parts. Part II considers the\\nchallenges to the primary standard, and Part III the challenges\\nto the secondary standard. Part IV reviews the cross-cutting\\n\\f 12\\nchallenges to EPA’s promulgation of the NAAQS, and Part V\\naddresses the challenge to the grandfathering provision.\\n\\nB. Jurisdiction and Standard of Review\\n\\n This court has jurisdiction to review EPA’s 2015 Rule\\npursuant to 42 U.S.C. § 7607(b)(1). Under section 7607,\\nagency action may be reversed if it is arbitrary, capricious, an\\nabuse of discretion, not in accordance with law, or in excess of\\nstatutory authority. Id. § 7607(d)(9)(A), (C); see Am.\\nPetroleum Inst. v. EPA, 684 F.3d 1342, 1347 (D.C. Cir. 2012).\\nTo withstand review, an agency must have examined all\\nrelevant facts and data, and articulated a rational explanation\\nfor its decision, including a reasonable connection between the\\nfacts and ultimate outcome. See Allied Local, 215 F.3d at 68.\\nWe cannot look at EPA’s decision as would a scientist, but\\ninstead must exercise our “narrowly defined duty of holding\\nagencies to certain minimal standards of rationality.”\\nMississippi, 744 F.3d at 1342 (quoting Nat’l Envtl. Dev.\\nAss’n’s Clean Air Project v. EPA, 686 F.3d 803, 810 (D.C. Cir.\\n2012)); see id. at 1348 (“We repeat: it is not our job to referee\\nbattles among experts; ours is only to evaluate the rationality\\nof EPA’s decision . . . .”).\\n\\n On questions of statutory interpretation, the court must\\nreview EPA’s actions in accordance with the standard set forth\\nin Chevron, U.S.A., Inc. v. Natural Resources Defense Council,\\nInc., 467 U.S. 837, 842-44 (1984). See Michigan v. EPA, 135\\nS. Ct. 2699, 2706-07 (2015); NRDC, 777 F.3d at 463. Chevron\\ndeference involves a two-step analysis. First, if a statute is\\nclear, the court must give effect to Congress’s unambiguous\\nlanguage and intent. Chevron, 467 U.S. at 842-43. Where a\\nstatute that an agency administers is ambiguous, the court must\\nturn to the second Chevron principle and give deference to the\\n\\f 13\\nagency’s reasonable interpretation of the statute. Id. at 843; see\\nalso S. Coast Air Quality Mgmt. Dist., 472 F.3d at 891.\\n\\n II. Primary Standard Challenges\\n\\n Industry and State Petitioners contend that EPA’s\\npromulgation of the 2015 Rule’s primary standard was\\narbitrary and capricious because, they say, EPA failed to\\nprovide a reasoned explanation for departing from the 2008\\nNAAQS. Environmental Petitioners argue that the primary\\nozone standard is too lenient because it occasionally permits\\nozone levels to exceed 0.07 ppm and will allegedly tolerate\\nadverse health effects. For the reasons below, we hold that\\nthese arguments lack merit.\\n\\nA. Industry and State Petitioners’ Challenge\\n\\n The Clean Air Act requires EPA to set primary NAAQS\\nthat are “requisite” to protect public health with an adequate\\nmargin of safety. 42 U.S.C. § 7409(b)(1). The term “requisite”\\nmeans “sufficient, but not more than necessary.” See Whitman\\nv. Am. Trucking Ass’ns, 531 U.S. 457, 473 (2001). While the\\ndetermination of what is “requisite” may require a contextual\\nassessment of acceptable risk, our precedent is clear that prior\\nNAAQS are not sacrosanct and are not granted presumptive\\nvalidity. See Mississippi, 744 F.3d at 1343. Prior NAAQS\\nneed not remain the governing standard until every aspect of\\nthose prior NAAQS is undermined. Id. To hold otherwise\\n“would bind EPA to potential deficiencies in past reviews\\nbecause discrepancies between past and current judgments as\\neasily reflect problems in the past as in the present.” Id. Thus,\\nwhen EPA reviews and revises the NAAQS, it does so against\\ncurrent policy considerations and existing scientific\\nknowledge. Id. Accordingly, “[t]he statutory framework\\nrequires us to ask only whether EPA’s proposed NAAQS is\\n\\f 14\\n‘requisite’; we need not ask why the prior NAAQS once was\\n‘requisite’ but is no longer up to the task.” Id.\\n\\n Given our decision in Mississippi, we must reject Industry\\nand State Petitioners’ argument that EPA departed from the\\n2008 NAAQS without adequate explanation. EPA has\\nadequately explained why on the record before it the revised\\nstandard is requisite to protect the public health. As the record\\nmakes clear, the Administrator considered the entire body of\\nscientific evidence available, including clinical studies,\\nepidemiologic evidence, human exposure and health risk\\nassessments, CASAC’s recommendations, and over 430,000\\npublic comments. See 80 Fed. Reg. at 65,293, 65,299, 65,301,\\n65,323, 65,326. And consistent with CASAC’s advice, the\\nAdministrator placed the most weight on clinical studies. See\\nid. at 65,302-03, 65,341, 65,352, 65,359. State Petitioners take\\nparticular issue with EPA’s reliance on the 2009 Schelegle\\nstudy, which, they argue, does not support EPA’s finding of\\nadverse effects. But the 2009 Schelegle study is merely one of\\nmany clinical studies that EPA relied on. State Petitioners\\nignore the fact that additional clinical studies and\\nepidemiologic evidence show substantial health effects at\\nozone levels as low as 0.06 ppm. See id. at 65,331, 65,334-35\\n& n.96, 65,344, 65,364.\\n\\n In addition to evaluating clinical studies, the Administrator\\nalso examined epidemiologic evidence and EPA’s exposure\\nand risk assessments. See 80 Fed. Reg. at 65,304, 65,314-17,\\n65,364. While the Administrator placed less weight on the\\nepidemiologic results than on clinical studies, see id. at 65,320,\\n65,324, 65,335, she nonetheless found that recent\\nepidemiologic studies suggested adverse health effects would\\noccur below the 2008 NAAQS standard. While Industry\\nPetitioners challenge the use of epidemiologic evidence given\\nthe uncertainties presented in these studies, even CASAC\\n\\f 15\\nconcluded that the epidemiologic evidence would have alone\\nbeen strong enough to justify revision of the 2008 NAAQS.\\n\\nB. Environmental Petitioners’ Challenges\\n\\n In contrast to Industry and State Petitioners,\\nEnvironmental Petitioners argue that the primary ozone\\nstandard is too lenient. They offer two reasons. First,\\nEnvironmental Petitioners contend that the form of the revised\\nstandard is not health-protective because it permits ozone\\nlevels to exceed 0.07 ppm on some days. Second, they argue\\nthat ozone exposures of 0.07 ppm will cause adverse health\\neffects, particularly in sensitive populations. For the reasons\\nbelow, we hold that these arguments lack merit.\\n\\n EPA reasonably explained its decision to retain\\n the form of the primary standard.\\n\\n Environmental Petitioners first take issue with the form of\\nthe revised standard—that the average, over three years, of the\\nannual fourth-highest maximum daily 8-hour average ozone\\nlevel must not exceed 0.07 ppm. By using this form, EPA\\n“allows ozone levels to exceed—multiple times in any year—\\nlevels that EPA itself agrees cause adverse health effects.”\\nEnvtl. Pet’rs Br. 20. Environmental Petitioners contend that\\nbecause areas with ozone levels above the 0.07 ppm threshold\\ncan still comply with EPA’s standards, the standard is unlawful\\nand arbitrary. We disagree.\\n\\n The Administrator’s decision to retain the same form used\\nin the 1997 and 2008 primary standards was in part based on\\nhow many people are estimated to experience unhealthy ozone\\nexposures. To this end, the Administrator utilized EPA’s\\nexposure assessment to gauge how often various population\\nsubgroups will be exposed to potentially health-impairing\\nozone concentrations while experiencing elevated breathing\\n\\f 16\\nrates. See 80 Fed. Reg. at 65,356. Rather than considering the\\nozone level of an area in isolation, the Administrator also\\n“consider[ed] activity patterns in the exposed population”\\nbecause adverse health responses to ozone exposure are\\ncritically dependent on breathing rates. Id. at 65,363; see also\\nid. (“Not considering activity patterns, and corresponding\\nventilation rates, can result in a standard that provides more\\nprotection than is requisite.”). This use of the exposure\\nassessment was rational.\\n\\n Further, the Administrator adequately interpreted the\\nexposure assessment. The Administrator correctly observed\\nthat the exposure assessment only chronicled how many people\\nwith an elevated breathing rate will be exposed to a specific\\nozone level, not how many will have an adverse effect. Given\\nthis limitation, the Administrator focused on estimates of two\\nor more exposures of concern to assess the potential for adverse\\neffects at and above an ozone concentration of 0.07 ppm. See\\nid. at 65,310, 65,325, 65,330, 65,343, 65,345-46, 65,354,\\n65,358, 65,361, 65,363. Based on the data, the Administrator\\nconcluded that a level of 0.07 ppm ozone would eliminate the\\noccasions on which school-age children experience two or\\nmore exposures of concern at ozone concentrations at and\\nabove 0.08 ppm and, even in the worst-case years and\\nlocations, would “virtually eliminate” the occasions on which\\nsuch children experience two or more exposures of concern at\\n0.07 ppm. Id. at 65,353. In particular, the Administrator noted\\nthat a primary level of 0.07 ppm would protect approximately\\n96% to 99% of children in the urban study areas from\\nexperiencing two or more exposures of concern above the 0.06\\nppm benchmark. Id. at 65,353; see also id. at 65,360-64. That\\nEnvironmental Petitioners cite data suggesting that 18,000\\nchildren would experience multiple exposures of concern at or\\nabove 0.07 ppm during the worst year and location is not\\ndeterminative because the primary standard for a non-threshold\\n\\f 17\\npollutant like ozone is not required to produce zero risk, and\\n“[t]he task of determining what standard is ‘requisite’ to protect\\nthe qualitative value of public health or what margin of safety\\nis ‘adequate’ to protect sensitive subpopulations necessarily\\nrequires the exercise of policy judgment.” Mississippi, 744\\nF.3d at 1351, 1358. EPA has reasonably exercised its\\njudgment.\\n\\n Given that the Administrator appropriately examined and\\nconsidered EPA’s exposure assessment, her decision to retain\\nthe form of the primary standard was also rational. The\\nAdministrator determined that, when combined with an ozone\\nlevel of 0.07 ppm, the form of the standard (three-year average\\nof the fourth-highest daily level) was requisite. In particular,\\nthe Administrator found that most areas that meet the revised\\nstandard will have an 8-hour ozone concentration below 0.07\\nppm. See 80 Fed. Reg. at 65,363. Environmental Petitioners,\\nhowever, criticize this form for allegedly failing to account for\\nindividual ozone days above 0.07 ppm. But Environmental\\nPetitioners elide a crucial detail of EPA’s methodology: the\\nexposure assessment study that “informed the Administrator’s\\nconsideration of the degree of public health protection provided\\nby various standard levels” considered the air quality at\\n“various standards with the current 8-hour averaging time and\\nfourth-high, 3-year average form.” Id. at 65,351-52 (emphasis\\nadded). As such, the Administrator reasonably accounted for\\ndays when ozone levels may exceed 0.07 ppm. Taken together,\\nwe conclude that the Administrator’s decision to retain the\\nform of the standard was appropriate and neither arbitrary nor\\ncapricious.\\n\\f 18\\n EPA reasonably set the primary standard at 0.07\\n ppm.\\n\\n Environmental Petitioners next argue that the\\nAdministrator’s decision to lower the primary standard from\\nfrom 0.075 to 0.07 ppm is an unexplained departure from\\nCASAC’s recommendation and from EPA’s prior position\\nregarding the adversity of certain lung function decrements.\\nWe reject both contentions.\\n\\n Environmental Petitioners argue that EPA failed to fulfill\\nits duty under the Act to provide “an explanation of the\\nreasons” for departing from CASAC’s scientific\\nrecommendations. 42 U.S.C. § 7607(d)(3); see also id.\\n§ 7607(d)(6). They contend CASAC told EPA that “[a]t [0.07\\nppm] there is substantial scientific certainty of a variety of\\nadverse effects, including decrease in lung function, increase\\nin respiratory symptoms, and increase in airway\\ninflammation,” Envtl. Pet’rs Br. 31 (quoting J.A. 531), and\\nEPA failed to rationally dispute or refute this finding.\\n\\n EPA adequately considered CASAC’s advice when setting\\nthe primary NAAQS to 0.07 ppm. While CASAC advised EPA\\nto “set the level of the standard lower than [0.07 ppm],” J.A.\\n531, that recommendation constituted policy—not scientific—\\nadvice. This distinction is crucial because we defer to EPA’s\\njudgment on issues of policy but require EPA to “fully explain\\nits reasons for any departure from” “CASAC’s expert scientific\\nrecommendations.” Mississippi, 744 F.3d at 1358, 1354-55\\n(emphasis added). CASAC’s letter to EPA makes clear that\\n“based on the scientific evidence” it “recommend[ed] a range\\nof levels for a revised primary ozone standard from [0.07 ppm]\\nto [0.06 ppm].” J.A. 531. CASAC then “acknowledge[d] that\\nthe choice of a level within the range recommended based on\\nscientific evidence is a policy judgment.” Id.\\n\\f 19\\n In an effort to influence EPA’s policy judgment, CASAC\\nnoted that given the “substantial scientific certainty of a variety\\nof adverse effects” at 0.07 ppm, that level may not be adequate\\nto protect public health. Id. So, CASAC’s “policy advice\\n[was] to set the level of the standard lower than [0.07 ppm]\\nwithin a range down to [0.06 ppm], taking into account [EPA’s]\\njudgment regarding the desired margin of safety to protect\\npublic health.” Id. EPA did not take CASAC’s policy advice\\nbecause the Administrator found that 0.07 ppm would still\\n“provide substantial protection against the broader range of\\n[ozone] exposure concentrations that have been shown in\\ncontrolled human exposure studies to result in respiratory\\neffects, including exposure concentrations below [0.07 ppm].”\\n80 Fed. Reg. at 65,363. EPA nonetheless chose a level for the\\nprimary standard that was within CASAC’s scientifically\\nrecommended range. In so doing, EPA did not abrogate its\\nduty under the Clean Air Act to consider CASAC’s scientific\\nrecommendations but instead made a valid policy decision.\\n\\n Environmental Petitioners also argue that EPA should\\nhave set the primary standard lower than 0.07 ppm, given the\\ncontrolled human exposure studies that had been published\\nsince the agency’s 2008 NAAQS review. Specifically,\\nEnvironmental Petitioners highlight two new studies\\nevaluating exposures to 0.06-0.063 ppm ozone that concluded\\nthat exposures at 0.06 ppm caused lung function decrements of\\n10% or more. Taken with EPA’s statement in the 2008 ozone\\nNAAQS rule that a lung function decrement of 10% or more\\n“should be considered adverse for sensitive populations,” 73\\nFed. Reg. at 16,454-55, Environmental Petitioners contend that\\nEPA acted arbitrarily by giving insufficient weight to those\\nnew studies. We are unpersuaded by this argument, which fails\\nto appreciate the dynamic nature of adversity determinations at\\neach NAAQS review.\\n\\f 20\\n In determining the appropriate level for the 2015 primary\\nstandard, the Administrator was not bound by adversity\\njudgments in the 2008 NAAQS review. Indeed, “as the\\ncontours and texture of scientific knowledge change, the\\nepistemological posture of EPA’s NAAQS review necessarily\\nchanges as well.” Mississippi, 744 F.3d at 1344. Thus, we\\nconsider only whether EPA’s 2015 Rule offers a rational\\nexplanation of why EPA chose a new adversity definition and\\nwhether the Administrator reasonably evaluated the evidence\\nof adversity. We hold that EPA’s actions were reasonable and\\nreasonably explained. Rather than applying a rigid test for\\ndetermining what level of decrement is adverse, the\\nAdministrator took a more comprehensive approach provided\\nby American Thoracic Society (ATS) guidelines. ATS\\nguidelines provide that “reversible loss of lung function in\\ncombination with the presence of [respiratory] symptoms\\nshould be considered adverse.” 80 Fed. Reg. at 65,357\\n(internal quotation marks omitted). EPA reasonably explained\\nthat it chose to adopt the ATS definition of adversity because\\n“the available evidence does not provide information on the\\nextent to which a short-term, transient decrease in lung function\\nin a population,” without more, could “change the risk profile\\nof the population,” and that CASAC was “conditional” about\\nwhether “the lung function decrements observed in some\\npeople at [0.06 ppm] . . . are adverse.” Id. at 65,358. The\\nclinical studies that Environmental Petitioners contend EPA\\ndismissed concluded that lung function decrements (such as\\n10%) occurred in some individuals at lower ozone\\nconcentrations, including 0.06 and 0.063 ppm, but not in\\n“combination [with] statistically significant increases in\\nrespiratory symptoms.” See id. at 65,357. Under ATS\\nguidance, the Administrator declined to designate such loss of\\nlung function as adverse. This decision was rational, and\\nEnvironmental Petitioners cannot show that the evidence\\nrequired EPA to decide differently.\\n\\f 21\\n III. Secondary Standard Challenges\\n\\n Next, we turn to the challenges to EPA’s secondary\\nstandard. The Environmental Petitioners contend that: (1) in\\nconsidering tree growth loss, EPA acted arbitrarily in setting\\nthe target level of air quality and therefore fell short of the\\nstatutory requirement to set a standard “requisite” to protect\\nagainst such harm; (2) EPA arbitrarily used a three-year\\naverage rather than a single-year, cumulative measurement of\\nozone exposure as a benchmark to gauge the protectiveness of\\nits standard, and arbitrarily declined to adopt the single-year\\ncumulative exposure index as the form and averaging time for\\nthe secondary standard; and (3) EPA unreasonably failed to\\nidentify a level of air quality requisite to protect the public\\nwelfare against adverse effects from visible leaf injury.\\nIndustry Petitioners argue only that EPA failed to justify its\\ndecision to lower the secondary standard from its 2008 level.\\nWe hold that EPA has not explained its decision to set a target\\nlevel of protection against tree growth loss based on a three-\\nyear average of cumulative, single-year ozone exposures, nor\\nhas it justified its decision not to specify any level of air quality\\nrequisite to protect against visible leaf injury. We reserve\\njudgment on whether EPA reasonably declined to adopt the\\ncumulative exposure index as the form and averaging time, and\\ndeny the remainder of the challenges.\\n\\n We begin by reviewing EPA’s secondary standard-setting\\nprocess. EPA concentrated its review on the association\\nbetween ozone exposure and “vegetation effects,” focusing on\\ntree growth loss (also referred to as “relative biomass loss”),\\ncrop yield loss, and visible leaf (or “foliar”) injury. 80 Fed.\\nReg. at 65,370-71. CASAC agreed that those three effects\\nwere “appropriate surrogates [for] a wide range of damage that\\nis adverse to public welfare” because ozone damage to trees,\\nleaves, and crops can directly affect numerous resources and\\n\\f 22\\necosystem services that are important to the public and\\nindirectly affect a wide array of ecosystem components and\\nfunctions. J.A. 532-33.\\n\\n In performing its analysis, EPA used an “exposure metric”\\ncalled the “W126 index” to gauge how differing levels of ozone\\nexposure correspond to effects on tree growth, crop yields, and\\nvisible leaf injury. 80 Fed. Reg. at 65,373. The W126 index,\\nwhich EPA deemed to be the “most biologically relevant\\nmetric[] for consideration of [ozone] exposures eliciting\\nvegetation-related effects,” measures the cumulative amount of\\nozone to which a plant is exposed over a single three-month\\ngrowing season. Id. at 65,373-74. W126 levels are expressed\\nas parts-per-million hours (ppm-hrs). Id.\\n\\n CASAC recommended that EPA use the single-year W126\\nexposure index as the form and averaging time for the\\nsecondary standard, J.A. 518, meaning that compliance with\\nthe standard would be measured based on a single growing\\nseason’s worth of cumulative ozone exposure. CASAC\\nadvised EPA to set the level for the secondary standard within\\na range of 7 ppm-hrs and 15 ppm-hrs. Based on the data\\nCASAC used, that corresponded to median annual tree growth\\nloss between 2% and 5.2%. J.A. 518, 534-36, 631. CASAC\\ncautioned that “at 17 ppm-hrs, the median tree species has 6%\\nrelative biomass loss,” which would be “unacceptably high.”\\nJ.A. 518. Regarding the other two surrogates, CASAC\\ncounseled that, to protect against loss of crop yield, “a level of\\n15 ppm-hrs for the highest 3-month sum in a single year is\\nrequisite,” and to reduce foliar injury, a “level below 10 ppm-\\nhrs is required.” Id. All of those levels were based on single-\\nyear measuring periods, so CASAC advised that if EPA were\\nto base its secondary standard on a three-year average of the\\nrelevant measurements, it should lower the level of the standard\\n\\f 23\\n“to protect against single unusually damaging years that will be\\nobscured in the average.” J.A. 536.\\n\\n EPA agreed with CASAC’s recommendation that 6% tree\\ngrowth loss would be “unacceptably high.” 80 Fed. Reg. at\\n65,406. However, as EPA explained, id. at 65,384, 65,392\\nn.197, CASAC’s advice on that was based in part on a study of\\ncottonwood trees, which are uniquely ozone-sensitive, and\\nCASAC itself advised EPA that the “cottonwood data . . .\\nreceive[d] too much emphasis” in EPA’s analysis, J.A. 533.\\nAfter excluding the cottonwood data, EPA concluded that\\nozone exposure of 19 ppm-hrs was associated with 6% tree\\ngrowth loss, and that 17 ppm-hrs of ozone exposure brought it\\ndown to 5.3% loss. 80 Fed. Reg. at 65,396, 65,407. EPA\\naccordingly chose to focus on a standard “somewhat below”\\nthe 19 ppm-hrs level associated with 6% median tree growth\\nloss. Id. at 65,406-07.\\n\\n EPA then departed from CASAC’s advice in several ways.\\nFirst, it chose not to use the W126 cumulative seasonal\\nexposure index as the form and averaging time of the standard,\\nopting instead to retain the averaging time (8 hours) and form\\nit had used in the 2008 rule (the three-year average of the\\nfourth-highest daily level). EPA purported to use as a\\n“benchmark,” or target level of protectiveness, the high end of\\nthe range of W126 levels CASAC had recommended. Recall\\nthat CASAC had recommended that the maximum seasonal\\nozone exposure not exceed (the cottonwood-adjusted) 17 ppm-\\nhrs in any single growing season. In deciding what exposure\\nlevel correlated to CASAC’s 17 ppm-hrs, however, EPA used\\na three-year average of anticipated seasonal exposure levels.\\nThus, EPA chose to set the standard at a level that it projected\\n(based on a statistical analysis of past ozone exposure data)\\nwould “in nearly all instances” going forward restrict the\\naverage of three growing seasons’ ozone exposures to the\\n\\f 24\\nequivalent of 17 ppm-hrs. Id. at 65,407. But by using that\\nmethod to set the level, EPA arrived at a standard that\\nstatistically tolerates cumulative ozone exposures in a single\\ngrowing season that are higher than CASAC’s maximum\\nrecommended level. EPA also did not specify any level of\\nozone to protect against visible leaf injury or crop loss. The\\nagency reasoned that data for those public-welfare harms were\\ntoo uncertain to permit it to discern a level that would provide\\nthe requisite protections, and that the standard it set to protect\\nagainst adverse effects from tree growth loss would at least\\nprovide “some increased protection” against visible leaf injury\\nand crop damage. Id. at 65,407-08.\\n\\n In sum, EPA ultimately chose to set the level of the\\nsecondary standard at 0.07 ppm while retaining the form and\\naveraging time it had previously used. Id. at 65,410.\\n\\nA. Environmental Petitioners’ Challenges\\n\\n The essence of the Environmental Petitioners’ petition is\\nthat EPA did not adequately explain its deviations from\\nCASAC’s advice. “In order to enable judicial review and to\\nsatisfy its statutory obligation to explain its reasons for\\ndeparting from CASAC, EPA must be precise in describing the\\nbasis for its disagreement with CASAC.” Mississippi, 744 F.3d\\nat 1355. Where EPA diverges from CASAC’s scientific\\nadvice, there must be “substantial evidence in the record when\\nconsidered as a whole which supports the Administrator’s\\ndeterminations.” Id. (quoting Lead Indus. Ass’n v. EPA, 647\\nF.2d 1130, 1146 (D.C. Cir. 1980)).\\n\\f 25\\n EPA reasonably used 17 ppm-hrs as the\\n benchmark level to protect against adverse effects\\n associated with tree growth loss.\\n\\n The Environmental Petitioners argue that, by establishing\\n17 ppm-hrs as the target level of protection against adverse\\nwelfare effects associated with tree growth loss, EPA\\nimpermissibly departed upward from CASAC’s advice in two\\nways—first, by excluding the data from the cottonwood tree\\nstudy from its tree growth loss analysis and, second, by\\nfocusing on limiting tree growth loss to under 6%, rather than\\nthe 2% target they say CASAC’s analysis required. Setting\\naside for the moment EPA’s decision to average the benchmark\\nover three years, we conclude that EPA adequately explained\\nits decision to exclude the cottonwood data and acted\\nconsistently with CASAC’s advice in choosing to limit tree\\ngrowth loss to under 6%.\\n\\n Cottonwood Data. EPA reviewed seedling studies of\\ntwelve different tree species to determine the median percentile\\nof tree growth loss at varying ozone levels. See 80 Fed. Reg.\\nat 65,372. One of those studies—of cottonwood seedlings—\\nsignificantly lowered the ozone level at which the twelve tree\\nspecies experience 6% median growth loss. See id. at 65,384,\\n65,391-92. CASAC had itself relied on that study, see J.A.\\n537, but advised EPA that the cottonwood data had received\\n“too much emphasis” because the study “did not control for\\nozone and climatic conditions,” and the results “show[ed]\\nextreme sensitivity to ozone compared to other studies,” J.A.\\n533. EPA accordingly excluded the cottonwood data from its\\ntree growth analysis, which increased the ozone level\\nassociated with 6% median growth tree loss from 17 ppm-hrs\\nup to 19 ppm-hrs. 80 Fed. Reg. at 65,384, 65,391-92.\\n\\f 26\\n The Environmental Petitioners object that EPA arbitrarily\\ndisregarded CASAC’s advice because, they say, CASAC never\\nexpressly advised EPA to disregard the cottonwood data\\naltogether, and CASAC relied on it in setting its recommended\\nexposure range. But we can only discern an unreasonable\\ndeviation where CASAC itself has been “precise about the\\nbasis for its recommendations.” Mississippi, 744 F.3d at 1358.\\nIt was far from precise on the disputed point. It warned EPA\\nto treat the cottonwood data with caution, but recommended a\\nrange of ozone levels in reliance on the median tree growth loss\\nestimates (outlined in Table 6-1 of EPA’s staff’s second draft\\nPolicy Assessment), which weigh the cottonwood data equally\\nwith other seedling studies. See J.A. 537 (citing EPA, Policy\\nAssessment for the Review of the Ozone NAAQS, Second\\nExternal Review Draft 6-19, tbl. 6-1 (2014), J.A. 631). The\\nEnvironmental Petitioners suggest that CASAC was only\\nwarning EPA to take care in describing the cottonwood data in\\nthe narrative portion of its Policy Assessment, not suggesting\\nanything about how the data was actually used, but that makes\\nlittle sense in the context of an exchange about quantitative\\ngrowth loss estimates. In light of CASAC’s mixed messages,\\nwe cannot say that EPA’s decision to exclude the cottonwood\\ndata was arbitrary.\\n\\n The 6% Target. Emphasizing CASAC’s admonition that\\n6% tree growth loss would be “unacceptably high,” EPA\\ndecided “to focus on a standard that would generally limit\\ncumulative exposures to those for which the median [tree\\ngrowth loss] estimate would be somewhat lower than 6%.” 80\\nFed. Reg. at 65,406-07. The Environmental Petitioners\\ncontend that CASAC “plainly specified” that 2% median\\ngrowth loss was the requisite level to avoid adverse welfare\\neffects related to tree growth loss, Envtl. Pet’rs Br. 42-44, but\\nthey misconstrue CASAC’s advice.\\n\\f 27\\n CASAC did not identify 2% growth loss as the only\\nsufficiently protective level, but as “an appropriate\\nscientifically based value to consider as a benchmark of\\nadverse impact,” elaborating that it would be “appropriate to\\nidentify a range of levels of alternative W126-based standards\\nthat includes levels that aim for not greater than 2% [growth\\nloss] for the median tree species.” J.A. 537 (emphases added).\\nCASAC in fact ultimately recommended that EPA set the\\nstandard between 7 ppm-hrs and 15 ppm-hrs, J.A. 518, which,\\nbased on the seedling data upon which CASAC relied\\n(including the cottonwood data), corresponds to median annual\\ntree growth loss between 2% and 5.2%, see J.A. 631. In other\\nwords, CASAC recommended that EPA consider limiting tree\\ngrowth loss to 2% as the lower end of a range of permissible\\ntarget levels. EPA followed CASAC’s advice by considering\\na 2% growth loss target as part of a range of growth loss targets\\nand determined that the studies underlying CASAC’s low-end\\nrecommendation were insufficiently reliable to base the\\nsecondary standard on that low level of tree growth loss. See\\n80 Fed. Reg. at 65,394.\\n\\n The Environmental Petitioners passingly disparage EPA’s\\nassessment of the studies associated with CASAC’s 2% growth\\nloss target, but that argument is equally unavailing. EPA\\nreasonably explained why it thought those studies were\\nunreliable, and its “evaluation of ‘scientific data within its\\ntechnical expertise’” is entitled significant deference. See\\nMiss. Comm’n on Envtl. Quality v. EPA, 790 F.3d 138, 150,\\n155-56 (D.C. Cir. 2015) (quoting City of Waukesha v. EPA,\\n320 F.3d 228, 247 (D.C. Cir. 2003)). For instance, in support\\nof its advice that EPA consider limiting median annual tree\\ngrowth loss to 2%, CASAC highlighted a study purporting to\\nshow that stands of aspen that experienced ozone-induced\\nbiomass loss of 2% annually had cumulative biomass loss of\\n21% over seven years, suggesting that even relatively minor\\n\\f 28\\nannual growth loss can lead to significant loss over longer\\nperiods of time. J.A. 537. But CASAC had misread the study.\\nAs EPA explained, the aspen in that study in fact experienced\\nmore than 20% biomass loss annually, suggesting that their\\ncumulative biomass loss was not much worse than their loss in\\nany single year. 80 Fed. Reg. at 65,394.\\n\\n EPA also clarified that the only other report CASAC relied\\non for its 2% target recommendation provided no explicit\\nrationale for selecting that level and did not identify any new\\ndata in support of its recommendations. See id. at 65,394-95 &\\nn.200. While it is true that EPA has cited that same report\\nfavorably in the past, see, e.g., National Ambient Air Quality\\nStandards for Ozone, 75 Fed. Reg. 2,938, 3,025 (Jan. 19, 2010),\\nthat alone does not make EPA’s current assessment arbitrary.\\nEPA must simply provide a “reasoned explanation” for its\\ndeparture from its past position—which it has done. See\\nEncino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125-26\\n(2016); see also Mississippi, 744 F.3d at 1351.\\n\\n EPA did not adequately explain its decision to use\\n a three-year average of the W126 index as the\\n benchmark.\\n\\n The Environmental Petitioners raise two challenges\\npertaining to the W126 index, both as a benchmark and as a\\nform and averaging time. First, they argue that EPA\\nimpermissibly departed from CASAC’s advice by setting the\\nsecondary standard level using a three-year average W126\\nbenchmark without lowering the level to protect against single-\\nyear exposures associated with median annual tree growth loss\\nof 6%, which CASAC had advised was “unacceptably high.”\\nJ.A. 518. They also contend that EPA arbitrarily disregarded\\nCASAC’s advice to adopt the W126 index as the form and\\n\\f 29\\naveraging time for the secondary standard. We remand to EPA\\non the first issue and do not reach the second.\\n\\n The Three-Year Average Benchmark. CASAC advised\\nEPA that basing the secondary standard on a “single-year\\nperiod” would provide “more protection for annual crops and\\nfor the anticipated cumulative effects on perennial species”\\nthan a three-year average. J.A. 518. It explained that EPA’s\\nproposal to use a three-year averaging period was “not\\nsupported by the available data,” J.A. 536, and that if EPA\\nchose to “base the secondary standard on a three-year\\naveraging period,” then “the level of the standard should be\\nrevised downward such that the level for the highest three-\\nmonth summation in any given year of the three-year period\\nwould not exceed [its] scientifically recommended” range of\\nsingle-year, W126 exposure levels, J.A. 518. This was\\nnecessary, CASAC explained, to “protect against single\\nunusually damaging years that will be obscured in the\\naverage.” J.A. 536.\\n\\n EPA argues it gave effect to CASAC’s recommendation\\nby using a three-year average benchmark of 17 ppm-hrs,\\nwhich, after adjusting for the cottonwood data, was “somewhat\\nbelow” the 19 ppm-hrs associated with 6% annual growth loss\\nthat CASAC had advised was “unacceptably high.” 80 Fed.\\nReg. at 65,406-07 (quoting J.A. 518). But CASAC had advised\\na maximum level associated with 5.2% annual biomass loss,\\nsee J.A. 631, and it expressly cautioned that 6% median growth\\nloss in a single year was unacceptable, see J.A. 518. EPA’s use\\nof a benchmark that averages out to less than 6% biomass loss\\nover three years does not accord with CASAC’s advice.\\nIndeed, as commenters informed EPA during the rulemaking,\\nsee J.A. 1836-40, EPA’s own air quality data suggests that\\nmany large national parks and wilderness areas that have met\\nEPA’s chosen three-year average 17 ppm-hrs benchmark—\\n\\f 30\\nareas that Congress considers significant to the public welfare,\\nsee 42 U.S.C. §§ 7470(2), 7472(a)—have meanwhile recorded\\nsingle-year W126 values at and above 19 ppm-hrs, which is\\nassociated with “unacceptably high” annual biomass loss of\\n6% and higher. See J.A. 1061-64; 80 Fed. Reg. at 65,391. EPA\\ncritiques that data as marred by outdated handling procedures,\\nbut the agency acknowledged that other data derived through\\nupdated procedures produced results “similar to” those\\nshowing harmful exposure spikes. J.A. 1213. Critically, EPA\\npoints us to no data or analysis (based on new or old\\nprocedures) suggesting that the chosen benchmark prevents\\nsingle seasonal exposures of 19 ppm-hrs or higher. In short,\\nEPA has not demonstrated how its chosen benchmark protects\\nagainst “unusually damaging years that will be obscured in the\\naverage.” J.A. 536.\\n\\n EPA alternatively defends its decision to use the three-\\nyear-average benchmark as providing a focus on public welfare\\neffects of “potentially greater” significance than effects\\n“associated with a single year” of exposure. 80 Fed. Reg. at\\n65,404. This position, however, is inconsistent with EPA’s\\nother actions. In establishing the secondary standard, for\\nexample, EPA heavily relied on data showing annual 6%\\nmedian tree growth loss at 19 ppm-hrs, see id. at 65,406, and\\nacknowledged the potential for a single season of high ozone\\nexposure to “alter biomass allocation and plant reproduction in\\nseasons subsequent to [that season’s] exposure,” thereby\\nleading to “a negative impact on species regeneration in\\nsubsequent years,” id. at 65,371-72; see also J.A. 740-41. EPA\\nadditionally recognized that “ozone effects in plants are\\ncumulative,” id. at 65,373 (quoting EPA, Integrated Science\\nAssessment 2-44 (2013)), meaning that the adverse vegetative\\neffects from single, high-ozone years are not offset by\\nsubsequent low-ozone years. EPA has identified no contrary\\n\\f 31\\nevidence in the record demonstrating why these single-year\\neffects matter less than a three-year average.\\n\\n We therefore remand this issue for EPA to either lower the\\nstandard to protect against unusually damaging cumulative\\nseasonal exposures that will be obscured in its three-year\\naverage, or explain its conclusion that the unadjusted average\\nis an appropriate benchmark notwithstanding CASAC’s\\ncontrary advice. Alternatively, EPA could adopt the single-\\nyear W126 exposure index as the form and averaging time,\\nwhich would presumably moot any problems with the way it\\ntranslated that index to use as a benchmark.\\n\\n The Form and Averaging Time. CASAC also\\nrecommended that EPA use the single-year W126 index as the\\nform and averaging time for the secondary standard. J.A. 518.\\nEPA instead chose to retain the existing form and averaging\\ntime—the three-year average of the fourth-highest daily\\nmaximum eight-hour concentration. Adopting the W126 index\\nas the form and averaging time was unnecessary, EPA\\nexplained, because the ozone exposure levels associated with\\nthe existing form and averaging time and a three-year average\\nof the W126 index are “highly correlated,” especially at lower\\nlevels, and “future control programs designed to help meet a\\nprimary [ozone] standard based on the” current form and\\naveraging time should “provide similar improvements in terms\\nof the 3-year average of the annual W126 metric.” J.A. 1253;\\nsee also 80 Fed Reg. at 65,400-01, 408-09.\\n\\n The Environmental Petitioners argue that EPA did not\\njustify its decision not to adopt the W126 index as the form and\\naveraging time. We lack any basis to assess the reasonableness\\nof EPA’s actions, however, because a critical piece of the\\npuzzle is missing. To review: EPA chose not to use the W126\\nindex as the form and averaging time because it found that\\n\\f 32\\nozone exposure levels associated with the existing form and\\naveraging time are “highly correlated” to a three-year average\\nof the W126 index. But, as discussed, EPA never explained\\nwhy it is reasonable to focus on a three-year average of the\\nW126 index in the first place. Therefore, we cannot assess the\\nrelevance of the claim that the two are “highly correlated.”\\nEPA’s reconsideration on remand of the three-year averaging\\nissue should supply us with the information necessary to\\nresolve this question, or, if EPA chooses to follow CASAC’s\\nadvice to lower the standard to control for unusually high single\\nyears, potentially moot the Environmental Petitioners’ concern\\nthat the current form tolerates even three-year average W126\\nlevels higher than 17 ppm-hrs during periods when a 0.07 ppm,\\n8-hour level is met. Accordingly, we decline to reach this\\nquestion.\\n\\n EPA arbitrarily failed to identify a level of air\\n quality requisite to protect against adverse effects\\n from visible leaf injury.\\n\\n The Environmental Petitioners also fault EPA for deciding\\nnot to specify a level of air quality to protect against adverse\\nwelfare effects from ozone-induced visible leaf injury.\\nBecause EPA failed to offer a reasoned explanation, we remand\\nfor it to reconsider.\\n\\n EPA has found that the “strongest evidence for effects\\nfrom [ozone] exposure on vegetation is from controlled\\nexposure studies, which ‘have clearly shown that exposure to\\n[ozone] is causally linked to visible foliar injury.’” 80 Fed.\\nReg. at 65,370 (quoting EPA, Integrated Science Assessment\\n1-15). Relying on that evidence, CASAC advised EPA of its\\n“scientific judgment” that “a level of 10 ppm-hrs is required to\\nreduce foliar injury.” J.A. 538. EPA nonetheless concluded\\nthat there were too many “uncertainties and complexities” in\\n\\f 33\\nthe evidence to specify a level of air quality to protect against\\nleaf injury. See 80 Fed. Reg. at 65,407-08. But the mere\\ninvocation of “substantial uncertainty” is not a justification for\\nthe agency’s failure to fulfill its statutory mandate. Motor\\nVehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins.\\nCo., 463 U.S. 29, 52 (1983) (internal quotation marks omitted).\\nEPA “must explain the evidence which is available, and must\\noffer a ‘rational connection between the facts found and the\\nchoice made.’” Id. (quoting Burlington Truck Lines, Inc. v.\\nUnited States, 371 U.S. 156, 168 (1962)). Where CASAC has\\n“reached a scientific conclusion that adverse [welfare] effects\\n[are] likely to occur,” EPA must “explain why the evidence on\\nwhich CASAC relied cannot support the degree of confidence\\nCASAC placed in it.” Mississippi, 744 F.3d at 1357; see also\\nAm. Farm Bureau Fed’n v. EPA, 559 F.3d 512, 529-30 (D.C.\\nCir. 2009). We defer to EPA’s judgment that the available\\nevidence is too uncertain only when the agency reasonably\\nexplains its decision, see State Farm, 463 U.S. at 51-53, and\\nEPA has failed to carry that burden here.\\n\\n None of the identified uncertainties justifies EPA’s\\ndecision not to set a standard to protect against visible leaf\\ninjury. For instance, EPA argues that it lacked criteria for\\nassessing the welfare effects of visible leaf injury, but that does\\nnot square with its conclusion in the Integrated Science\\nAssessment (compiling all available scientific criteria) that\\nvisible leaf injury “has been well characterized and\\ndocumented over several decades.” J.A. 985-86. EPA had at\\nits fingertips a wealth of new data regarding visible leaf injury,\\nincluding new controlled exposure studies, multiyear field\\nsurveys, and biomonitoring data of ozone-induced visible leaf\\ninjury on public lands. J.A. 749. Using that data, EPA\\nanalyzed how visible leaf injury affects “ecosystem services”\\nlike aesthetic value and recreation on “Class I” lands—lands\\nthat “have particular public welfare significance.” J.A. 749-50,\\n\\f 34\\n1036-44; see also 42 U.S.C. §§ 7470(2), 7472(a). CASAC\\npointed out that those same ecosystem services are relevant in\\nidentifying “damage that is adverse to public welfare.” J.A.\\n532-33.\\n\\n CASAC concluded, based on the same data available to\\nEPA, that “a level of 10 ppm-hrs is required to reduce foliar\\ninjury.” J.A. 538. EPA counters that “decreases in leaf injury\\n[also occur] with decreasing ozone exposures across a range of\\nvalues well above 10 ppm-hrs,” and that CASAC\\nrecommended a range of ozone levels that included values\\nabove 10 ppm-hrs. See EPA Br. 89; 80 Fed. Reg. at 65,395-\\n96. Be that as it may, EPA has not explained why, unlike\\nCASAC, it was unable to choose a level at all. Indeed,\\n“[b]ecause the EPA failed to identify any target level, we need\\nnot decide whether it was reasonable for the agency to reject\\nthe target recommended by . . . CASAC.” Am. Farm Bureau,\\n559 F.3d at 530.\\n\\n EPA objects that it only found that visible leaf injury has\\nthe “potential to be adverse to the public welfare,” and that the\\nClean Air Act does not require EPA “to identify a precise,\\nquantified level of public welfare protection for every\\npotentially adverse public welfare impact” it considers when\\nrevising the secondary standard. EPA Br. 87-88 (first quoting\\n80 Fed. Reg. at 65,388). But EPA has failed to justify its\\ninaction where CASAC unequivocally found that “damage to\\nresource use from foliar injury” was an “adverse welfare\\neffect.” J.A. 518. In light of CASAC’s “scientific conclusion\\nthat adverse [welfare] effects were likely to occur” from visible\\nleaf injury, EPA’s inaction in the face of such effects is\\n“unacceptable.” Mississippi, 744 F.3d at 1357. Nor was it\\nsufficient for EPA simply to conclude that the standard set to\\nprotect against tree growth loss sufficed because it provided\\n“additional” protection against leaf injury. 80 Fed. Reg. at\\n\\f 35\\n65,407-08. Nothing in the record suggests that tree growth loss\\nis a surrogate for leaf injury. EPA’s assurance that the standard\\nset to protect against tree growth loss will provide\\n“incidental[]” protection against visible leaf injury “cannot\\nsave its decision.” Am. Farm Bureau, 559 F.3d at 529-30.\\n\\n Other purported uncertainties do not support EPA’s\\ninaction. For instance, EPA asserts that the Administrator\\n“lacked evidence that would allow her to measure the\\nrelationship between leaf injury and other vegetation effects\\nthat she might find adverse,” EPA Br. 87, but EPA and CASAC\\nboth determined that visible leaf injury itself can “impact the\\npublic welfare” by harming “aesthetic value and outdoor\\nrecreation” in public lands, 80 Fed. Reg. at 65,379; J.A. 533.\\nVisible leaf injury impairs a variety of outdoor activities,\\nincluding “scenic viewing, wildlife watching, hiking, and\\ncamping, that are of significance to the public welfare” and\\ngenerate “millions of dollars in economic value” annually. 80\\nFed. Reg. at 65,381; J.A. 749. Much of the documented visible\\nleaf injury is in Class I areas, see 80 Fed. Reg. at 65,378, the\\npreservation and enjoyment of which is a “clear public\\ninterest,” id. at 65,377 (quoting 73 Fed. Reg. at 16,496). Those\\ninterests are manifest whether or not the relationship between\\nleaf injury and other vegetative effects has been well\\ncharacterized.\\n\\n This case is a far cry from Center for Biological Diversity\\nv. EPA, where we upheld EPA’s determination that the\\navailable data was too uncertain to support setting a standard\\nfor acid rain precursors. 749 F.3d 1079 (D.C. Cir. 2014). In\\nthe administrative record in that case, EPA “explained in great\\ndetail” why the scientific uncertainties were so “unusually\\nprofound” that EPA “could not form” a reasoned judgment as\\nto a requisite level of protection. Id. at 1088, 1090-91. CASAC\\nconcurred with that assessment. Id. at 1086 n.11. Here, in\\n\\f 36\\ncontrast, CASAC expressed its scientific judgment that a target\\nlevel of 10 ppm-hrs was “required to reduce” leaf injury. J.A.\\n538. EPA’s failure to explain why it could not accept data that\\nCASAC deemed informative falls short of reasoned\\ndecisionmaking.\\n\\nB. Industry Petitioners’ Challenge\\n\\n The Industry Petitioners glancingly claim that EPA failed\\nto explain why, when faced with the same scientific evidence\\nthat was available in 2008, it revised the secondary standard\\ndownward. As we have explained in reference to the primary\\nstandard, it is a “conceptual error” to assume that “EPA is\\nsomehow bound by” a prior standard in a subsequent review.\\nMississippi, 844 F.3d at 1344. What is more, EPA did in fact\\nrely on new scientific evidence, including more than 400 new\\nstudies and various new analyses of existing data, which\\n“strengthen[ed] [its] confidence” that the 2008 secondary\\nstandard was no longer as demanding as was “requisite” to\\nprotect the public welfare. See 80 Fed. Reg. at 65,369-70,\\n65,384. “[A]dditional certainty . . . that the line marked by the\\nterm ‘requisite’ has shifted” is a reasonable—if not\\nparadigmatic—basis for EPA to revise its prior standards.\\nMississippi, 844 F.3d at 1344.\\n\\n IV. Cross-Cutting NAAQS Challenges\\n\\n We now turn to the arguments of State and Industry\\nPetitioners that EPA failed to take into account all the factors\\nrequired by law when setting the primary and secondary\\nstandards. First, they fault EPA for not considering the “overall\\nadverse economic, social, and energy impacts” of the revised\\nNAAQS. Second, they argue that the Clean Air Act requires\\nEPA to consider the impact of background ozone on the ability\\nof states to attain the revised NAAQS. And, lastly, they argue\\nthat EPA’s interpretation of the Act, which excluded\\n\\f 37\\nconsideration of background ozone, leaves no “intelligible\\nprinciple” by which to set NAAQS, creating a constitutional\\nnondelegation issue. None of these arguments has merit.\\n\\nA. Adverse Impacts\\n\\n According to section 109(b) of the Clean Air Act, EPA\\nmust set the NAAQS at a level “requisite to protect the public\\nhealth” and “the public welfare.” 42 U.S.C. § 7409(b). The\\nNAAQS should be revised periodically “as may be\\nappropriate.” Id. § 7409(d)(1). Industry Petitioners invoke the\\nterm “appropriate” and claim that “an evaluation of\\n‘appropriateness’ must take into account the adverse\\nsocioeconomic and energy impacts” of the revised NAAQS.\\nIndus. Pet’rs Br. 33. To bolster this argument, they point to\\nMichigan v. EPA, which held that the phrase “appropriate and\\nnecessary” in section 112(n) of the Act—governing power\\nplants’ hazardous air pollution—“requires at least some\\nattention to cost” because “‘appropriate’ is ‘the classic broad\\nand all-encompassing term that naturally and traditionally\\nincludes consideration of all the relevant factors.’” 135 S. Ct.\\nat 2707 (quoting White Stallion Energy Ctr., LLC v. EPA, 748\\nF.3d 1222, 1266 (D.C. Cir. 2014) (Kavanaugh, J., concurring\\nin part and dissenting in part)).\\n\\n In Whitman, the Supreme Court considered a similar\\nargument regarding NAAQS-setting, with challengers\\nasserting that the words “requisite” and “adequate” in section\\n109(b)(1) meant that EPA must consider the economic costs of\\nimplementing revised NAAQS. 531 U.S. at 468. The Whitman\\nCourt rejected this reading, finding that the plain text of the Act\\n“unambiguously bars cost considerations from the NAAQS-\\nsetting process.” Id. at 471. The text of the Act refers only to\\nthe “public health,” and though it uses the terms “requisite” and\\n“adequate,” the Whitman Court found it “implausible that\\n\\f 38\\nCongress would give to the EPA through these modest words\\nthe power to determine whether implementation costs should\\nmoderate national air quality standards,” especially given that\\nother Clean Air Act provisions “explicitly permit[] or require[]\\neconomic costs to be taken into account in implementing\\n[other] air quality standards.” Id. at 467-68.\\n\\n According to Whitman, the Act commands EPA to follow\\na specific process when setting primary NAAQS: “[I]dentify\\nthe maximum airborne concentration of a pollutant that the\\npublic health can tolerate, decrease the concentration to\\nprovide an ‘adequate’ margin of safety, and set the standard at\\nthat level.” Id. at 465. Industry Petitioners argue that this\\naccount of the NAAQS-setting process is incomplete, and that\\nconsideration of other “adverse economic, social, and energy\\nimpacts” is required. But at bottom this is the same argument\\nrejected in Whitman, with the “costs” of the revised and more\\nstringent NAAQS merely reframed as “impacts.” In listing\\nthose impacts, Industry Petitioners emphasize that stricter\\nstandards could reduce gross domestic product, drive up energy\\ncosts, “stymie economic growth by forcing the early retirement\\nof facilities unable to implement controls, contributing to job\\nlosses[,] discourag[e] existing businesses from expanding in\\nnonattainment regions[,] and driv[e] away potential new\\ninvestments,” Indus. Pet’rs Br. 35, but these “impacts” are no\\ndifferent than the “economic costs” that the petitioners in\\nWhitman worried “might produce health losses sufficient to\\noffset the health gains achieved in cleaning the air—for\\nexample, by closing down whole industries and thereby\\nimpoverishing the workers and consumers dependent upon\\nthose industries,” 531 U.S. at 466. Whitman forbids EPA from\\ntaking these considerations into account, however\\ndenominated.\\n\\f 39\\n As for Petitioners’ reliance on Michigan and the Act’s use\\nof the word “appropriate,” that argument fails twice over. We\\nhave already rejected the idea that “appropriate” in section\\n109(d) requires consideration of economic costs. Am. Trucking\\nAss’ns, Inc. v. EPA, 175 F.3d 1027, 1040-41 (D.C. Cir. 1999).\\nFurther, Michigan involved a different provision of the Clean\\nAir Act, and the Court was careful to emphasize that its reading\\nof “appropriate” was dependent on the statutory context,\\nexplaining that “[t]here are undoubtedly settings in which the\\nphrase ‘appropriate and necessary’ does not encompass cost.”\\n135 S. Ct. at 2707. Indeed, Michigan explicitly distinguished\\nsection 112(n) from section 109(b)(1), explaining yet again that\\nthe criteria for setting NAAQS “do[] not encompass cost.” Id.\\nat 2709.\\n\\n Industry Petitioners also point to section 109(d)(2)(C) of\\nthe Act, which requires CASAC to advise EPA “of any adverse\\npublic health, welfare, social, economic, or energy effects\\nwhich may result from various strategies for attainment and\\nmaintenance” of revised NAAQS. 42 U.S.C. § 7409(d)(2)(C).\\nAccording to Petitioners, the fact that CASAC is required to\\nsupply information to EPA about the “social, economic, or\\nenergy effects” of the revised NAAQS implies that EPA is\\nobliged to consider that information in setting the NAAQS.\\nThat argument, however, was also raised and rejected in\\nWhitman, where the Supreme Court concluded that this\\nprovision was intended to “enable the [EPA] to assist the States\\nin carrying out their statutory role as primary implementers of\\nthe NAAQS,” but had “no bearing upon whether cost\\nconsiderations are to be taken into account in formulating the\\n[NAAQS].” 531 U.S. at 470-71 (second emphasis added).\\n\\n Petitioners’ remaining arguments rely heavily on Justice\\nBreyer’s concurrence in Whitman, in which he urged that the\\nAct “does not require the EPA to eliminate every health risk,\\n\\f 40\\nhowever slight, at any economic cost, however great, to the\\npoint of hurtling industry over the brink of ruin, or even forcing\\ndeindustrialization.” Id. at 494 (Breyer, J., concurring in part\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4186,"cells":{"id":{"kind":"number","value":4655415,"string":"4,655,415"},"name":{"kind":"string","value":"HUMANE SOCIETY OF THE UNITED STATES, Et Al., Appellees/Cross-Appellants v. Sonny PERDUE, Secretary of the United States Department of Agriculture, Appellant/Cross-Appellee National Pork Producers Council, Appellant/Cross-Appellee"},"name_abbreviation":{"kind":"string","value":"Humane Soc'y of the United States v. Perdue"},"decision_date":{"kind":"string","value":"2019-08-23"},"docket_number":{"kind":"string","value":"18-5188; C/w 18-5197, 18-5201"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Griffith, Katsas, Rao\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge KATSAS\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued May 9, 2019 Decided August 23, 2019\\n\\n No. 18-5188\\n\\n HUMANE SOCIETY OF THE UNITED STATES, ET AL.,\\n APPELLEES/CROSS-APPELLANTS\\n\\n v.\\n\\n SONNY PERDUE, SECRETARY OF THE UNITED STATES\\n DEPARTMENT OF AGRICULTURE,\\n APPELLANT/CROSS-APPELLEE\\n\\n NATIONAL PORK PRODUCERS COUNCIL,\\n APPELLANT/CROSS-APPELLEE\\n\\n\\n Consolidated with 18-5197, 18-5201\\n\\n\\n Appeals from the United States District Court\\n for the District of Columbia\\n (No. 1:12-cv-01582)\\n\\n\\n Lewis S. Yelin, Attorney, U.S. Department of Justice,\\nargued the cause for appellant/cross-appellee Sonny Perdue.\\nWith him on the briefs were Scott R. McIntosh, Attorney, and\\nStephen A. Vaden, General Counsel, U.S. Department of\\nAgriculture.\\n\\f 2\\n Rebecca E. Bazan argued the cause for appellant/cross-\\nappellee National Pork Producers Council. With her on the\\nbriefs was John M. Simpson. Michelle C. Pardo entered an\\nappearance.\\n\\n Matthew E. Penzer argued the cause and filed the briefs\\nfor appellees/cross-appellants. Ralph E. Henry and Peter A.\\nBrandt entered appearances.\\n\\n Before: GRIFFITH, KATSAS, and RAO, Circuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge KATSAS.\\n\\n KATSAS, Circuit Judge: In this case, a pork farmer\\ncontends that the government unlawfully has permitted funds\\nfor promoting the pork industry to be used instead for lobbying\\non the industry’s behalf. We consider whether, on summary\\njudgment, the farmer has proven his constitutional standing to\\nmaintain this lawsuit.\\n\\n I\\n\\n The Pork Promotion, Research, and Consumer\\nInformation Act requires the federal government to promote\\nthe American pork industry. The Act authorizes the\\ngovernment to collect assessments (often called “checkoffs”)\\nfrom pork producers to finance efforts to “strengthen the\\nposition of the pork industry in the marketplace,” 7 U.S.C.\\n§ 4801(b)(1)(A), and “maintain, develop, and expand markets\\nfor pork and pork products,” id. § 4801(b)(1)(B). These\\nassessments are paid to the National Pork Board, an entity\\nestablished to “develop … proposals for promotion, research,\\nand consumer information plans and projects.” Id.\\n§ 4808(b)(1)(A). The Department of Agriculture must approve\\nthe Board’s promotional efforts and its annual expenses. Id.\\n\\f 3\\n§ 4808(b)(1)–(3). The Board may contract with private entities\\nto carry out those efforts, again with USDA’s approval. Id.\\n§ 4808(b)(4)(A). The Act prohibits using checkoff funds “for\\nthe purpose of influencing legislation.” Id. § 4809(e).\\n\\n In 2006, the Board purchased four trademarks from the\\nNational Pork Producers Council, a private lobbying\\norganization for the pork industry. The trademarks were\\nassociated with the slogan “Pork: The Other White Meat.” The\\nagreement required the Board to pay $3 million per year for\\ntwenty years, but permitted the Board, with advance notice, to\\ncancel the contract for any reason. In 2011, the Board adopted\\na new marketing campaign and stopped using three of the four\\ntrademarks, but it declined to end the contract and continued to\\nmake the annual payments.\\n\\n In 2012, the Humane Society of the United States, Iowa\\nCitizens for Community Improvement, and Harvey Dillenburg\\nfiled this lawsuit against the Secretary of Agriculture. The\\nHumane Society is an animal-protection organization; Iowa\\nCitizens is an organization of farmers; and Dillenburg is an\\nindividual pork farmer. They sought to challenge both\\nUSDA’s initial approval of the contract and its later approvals\\nof the annual payments. They contend that the contract\\nimpermissibly funds the Council’s lobbying activities, which\\nhas become more apparent as the Board continues to pay for\\ntrademarks that it no longer uses. The complaint alleges that\\nthis misuse of checkoff funds “diminishes the resources\\navailable for promotions or other legitimate programs” under\\nthe Act, thus “diminishing Mr. Dillenburg’s return on his\\ncompelled checkoff investment.” J.A. 51. The complaint\\nfurther alleges that Dillenburg is harmed by the payment of\\nfunds to the Council, “a lobbying organization that pushes for\\npolicies” he opposes. Id.\\n\\f 4\\n On a motion to dismiss, the district court held that none of\\nthe plaintiffs adequately alleged standing. Humane Soc’y v.\\nVilsack, 19 F. Supp. 3d 24 (D.D.C. 2013) (Humane Society I).\\nIn part, the court reasoned that Dillenburg’s “claimed reduced\\nreturn on investment” was “unsupported by facts,” because\\nstudies showed that the checkoff program yielded increasing\\nreturns to farmers over the periods at issue. Id. at 35–36.\\n\\n This Court reversed. Humane Soc’y v. Vilsack, 797 F.3d\\n4 (D.C. Cir. 2015) (Humane Society II). We held that\\nDillenburg’s first alleged injury—reduced “return on his\\ninvestment” from the misuse of checkoff funds—described a\\n“classic form of concrete and particularized harm: actual\\neconomic loss.” Id. at 8–9. We credited allegations that the\\nBoard was paying too much for the trademarks and that other\\nadvertising would have better propped up demand for pork.\\nSee id. We therefore assumed that “the price at which pork\\nproducers can sell their hogs is lower than it would be if the\\nBoard were spending those funds on legitimate promotions.”\\nId. at 9. After concluding that Dillenburg thus adequately had\\nalleged standing, we declined to consider whether the Humane\\nSociety or Iowa Citizens had done so. Id. at 10.\\n\\n On remand, the parties entered a joint stipulation. USDA\\nagreed, in deciding whether to approve the 2016 and future\\ncontract payments, to re-value the trademarks and to consider\\nevidence submitted by the plaintiffs. At the same time, the\\nplaintiffs agreed to dismiss their claims for recoupment of\\nfunds already spent. The Council intervened as a defendant\\nshortly thereafter. After completing its review, USDA decided\\nto authorize future contract payments.\\n\\n The parties then filed cross-motions for summary\\njudgment. In support of standing, Dillenburg submitted a\\ndeclaration stating that the contested payments prevented\\n\\f 5\\ncheckoff funds “from being used for promotions and other\\nlegitimate demand-enhancement activities” required by the\\nPork Act, thus depriving him “of the direct economic benefit\\nof the lawful and effective promotions to which [he is] entitled\\nas a statutory beneficiary.” J.A. 147. Dillenburg further stated\\nthat he opposed using checkoff funds to lobby for “policy\\nactivities that [he] believe[s] are harmful to [him].” Id. Neither\\nthe Humane Society nor Iowa Citizens filed standing affidavits.\\n\\n The district court held that Dillenburg adequately proved\\nstanding. Humane Soc’y v. Perdue, 290 F. Supp. 3d 5, 15–18\\n(D.D.C. 2018) (Humane Society III). The court reasoned that\\nDillenburg, as a pork farmer, “is affected by the market price\\nfor pork.” Id. at 16. Moreover, it concluded that the alleged\\ndiversion of “millions of checkoff dollars” from legitimate\\npork promotion was enough to show “economic loss.” Id. at\\n17. After finding standing, the court rejected challenges to\\napprovals made by USDA before 2016 as either untimely or\\nmoot, but it held that the 2016 decision to continue paying for\\nthe trademarks was arbitrary and capricious. See id. at 18–30.\\n\\n All parties have appealed. Our review is de novo. Trudel\\nv. SunTrust Bank, 924 F.3d 1281, 1285 (D.C. Cir. 2019).\\n\\n II\\n\\n The Constitution limits the “judicial Power of the United\\nStates” to “Cases” or “Controversies.” U.S. Const. art. III,\\n§§ 1–2. “To state a case or controversy under Article III, a\\nplaintiff must establish standing.” Ariz. Christian Sch. Tuition\\nOrg. v. Winn, 563 U.S. 125, 133 (2011). The test for standing\\nis settled: a plaintiff must have “(1) suffered an injury in fact,\\n(2) that is fairly traceable to the challenged conduct of the\\ndefendant, and (3) that is likely to be redressed by a favorable\\njudicial decision.” Spokeo, Inc. v. Robins, 136 S. Ct. 1540,\\n1547 (2016).\\n\\f 6\\n In Lujan v. Defenders of Wildlife, 504 U.S. 555 (1992), the\\nSupreme Court addressed the plaintiff’s burden to prove\\nstanding. The Court explained that “each element must be\\nsupported in the same way as any other matter on which the\\nplaintiff bears the burden of proof, i.e., with the manner and\\ndegree of evidence required at the successive stages of the\\nlitigation.” Id. at 561. Thus, at the pleading stage, the\\ncomplaint must “‘state a plausible claim’ that each element of\\nstanding is satisfied.” Hancock v. Urban Outfitters, Inc., 830\\nF.3d 511, 513 (D.C. Cir. 2016) (quoting Ashcroft v. Iqbal, 556\\nU.S. 662, 678–79 (2009)). But on summary judgment, “the\\nplaintiff can no longer rest on such mere allegations.” Defs. of\\nWildlife, 504 U.S. at 561 (quotation marks omitted); see\\nChamber of Commerce v. EPA, 642 F.3d 192, 201–02 (D.C.\\nCir. 2011). Parties must support factual assertions by “citing\\nto particular parts of materials in the record.” Fed. R. Civ. P.\\n56(c)(1)(A). So the plaintiff “must set forth by affidavit or\\nother evidence specific facts” that prove standing. Defs. of\\nWildlife, 504 U.S. at 561 (quotation marks omitted); see Scenic\\nAm., Inc. v. DOT, 836 F.3d 42, 49 n.3 (D.C. Cir. 2016) (“the\\nplaintiff has the burden to establish the evidentiary basis for its\\nstanding at the summary judgment stage”).\\n\\n The prior appeal in this case arose from USDA’s motion\\nto dismiss. In that context, we held that Dillenburg “alleged a\\nplausible claim to Article III standing” because injury to his\\n“bottom line” constituted a sufficient injury in fact, and\\nallegations that we were bound to credit “plausibly” showed\\nsuch an injury. Humane Society II, 797 F.3d at 8–10.\\n\\n Now, on summary judgment, the plaintiffs must prove\\ninjury in fact with “specific facts” in the record. Defs. of\\nWildlife, 504 U.S. at 561. But while the plaintiffs’ burden has\\ngrown, their allegations have shrunk. Dillenburg’s declaration\\nnowhere asserts a diminished return on investment, a reduced\\n\\f 7\\nbottom line, or any similar economic injury. Nor does it\\nprovide evidence that the Board’s alleged misadventures have\\nreduced the price of pork. Rather, Dillenburg declares that the\\nmisuse of checkoff funds robs him of the “direct economic\\nbenefit of the lawful and effective promotions to which [he is]\\nentitled as a statutory beneficiary.” J.A. 147. The briefing\\ndoubles down on this point, in contending that a bottom-line,\\npocketbook injury is unnecessary: “All that is needed to\\nestablish standing here is evidence that Mr. Dillenburg is a\\npaying member of the checkoff program, who is therefore\\nentitled to the statutorily required promotional benefit for his\\ncontributions.” Humane Soc’y Br. at 30; see also id. at 31\\n(“Unlawful uses of checkoff funds cause injury to …\\nassessment-paying producers by denying them lawful\\npromotional programming.”). The plaintiffs also pressed this\\ntheory of injury during oral argument. See, e.g., Oral Arg. at\\n49:42–47 (“he’s not getting the effective promotional services\\nthat the law requires”), 52:41–48 (“producers … are not getting\\nthe lawful and effective promotional services”).\\n\\n Dillenburg’s declaration does not prove an injury in fact.\\nRather than show harm to his bottom line, Dillenburg instead\\ncomplains that he has been deprived of a statutory entitlement\\nto lawful promotions. But “Article III standing requires a\\nconcrete injury even in the context of a statutory violation.”\\nSpokeo, 136 S. Ct. at 1549. The plaintiffs must show that the\\nstatute protected “a right to be free of a harm capable of\\nsatisfying Article III.” Jeffries v. Volume Servs. Am., Inc., 928\\nF.3d 1059, 1064 (D.C. Cir. 2019). On that score, Dillenburg’s\\ndeclaration says little. Without elaboration, it speaks of a\\n“direct economic benefit” that other promotions would have\\nprovided. J.A. 147. But on summary judgment, a party cannot\\nestablish standing with “conclusory allegations of an affidavit.”\\nLujan v. Nat’l Wildlife Fed’n, 497 U.S. 871, 888 (1990). For\\nthat reason, in Swanson Group Manufacturing LLC v. Jewell,\\n\\f 8\\n790 F.3d 235 (D.C. Cir. 2015), we held that a declaration\\nasserting only “economic loss and hardship,” which “told\\nnothing about the nature” of the loss, was insufficient to defeat\\na motion for summary judgment. Id. at 242–43 (quotation\\nmarks omitted). The same is true here—Dillenburg fails to\\nelaborate on the “direct economic benefit” that he allegedly has\\nlost because of the Board’s misdeeds.\\n\\n In concluding that Dillenburg proved standing, the district\\ncourt reasoned that Dillenburg, as a pork producer, “is affected\\nby the market price for pork.” Humane Society III, 290 F.\\nSupp. 3d at 16. Perhaps so, but Dillenburg’s declaration lacks\\nany indication that the “price for pork” was “affected” by the\\nalleged misuse of checkoff funds. The court assumed this\\ncritical fact based on Carpenters Industrial Council v. Zinke,\\n854 F.3d 1 (D.C. Cir. 2017). See Humane Society III, 290 F.\\nSupp. 3d at 16–17. Yet Carpenters did not displace settled law\\nthat courts cannot “presume the missing facts necessary to\\nestablish an element of standing.” Swanson, 790 F.3d at 240\\n(cleaned up). To the contrary, our decision carefully combed\\n“the complaint and declarations to assess whether the\\nplaintiff’s assertions suffice to show the elements of standing.”\\nCarpenters, 854 F.3d at 5. We explained that the declarations\\nin that case identified specific “harms ranging from lost sales\\nand diminished production to closures and layoffs.” Id. at 9.\\nAs discussed above, Dillenburg’s one-page declaration makes\\nno similar showing.\\n\\n The declaration suggests a second injury: Dillenburg\\n“oppose[s] the misuse of checkoff funds … for the purpose of\\ninfluencing legislation,” because he “believe[s]” those efforts\\nare “harmful to [him] and other independent farmers.” J.A.\\n147. This theory of injury also falls short. Plaintiffs cannot\\npredicate Article III standing on a “mere interest in a problem\\nor an ideological injury.” PETA v. USDA, 797 F.3d 1087, 1094\\n\\f 9\\n(D.C. Cir. 2015) (cleaned up). An “interest in the proper\\nadministration of the laws” is canonically “nonconcrete.”\\nSummers v. Earth Island Inst., 555 U.S. 488, 497 (2009)\\n(quotation marks omitted). Here, Dillenburg does not explain\\nhow his pork business is harmed by lobbying to promote the\\npork industry, let alone identify a specific lobbying activity that\\ncauses him a concrete injury. Instead, Dillenburg highlights\\nhis general opposition to the Council’s lobbying and the\\ngovernment’s funding of it. This declares “only a generally\\navailable grievance about government,” which “does not state\\nan Article III case or controversy.” Defs. of Wildlife, 504 U.S.\\nat 573–74.\\n\\n Finally, the plaintiffs cannot salvage this case based on the\\npossible standing of the Humane Society or Iowa Citizens. On\\nsummary judgment, neither organization filed any standing\\ndeclarations or affidavits. And now in this Court, the plaintiffs\\ndo not assert the standing of these organizations as an\\nalternative ground for reaching the merits. The plaintiffs thus\\nhave twice forfeited any claim that these organizations have\\nstanding. See Scenic Am., 836 F.3d at 53 n.4 (“Although a\\nparty cannot forfeit a claim that we lack jurisdiction, it can\\nforfeit a claim that we possess jurisdiction.”).\\n\\n * * * *\\n\\n Because the plaintiffs offer no evidence that the Board’s\\nalleged misuse of checkoff funds caused them to suffer an\\ninjury in fact, we vacate the district court’s order and remand\\nwith instructions to dismiss the case for lack of standing.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4432668/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge KATSAS\"}]}"},"judge1":{"kind":"string","value":"GRIFFITH"},"judge2":{"kind":"string","value":"KATSAS"},"judge3":{"kind":"string","value":"RAO"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4655415/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: HUMANE SOCIETY OF THE UNITED STATES, Et Al., Appellees/Cross-Appellants v. Sonny PERDUE, Secretary of the United States Department of Agriculture, Appellant/Cross-Appellee National Pork Producers Council, Appellant/Cross-Appellee\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-23\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued May 9, 2019 Decided August 23, 2019\n\n No. 18-5188\n\n HUMANE SOCIETY OF THE UNITED STATES, ET AL.,\n APPELLEES/CROSS-APPELLANTS\n\n v.\n\n SONNY PERDUE, SECRETARY OF THE UNITED STATES\n DEPARTMENT OF AGRICULTURE,\n APPELLANT/CROSS-APPELLEE\n\n NATIONAL PORK PRODUCERS COUNCIL,\n APPELLANT/CROSS-APPELLEE\n\n\n Consolidated with 18-5197, 18-5201\n\n\n Appeals from the United States District Court\n for the District of Columbia\n (No. 1:12-cv-01582)\n\n\n Lewis S. Yelin, Attorney, U.S. Department of Justice,\nargued the cause for appellant/cross-appellee Sonny Perdue.\nWith him on the briefs were Scott R. McIntosh, Attorney, and\nStephen A. Vaden, General Counsel, U.S. Department of\nAgriculture.\n\f 2\n Rebecca E. Bazan argued the cause for appellant/cross-\nappellee National Pork Producers Council. With her on the\nbriefs was John M. Simpson. Michelle C. Pardo entered an\nappearance.\n\n Matthew E. Penzer argued the cause and filed the briefs\nfor appellees/cross-appellants. Ralph E. Henry and Peter A.\nBrandt entered appearances.\n\n Before: GRIFFITH, KATSAS, and RAO, Circuit Judges.\n\n Opinion for the Court filed by Circuit Judge KATSAS.\n\n KATSAS, Circuit Judge: In this case, a pork farmer\ncontends that the government unlawfully has permitted funds\nfor promoting the pork industry to be used instead for lobbying\non the industry’s behalf. We consider whether, on summary\njudgment, the farmer has proven his constitutional standing to\nmaintain this lawsuit.\n\n I\n\n The Pork Promotion, Research, and Consumer\nInformation Act requires the federal government to promote\nthe American pork industry. The Act authorizes the\ngovernment to collect assessments (often called “checkoffs”)\nfrom pork producers to finance efforts to “strengthen the\nposition of the pork industry in the marketplace,” 7 U.S.C.\n§ 4801(b)(1)(A), and “maintain, develop, and expand markets\nfor pork and pork products,” id. § 4801(b)(1)(B). These\nassessments are paid to the National Pork Board, an entity\nestablished to “develop … proposals for promotion, research,\nand consumer information plans and projects.” Id.\n§ 4808(b)(1)(A). The Department of Agriculture must approve\nthe Board’s promotional efforts and its annual expenses. Id.\n\f 3\n§ 4808(b)(1)–(3). The Board may contract with private entities\nto carry out those efforts, again with USDA’s approval. Id.\n§ 4808(b)(4)(A). The Act prohibits using checkoff funds “for\nthe purpose of influencing legislation.” Id. § 4809(e).\n\n In 2006, the Board purchased four trademarks from the\nNational Pork Producers Council, a private lobbying\norganization for the pork industry. The trademarks were\nassociated with the slogan “Pork: The Other White Meat.” The\nagreement required the Board to pay $3 million per year for\ntwenty years, but permitted the Board, with advance notice, to\ncancel the contract for any reason. In 2011, the Board adopted\na new marketing campaign and stopped using three of the four\ntrademarks, but it declined to end the contract and continued to\nmake the annual payments.\n\n In 2012, the Humane Society of the United States, Iowa\nCitizens for Community Improvement, and Harvey Dillenburg\nfiled this lawsuit against the Secretary of Agriculture. The\nHumane Society is an animal-protection organization; Iowa\nCitizens is an organization of farmers; and Dillenburg is an\nindividual pork farmer. They sought to challenge both\nUSDA’s initial approval of the contract and its later approvals\nof the annual payments. They contend that the contract\nimpermissibly funds the Council’s lobbying activities, which\nhas become more apparent as the Board continues to pay for\ntrademarks that it no longer uses. The complaint alleges that\nthis misuse of checkoff funds “diminishes the resources\navailable for promotions or other legitimate programs” under\nthe Act, thus “diminishing Mr. Dillenburg’s return on his\ncompelled checkoff investment.” J.A. 51. The complaint\nfurther alleges that Dillenburg is harmed by the payment of\nfunds to the Council, “a lobbying organization that pushes for\npolicies” he opposes. Id.\n\f 4\n On a motion to dismiss, the district court held that none of\nthe plaintiffs adequately alleged standing. Humane Soc’y v.\nVilsack, 19 F. Supp. 3d 24 (D.D.C. 2013) (Humane Society I).\nIn part, the court reasoned that Dillenburg’s “claimed reduced\nreturn on investment” was “unsupported by facts,” because\nstudies showed that the checkoff program yielded increasing\nreturns to farmers over the periods at issue. Id. at 35–36.\n\n This Court reversed. Humane Soc’y v. Vilsack, 797 F.3d\n4 (D.C. Cir. 2015) (Humane Society II). We held that\nDillenburg’s first alleged injury—reduced “return on his\ninvestment” from the misuse of checkoff funds—described a\n“classic form of concrete and particularized harm: actual\neconomic loss.” Id. at 8–9. We credited allegations that the\nBoard was paying too much for the trademarks and that other\nadvertising would have better propped up demand for pork.\nSee id. We therefore assumed that “the price at which pork\nproducers can sell their hogs is lower than it would be if the\nBoard were spending those funds on legitimate promotions.”\nId. at 9. After concluding that Dillenburg thus adequately had\nalleged standing, we declined to consider whether the Humane\nSociety or Iowa Citizens had done so. Id. at 10.\n\n On remand, the parties entered a joint stipulation. USDA\nagreed, in deciding whether to approve the 2016 and future\ncontract payments, to re-value the trademarks and to consider\nevidence submitted by the plaintiffs. At the same time, the\nplaintiffs agreed to dismiss their claims for recoupment of\nfunds already spent. The Council intervened as a defendant\nshortly thereafter. After completing its review, USDA decided\nto authorize future contract payments.\n\n The parties then filed cross-motions for summary\njudgment. In support of standing, Dillenburg submitted a\ndeclaration stating that the contested payments prevented\n\f 5\ncheckoff funds “from being used for promotions and other\nlegitimate demand-enhancement activities” required by the\nPork Act, thus depriving him “of the direct economic benefit\nof the lawful and effective promotions to which [he is] entitled\nas a statutory beneficiary.” J.A. 147. Dillenburg further stated\nthat he opposed using checkoff funds to lobby for “policy\nactivities that [he] believe[s] are harmful to [him].” Id. Neither\nthe Humane Society nor Iowa Citizens filed standing affidavits.\n\n The district court held that Dillenburg adequately proved\nstanding. Humane Soc’y v. Perdue, 290 F. Supp. 3d 5, 15–18\n(D.D.C. 2018) (Humane Society III). The court reasoned that\nDillenburg, as a pork farmer, “is affected by the market price\nfor pork.” Id. at 16. Moreover, it concluded that the alleged\ndiversion of “millions of checkoff dollars” from legitimate\npork promotion was enough to show “economic loss.” Id. at\n17. After finding standing, the court rejected challenges to\napprovals made by USDA before 2016 as either untimely or\nmoot, but it held that the 2016 decision to continue paying for\nthe trademarks was arbitrary and capricious. See id. at 18–30.\n\n All parties have appealed. Our review is de novo. Trudel\nv. SunTrust Bank, 924 F.3d 1281, 1285 (D.C. Cir. 2019).\n\n II\n\n The Constitution limits the “judicial Power of the United\nStates” to “Cases” or “Controversies.” U.S. Const. art. III,\n§§ 1–2. “To state a case or controversy under Article III, a\nplaintiff must establish standing.” Ariz. Christian Sch. Tuition\nOrg. v. Winn, 563 U.S. 125, 133 (2011). The test for standing\nis settled: a plaintiff must have “(1) suffered an injury in fact,\n(2) that is fairly traceable to the challenged conduct of the\ndefendant, and (3) that is likely to be redressed by a favorable\njudicial decision.” Spokeo, Inc. v. Robins, 136 S. Ct. 1540,\n1547 (2016).\n\f 6\n In Lujan v. Defenders of Wildlife, 504 U.S. 555 (1992), the\nSupreme Court addressed the plaintiff’s burden to prove\nstanding. The Court explained that “each element must be\nsupported in the same way as any other matter on which the\nplaintiff bears the burden of proof, i.e., with the manner and\ndegree of evidence required at the successive stages of the\nlitigation.” Id. at 561. Thus, at the pleading stage, the\ncomplaint must “‘state a plausible claim’ that each element of\nstanding is satisfied.” Hancock v. Urban Outfitters, Inc., 830\nF.3d 511, 513 (D.C. Cir. 2016) (quoting Ashcroft v. Iqbal, 556\nU.S. 662, 678–79 (2009)). But on summary judgment, “the\nplaintiff can no longer rest on such mere allegations.” Defs. of\nWildlife, 504 U.S. at 561 (quotation marks omitted); see\nChamber of Commerce v. EPA, 642 F.3d 192, 201–02 (D.C.\nCir. 2011). Parties must support factual assertions by “citing\nto particular parts of materials in the record.” Fed. R. Civ. P.\n56(c)(1)(A). So the plaintiff “must set forth by affidavit or\nother evidence specific facts” that prove standing. Defs. of\nWildlife, 504 U.S. at 561 (quotation marks omitted); see Scenic\nAm., Inc. v. DOT, 836 F.3d 42, 49 n.3 (D.C. Cir. 2016) (“the\nplaintiff has the burden to establish the evidentiary basis for its\nstanding at the summary judgment stage”).\n\n The prior appeal in this case arose from USDA’s motion\nto dismiss. In that context, we held that Dillenburg “alleged a\nplausible claim to Article III standing” because injury to his\n“bottom line” constituted a sufficient injury in fact, and\nallegations that we were bound to credit “plausibly” showed\nsuch an injury. Humane Society II, 797 F.3d at 8–10.\n\n Now, on summary judgment, the plaintiffs must prove\ninjury in fact with “specific facts” in the record. Defs. of\nWildlife, 504 U.S. at 561. But while the plaintiffs’ burden has\ngrown, their allegations have shrunk. Dillenburg’s declaration\nnowhere asserts a diminished return on investment, a reduced\n\f 7\nbottom line, or any similar economic injury. Nor does it\nprovide evidence that the Board’s alleged misadventures have\nreduced the price of pork. Rather, Dillenburg declares that the\nmisuse of checkoff funds robs him of the “direct economic\nbenefit of the lawful and effective promotions to which [he is]\nentitled as a statutory beneficiary.” J.A. 147. The briefing\ndoubles down on this point, in contending that a bottom-line,\npocketbook injury is unnecessary: “All that is needed to\nestablish standing here is evidence that Mr. Dillenburg is a\npaying member of the checkoff program, who is therefore\nentitled to the statutorily required promotional benefit for his\ncontributions.” Humane Soc’y Br. at 30; see also id. at 31\n(“Unlawful uses of checkoff funds cause injury to …\nassessment-paying producers by denying them lawful\npromotional programming.”). The plaintiffs also pressed this\ntheory of injury during oral argument. See, e.g., Oral Arg. at\n49:42–47 (“he’s not getting the effective promotional services\nthat the law requires”), 52:41–48 (“producers … are not getting\nthe lawful and effective promotional services”).\n\n Dillenburg’s declaration does not prove an injury in fact.\nRather than show harm to his bottom line, Dillenburg instead\ncomplains that he has been deprived of a statutory entitlement\nto lawful promotions. But “Article III standing requires a\nconcrete injury even in the context of a statutory violation.”\nSpokeo, 136 S. Ct. at 1549. The plaintiffs must show that the\nstatute protected “a right to be free of a harm capable of\nsatisfying Article III.” Jeffries v. Volume Servs. Am., Inc., 928\nF.3d 1059, 1064 (D.C. Cir. 2019). On that score, Dillenburg’s\ndeclaration says little. Without elaboration, it speaks of a\n“direct economic benefit” that other promotions would have\nprovided. J.A. 147. But on summary judgment, a party cannot\nestablish standing with “conclusory allegations of an affidavit.”\nLujan v. Nat’l Wildlife Fed’n, 497 U.S. 871, 888 (1990). For\nthat reason, in Swanson Group Manufacturing LLC v. Jewell,\n\f 8\n790 F.3d 235 (D.C. Cir. 2015), we held that a declaration\nasserting only “economic loss and hardship,” which “told\nnothing about the nature” of the loss, was insufficient to defeat\na motion for summary judgment. Id. at 242–43 (quotation\nmarks omitted). The same is true here—Dillenburg fails to\nelaborate on the “direct economic benefit” that he allegedly has\nlost because of the Board’s misdeeds.\n\n In concluding that Dillenburg proved standing, the district\ncourt reasoned that Dillenburg, as a pork producer, “is affected\nby the market price for pork.” Humane Society III, 290 F.\nSupp. 3d at 16. Perhaps so, but Dillenburg’s declaration lacks\nany indication that the “price for pork” was “affected” by the\nalleged misuse of checkoff funds. The court assumed this\ncritical fact based on Carpenters Industrial Council v. Zinke,\n854 F.3d 1 (D.C. Cir. 2017). See Humane Society III, 290 F.\nSupp. 3d at 16–17. Yet Carpenters did not displace settled law\nthat courts cannot “presume the missing facts necessary to\nestablish an element of standing.” Swanson, 790 F.3d at 240\n(cleaned up). To the contrary, our decision carefully combed\n“the complaint and declarations to assess whether the\nplaintiff’s assertions suffice to show the elements of standing.”\nCarpenters, 854 F.3d at 5. We explained that the declarations\nin that case identified specific “harms ranging from lost sales\nand diminished production to closures and layoffs.” Id. at 9.\nAs discussed above, Dillenburg’s one-page declaration makes\nno similar showing.\n\n The declaration suggests a second injury: Dillenburg\n“oppose[s] the misuse of checkoff funds … for the purpose of\ninfluencing legislation,” because he “believe[s]” those efforts\nare “harmful to [him] and other independent farmers.” J.A.\n147. This theory of injury also falls short. Plaintiffs cannot\npredicate Article III standing on a “mere interest in a problem\nor an ideological injury.” PETA v. USDA, 797 F.3d 1087, 1094\n\f 9\n(D.C. Cir. 2015) (cleaned up). An “interest in the proper\nadministration of the laws” is canonically “nonconcrete.”\nSummers v. Earth Island Inst., 555 U.S. 488, 497 (2009)\n(quotation marks omitted). Here, Dillenburg does not explain\nhow his pork business is harmed by lobbying to promote the\npork industry, let alone identify a specific lobbying activity that\ncauses him a concrete injury. Instead, Dillenburg highlights\nhis general opposition to the Council’s lobbying and the\ngovernment’s funding of it. This declares “only a generally\navailable grievance about government,” which “does not state\nan Article III case or controversy.” Defs. of Wildlife, 504 U.S.\nat 573–74.\n\n Finally, the plaintiffs cannot salvage this case based on the\npossible standing of the Humane Society or Iowa Citizens. On\nsummary judgment, neither organization filed any standing\ndeclarations or affidavits. And now in this Court, the plaintiffs\ndo not assert the standing of these organizations as an\nalternative ground for reaching the merits. The plaintiffs thus\nhave twice forfeited any claim that these organizations have\nstanding. See Scenic Am., 836 F.3d at 53 n.4 (“Although a\nparty cannot forfeit a claim that we lack jurisdiction, it can\nforfeit a claim that we possess jurisdiction.”).\n\n * * * *\n\n Because the plaintiffs offer no evidence that the Board’s\nalleged misuse of checkoff funds caused them to suffer an\ninjury in fact, we vacate the district court’s order and remand\nwith instructions to dismiss the case for lack of standing.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: HUMANE SOCIETY OF THE UNITED STATES, Et Al., Appellees/Cross-Appellants v. Sonny PERDUE, Secretary of the United States Department of Agriculture, Appellant/Cross-Appellee National Pork Producers Council, Appellant/Cross-Appellee\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-23\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued May 9, 2019 Decided August 23, 2019\\n\\n No. 18-5188\\n\\n HUMANE SOCIETY OF THE UNITED STATES, ET AL.,\\n APPELLEES/CROSS-APPELLANTS\\n\\n v.\\n\\n SONNY PERDUE, SECRETARY OF THE UNITED STATES\\n DEPARTMENT OF AGRICULTURE,\\n APPELLANT/CROSS-APPELLEE\\n\\n NATIONAL PORK PRODUCERS COUNCIL,\\n APPELLANT/CROSS-APPELLEE\\n\\n\\n Consolidated with 18-5197, 18-5201\\n\\n\\n Appeals from the United States District Court\\n for the District of Columbia\\n (No. 1:12-cv-01582)\\n\\n\\n Lewis S. Yelin, Attorney, U.S. Department of Justice,\\nargued the cause for appellant/cross-appellee Sonny Perdue.\\nWith him on the briefs were Scott R. McIntosh, Attorney, and\\nStephen A. Vaden, General Counsel, U.S. Department of\\nAgriculture.\\n\\f 2\\n Rebecca E. Bazan argued the cause for appellant/cross-\\nappellee National Pork Producers Council. With her on the\\nbriefs was John M. Simpson. Michelle C. Pardo entered an\\nappearance.\\n\\n Matthew E. Penzer argued the cause and filed the briefs\\nfor appellees/cross-appellants. Ralph E. Henry and Peter A.\\nBrandt entered appearances.\\n\\n Before: GRIFFITH, KATSAS, and RAO, Circuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge KATSAS.\\n\\n KATSAS, Circuit Judge: In this case, a pork farmer\\ncontends that the government unlawfully has permitted funds\\nfor promoting the pork industry to be used instead for lobbying\\non the industry’s behalf. We consider whether, on summary\\njudgment, the farmer has proven his constitutional standing to\\nmaintain this lawsuit.\\n\\n I\\n\\n The Pork Promotion, Research, and Consumer\\nInformation Act requires the federal government to promote\\nthe American pork industry. The Act authorizes the\\ngovernment to collect assessments (often called “checkoffs”)\\nfrom pork producers to finance efforts to “strengthen the\\nposition of the pork industry in the marketplace,” 7 U.S.C.\\n§ 4801(b)(1)(A), and “maintain, develop, and expand markets\\nfor pork and pork products,” id. § 4801(b)(1)(B). These\\nassessments are paid to the National Pork Board, an entity\\nestablished to “develop … proposals for promotion, research,\\nand consumer information plans and projects.” Id.\\n§ 4808(b)(1)(A). The Department of Agriculture must approve\\nthe Board’s promotional efforts and its annual expenses. Id.\\n\\f 3\\n§ 4808(b)(1)–(3). The Board may contract with private entities\\nto carry out those efforts, again with USDA’s approval. Id.\\n§ 4808(b)(4)(A). The Act prohibits using checkoff funds “for\\nthe purpose of influencing legislation.” Id. § 4809(e).\\n\\n In 2006, the Board purchased four trademarks from the\\nNational Pork Producers Council, a private lobbying\\norganization for the pork industry. The trademarks were\\nassociated with the slogan “Pork: The Other White Meat.” The\\nagreement required the Board to pay $3 million per year for\\ntwenty years, but permitted the Board, with advance notice, to\\ncancel the contract for any reason. In 2011, the Board adopted\\na new marketing campaign and stopped using three of the four\\ntrademarks, but it declined to end the contract and continued to\\nmake the annual payments.\\n\\n In 2012, the Humane Society of the United States, Iowa\\nCitizens for Community Improvement, and Harvey Dillenburg\\nfiled this lawsuit against the Secretary of Agriculture. The\\nHumane Society is an animal-protection organization; Iowa\\nCitizens is an organization of farmers; and Dillenburg is an\\nindividual pork farmer. They sought to challenge both\\nUSDA’s initial approval of the contract and its later approvals\\nof the annual payments. They contend that the contract\\nimpermissibly funds the Council’s lobbying activities, which\\nhas become more apparent as the Board continues to pay for\\ntrademarks that it no longer uses. The complaint alleges that\\nthis misuse of checkoff funds “diminishes the resources\\navailable for promotions or other legitimate programs” under\\nthe Act, thus “diminishing Mr. Dillenburg’s return on his\\ncompelled checkoff investment.” J.A. 51. The complaint\\nfurther alleges that Dillenburg is harmed by the payment of\\nfunds to the Council, “a lobbying organization that pushes for\\npolicies” he opposes. Id.\\n\\f 4\\n On a motion to dismiss, the district court held that none of\\nthe plaintiffs adequately alleged standing. Humane Soc’y v.\\nVilsack, 19 F. Supp. 3d 24 (D.D.C. 2013) (Humane Society I).\\nIn part, the court reasoned that Dillenburg’s “claimed reduced\\nreturn on investment” was “unsupported by facts,” because\\nstudies showed that the checkoff program yielded increasing\\nreturns to farmers over the periods at issue. Id. at 35–36.\\n\\n This Court reversed. Humane Soc’y v. Vilsack, 797 F.3d\\n4 (D.C. Cir. 2015) (Humane Society II). We held that\\nDillenburg’s first alleged injury—reduced “return on his\\ninvestment” from the misuse of checkoff funds—described a\\n“classic form of concrete and particularized harm: actual\\neconomic loss.” Id. at 8–9. We credited allegations that the\\nBoard was paying too much for the trademarks and that other\\nadvertising would have better propped up demand for pork.\\nSee id. We therefore assumed that “the price at which pork\\nproducers can sell their hogs is lower than it would be if the\\nBoard were spending those funds on legitimate promotions.”\\nId. at 9. After concluding that Dillenburg thus adequately had\\nalleged standing, we declined to consider whether the Humane\\nSociety or Iowa Citizens had done so. Id. at 10.\\n\\n On remand, the parties entered a joint stipulation. USDA\\nagreed, in deciding whether to approve the 2016 and future\\ncontract payments, to re-value the trademarks and to consider\\nevidence submitted by the plaintiffs. At the same time, the\\nplaintiffs agreed to dismiss their claims for recoupment of\\nfunds already spent. The Council intervened as a defendant\\nshortly thereafter. After completing its review, USDA decided\\nto authorize future contract payments.\\n\\n The parties then filed cross-motions for summary\\njudgment. In support of standing, Dillenburg submitted a\\ndeclaration stating that the contested payments prevented\\n\\f 5\\ncheckoff funds “from being used for promotions and other\\nlegitimate demand-enhancement activities” required by the\\nPork Act, thus depriving him “of the direct economic benefit\\nof the lawful and effective promotions to which [he is] entitled\\nas a statutory beneficiary.” J.A. 147. Dillenburg further stated\\nthat he opposed using checkoff funds to lobby for “policy\\nactivities that [he] believe[s] are harmful to [him].” Id. Neither\\nthe Humane Society nor Iowa Citizens filed standing affidavits.\\n\\n The district court held that Dillenburg adequately proved\\nstanding. Humane Soc’y v. Perdue, 290 F. Supp. 3d 5, 15–18\\n(D.D.C. 2018) (Humane Society III). The court reasoned that\\nDillenburg, as a pork farmer, “is affected by the market price\\nfor pork.” Id. at 16. Moreover, it concluded that the alleged\\ndiversion of “millions of checkoff dollars” from legitimate\\npork promotion was enough to show “economic loss.” Id. at\\n17. After finding standing, the court rejected challenges to\\napprovals made by USDA before 2016 as either untimely or\\nmoot, but it held that the 2016 decision to continue paying for\\nthe trademarks was arbitrary and capricious. See id. at 18–30.\\n\\n All parties have appealed. Our review is de novo. Trudel\\nv. SunTrust Bank, 924 F.3d 1281, 1285 (D.C. Cir. 2019).\\n\\n II\\n\\n The Constitution limits the “judicial Power of the United\\nStates” to “Cases” or “Controversies.” U.S. Const. art. III,\\n§§ 1–2. “To state a case or controversy under Article III, a\\nplaintiff must establish standing.” Ariz. Christian Sch. Tuition\\nOrg. v. Winn, 563 U.S. 125, 133 (2011). The test for standing\\nis settled: a plaintiff must have “(1) suffered an injury in fact,\\n(2) that is fairly traceable to the challenged conduct of the\\ndefendant, and (3) that is likely to be redressed by a favorable\\njudicial decision.” Spokeo, Inc. v. Robins, 136 S. Ct. 1540,\\n1547 (2016).\\n\\f 6\\n In Lujan v. Defenders of Wildlife, 504 U.S. 555 (1992), the\\nSupreme Court addressed the plaintiff’s burden to prove\\nstanding. The Court explained that “each element must be\\nsupported in the same way as any other matter on which the\\nplaintiff bears the burden of proof, i.e., with the manner and\\ndegree of evidence required at the successive stages of the\\nlitigation.” Id. at 561. Thus, at the pleading stage, the\\ncomplaint must “‘state a plausible claim’ that each element of\\nstanding is satisfied.” Hancock v. Urban Outfitters, Inc., 830\\nF.3d 511, 513 (D.C. Cir. 2016) (quoting Ashcroft v. Iqbal, 556\\nU.S. 662, 678–79 (2009)). But on summary judgment, “the\\nplaintiff can no longer rest on such mere allegations.” Defs. of\\nWildlife, 504 U.S. at 561 (quotation marks omitted); see\\nChamber of Commerce v. EPA, 642 F.3d 192, 201–02 (D.C.\\nCir. 2011). Parties must support factual assertions by “citing\\nto particular parts of materials in the record.” Fed. R. Civ. P.\\n56(c)(1)(A). So the plaintiff “must set forth by affidavit or\\nother evidence specific facts” that prove standing. Defs. of\\nWildlife, 504 U.S. at 561 (quotation marks omitted); see Scenic\\nAm., Inc. v. DOT, 836 F.3d 42, 49 n.3 (D.C. Cir. 2016) (“the\\nplaintiff has the burden to establish the evidentiary basis for its\\nstanding at the summary judgment stage”).\\n\\n The prior appeal in this case arose from USDA’s motion\\nto dismiss. In that context, we held that Dillenburg “alleged a\\nplausible claim to Article III standing” because injury to his\\n“bottom line” constituted a sufficient injury in fact, and\\nallegations that we were bound to credit “plausibly” showed\\nsuch an injury. Humane Society II, 797 F.3d at 8–10.\\n\\n Now, on summary judgment, the plaintiffs must prove\\ninjury in fact with “specific facts” in the record. Defs. of\\nWildlife, 504 U.S. at 561. But while the plaintiffs’ burden has\\ngrown, their allegations have shrunk. Dillenburg’s declaration\\nnowhere asserts a diminished return on investment, a reduced\\n\\f 7\\nbottom line, or any similar economic injury. Nor does it\\nprovide evidence that the Board’s alleged misadventures have\\nreduced the price of pork. Rather, Dillenburg declares that the\\nmisuse of checkoff funds robs him of the “direct economic\\nbenefit of the lawful and effective promotions to which [he is]\\nentitled as a statutory beneficiary.” J.A. 147. The briefing\\ndoubles down on this point, in contending that a bottom-line,\\npocketbook injury is unnecessary: “All that is needed to\\nestablish standing here is evidence that Mr. Dillenburg is a\\npaying member of the checkoff program, who is therefore\\nentitled to the statutorily required promotional benefit for his\\ncontributions.” Humane Soc’y Br. at 30; see also id. at 31\\n(“Unlawful uses of checkoff funds cause injury to …\\nassessment-paying producers by denying them lawful\\npromotional programming.”). The plaintiffs also pressed this\\ntheory of injury during oral argument. See, e.g., Oral Arg. at\\n49:42–47 (“he’s not getting the effective promotional services\\nthat the law requires”), 52:41–48 (“producers … are not getting\\nthe lawful and effective promotional services”).\\n\\n Dillenburg’s declaration does not prove an injury in fact.\\nRather than show harm to his bottom line, Dillenburg instead\\ncomplains that he has been deprived of a statutory entitlement\\nto lawful promotions. But “Article III standing requires a\\nconcrete injury even in the context of a statutory violation.”\\nSpokeo, 136 S. Ct. at 1549. The plaintiffs must show that the\\nstatute protected “a right to be free of a harm capable of\\nsatisfying Article III.” Jeffries v. Volume Servs. Am., Inc., 928\\nF.3d 1059, 1064 (D.C. Cir. 2019). On that score, Dillenburg’s\\ndeclaration says little. Without elaboration, it speaks of a\\n“direct economic benefit” that other promotions would have\\nprovided. J.A. 147. But on summary judgment, a party cannot\\nestablish standing with “conclusory allegations of an affidavit.”\\nLujan v. Nat’l Wildlife Fed’n, 497 U.S. 871, 888 (1990). For\\nthat reason, in Swanson Group Manufacturing LLC v. Jewell,\\n\\f 8\\n790 F.3d 235 (D.C. Cir. 2015), we held that a declaration\\nasserting only “economic loss and hardship,” which “told\\nnothing about the nature” of the loss, was insufficient to defeat\\na motion for summary judgment. Id. at 242–43 (quotation\\nmarks omitted). The same is true here—Dillenburg fails to\\nelaborate on the “direct economic benefit” that he allegedly has\\nlost because of the Board’s misdeeds.\\n\\n In concluding that Dillenburg proved standing, the district\\ncourt reasoned that Dillenburg, as a pork producer, “is affected\\nby the market price for pork.” Humane Society III, 290 F.\\nSupp. 3d at 16. Perhaps so, but Dillenburg’s declaration lacks\\nany indication that the “price for pork” was “affected” by the\\nalleged misuse of checkoff funds. The court assumed this\\ncritical fact based on Carpenters Industrial Council v. Zinke,\\n854 F.3d 1 (D.C. Cir. 2017). See Humane Society III, 290 F.\\nSupp. 3d at 16–17. Yet Carpenters did not displace settled law\\nthat courts cannot “presume the missing facts necessary to\\nestablish an element of standing.” Swanson, 790 F.3d at 240\\n(cleaned up). To the contrary, our decision carefully combed\\n“the complaint and declarations to assess whether the\\nplaintiff’s assertions suffice to show the elements of standing.”\\nCarpenters, 854 F.3d at 5. We explained that the declarations\\nin that case identified specific “harms ranging from lost sales\\nand diminished production to closures and layoffs.” Id. at 9.\\nAs discussed above, Dillenburg’s one-page declaration makes\\nno similar showing.\\n\\n The declaration suggests a second injury: Dillenburg\\n“oppose[s] the misuse of checkoff funds … for the purpose of\\ninfluencing legislation,” because he “believe[s]” those efforts\\nare “harmful to [him] and other independent farmers.” J.A.\\n147. This theory of injury also falls short. Plaintiffs cannot\\npredicate Article III standing on a “mere interest in a problem\\nor an ideological injury.” PETA v. USDA, 797 F.3d 1087, 1094\\n\\f 9\\n(D.C. Cir. 2015) (cleaned up). An “interest in the proper\\nadministration of the laws” is canonically “nonconcrete.”\\nSummers v. Earth Island Inst., 555 U.S. 488, 497 (2009)\\n(quotation marks omitted). Here, Dillenburg does not explain\\nhow his pork business is harmed by lobbying to promote the\\npork industry, let alone identify a specific lobbying activity that\\ncauses him a concrete injury. Instead, Dillenburg highlights\\nhis general opposition to the Council’s lobbying and the\\ngovernment’s funding of it. This declares “only a generally\\navailable grievance about government,” which “does not state\\nan Article III case or controversy.” Defs. of Wildlife, 504 U.S.\\nat 573–74.\\n\\n Finally, the plaintiffs cannot salvage this case based on the\\npossible standing of the Humane Society or Iowa Citizens. On\\nsummary judgment, neither organization filed any standing\\ndeclarations or affidavits. And now in this Court, the plaintiffs\\ndo not assert the standing of these organizations as an\\nalternative ground for reaching the merits. The plaintiffs thus\\nhave twice forfeited any claim that these organizations have\\nstanding. See Scenic Am., 836 F.3d at 53 n.4 (“Although a\\nparty cannot forfeit a claim that we lack jurisdiction, it can\\nforfeit a claim that we possess jurisdiction.”).\\n\\n * * * *\\n\\n Because the plaintiffs offer no evidence that the Board’s\\nalleged misuse of checkoff funds caused them to suffer an\\ninjury in fact, we vacate the district court’s order and remand\\nwith instructions to dismiss the case for lack of standing.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4187,"cells":{"id":{"kind":"number","value":4657748,"string":"4,657,748"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"In re: Sealed Case"},"decision_date":{"kind":"string","value":"2019-08-30"},"docket_number":{"kind":"string","value":"14-3043"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before: HENDERSON and GRIFFITH, Circuit Judges, and SENTELLE, Senior Circuit Judge.\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge GRIFFITH\", \"type\": \"010combined\", \"text\": \"PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n tnitrb $Iati !Inurt nf ta1\\n FOR TIlE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued October 6, 2016 Decided June 25, 2019\\n Reissued August 30, 2019\\n\\n No. 14-3043\\n\\n IN RE: SEALED CASE\\n\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No:1 : 10-cr-00334-1)\\n\\n\\n Before: HENDERSON and GRIFFITH, Circuit Judges, and\\nSENTELLE, Senior Circuit Judge.\\n\\n Opinion for the Court filed by Circuit Judge GRIFFITH.\\n\\n Opinion concurring in part and concurring in the judgment\\nfiled by Circuit Judge HENDERSON.\\n\\n GRIFFITH, Circuit Judge: Appellant pleaded guilty to\\nviolating 21 U.S.C. § 960a, which prohibits using the proceeds\\nof drug trafficking to support foreign terrorist groups. He now\\nappeals his conviction. This case requires us to address how the\\nForeign Commerce and Due Process Clauses of the\\nConstitution limit the extraterritorial application of the statute,\\nand to consider the effect of Appellant’s guilty plea on his\\n\\n NOTE: Portions of this opinion contain Sealed information,\\nwhich has been redacted.\\n\\n 2\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nability to appeal his conviction. We conclude that Congress had\\nthe authority to criminalize Appellant’s conduct even though\\nhis actions occurred outside of the United States, and that his\\nplea agreement precludes his other arguments on appeal.\\nAccordingly, we affirm Appellant’s conviction.\\n\\n I\\n\\n Because of the threat they posed to the security of the\\n United States and its citizens, between Year and Year the State\\n Department designated terrorist ‘zation\\n 1 and Orc’”’ization 2\\n :h groups\\n I\\nI •\\n\\n controlled terrItory In Country A involved in the drug trade.\\nAppellant was a key leader of an extensive criminal enterprise\\nthat produced and transported drugs in those areas, flying large\\n amounts of drugs from Country A to Country B, Country C,\\n Country D, and Country E. For example, while in Country B,\\nAppellant supervised the delivery of large quantities of drugs,\\n reloading the planes used to drop off the drugs with millions of\\n U.S. dollars and other currency that were then flown to Country\\nA. He knew some of that money was being paid as a “tax” to\\n Organization 1 and Organization 2 in exchange for permission\\nto operate in their territory, and that those groups were actively\\n engaging in terrorism. At one point, he also personally\\nparticipated in the delivery of money and weapons to\\n Organization 1.\\n\\n A federal grand jury indicted Appellant on a single count\\nof violating 21 U.S.C. § 960a. Broadly speaking, the statute\\ncriminalizes supporting terrorist groups with the proceeds of\\nillegal drttg trafficking. The statute has three parts. First, it\\nidentifies a set of drug-related activities that would be punished\\nunder a separate statute if they were committed within the\\n\\n 3\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\njurisdiction of the United States. The statute then specifies that\\nit prohibits engaging in those drug-related activities “knowing\\nor intending to provide, directly or indirectly, anything of\\npecuniary value to any person or organization that has engaged\\nor engages in terrorist activity.” 21 U.S.C. § 960a(a).\\n\\n Finally, the statute contains five jurisdictional provisions.\\nThe government need prove only one, and three are relevant\\nhere. Jurisdiction exists when either the terrorist offense or\\ndrug activity involved in the charged conduct would\\nindependently violate U.S. criminal law; when “the offense, the\\nprohibited drug activity, or the terrorist offense occurs in or\\naffects interstate or foreign commerce”; or when a defendant,\\nafter violating the substantive provisions of § 960a, “is brought\\ninto or found in the United States, even if the conduct required\\nfor the offense occurs outside the United States.” Id.\\n§ 960a(b)(1), (2), (5).2\\n Here, the government initially alleged that Appellant had\\nconspired to manufacture, distribute, and possess with intent to\\ndistribute a quantity of drugs. The government charged that\\nAppellant then paid Organization 1 and Organization 2 for\\npermission to operate in their territory using proceeds of that\\nconspiracy. The government asserted that three of the statute’s\\njurisdictional elements were met because “the prohibited drug\\n\\n 2\\n Jurisdiction also exists when the support given is for “a\\nterrorist offense that causes or is designed to cause death or serious\\nbodily injury to a national of the United States while that national is\\noutside the United States, or substantial damage to the property of a\\nlegal entity organized under the laws of the United States . while\\n . .\\n\\n\\n\\n\\nthat property is outside of the United States,” or when the drug or\\nterrorist activity giving rise to the offense occurs outside of the\\nUnited States and the defendant is a U.S. national or organization. 21\\nU.S.C. § 960a(b)(3), (4).\\n\\n 4\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nactivity and terrorist offense violate[d] the criminal laws of the\\nUnited States, specifically” 18 U.S.C. § 23393, which\\ncriminalizes the material support of foreign terrorist\\norganizations; “the prohibited drug activity and the terrorist\\noffenses occur[edj in or affect[ed] interstate or foreign\\ncommerce”; and Appellant was brought into the United States\\nafter the conduct that violated the statute occurred abroad.\\n\\n Appellant was arrested on Date and extradited from\\nCountry A. He was arraigned in federal district court in the\\nDistrict of Columbia on Date. In Month, Appellant moved to\\ndismiss the indictment, arguing that (1) the indictment did not\\nallege all the elements of the offense charged; (2) the\\nindictment failed to put him on adequate notice of the conduct\\nthat was allegedly criminal; (3) no nexus between the charged\\nconduct and the United States existed, meaning that his\\nprosecution violated the Due Process Clause; and (4) the\\njurisdictional reach of 960a exceeded Congress’s enumerated\\npowers, both facially and as-applied in this case.\\n\\n 5\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nnever moved to dismiss the information to which he\\npleaded guilty.\\n\\n As part of the plea agreement, Appellant also stipulated to\\na statement of facts detailing his participation in the scheme\\nand agreed that those facts “satisfljedJ each of the essential\\nelements of the charge to which [he] [was] entering [Jhis plea.”\\nHe expressly waived the right to appeal his sentence in all but\\na few limited circumstances not applicable here.\\n the district court sentenced Appellant to time in\\nprison.\\n\\n Appellant did not expressly waive his right to appeal his\\nconviction, and we must first determine whether his guilty plea\\nbars his appeal. After we held oral argument, the Supreme\\nCourt granted certiorari in Class v. United States, 137 5. Ct.\\n1065 (2017), a case that raised the same issue. We held\\nAppellant’s appeal in abeyance pending a decision in Class,\\nwhich has been announced. We requested supplemental\\nbriefing regarding the effect of that decision on Appellant’s\\nappeal.\\n\\n II\\n\\n Appellant challenges his conviction on three grounds.\\nfirst, he argues that Article I of the Constitution does not\\nauthorize Congress to pass a criminal statute with the\\nextraterritorial reach of § 960a. Second, he claims that due\\nprocess requires a greater connection between the punished\\nconduct and the United States than existed here. finally,\\n\\n 6\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nAppellant contends that his behavior does not satisfy some of\\nthe statute’s elements.3 We reject each of these challenges.\\n\\n A\\n\\n Generally, “[u]nconditional guilty pleas” like Appellant’s\\nthat “are knowing and intelligent” and do not reserve some\\nground for appeal in writing serve to “waive the pleading\\ndefendant[’s] claims of error on appeal, even constitutional\\nclaims.” United States v. Delgado—Garcia, 374 F.3d 1337,\\n1341 (D.C. Cir. 2004); see also FED. R. CRIM. P. 1 1(a)(2).\\nThere are limited exceptions to this rule, and Appellant\\nmaintains this case falls within one: the right to argue on appeal\\nthat a prosecution was defective from its inception, because the\\ndefendant could not lawfully have been “haled into court at\\nall.” Blackledge v. Perry, 417 U.S. 21, 30-31 (1974).\\n\\n The Supreme Court first articulated this exception in\\nBlackledge v. Periy and Menna v. New York, 423 U.S. 61\\n(1975). The Blackledge-Menna doctrine recognizes that there\\nare some cases in which the Constitution prohibits a\\nprosecution even though the defendant has admitted in a plea\\nagreement that he did exactly what the government alleged.\\n\\n In his opening brief, Appellant also argued that “[t]he\\nsuperseding information in this case appears to equate the statutory\\nlanguage ‘has engaged or engages in terrorist activity . or\\n . .\\n\\n\\n\\n\\nterrorism’ with presence on the State Department list of [Foreign\\nTerrorist Organizations].” Appellant Br. 21. Appellant suggested\\nthat constitutes a bill of attainder and “a violation of separation of\\npowers.” Id. at 2 1-22. After the government clarified in its brief that\\nit does not, in fact, consider listing by the State Department to be\\nconclusive of that particular statutory element, Gov’t Br. 32,\\nAppellant withdrew this aspect of his appeal, Appellant Reply Br.\\n28.\\n\\n 7\\n PUBLIC COPY - SEALED INFOR1\\\\’IATION DELETED\\n\\nMenna, 423 U.S. at 62 n.2. In such a case, a guilty plea will not\\nbar a constitutional challenge to the prosecution. Id For\\nexample, a guilty plea does not preclude “a claim that the\\ncharged offense violates the double jeopardy clause,”\\nDelgado—Garcia, 374 F.3d at 1341, or bar the argument that\\nthe prosecution was vindictive, see Blackledge, 417 U.S. at 30.\\n\\n The boundaries of the Blackledge-Menna doctrine were\\nrecently clarified in Class v. United States, 13$ S. Ct. 79$\\n(2018). In that case, the Supreme Court held that a guilty plea\\nwill never, “by itself,” bar a defendant from challenging the\\nconstitutionality of his statute of conviction, Id. at $03, because\\nsuch an argument “call[s] into question the {g]overnment’s\\npower to ‘constitutionally prosecute’ him,” Id. at $05 (quoting\\nUnited States v. Broce, 48$ U.S. 563, 575 (1989)). A guilty\\nplea does, however, bar a defendant from making arguments\\non appeal that contradict admissions made in his plea\\nagreement. Id. at 805 (holding that a defendant who pleads\\nguilty “relinquishes any claim that would contradict the\\n‘admissions necessarily made upon entry of a voluntary plea of\\nguilty” (quoting Broce, 488 U.S. at 573-74)).\\n\\n As the government concedes, Appellant’s constitutional\\narguments are largely of the type that, as Class made clear,\\nsurvive a guilty plea under the Blackledge-Menna doctrine.\\nGov’t Suppi. Br. 3-4. Appellant’s challenges to the potential\\nreach of Congress’s power under Article I and the restraints on\\nthat power under the Due Process Clause “call into question the\\n[g]overnment’s power to constitutionally prosecute him.”\\nClass, 138 5. Ct. at 805 (internal quotation marks omitted).\\nAnd, critically, his challenge to Congress’s authority to enact\\n§ 960a and his due process claim do not contradict any of the\\nadmissions made in his plea agreement. See id. at 804. Instead,\\n“[t]hey are consistent with Appellant’s knowing, voluntary,\\n\\n 8\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nand intelligent admission that he did what the [charging\\ndocument] alleged,” and “can be resolved without any need to\\nventure beyond that record.” Id. (internal quotation marks\\nomitted). That means that Appellant may still raise at least\\nthese arguments despite pleading guilty. Id. at 807.\\n\\n Appellant argues that our review is de novo. The\\ngovernment counters that his failure to obtain a ruling on his\\nmotion to dismiss means that our review is for plain error.\\nBecause the challenges not barred by Appellant’s guilty plea\\nfail even under de novo review, we need not resolve which\\nstandard applies.\\n\\n B\\n\\n The parties agree that congressional authority to enact\\n§ 960a must be found among the powers enumerated in Article\\nI, section 8. Appellant Suppl. Br. 9; Gov’t Suppl. Br. 5. They\\ndisagree whether such authority exists, although Appellant’s\\nreasons for doubting Congress’s power are not always clear. At\\ntimes he characterizes his Article I claim as a facial challenge\\nto the statute, arguing that § 960a “applies only\\nextraterritorially” and “[i]n every case where the statute is\\napplied extraterritorially, it is invalid as beyond the power of\\nCongress to legislate.” Appellant Suppi. Br. 7. Elsewhere he\\nframes his argument as turning on “the facts of the instant\\ncase,” id. at 11, arguing that Congress lacked the power to\\ncriminalize the funding of terrorist organizations with the\\nproceeds of drug trafficking when both the trafficking and\\nfunding “occur[] solely on foreign soil” and “do[] not affect\\nAmerican citizens or property,” Id. at 5. This second version of\\nhis argument suggests that the statute may constitutionally be\\napplied to some defendants—just not to Appellant, given what\\n\\n 9\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nhe characterizes as the attenuated connection between his\\nconduct and the United States.\\n\\n In light of this uncertainty, we assume for purposes of this\\nappeal that Appellant intended to bring both a facial and an as-\\napplied challenge to § 960a.4 The government responds to the\\nformer by disputing that every application of the statute is\\nunconstitutional, and to the latter by arguing that, even\\naccepting as true Appellant’s characterization of his actions’\\nconnection to the United States, Congress had the power to\\ncriminalize his behavior.\\n\\n 1\\n\\n\\n ‘\\n Our colleague understands Appellant to bring only a facial\\nchallenge to § 960a. We agree that Appellant frequently describes\\nhis argument in those terms. But, for the reasons given in Parts\\n11.3.1-2, Appellant disputes only certain applications of § 960a:\\nspecifically, applications of the statute to conduct that occurs in\\ncommerce between foreign countries that affects commerce with the\\nUnited States. Put another way, although the statute contains\\nnumerous jurisdictional provisions, the substance of Appellant’s\\nbriefing addresses part of one of them. This sort of “separate attack\\non a defined subset of the statute’s applications” is better read as an\\nas-applied challenge. United States v. Stevens, 559 U.S. 460, 473 n.3\\n(2010). Contrary to our colleague’s suggestion, in reaching this\\nconclusion we have not forgotten “the cardinal principle ofjudicial\\nrestraint” and decide no more than is necessary to resolve the case.\\nPost, at 2-3 (Henderson, J., concurring in part and concurring in the\\njudgment). We have a simple disagreement about how best to\\ncharacterize Appellant’s arguments. That disagreement implicates\\nno grand constitutional values, only different approaches to reading\\nbriefs and arguments. We simply see an as-applied challenge amidst\\nthe sometimes confusing jumble of Appellant’s arguments when our\\ncolleague does not.\\n\\n 10\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n To succeed on a facial challenge, a defendant must\\ngenerally “establish[] that no set of circumstances exists under\\nwhich the [statute] would be valid,’ i.e., that the law is\\nunconstitutional in all of its applications.” Wash. State Grange\\n Wash. State Republican Party, 552 U.S. 442, 449 (2008)\\n(quoting United States v. Salerno, 481 U.S. 739, 745 (1987)).\\nAlthough there has been some debate whether every facial\\nchallenge must meet the Salerno standard, “all agree that a\\nfacial challenge must fail where the statute has a plainly\\nlegitimate sweep.” Id. (internal quotation marks omitted);\\nHodge v. Talkin, 799 f.3d 1145, 1156 (D.C. Cir. 2015).\\n\\n That means that Appellant’s facial challenge cannot\\nsucceed if § 960a has a “plainly legitimate sweep.” We\\nconclude that it does. Indeed, Appellant does not contest (nor\\ncould he) that Congress acted within its Article I power when\\nproviding for the prosecution of persons meeting many of the\\ncriteria set forth in § 960a(b). For example, courts have long\\nrecognized that Congress may criminalize many actions that\\ntarget U.S. nationals or interests abroad, or provide for the\\nprosecution of much conduct by U.S. nationals or organizations\\noverseas. 21 U.S.C. § 960a(b)(3), (4); see, e.g., United States\\nv. Bowman, 260 U.S. 94, 98-99 (1922). Prior prosecutions\\nunder § 960a involved defendants who had engaged in such\\nconduct. See, e.g., United States v. Mohammed, 693 F.3d 192,\\n195-96 (D.C. Cir. 2012) (defendant trafficked drugs to support\\nthe Taliban, which was engaged in a conflict with the U.S.\\nmilitary in Afghanistan); United States v. Viglakis, No. 12-cr-\\n585, 2013 WL 4477023, at *2 (S.D.N.Y. Aug. 14, 2013)\\n(defendant supplied weapons to FARC, which “has engaged in\\nviolence against U.S. persons, commercial [interests,] and\\nproperty interests” located abroad); United States v. Saade, No.\\n11-cr-Ill, 2012 WL 2878087, at *2 (S.D.N.Y. July 11,2012)\\n(denying a motion to dismiss an indictment on due process\\n\\n 11\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ngrounds when the proceeds of the defendant’s drug trafficking\\n“would be used by the Taliban to buy weapons with which to\\nfight Americans”). Because Article I permits Congress to\\nauthorize these applications of § 960a, Appellant’s facial claim\\nfails.\\n\\n 2\\n\\n Appellant next argues that Article I does not authorize\\nCongress to criminalize the behavior in which he engaged.\\nUnder the Blackledge-Menna doctrine, that argument can only\\nsurvive Appellant’s guilty plea if it is consistent with the\\nfactual admissions Appellant made as part of his plea\\nagreement. Class, 138 5. Ct. at 804. Put another way, because\\nof what he already agreed to in his plea, Appellant may not\\nrelitigate before us what he did or did not do. His arguments on\\nappeal must be based on the facts established through his plea.\\nThe government and Appellant believe that to be the case here,\\nbut we disagree in part. Aspects of Appellant’s as-applied\\nchallenge ignore or revise facts to which he stipulated.\\n\\n As we noted in discussing Appellant’s facial challenge,\\nelements of § 960a(b) limit the extraterritorial reach of the\\nstatute. The government must find jurisdiction to prosecute\\nAppellant under at least one of those elements. Appellant\\nstipulated that the facts in this case fulfill three:\\n(1) Section 960a(b)(1), because the prohibited drug activity\\nand terrorist offense violated the criminal law of the United\\nStates that prohibits providing material support to foreign\\nterrorist groups, 1$ U.S.C. § 23393; (2) Section 960a(b)(2),\\nbecause the drug activity and terrorist offenses connected to\\nAppellant occurred in or affected interstate or foreign\\ncommerce; and (3) Section 960a(b)(5), because Appellant was\\nbrought to the United States for prosecution.\\n\\n 12\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n\\n Because parties may not stipulate to legal conclusions in\\nplea agreements, see United States v. Flores, 912 f.3d 613, 619\\n(D.C. Cir. 2019), Appellant’s plea then sets forth the specific\\nfacts that establish how his conduct fulfilled these elements.\\nRelevant here, the agreement describes how he provided\\nmoney and weapons to Organization 1 and Organization 2,\\n“which were designated Foreign Terrorist Organizations by the\\nUnited States Department of State during the period of the\\nconspiracy.” A Foreign Terrorist Organization is a “foreign\\norganization” that “engages in terrorist activity.. or terrorism\\n .\\n\\n\\n\\n\\n• . or retains the capability and intent to engage in terrorist\\n .\\n\\n\\n\\n\\nactivity or terrorism” that “threatens the security of United\\nStates nationals or the national security of the United States.”\\n8 U.S.C. § 1189. Indeed, Appellant stipulated that he knew that\\nOrganization 1 and Organization 2 were actively engaging in\\nterrorism while he was facilitating payments to them,\\nsometimes in weapons. Appellant also stipulated that the\\nmoney and weapons he supplied to Organization 1 and\\nOrganization 2 came from the large quantities of drugs he\\nmoved around the region. Taken together, Appellant admitted\\nthat he had trafficked millions of dollars’ worth of drugs and\\nweapons between multiple countries to fund groups engaging\\nin terrorism that threatened the security of U.S. nationals or the\\nnational security of the United States.\\n\\n Appellant now argues that Congress lacks the power under\\nArticle Ito authorize the application of § 960a to this admitted\\nconduct. In light of the admissions in Appellant’s plea\\nagreement, we disagree, and conclude that Congress had\\n\\n 13\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nauthority under the Foreign Commerce Clause to authorize\\nAppellant’s prosecution.5\\n\\n Appellant argues that his conduct was beyond the reach of\\nCongress’s power “[t]o regulate Commerce with foreign\\nNations” because it involved only “matters between two or\\nmore foreign countries not involving the United States,”\\nAppellant Suppl. Br. 9, and that his actions at most\\n“affect[ed]’ American commerce,” Appellant Reply Br. 18.\\nDoubt certainly exists as to Congress’s ability under the\\nForeign Commerce Clause to criminalize conduct with no\\neffect on the United States. See, e.g., United States v. Baston,\\n81$ f.3d 651, 667 (11th Cir. 2016) (observing that the Supreme\\nCourt has never “thoroughly explored the scope of the Foreign\\nCommerce Clause”); Baston v. United States, 137 5. Ct. $50,\\n$51 (2017) (Thomas, J., dissenting from denial of certiorari)\\n(observing that the Supreme Court has only evaluated laws\\ninvoking the foreign Commerce Clause to “regulat[e] conduct\\nwith a significant connection to the United States”). But see\\nHartford fire Ins. Co. v. Catfornia, 509 U.S. 764, 813-14\\n(1993) (Scalia, J., dissenting) (“Congress has broad power\\nunder [the Foreign Commerce Clause], and this Court has\\nrepeatedly upheld its power to make laws applicable to persons\\nor activities beyond our territorial boundaries where United\\nStates interests are affected.”); 3d. of Trs. of Univ. of Ill. v.\\nUnited States, 289 U.S. 48, 56 (1933) (describing as an\\n“essential attribute of the power” granted to Congress under the\\n\\n\\n Accordingly, we need not reach the government’s contention\\nthat Congress’s power under Article I, section 8 “[tb define and\\npunish Piracies and Felonies committed on the high Seas, and\\nOffenses against the Law of Nations” and “[t]o make all Laws which\\nshall be necessary and proper” to execute treaties justified\\napplication of 96Oato Appellant’s conduct.\\n\\n 14\\n PUBLIC COPY - SEALED INFORT’IATION DELETED\\n\\nForeign Commerce Clause “that it is exclusive and plenary”).\\nAfter all, the clause refers to “Commerce with foreign\\nNations,” lending support to the notion that some connection\\nto the United States must be established for Congress to\\nlegislate pursuant to this power. U.S. C0NsT. art. I, § 8, cI. 3\\n(emphasis added); Gibbons v. Ogden, 22 U.S. (9 Wheat) 1, 3\\n(1824) (“The power to regulate commerce extends to every\\nspecies of commercial intercourse between the United States\\nand foreign nations.” (emphasis added)); Anthony I.\\nColangelo, The foreign Commerce Clause, 96 VA. L. REv. 949\\n(2010) (arguing that the Foreign Commerce Clause authorizes\\nonly regulation of commerce with some nexus to the United\\nStates and that it may never authorize greater regulation of\\nactivities within foreign nations than would be authorized by\\nthe Interstate Commerce Clause within states).\\n\\n But Appellant’s argument goes further, asking us to hold\\nthat Congress can only make criminal a foreign commercial\\nactivity that takes place in the course of commerce with the\\nUnited States. Such a narrow construction would bar Congress\\nfrom outlawing drug trafficking that lends financial support to\\nterrorist organizations in foreign countries when that support\\nmerely “affects” commerce with the United States. Although\\nour sister circuits disagree on the precise test that should be\\nused to determine when overseas actions may be proscribed\\nunder the Foreign Commerce Clause, every Court of Appeals\\nto have considered the question in the context of commercial\\nactivity such as Appellant’s has disavowed his cramped\\nreading. Some circuits have used the familiar Interstate\\nCommerce Clause framework from United States v. Lopez, 514\\nU.S. 549, 558-59 (1995), and concluded that Congress has\\nbroad power to regulate overseas commercial conduct that has\\na “substantial effect” on commerce with the United States. See,\\ne.g., United States v. Durham, 902 F.3d 1180, 1192-93 (10th\\n\\n 15\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nCir. 201$); Baston, $18 f.3d at 667-68; United States v.\\nFendleton, 65$ F.3d 299, 308 (3d Cir. 2011). Some permit the\\nregulation of foreign conduct with less of an effect on the\\nUnited States. See, e.g., United States v. Bollinger, 79$ F.3d\\n201, 215-16 (4th Cir. 2015) (holding that Congress may\\nregulate conduct abroad that “demonstrably” rather than\\n“substantially” affects commerce with the United States);\\nUnited States v. Clark, 435 f.3d 1100, 1114 (9th Cir. 2006)\\n(requiring only a “constitutionally tenable nexus with foreign\\ncommerce”), partially superseded as noted in United States v.\\nFepe, 895 f.3d 679, 629 (9th Cir. 201$). But no court has\\nadopted the narrow view of congressional power Appellant\\nchampions. We find our sister circuits’ consistent adoption of\\nat least some sort of effects test to be persuasive. Cf RJR\\nNabisco, Inc. v. European Cmty., 136 S. Ct. 2090, 2105 (2016)\\n(holding that the term “foreign commerce” in the RICO statute\\nrestricts the reach of the law to enterprises that “engage in, or\\naffect in some significant way, commerce directly involving\\nthe United States—e.g., commerce between the United States\\nand a foreign country”).\\n\\n We need not, however, define today the precise level of\\n“effect” necessary, because the stipulations in Appellant’s plea\\nagreement place his conduct abroad within Congress’s reach\\nunder any version of the test. As already noted, his was not a\\nminor offense, the effects of which were only tangentially\\nrelated to the United States. Appellant was a leader of a large,\\ntransnational drug trafficking conspiracy that supported\\nestablished and powerful terrorist groups. Those groups\\nmurdered and kidnapped civilians, and the State Department\\nhad publicly concluded that they threatened U.S. nationals and\\ninterests. Appellant’s plea agreement does not specify where\\nthe drugs he trafficked were ultimately headed, but given the\\nquantity of drugs, the involvement of Organization 1, and the\\n\\n 16\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nnature of the Country A drug export markets in which\\nAppellant participated, his actions had a significant effect on\\nthe drug trade between the United States and the countries in\\n\\n\\n\\n I ant stipulated, moreover, that the terrorist\\norganizations he supported through his international drug\\ntrafficking threatened U.S. nationals or the national security of\\nthe United States. See Superseding Information at 2-3, No. 10-\\ncr-334 (D.D.C. Dec. 21, 2010), Dkt. No. 6 (stating that\\nOrganization 2 and Organization 1 “were designated Foreign\\nTerrorist Organizations by the United States Department of\\nState during the period of the conspiracy”); 2 U.S.C.\\n§ 1 189(a)(l)(C) (requiring that “the terrorist activity or\\nterrorism of’ an organization named as a foreign Terrorist\\nOrganization by the State Department “threaten[] the security\\nof United States nationals or the national security of the United\\nStates”). That magnifies the effect of his conduct on commerce\\nbetween the countries where he was operating and the United\\nStates. Cf $trassheim v. Daily, 221 U.S. 280, 225 (1911)\\n(“Acts done outside a jurisdiction, but intended to produce and\\nproducing detrimental effects within it, justify a state in\\npunishing the cause of the harm as if he had been present at the\\neffect, if the state should succeed in getting him within its\\npower.”).\\n\\n Appellant’s argument for why we should not apply any\\nversion of an effects test when evaluating Congress’s power\\nunder the foreign Commerce Clause is conclusory and without\\nmerit. He fails to address the reasoning given by those circuits\\nthat have considered the question or to explore the history of\\nthe clause. Instead, he simply relies on the bare assertion that a\\npower with such a weak limit would be too broad. Appellant\\n\\n 17\\n PUBLIC COPY - SEALED INFOR1vIATION DELETED\\n\\nReply Br. 18-19; Appellant Suppi. Br. 9. Given the Supreme\\nCourt’s longstanding approach to the Interstate Commerce\\nClause and the approach of our sister circuits to the Foreign\\nCommerce Clause, the need for the strict construction\\nAppellant urges is by no means self-evident. And once we\\nconclude that some version of an “effects test” applies,\\nAppellant’s remaining arguments are simply attempts to\\nrecharacterize the facts to which he stipulated, minimizing his\\nrole in the conspiracy and its magnitude to suggest that the\\neffect of his behavior on commerce with the United States was\\ntoo small for Congress to reach. Even after Class, we may not\\nconsider such arguments when, as here, they contradict the\\nadmissions made in the defendant’s plea agreement.\\n\\n The facts of Class illustrate why this aspect of Appellant’s\\nas-applied challenge is foreclosed by his plea agreement.\\nRodney Class pleaded guilty to possessing a firearm on the\\ngrounds of the U.S. Capitol. Class, 138 S. Ct. at 802; see also\\n40 U.S.C. § 5104(e)(1)(A)(i) (making it a crime to “carry on or\\nhave readily accessible to any individual on the Grounds or in\\nany of the Capitol Buildings a firearm”). The statute under\\nwhich he was convicted required proof of (1) carrying or\\nhaving readily accessible a firearm, dangerous weapon,\\nexplosives, or an incendiary device; and (2) presence “on the\\nGrounds or in any of the Capitol Buildings.” 40 U.S.C.\\n§ 5104(e)(1)(A)(i). As part of his plea, Class stipulated that\\nboth of these elements had been met because he had, in fact,\\npossessed a firearm on the Capitol Grounds when he drove his\\nvehicle into a parking lot near the Capitol. Class, 138 S. Ct. at\\n802, 804. On appeal, Class argued in relevant part “that he was\\ndenied fair notice that weapons were banned in the parking lot”\\nin violation of the Due Process Clause, both because no signs\\nexplained that the parking lot was an area in which firearms\\nwere prohibited and because the statute itself was so vague as\\n\\n 1$\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nto provide insufficient notice of what constituted the “Capitol\\nGrounds.” Id at $02; Id at 80$, $13 n.4 (Auto, J., dissenting).\\nCritically, neither his plea agreement nor the statute said\\nanything about his being aware that the parking lot was part of\\nthe Capitol Grounds.\\n\\n Because of that omission, the Supreme Court concluded\\nthat Class could press his due process claim on appeal because\\nit was “consistent with [his] knowing, voluntary, and intelligent\\nadmission that he did what the indictment alleged,” and could\\nbe resolved on appeal by reference to the existing record. Id. at\\n$04 (majority opinion) (internal quotation marks omitted).\\nAlthough Class had agreed that he was carrying weapons in a\\nprohibited place, he had not admitted to knowledge of the\\nnature of that place. Id. His argument that it violated the Due\\nProcess Clause to prosecute someone for unknowingly\\npossessing firearms in a prohibited location thus did not\\ncontradict any of the “admissions necessarily made” when he\\npleaded guilty. Id. at $05 (quoting Broce, 48$ U.S. at 573-74).\\n\\n This case more closely resembles a case that Class was\\nnot. Had Class admitted in his plea agreement that he knew he\\nwas in a place where possession of firearms was prohibited, he\\nwould not have been able to argue on appeal that he lacked\\nsufficient notice. Appellant made that kind of contradictory\\nadmission when he stipulated to the facts described above,\\nwhich demonstrate that his conduct affected commerce with\\nthe United States. As we concluded above, Appellant may\\nchallenge whether conduct that only “affects” foreign\\ncommerce with the United States may be prohibited by\\nCongress—just as Class, in our hypothetical, would remain\\nfree to claim that a statute that does not require notice violates\\nthe Due Process Clause. Appellant may not, however, argue\\n\\n 19\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nthat his conduct failed to affect foreign commerce without\\ncontradicting the terms of his plea agreement.6\\n\\n C\\n\\n Appellant next argues that his prosecution under § 960a,\\neven if authorized by Article I, violates the Due Process Clause\\nbecause the government has not established a sufficient nexus\\nbetween his conduct and the United States. We disagree.\\n\\n Appellant contends that Congress, subject to exceptions\\nnot applicable here, may only proscribe overseas conduct when\\na nexus exists between the conduct and the United States. We\\nhave repeatedly declined, however, to hold that the Due\\nProcess Clause demands such a nexus—or to even resolve\\n“whether the Due Process Clause constrains the extraterritorial\\n\\n\\n 6\\n Appellant’s claim reveals the difficulty faced by prosecutors\\nnegotiating plea agreements to resolve charges brought under\\nstatutes with jurisdictional elements such as § 960a. By pleading\\nguilty, Appellant prevented the government from putting on\\nevidence of his crime’s effect on foreign commerce with the United\\nStates. His appeal then posits that the tack of such evidence renders\\nhis prosecution unconstitutional. Prosecutors would be well-served\\nby ensuring that stipulations of fact entered into by pleading\\ndefendants include the factual bases for elements that bring the\\ndefendants’ conduct within the scope of Congress’s power to\\nlegislate, rather than simply reciting that such elements have been\\nmet. Cf Weston v. Wash. Metro. Area Transit Auth., 78 F.3d 682,\\n625 (D.C. Cir. 1996) (“While parties may enter into stipulations of\\nfact that are binding upon them unless they can show manifest\\ninjustice, parties may not stipulate to the legal conclusions to be\\nreached by the court.” (internal quotation marks and citations\\nomitted)). The stipulations of fact here, however, contained such\\nfacts.\\n\\n 20\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\napplication of federal criminal laws” at all. United States v.\\nBallestas, 795 F.3d 13$, 148 (D.C. Cir. 2015); see also United\\nStates v. Au, 718 f.3d 929, 943-44 & n.7 (D.C. Cir. 2013)\\n(declining to reach the question and observing that the court\\nhad “not found.. any case in which extraterritorial application\\n .\\n\\n\\n\\n\\nof a federal criminal statute was actually deemed a due process\\nviolation . .suggest[ing] [the defendant’s] burden is a heavy\\n .\\n\\n\\n\\n\\none”). Instead, “[t]he ultimate question under the Due Process\\nClause is not nexus, but is whether application of the statute to\\nthe defendant would be arbitrary or fundamentally unfair.”\\nBallestas, 795 F.3d at 148 (internal alterations and quotation\\nmarks omitted). “[T]he animating principle governing the due\\nprocess limits of extraterritorial jurisdiction,” to the extent such\\nlimits exist, “is the idea that ‘no man shall be held criminally\\nresponsible for conduct which he could not reasonably\\nunderstand to be proscribed.” Ali, 712 F.3d at 944 (quoting\\nBouie v. City of Columbia, 37$ U.S. 347, 351 (1964)).\\n\\n Appellant was a leader of an international drug trafficking\\norganization that funded and provided weapons to terrorist\\ngroups that openly murdered and kidnapped civilians. These\\ngroups, moreover, were publicly designated by the State\\nDepartment as directly threatening U.S. nationals or the\\nnational security of the United States. And various\\ninternational agreements specifically prohibit Appellant’s\\nconduct. See International Convention for the Suppression of\\nthe Financing of Terrorism, openedfor signature Dec. 9, 1999,\\n2178 U.N.T.S. 197 (entered into force Apr. 10, 2002); United\\nNations Convention Against Illicit Traffic in Narcotic Drugs\\nand Psychotropic Substances, opened for signature Dec. 20,\\n198$, 1582 U.N.T.S. 95 (entered into force Nov. 11, 1990).\\nAppellant can find no refuge in the Due Process Clause. He\\nadmitted, after all, that he engaged in conduct, proscribed by\\ninternational agreement, that supported groups that threaten\\n\\n 21\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nU.S. nationals or the national security of the United States. See\\nAll, 71$ F.3d at 944-45 (describing, on the basis of similar\\ninternational agreements, hostage taking as “an offense whose\\nproscription ‘is a result of universal condemnation of those\\nactivities and general interest in cooperating to suppress them”\\n(quoting RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW\\n§ 404 cmt. a (AM. LAW INST. 1987))). To apply § 960a to\\nAppellant “was neither arbitrary nor fundamentally unfair.”\\nBallestas, 795 f.3d at 148.\\n\\n III\\n\\n Appellant also argues that his conduct did not satisfy the\\njurisdictional elements of 21 U.S.C. § 960a(b), a defect in the\\ndistrict court’s subject mafterjurisdiction that cannot be waived\\nby his subsequent guilty plea. But Appellant misunderstands\\nthe work these elements of the statute perform.\\n\\n When used in statutory provisions like § 960a(b),\\njurisdiction can refer either to a “so-called ‘jurisdictional\\nelement’ addressing the reach of [Congress’s] legislative\\nauthority,” or “subject[ Jmatterj urisdiction—the jurisdiction of\\nthe ‘district courts.” United States v. Miranda, 780 F.3d 1185,\\n 1195 (D.C. Cir. 2015). The former, “consistent with the\\ndescription ‘jurisdictional element,’ treat[s] the relevant\\ncondition as an element of the offense to be found by a jury,”\\nand proof of such an element “is no different from proof of any\\nother element of a federal crime.” Id. (quoting Hugi v. United\\nStates, 164 F.3d 378, 381 (7th Cir. 1999)); see also Id. at 1191\\n(“[T]o ask ‘what conduct [a statute] reaches is to ask what\\nconduct [the statute] prohibits, which is a merits question,’ not\\na question of subject[ ]matter jurisdiction.” (quoting Morrison\\nv. Nat’l Austi. BankLtd., 561 U.S. 247, 254 (2010))); United\\nStates v. Gonzalez, 311 F.3d 440, 443 (1st Cir. 2002) (noting\\n\\n 22\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nthat Congress sometimes “use[s] the word ‘jurisdiction’\\nloosely to describe its own assertion of authority to regulate,”\\nsuch as when “it fixes an ‘affects interstate commerce’ . as a\\n . .\\n\\n\\n\\n\\ncondition of the crime,” and that these “conditions have\\nnothing whatever to do with the subject matter Uurisdiction] of\\nthe federal district court”). In contrast, subject matter\\njurisdiction “refers to a tribunal’s power to hear a case” and\\npresents a legal question. Miranda, 780 F.3d at 1191 (quoting\\nMorrison, 561 U.S. at 254).\\n\\n As discussed in Part 11.3, § 960a(a) refers to much conduct\\noccurring outside of the United States, and § 960a(b) puts\\nlimits on that extraterritorial sweep. “The extraterritorial reach\\nof a statute ordinarily presents a merits question, not a\\njurisdictional question.” Miranda, 780 F.3d at 1191. We\\ntherefore begin with the presumption that the provision does\\nnot speak to the subject matterjurisdiction of the federal courts.\\nWe will not, moreover, interpret “a threshold limitation on a\\nstatute’s scope” as referring to a court’s subject matter\\njurisdiction unless Congress “clearly states” as much. Arbaugh\\nv. Y&H Corp., 546 U.S. 500, 515 (2006). In arguing that\\nCongress has done so here, Appellant relies on Miranda, in\\nwhich this Court considered the Maritime Drug Law\\nEnforcement Act (MDLEA). The MDLEA prohibits drug\\nactivity on two categories of ships: “(1) a vessel of the United\\n States or a vessel subject to the jurisdiction of the United\\n States; or (2) any other vessel if the individual is a citizen of\\nthe United States or a resident alien of the United States.” 46\\nU.S.C. § 70503(a), (e). We concluded in Miranda that “subject\\nto the jurisdiction of the United States” represented a limit on\\nour subject matter jurisdiction, rather than an element of the\\noffense. 780 F.3d at 1191. Appellant argues that Miranda\\ncontrols the present case, and because § 960a(b) defines our\\n\\n 23\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nsubject matter jurisdiction, his challenge on this point cannot\\nbe waived by his plea.\\n\\n Miranda, however, is distinguishable based on the text of\\nthe statute at issue. The MDLEA provision defining\\n“[j]urisdiction of the United States with respect to a vessel\\nsubject to this chapter” explicitly states that such jurisdiction\\n“is not an element of an offense. Jurisdictional issues arising\\nunder this chapter are preliminary questions of law to be\\ndetermined solely by the trial judge.” 46 U.S.C. § 70504(a)\\n (emphasis added). Thus, the plain text prevented us from\\nfinding that “vessels subject to the jurisdiction of the United\\n States” represented ajurisdictional element because the inquiry\\nwas a “preliminary” or threshold issue—a key characteristic of\\nsubject matter jurisdiction. Miranda, 780 f.3d at 1193.\\nfurthermore, the text stated that the issue was “not an element\\nof the offense,” “fortifying its jurisdictional character.” Id. And\\nfinally, the text provided for the issue to be determined solely\\nby the trial judge as a question of law, just as “subject[ ]matter\\njurisdiction presents a question of law for resolution by the\\ncourt.” Id. Taken together, the MDLEA satisfied Arbaugh’s\\n“readily administrable bright line” rule that Congress “clearly\\nstate[] that a threshold limitation on a statute’s scope [would]\\ncount as jurisdictional.” 546 U.S. at 5 15-16.\\n\\n Section 960a(b), in contrast, lacks any of these features,\\nstating only that “{t]here is jurisdiction over an offense under\\nthis section if’ one of several conditions is met. And each of\\nthose conditions resembles the jurisdictional elements\\ncommonly found to restrict the scope of congressional, rather\\nthan judicial, power. Accordingly, “to ask what conduct\\n [ 960a] reaches is to ask what conduct [the statute] prohibits,\\nwhich is a merits question, not a question of subject matter\\njurisdiction.” Miranda, 780 F.3d at 1191 (internal quotation\\n\\n 24\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nmarks omitted); see also United States v. Harrington, 108 F.3d\\n1460, 1465 n.2 (D.C. Cir. 1997) (describing jurisdictional\\nelements that restrict the reach of statutes to Congress’s\\nenumerated powers by requiring factual findings, such as an\\neffect on interstate commerce). Appellant’s guilty plea thus\\nwaives his claim that his conduct does not meet the\\nrequirements of § 960a(b).\\n\\n\\n Iv\\n\\n For these reasons, we affirm Appellant’s conviction.\\n\\n So ordered.\\n\\n PUBLIC COPY - SEALED ENFORMATION DELETED\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435001/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge GRIFFITH\"}, {\"author\": \"KAREN LECRAFT HENDERSON, Circuit Judge, concurring\", \"type\": \"concurrence\", \"text\": \"KAREN LECRAFT HENDERSON, Circuit Judge, concurring\\nin part and concurring in the judgment:* I concur in all but\\nSection II.B.2 of the majority opinion. I cannot join Section\\nII.B.2 because I am convinced Appellant has not mounted an\\nas-applied Article I challenge.\\n\\n Appellant categorizes his Article I challenge as facial. See\\nAppellant Op. Br. 3, 9, 15, 16; Appellant Rep. Br. 2, 17;\\nAppellant Supp. Br. 2, 5, 7. So does the government. See\\nGov’t Resp. Br. iii, 21, 23; Gov’t Supp. Br. i, 5. And the\\nsubstance of Appellant’s challenge tracks the formal\\ncharacterization. Indeed, the challenge in his opening brief is\\nfacial. See Appellant Op. Br. 18 (“Since there is no power\\nenumerated in the Constitution which authorizes Congress to\\nenact 21 U.S.C. sec. 960a, it is unconstitutional on its face.”);\\nid. at 17 (“The Foreign commerce clause permits Congress to\\nregulate commerce with foreign nations, not between foreign\\nnations. 21 U.S.C. sec. 960a does not purport to regulate\\ncommerce between the United States and a foreign country.”\\n(internal citation omitted)); id. (“[T]he general foreign\\nauthority of the legislative branch does not authorize the\\nenactment of 21 U.S.C. sec. 960a.”); id. at 18 (“The Offenses\\nClause empowers Congress ‘to define and punish. Offenses\\n . .\\n\\n\\n\\nAgainst the Law of Nations.’ . To the extent that the term\\n . .\\n\\n\\n\\n‘Law of Nations’ has been expanded beyond piracy, that\\nexpansion [does not extend to] financing terrorist\\norganizations,” the conduct the statute prohibits.). And, of\\ncourse, Appellant was required to raise all of his arguments in\\nhis opening brief. See Al-Tarnimi v. Adelson, 916 f.3d 1, 6\\n(D.C. Cir. 2019) (“A party forfeits an argument by failing to\\nraise it in his opening brief”). Even were we willing to\\nconsider an as-applied challenge raised for the first time in his\\n\\n\\n *\\n NOTE: Portions of this opinion contain Sealed information,\\nwhich has been redacted.\\n\\n 2\\n PUBLIC COPY - SEALED INFORI\\\\’IATION DELETED\\n\\n supplemental brief—which, I submit, we should not be—there\\n is no as-applied challenge raised there. In his supplemental\\n brief, Appellant continues to argue that there is no\\n constitutional application of 21 U.S.C. § 960a. See Appellant\\n Supp. Br. at 7 (“In every case where the statute is applied\\n extraterritorially, it is invalid as beyond the power of Congress\\n to legislate. [And] the statute applies only extraterritorially.”).\\n The majority opinion references two pages in his supplemental\\n brief that might be read, at first blush, as unintentionally raising\\n an as-applied challenge. See Maj. Op. $ (citing Appellant\\n Supp. Br. 5, 11). Appellant’s language in both instances,\\n however, is consistent with a facial challenge. See Appellant\\n Supp. Br. 5 (Appellant’s argument that “21 U.S.C. sec. 960a is\\nfacially unconstitutional because there is no power granted to\\n Congress under Article I, Section 8 of the Constitution which\\n authorizes the extraterritorial reach of conduct occurring solely\\n on foreign soil and which does not affect American citizens or\\nproperty” is based on his (erroneous) belief that the statute\\nprohibits only conduct occurring solely on foreign soil, which\\n does not affect American citizens or property (emphasis\\n added)); Id. at 11 (Appellant’s reference to “the facts of the\\n instant case” is a response to the government’s attempt to\\n defeat his facial challenge by demonstrating that the statute is\\n constitutional as applied to him). Plainly, there are\\n constitutional applications of the statute. That is why we reject\\n Appellant’s facial constitutional challenge. See Maj. Op.\\n Section II.B. 1. But it is not a reason to treat his facial challenge\\n as as-applied. See Gov’t Supp. Br. 5 (arguing that Appellant\\n “did not” raise an as-applied challenge and that existence of\\n constitutional applications of statute defeats Appellant’s facial\\n challenge).\\n\\n Resolving only the facial Article I challenge would allow\\nus to avoid weighing in on the important constitutional\\n\\n 3\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nquestion discussed in Section II.B.2, which should be left for a\\nday when it is properly presented. I fear that, in their haste to\\nexplain how the foreign commerce clause works in the criminal\\ncontext, my colleagues have forgotten “the cardinal principle\\nof judicial restraint—if it is not necessary to decide more, it is\\nnecessary not to decide more.” Cohen v. Bd. of Trs. of Univ. of\\nD.C., 819 F.3d 476, 485 (D.C. Cir. 2016) (Griffith, I.)\\n(alteration omitted) (quoting FDK Labs, Inc. v. DEA, 362 F.3d\\n786, 799 (D.C. Cir. 2004) (Roberts, J., concurring in part and\\nconcurring in the judgment)). I have not forgotten that\\nprinciple and therefore do not join Section II.B.2 of the\\nmajority opinion.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435001/\", \"author_raw\": \"KAREN LECRAFT HENDERSON, Circuit Judge, concurring\"}]}"},"judge1":{"kind":"string","value":"HENDERSON"},"judge2":{"kind":"string","value":"GRIFFITH"},"judge3":{"kind":"string","value":"SENTELLE"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"HENDERSON\": \", Circuit\", \"GRIFFITH\": \", Circuit\", \"SENTELLE\": \", Senior Circuit\"}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":1,"string":"1"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4657748/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: In re: Sealed Case\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-30\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: Yes\nAdditional notes: \n\n[CASE_TEXT]\nPUBLIC COPY - SEALED INFORMATION DELETED\n\n\n tnitrb $Iati !Inurt nf ta1\n FOR TIlE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued October 6, 2016 Decided June 25, 2019\n Reissued August 30, 2019\n\n No. 14-3043\n\n IN RE: SEALED CASE\n\n\n\n Appeal from the United States District Court\n for the District of Columbia\n (No:1 : 10-cr-00334-1)\n\n\n Before: HENDERSON and GRIFFITH, Circuit Judges, and\nSENTELLE, Senior Circuit Judge.\n\n Opinion for the Court filed by Circuit Judge GRIFFITH.\n\n Opinion concurring in part and concurring in the judgment\nfiled by Circuit Judge HENDERSON.\n\n GRIFFITH, Circuit Judge: Appellant pleaded guilty to\nviolating 21 U.S.C. § 960a, which prohibits using the proceeds\nof drug trafficking to support foreign terrorist groups. He now\nappeals his conviction. This case requires us to address how the\nForeign Commerce and Due Process Clauses of the\nConstitution limit the extraterritorial application of the statute,\nand to consider the effect of Appellant’s guilty plea on his\n\n NOTE: Portions of this opinion contain Sealed information,\nwhich has been redacted.\n\n 2\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nability to appeal his conviction. We conclude that Congress had\nthe authority to criminalize Appellant’s conduct even though\nhis actions occurred outside of the United States, and that his\nplea agreement precludes his other arguments on appeal.\nAccordingly, we affirm Appellant’s conviction.\n\n I\n\n Because of the threat they posed to the security of the\n United States and its citizens, between Year and Year the State\n Department designated terrorist ‘zation\n 1 and Orc’”’ization 2\n :h groups\n I\nI •\n\n controlled terrItory In Country A involved in the drug trade.\nAppellant was a key leader of an extensive criminal enterprise\nthat produced and transported drugs in those areas, flying large\n amounts of drugs from Country A to Country B, Country C,\n Country D, and Country E. For example, while in Country B,\nAppellant supervised the delivery of large quantities of drugs,\n reloading the planes used to drop off the drugs with millions of\n U.S. dollars and other currency that were then flown to Country\nA. He knew some of that money was being paid as a “tax” to\n Organization 1 and Organization 2 in exchange for permission\nto operate in their territory, and that those groups were actively\n engaging in terrorism. At one point, he also personally\nparticipated in the delivery of money and weapons to\n Organization 1.\n\n A federal grand jury indicted Appellant on a single count\nof violating 21 U.S.C. § 960a. Broadly speaking, the statute\ncriminalizes supporting terrorist groups with the proceeds of\nillegal drttg trafficking. The statute has three parts. First, it\nidentifies a set of drug-related activities that would be punished\nunder a separate statute if they were committed within the\n\n 3\n PUBLIC COPY - SEALED INFORMATION DELETED\n\njurisdiction of the United States. The statute then specifies that\nit prohibits engaging in those drug-related activities “knowing\nor intending to provide, directly or indirectly, anything of\npecuniary value to any person or organization that has engaged\nor engages in terrorist activity.” 21 U.S.C. § 960a(a).\n\n Finally, the statute contains five jurisdictional provisions.\nThe government need prove only one, and three are relevant\nhere. Jurisdiction exists when either the terrorist offense or\ndrug activity involved in the charged conduct would\nindependently violate U.S. criminal law; when “the offense, the\nprohibited drug activity, or the terrorist offense occurs in or\naffects interstate or foreign commerce”; or when a defendant,\nafter violating the substantive provisions of § 960a, “is brought\ninto or found in the United States, even if the conduct required\nfor the offense occurs outside the United States.” Id.\n§ 960a(b)(1), (2), (5).2\n Here, the government initially alleged that Appellant had\nconspired to manufacture, distribute, and possess with intent to\ndistribute a quantity of drugs. The government charged that\nAppellant then paid Organization 1 and Organization 2 for\npermission to operate in their territory using proceeds of that\nconspiracy. The government asserted that three of the statute’s\njurisdictional elements were met because “the prohibited drug\n\n 2\n Jurisdiction also exists when the support given is for “a\nterrorist offense that causes or is designed to cause death or serious\nbodily injury to a national of the United States while that national is\noutside the United States, or substantial damage to the property of a\nlegal entity organized under the laws of the United States . while\n . .\n\n\n\n\nthat property is outside of the United States,” or when the drug or\nterrorist activity giving rise to the offense occurs outside of the\nUnited States and the defendant is a U.S. national or organization. 21\nU.S.C. § 960a(b)(3), (4).\n\n 4\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nactivity and terrorist offense violate[d] the criminal laws of the\nUnited States, specifically” 18 U.S.C. § 23393, which\ncriminalizes the material support of foreign terrorist\norganizations; “the prohibited drug activity and the terrorist\noffenses occur[edj in or affect[ed] interstate or foreign\ncommerce”; and Appellant was brought into the United States\nafter the conduct that violated the statute occurred abroad.\n\n Appellant was arrested on Date and extradited from\nCountry A. He was arraigned in federal district court in the\nDistrict of Columbia on Date. In Month, Appellant moved to\ndismiss the indictment, arguing that (1) the indictment did not\nallege all the elements of the offense charged; (2) the\nindictment failed to put him on adequate notice of the conduct\nthat was allegedly criminal; (3) no nexus between the charged\nconduct and the United States existed, meaning that his\nprosecution violated the Due Process Clause; and (4) the\njurisdictional reach of 960a exceeded Congress’s enumerated\npowers, both facially and as-applied in this case.\n\n 5\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nnever moved to dismiss the information to which he\npleaded guilty.\n\n As part of the plea agreement, Appellant also stipulated to\na statement of facts detailing his participation in the scheme\nand agreed that those facts “satisfljedJ each of the essential\nelements of the charge to which [he] [was] entering [Jhis plea.”\nHe expressly waived the right to appeal his sentence in all but\na few limited circumstances not applicable here.\n the district court sentenced Appellant to time in\nprison.\n\n Appellant did not expressly waive his right to appeal his\nconviction, and we must first determine whether his guilty plea\nbars his appeal. After we held oral argument, the Supreme\nCourt granted certiorari in Class v. United States, 137 5. Ct.\n1065 (2017), a case that raised the same issue. We held\nAppellant’s appeal in abeyance pending a decision in Class,\nwhich has been announced. We requested supplemental\nbriefing regarding the effect of that decision on Appellant’s\nappeal.\n\n II\n\n Appellant challenges his conviction on three grounds.\nfirst, he argues that Article I of the Constitution does not\nauthorize Congress to pass a criminal statute with the\nextraterritorial reach of § 960a. Second, he claims that due\nprocess requires a greater connection between the punished\nconduct and the United States than existed here. finally,\n\n 6\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nAppellant contends that his behavior does not satisfy some of\nthe statute’s elements.3 We reject each of these challenges.\n\n A\n\n Generally, “[u]nconditional guilty pleas” like Appellant’s\nthat “are knowing and intelligent” and do not reserve some\nground for appeal in writing serve to “waive the pleading\ndefendant[’s] claims of error on appeal, even constitutional\nclaims.” United States v. Delgado—Garcia, 374 F.3d 1337,\n1341 (D.C. Cir. 2004); see also FED. R. CRIM. P. 1 1(a)(2).\nThere are limited exceptions to this rule, and Appellant\nmaintains this case falls within one: the right to argue on appeal\nthat a prosecution was defective from its inception, because the\ndefendant could not lawfully have been “haled into court at\nall.” Blackledge v. Perry, 417 U.S. 21, 30-31 (1974).\n\n The Supreme Court first articulated this exception in\nBlackledge v. Periy and Menna v. New York, 423 U.S. 61\n(1975). The Blackledge-Menna doctrine recognizes that there\nare some cases in which the Constitution prohibits a\nprosecution even though the defendant has admitted in a plea\nagreement that he did exactly what the government alleged.\n\n In his opening brief, Appellant also argued that “[t]he\nsuperseding information in this case appears to equate the statutory\nlanguage ‘has engaged or engages in terrorist activity . or\n . .\n\n\n\n\nterrorism’ with presence on the State Department list of [Foreign\nTerrorist Organizations].” Appellant Br. 21. Appellant suggested\nthat constitutes a bill of attainder and “a violation of separation of\npowers.” Id. at 2 1-22. After the government clarified in its brief that\nit does not, in fact, consider listing by the State Department to be\nconclusive of that particular statutory element, Gov’t Br. 32,\nAppellant withdrew this aspect of his appeal, Appellant Reply Br.\n28.\n\n 7\n PUBLIC COPY - SEALED INFOR1\\’IATION DELETED\n\nMenna, 423 U.S. at 62 n.2. In such a case, a guilty plea will not\nbar a constitutional challenge to the prosecution. Id For\nexample, a guilty plea does not preclude “a claim that the\ncharged offense violates the double jeopardy clause,”\nDelgado—Garcia, 374 F.3d at 1341, or bar the argument that\nthe prosecution was vindictive, see Blackledge, 417 U.S. at 30.\n\n The boundaries of the Blackledge-Menna doctrine were\nrecently clarified in Class v. United States, 13$ S. Ct. 79$\n(2018). In that case, the Supreme Court held that a guilty plea\nwill never, “by itself,” bar a defendant from challenging the\nconstitutionality of his statute of conviction, Id. at $03, because\nsuch an argument “call[s] into question the {g]overnment’s\npower to ‘constitutionally prosecute’ him,” Id. at $05 (quoting\nUnited States v. Broce, 48$ U.S. 563, 575 (1989)). A guilty\nplea does, however, bar a defendant from making arguments\non appeal that contradict admissions made in his plea\nagreement. Id. at 805 (holding that a defendant who pleads\nguilty “relinquishes any claim that would contradict the\n‘admissions necessarily made upon entry of a voluntary plea of\nguilty” (quoting Broce, 488 U.S. at 573-74)).\n\n As the government concedes, Appellant’s constitutional\narguments are largely of the type that, as Class made clear,\nsurvive a guilty plea under the Blackledge-Menna doctrine.\nGov’t Suppi. Br. 3-4. Appellant’s challenges to the potential\nreach of Congress’s power under Article I and the restraints on\nthat power under the Due Process Clause “call into question the\n[g]overnment’s power to constitutionally prosecute him.”\nClass, 138 5. Ct. at 805 (internal quotation marks omitted).\nAnd, critically, his challenge to Congress’s authority to enact\n§ 960a and his due process claim do not contradict any of the\nadmissions made in his plea agreement. See id. at 804. Instead,\n“[t]hey are consistent with Appellant’s knowing, voluntary,\n\n 8\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nand intelligent admission that he did what the [charging\ndocument] alleged,” and “can be resolved without any need to\nventure beyond that record.” Id. (internal quotation marks\nomitted). That means that Appellant may still raise at least\nthese arguments despite pleading guilty. Id. at 807.\n\n Appellant argues that our review is de novo. The\ngovernment counters that his failure to obtain a ruling on his\nmotion to dismiss means that our review is for plain error.\nBecause the challenges not barred by Appellant’s guilty plea\nfail even under de novo review, we need not resolve which\nstandard applies.\n\n B\n\n The parties agree that congressional authority to enact\n§ 960a must be found among the powers enumerated in Article\nI, section 8. Appellant Suppl. Br. 9; Gov’t Suppl. Br. 5. They\ndisagree whether such authority exists, although Appellant’s\nreasons for doubting Congress’s power are not always clear. At\ntimes he characterizes his Article I claim as a facial challenge\nto the statute, arguing that § 960a “applies only\nextraterritorially” and “[i]n every case where the statute is\napplied extraterritorially, it is invalid as beyond the power of\nCongress to legislate.” Appellant Suppi. Br. 7. Elsewhere he\nframes his argument as turning on “the facts of the instant\ncase,” id. at 11, arguing that Congress lacked the power to\ncriminalize the funding of terrorist organizations with the\nproceeds of drug trafficking when both the trafficking and\nfunding “occur[] solely on foreign soil” and “do[] not affect\nAmerican citizens or property,” Id. at 5. This second version of\nhis argument suggests that the statute may constitutionally be\napplied to some defendants—just not to Appellant, given what\n\n 9\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nhe characterizes as the attenuated connection between his\nconduct and the United States.\n\n In light of this uncertainty, we assume for purposes of this\nappeal that Appellant intended to bring both a facial and an as-\napplied challenge to § 960a.4 The government responds to the\nformer by disputing that every application of the statute is\nunconstitutional, and to the latter by arguing that, even\naccepting as true Appellant’s characterization of his actions’\nconnection to the United States, Congress had the power to\ncriminalize his behavior.\n\n 1\n\n\n ‘\n Our colleague understands Appellant to bring only a facial\nchallenge to § 960a. We agree that Appellant frequently describes\nhis argument in those terms. But, for the reasons given in Parts\n11.3.1-2, Appellant disputes only certain applications of § 960a:\nspecifically, applications of the statute to conduct that occurs in\ncommerce between foreign countries that affects commerce with the\nUnited States. Put another way, although the statute contains\nnumerous jurisdictional provisions, the substance of Appellant’s\nbriefing addresses part of one of them. This sort of “separate attack\non a defined subset of the statute’s applications” is better read as an\nas-applied challenge. United States v. Stevens, 559 U.S. 460, 473 n.3\n(2010). Contrary to our colleague’s suggestion, in reaching this\nconclusion we have not forgotten “the cardinal principle ofjudicial\nrestraint” and decide no more than is necessary to resolve the case.\nPost, at 2-3 (Henderson, J., concurring in part and concurring in the\njudgment). We have a simple disagreement about how best to\ncharacterize Appellant’s arguments. That disagreement implicates\nno grand constitutional values, only different approaches to reading\nbriefs and arguments. We simply see an as-applied challenge amidst\nthe sometimes confusing jumble of Appellant’s arguments when our\ncolleague does not.\n\n 10\n PUBLIC COPY - SEALED INFORMATION DELETED\n\n To succeed on a facial challenge, a defendant must\ngenerally “establish[] that no set of circumstances exists under\nwhich the [statute] would be valid,’ i.e., that the law is\nunconstitutional in all of its applications.” Wash. State Grange\n Wash. State Republican Party, 552 U.S. 442, 449 (2008)\n(quoting United States v. Salerno, 481 U.S. 739, 745 (1987)).\nAlthough there has been some debate whether every facial\nchallenge must meet the Salerno standard, “all agree that a\nfacial challenge must fail where the statute has a plainly\nlegitimate sweep.” Id. (internal quotation marks omitted);\nHodge v. Talkin, 799 f.3d 1145, 1156 (D.C. Cir. 2015).\n\n That means that Appellant’s facial challenge cannot\nsucceed if § 960a has a “plainly legitimate sweep.” We\nconclude that it does. Indeed, Appellant does not contest (nor\ncould he) that Congress acted within its Article I power when\nproviding for the prosecution of persons meeting many of the\ncriteria set forth in § 960a(b). For example, courts have long\nrecognized that Congress may criminalize many actions that\ntarget U.S. nationals or interests abroad, or provide for the\nprosecution of much conduct by U.S. nationals or organizations\noverseas. 21 U.S.C. § 960a(b)(3), (4); see, e.g., United States\nv. Bowman, 260 U.S. 94, 98-99 (1922). Prior prosecutions\nunder § 960a involved defendants who had engaged in such\nconduct. See, e.g., United States v. Mohammed, 693 F.3d 192,\n195-96 (D.C. Cir. 2012) (defendant trafficked drugs to support\nthe Taliban, which was engaged in a conflict with the U.S.\nmilitary in Afghanistan); United States v. Viglakis, No. 12-cr-\n585, 2013 WL 4477023, at *2 (S.D.N.Y. Aug. 14, 2013)\n(defendant supplied weapons to FARC, which “has engaged in\nviolence against U.S. persons, commercial [interests,] and\nproperty interests” located abroad); United States v. Saade, No.\n11-cr-Ill, 2012 WL 2878087, at *2 (S.D.N.Y. July 11,2012)\n(denying a motion to dismiss an indictment on due process\n\n 11\n PUBLIC COPY - SEALED INFORMATION DELETED\n\ngrounds when the proceeds of the defendant’s drug trafficking\n“would be used by the Taliban to buy weapons with which to\nfight Americans”). Because Article I permits Congress to\nauthorize these applications of § 960a, Appellant’s facial claim\nfails.\n\n 2\n\n Appellant next argues that Article I does not authorize\nCongress to criminalize the behavior in which he engaged.\nUnder the Blackledge-Menna doctrine, that argument can only\nsurvive Appellant’s guilty plea if it is consistent with the\nfactual admissions Appellant made as part of his plea\nagreement. Class, 138 5. Ct. at 804. Put another way, because\nof what he already agreed to in his plea, Appellant may not\nrelitigate before us what he did or did not do. His arguments on\nappeal must be based on the facts established through his plea.\nThe government and Appellant believe that to be the case here,\nbut we disagree in part. Aspects of Appellant’s as-applied\nchallenge ignore or revise facts to which he stipulated.\n\n As we noted in discussing Appellant’s facial challenge,\nelements of § 960a(b) limit the extraterritorial reach of the\nstatute. The government must find jurisdiction to prosecute\nAppellant under at least one of those elements. Appellant\nstipulated that the facts in this case fulfill three:\n(1) Section 960a(b)(1), because the prohibited drug activity\nand terrorist offense violated the criminal law of the United\nStates that prohibits providing material support to foreign\nterrorist groups, 1$ U.S.C. § 23393; (2) Section 960a(b)(2),\nbecause the drug activity and terrorist offenses connected to\nAppellant occurred in or affected interstate or foreign\ncommerce; and (3) Section 960a(b)(5), because Appellant was\nbrought to the United States for prosecution.\n\n 12\n PUBLIC COPY - SEALED INFORMATION DELETED\n\n\n\n Because parties may not stipulate to legal conclusions in\nplea agreements, see United States v. Flores, 912 f.3d 613, 619\n(D.C. Cir. 2019), Appellant’s plea then sets forth the specific\nfacts that establish how his conduct fulfilled these elements.\nRelevant here, the agreement describes how he provided\nmoney and weapons to Organization 1 and Organization 2,\n“which were designated Foreign Terrorist Organizations by the\nUnited States Department of State during the period of the\nconspiracy.” A Foreign Terrorist Organization is a “foreign\norganization” that “engages in terrorist activity.. or terrorism\n .\n\n\n\n\n• . or retains the capability and intent to engage in terrorist\n .\n\n\n\n\nactivity or terrorism” that “threatens the security of United\nStates nationals or the national security of the United States.”\n8 U.S.C. § 1189. Indeed, Appellant stipulated that he knew that\nOrganization 1 and Organization 2 were actively engaging in\nterrorism while he was facilitating payments to them,\nsometimes in weapons. Appellant also stipulated that the\nmoney and weapons he supplied to Organization 1 and\nOrganization 2 came from the large quantities of drugs he\nmoved around the region. Taken together, Appellant admitted\nthat he had trafficked millions of dollars’ worth of drugs and\nweapons between multiple countries to fund groups engaging\nin terrorism that threatened the security of U.S. nationals or the\nnational security of the United States.\n\n Appellant now argues that Congress lacks the power under\nArticle Ito authorize the application of § 960a to this admitted\nconduct. In light of the admissions in Appellant’s plea\nagreement, we disagree, and conclude that Congress had\n\n 13\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nauthority under the Foreign Commerce Clause to authorize\nAppellant’s prosecution.5\n\n Appellant argues that his conduct was beyond the reach of\nCongress’s power “[t]o regulate Commerce with foreign\nNations” because it involved only “matters between two or\nmore foreign countries not involving the United States,”\nAppellant Suppl. Br. 9, and that his actions at most\n“affect[ed]’ American commerce,” Appellant Reply Br. 18.\nDoubt certainly exists as to Congress’s ability under the\nForeign Commerce Clause to criminalize conduct with no\neffect on the United States. See, e.g., United States v. Baston,\n81$ f.3d 651, 667 (11th Cir. 2016) (observing that the Supreme\nCourt has never “thoroughly explored the scope of the Foreign\nCommerce Clause”); Baston v. United States, 137 5. Ct. $50,\n$51 (2017) (Thomas, J., dissenting from denial of certiorari)\n(observing that the Supreme Court has only evaluated laws\ninvoking the foreign Commerce Clause to “regulat[e] conduct\nwith a significant connection to the United States”). But see\nHartford fire Ins. Co. v. Catfornia, 509 U.S. 764, 813-14\n(1993) (Scalia, J., dissenting) (“Congress has broad power\nunder [the Foreign Commerce Clause], and this Court has\nrepeatedly upheld its power to make laws applicable to persons\nor activities beyond our territorial boundaries where United\nStates interests are affected.”); 3d. of Trs. of Univ. of Ill. v.\nUnited States, 289 U.S. 48, 56 (1933) (describing as an\n“essential attribute of the power” granted to Congress under the\n\n\n Accordingly, we need not reach the government’s contention\nthat Congress’s power under Article I, section 8 “[tb define and\npunish Piracies and Felonies committed on the high Seas, and\nOffenses against the Law of Nations” and “[t]o make all Laws which\nshall be necessary and proper” to execute treaties justified\napplication of 96Oato Appellant’s conduct.\n\n 14\n PUBLIC COPY - SEALED INFORT’IATION DELETED\n\nForeign Commerce Clause “that it is exclusive and plenary”).\nAfter all, the clause refers to “Commerce with foreign\nNations,” lending support to the notion that some connection\nto the United States must be established for Congress to\nlegislate pursuant to this power. U.S. C0NsT. art. I, § 8, cI. 3\n(emphasis added); Gibbons v. Ogden, 22 U.S. (9 Wheat) 1, 3\n(1824) (“The power to regulate commerce extends to every\nspecies of commercial intercourse between the United States\nand foreign nations.” (emphasis added)); Anthony I.\nColangelo, The foreign Commerce Clause, 96 VA. L. REv. 949\n(2010) (arguing that the Foreign Commerce Clause authorizes\nonly regulation of commerce with some nexus to the United\nStates and that it may never authorize greater regulation of\nactivities within foreign nations than would be authorized by\nthe Interstate Commerce Clause within states).\n\n But Appellant’s argument goes further, asking us to hold\nthat Congress can only make criminal a foreign commercial\nactivity that takes place in the course of commerce with the\nUnited States. Such a narrow construction would bar Congress\nfrom outlawing drug trafficking that lends financial support to\nterrorist organizations in foreign countries when that support\nmerely “affects” commerce with the United States. Although\nour sister circuits disagree on the precise test that should be\nused to determine when overseas actions may be proscribed\nunder the Foreign Commerce Clause, every Court of Appeals\nto have considered the question in the context of commercial\nactivity such as Appellant’s has disavowed his cramped\nreading. Some circuits have used the familiar Interstate\nCommerce Clause framework from United States v. Lopez, 514\nU.S. 549, 558-59 (1995), and concluded that Congress has\nbroad power to regulate overseas commercial conduct that has\na “substantial effect” on commerce with the United States. See,\ne.g., United States v. Durham, 902 F.3d 1180, 1192-93 (10th\n\n 15\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nCir. 201$); Baston, $18 f.3d at 667-68; United States v.\nFendleton, 65$ F.3d 299, 308 (3d Cir. 2011). Some permit the\nregulation of foreign conduct with less of an effect on the\nUnited States. See, e.g., United States v. Bollinger, 79$ F.3d\n201, 215-16 (4th Cir. 2015) (holding that Congress may\nregulate conduct abroad that “demonstrably” rather than\n“substantially” affects commerce with the United States);\nUnited States v. Clark, 435 f.3d 1100, 1114 (9th Cir. 2006)\n(requiring only a “constitutionally tenable nexus with foreign\ncommerce”), partially superseded as noted in United States v.\nFepe, 895 f.3d 679, 629 (9th Cir. 201$). But no court has\nadopted the narrow view of congressional power Appellant\nchampions. We find our sister circuits’ consistent adoption of\nat least some sort of effects test to be persuasive. Cf RJR\nNabisco, Inc. v. European Cmty., 136 S. Ct. 2090, 2105 (2016)\n(holding that the term “foreign commerce” in the RICO statute\nrestricts the reach of the law to enterprises that “engage in, or\naffect in some significant way, commerce directly involving\nthe United States—e.g., commerce between the United States\nand a foreign country”).\n\n We need not, however, define today the precise level of\n“effect” necessary, because the stipulations in Appellant’s plea\nagreement place his conduct abroad within Congress’s reach\nunder any version of the test. As already noted, his was not a\nminor offense, the effects of which were only tangentially\nrelated to the United States. Appellant was a leader of a large,\ntransnational drug trafficking conspiracy that supported\nestablished and powerful terrorist groups. Those groups\nmurdered and kidnapped civilians, and the State Department\nhad publicly concluded that they threatened U.S. nationals and\ninterests. Appellant’s plea agreement does not specify where\nthe drugs he trafficked were ultimately headed, but given the\nquantity of drugs, the involvement of Organization 1, and the\n\n 16\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nnature of the Country A drug export markets in which\nAppellant participated, his actions had a significant effect on\nthe drug trade between the United States and the countries in\n\n\n\n I ant stipulated, moreover, that the terrorist\norganizations he supported through his international drug\ntrafficking threatened U.S. nationals or the national security of\nthe United States. See Superseding Information at 2-3, No. 10-\ncr-334 (D.D.C. Dec. 21, 2010), Dkt. No. 6 (stating that\nOrganization 2 and Organization 1 “were designated Foreign\nTerrorist Organizations by the United States Department of\nState during the period of the conspiracy”); 2 U.S.C.\n§ 1 189(a)(l)(C) (requiring that “the terrorist activity or\nterrorism of’ an organization named as a foreign Terrorist\nOrganization by the State Department “threaten[] the security\nof United States nationals or the national security of the United\nStates”). That magnifies the effect of his conduct on commerce\nbetween the countries where he was operating and the United\nStates. Cf $trassheim v. Daily, 221 U.S. 280, 225 (1911)\n(“Acts done outside a jurisdiction, but intended to produce and\nproducing detrimental effects within it, justify a state in\npunishing the cause of the harm as if he had been present at the\neffect, if the state should succeed in getting him within its\npower.”).\n\n Appellant’s argument for why we should not apply any\nversion of an effects test when evaluating Congress’s power\nunder the foreign Commerce Clause is conclusory and without\nmerit. He fails to address the reasoning given by those circuits\nthat have considered the question or to explore the history of\nthe clause. Instead, he simply relies on the bare assertion that a\npower with such a weak limit would be too broad. Appellant\n\n 17\n PUBLIC COPY - SEALED INFOR1vIATION DELETED\n\nReply Br. 18-19; Appellant Suppi. Br. 9. Given the Supreme\nCourt’s longstanding approach to the Interstate Commerce\nClause and the approach of our sister circuits to the Foreign\nCommerce Clause, the need for the strict construction\nAppellant urges is by no means self-evident. And once we\nconclude that some version of an “effects test” applies,\nAppellant’s remaining arguments are simply attempts to\nrecharacterize the facts to which he stipulated, minimizing his\nrole in the conspiracy and its magnitude to suggest that the\neffect of his behavior on commerce with the United States was\ntoo small for Congress to reach. Even after Class, we may not\nconsider such arguments when, as here, they contradict the\nadmissions made in the defendant’s plea agreement.\n\n The facts of Class illustrate why this aspect of Appellant’s\nas-applied challenge is foreclosed by his plea agreement.\nRodney Class pleaded guilty to possessing a firearm on the\ngrounds of the U.S. Capitol. Class, 138 S. Ct. at 802; see also\n40 U.S.C. § 5104(e)(1)(A)(i) (making it a crime to “carry on or\nhave readily accessible to any individual on the Grounds or in\nany of the Capitol Buildings a firearm”). The statute under\nwhich he was convicted required proof of (1) carrying or\nhaving readily accessible a firearm, dangerous weapon,\nexplosives, or an incendiary device; and (2) presence “on the\nGrounds or in any of the Capitol Buildings.” 40 U.S.C.\n§ 5104(e)(1)(A)(i). As part of his plea, Class stipulated that\nboth of these elements had been met because he had, in fact,\npossessed a firearm on the Capitol Grounds when he drove his\nvehicle into a parking lot near the Capitol. Class, 138 S. Ct. at\n802, 804. On appeal, Class argued in relevant part “that he was\ndenied fair notice that weapons were banned in the parking lot”\nin violation of the Due Process Clause, both because no signs\nexplained that the parking lot was an area in which firearms\nwere prohibited and because the statute itself was so vague as\n\n 1$\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nto provide insufficient notice of what constituted the “Capitol\nGrounds.” Id at $02; Id at 80$, $13 n.4 (Auto, J., dissenting).\nCritically, neither his plea agreement nor the statute said\nanything about his being aware that the parking lot was part of\nthe Capitol Grounds.\n\n Because of that omission, the Supreme Court concluded\nthat Class could press his due process claim on appeal because\nit was “consistent with [his] knowing, voluntary, and intelligent\nadmission that he did what the indictment alleged,” and could\nbe resolved on appeal by reference to the existing record. Id. at\n$04 (majority opinion) (internal quotation marks omitted).\nAlthough Class had agreed that he was carrying weapons in a\nprohibited place, he had not admitted to knowledge of the\nnature of that place. Id. His argument that it violated the Due\nProcess Clause to prosecute someone for unknowingly\npossessing firearms in a prohibited location thus did not\ncontradict any of the “admissions necessarily made” when he\npleaded guilty. Id. at $05 (quoting Broce, 48$ U.S. at 573-74).\n\n This case more closely resembles a case that Class was\nnot. Had Class admitted in his plea agreement that he knew he\nwas in a place where possession of firearms was prohibited, he\nwould not have been able to argue on appeal that he lacked\nsufficient notice. Appellant made that kind of contradictory\nadmission when he stipulated to the facts described above,\nwhich demonstrate that his conduct affected commerce with\nthe United States. As we concluded above, Appellant may\nchallenge whether conduct that only “affects” foreign\ncommerce with the United States may be prohibited by\nCongress—just as Class, in our hypothetical, would remain\nfree to claim that a statute that does not require notice violates\nthe Due Process Clause. Appellant may not, however, argue\n\n 19\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nthat his conduct failed to affect foreign commerce without\ncontradicting the terms of his plea agreement.6\n\n C\n\n Appellant next argues that his prosecution under § 960a,\neven if authorized by Article I, violates the Due Process Clause\nbecause the government has not established a sufficient nexus\nbetween his conduct and the United States. We disagree.\n\n Appellant contends that Congress, subject to exceptions\nnot applicable here, may only proscribe overseas conduct when\na nexus exists between the conduct and the United States. We\nhave repeatedly declined, however, to hold that the Due\nProcess Clause demands such a nexus—or to even resolve\n“whether the Due Process Clause constrains the extraterritorial\n\n\n 6\n Appellant’s claim reveals the difficulty faced by prosecutors\nnegotiating plea agreements to resolve charges brought under\nstatutes with jurisdictional elements such as § 960a. By pleading\nguilty, Appellant prevented the government from putting on\nevidence of his crime’s effect on foreign commerce with the United\nStates. His appeal then posits that the tack of such evidence renders\nhis prosecution unconstitutional. Prosecutors would be well-served\nby ensuring that stipulations of fact entered into by pleading\ndefendants include the factual bases for elements that bring the\ndefendants’ conduct within the scope of Congress’s power to\nlegislate, rather than simply reciting that such elements have been\nmet. Cf Weston v. Wash. Metro. Area Transit Auth., 78 F.3d 682,\n625 (D.C. Cir. 1996) (“While parties may enter into stipulations of\nfact that are binding upon them unless they can show manifest\ninjustice, parties may not stipulate to the legal conclusions to be\nreached by the court.” (internal quotation marks and citations\nomitted)). The stipulations of fact here, however, contained such\nfacts.\n\n 20\n PUBLIC COPY - SEALED INFORMATION DELETED\n\napplication of federal criminal laws” at all. United States v.\nBallestas, 795 F.3d 13$, 148 (D.C. Cir. 2015); see also United\nStates v. Au, 718 f.3d 929, 943-44 & n.7 (D.C. Cir. 2013)\n(declining to reach the question and observing that the court\nhad “not found.. any case in which extraterritorial application\n .\n\n\n\n\nof a federal criminal statute was actually deemed a due process\nviolation . .suggest[ing] [the defendant’s] burden is a heavy\n .\n\n\n\n\none”). Instead, “[t]he ultimate question under the Due Process\nClause is not nexus, but is whether application of the statute to\nthe defendant would be arbitrary or fundamentally unfair.”\nBallestas, 795 F.3d at 148 (internal alterations and quotation\nmarks omitted). “[T]he animating principle governing the due\nprocess limits of extraterritorial jurisdiction,” to the extent such\nlimits exist, “is the idea that ‘no man shall be held criminally\nresponsible for conduct which he could not reasonably\nunderstand to be proscribed.” Ali, 712 F.3d at 944 (quoting\nBouie v. City of Columbia, 37$ U.S. 347, 351 (1964)).\n\n Appellant was a leader of an international drug trafficking\norganization that funded and provided weapons to terrorist\ngroups that openly murdered and kidnapped civilians. These\ngroups, moreover, were publicly designated by the State\nDepartment as directly threatening U.S. nationals or the\nnational security of the United States. And various\ninternational agreements specifically prohibit Appellant’s\nconduct. See International Convention for the Suppression of\nthe Financing of Terrorism, openedfor signature Dec. 9, 1999,\n2178 U.N.T.S. 197 (entered into force Apr. 10, 2002); United\nNations Convention Against Illicit Traffic in Narcotic Drugs\nand Psychotropic Substances, opened for signature Dec. 20,\n198$, 1582 U.N.T.S. 95 (entered into force Nov. 11, 1990).\nAppellant can find no refuge in the Due Process Clause. He\nadmitted, after all, that he engaged in conduct, proscribed by\ninternational agreement, that supported groups that threaten\n\n 21\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nU.S. nationals or the national security of the United States. See\nAll, 71$ F.3d at 944-45 (describing, on the basis of similar\ninternational agreements, hostage taking as “an offense whose\nproscription ‘is a result of universal condemnation of those\nactivities and general interest in cooperating to suppress them”\n(quoting RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW\n§ 404 cmt. a (AM. LAW INST. 1987))). To apply § 960a to\nAppellant “was neither arbitrary nor fundamentally unfair.”\nBallestas, 795 f.3d at 148.\n\n III\n\n Appellant also argues that his conduct did not satisfy the\njurisdictional elements of 21 U.S.C. § 960a(b), a defect in the\ndistrict court’s subject mafterjurisdiction that cannot be waived\nby his subsequent guilty plea. But Appellant misunderstands\nthe work these elements of the statute perform.\n\n When used in statutory provisions like § 960a(b),\njurisdiction can refer either to a “so-called ‘jurisdictional\nelement’ addressing the reach of [Congress’s] legislative\nauthority,” or “subject[ Jmatterj urisdiction—the jurisdiction of\nthe ‘district courts.” United States v. Miranda, 780 F.3d 1185,\n 1195 (D.C. Cir. 2015). The former, “consistent with the\ndescription ‘jurisdictional element,’ treat[s] the relevant\ncondition as an element of the offense to be found by a jury,”\nand proof of such an element “is no different from proof of any\nother element of a federal crime.” Id. (quoting Hugi v. United\nStates, 164 F.3d 378, 381 (7th Cir. 1999)); see also Id. at 1191\n(“[T]o ask ‘what conduct [a statute] reaches is to ask what\nconduct [the statute] prohibits, which is a merits question,’ not\na question of subject[ ]matter jurisdiction.” (quoting Morrison\nv. Nat’l Austi. BankLtd., 561 U.S. 247, 254 (2010))); United\nStates v. Gonzalez, 311 F.3d 440, 443 (1st Cir. 2002) (noting\n\n 22\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nthat Congress sometimes “use[s] the word ‘jurisdiction’\nloosely to describe its own assertion of authority to regulate,”\nsuch as when “it fixes an ‘affects interstate commerce’ . as a\n . .\n\n\n\n\ncondition of the crime,” and that these “conditions have\nnothing whatever to do with the subject matter Uurisdiction] of\nthe federal district court”). In contrast, subject matter\njurisdiction “refers to a tribunal’s power to hear a case” and\npresents a legal question. Miranda, 780 F.3d at 1191 (quoting\nMorrison, 561 U.S. at 254).\n\n As discussed in Part 11.3, § 960a(a) refers to much conduct\noccurring outside of the United States, and § 960a(b) puts\nlimits on that extraterritorial sweep. “The extraterritorial reach\nof a statute ordinarily presents a merits question, not a\njurisdictional question.” Miranda, 780 F.3d at 1191. We\ntherefore begin with the presumption that the provision does\nnot speak to the subject matterjurisdiction of the federal courts.\nWe will not, moreover, interpret “a threshold limitation on a\nstatute’s scope” as referring to a court’s subject matter\njurisdiction unless Congress “clearly states” as much. Arbaugh\nv. Y&H Corp., 546 U.S. 500, 515 (2006). In arguing that\nCongress has done so here, Appellant relies on Miranda, in\nwhich this Court considered the Maritime Drug Law\nEnforcement Act (MDLEA). The MDLEA prohibits drug\nactivity on two categories of ships: “(1) a vessel of the United\n States or a vessel subject to the jurisdiction of the United\n States; or (2) any other vessel if the individual is a citizen of\nthe United States or a resident alien of the United States.” 46\nU.S.C. § 70503(a), (e). We concluded in Miranda that “subject\nto the jurisdiction of the United States” represented a limit on\nour subject matter jurisdiction, rather than an element of the\noffense. 780 F.3d at 1191. Appellant argues that Miranda\ncontrols the present case, and because § 960a(b) defines our\n\n 23\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nsubject matter jurisdiction, his challenge on this point cannot\nbe waived by his plea.\n\n Miranda, however, is distinguishable based on the text of\nthe statute at issue. The MDLEA provision defining\n“[j]urisdiction of the United States with respect to a vessel\nsubject to this chapter” explicitly states that such jurisdiction\n“is not an element of an offense. Jurisdictional issues arising\nunder this chapter are preliminary questions of law to be\ndetermined solely by the trial judge.” 46 U.S.C. § 70504(a)\n (emphasis added). Thus, the plain text prevented us from\nfinding that “vessels subject to the jurisdiction of the United\n States” represented ajurisdictional element because the inquiry\nwas a “preliminary” or threshold issue—a key characteristic of\nsubject matter jurisdiction. Miranda, 780 f.3d at 1193.\nfurthermore, the text stated that the issue was “not an element\nof the offense,” “fortifying its jurisdictional character.” Id. And\nfinally, the text provided for the issue to be determined solely\nby the trial judge as a question of law, just as “subject[ ]matter\njurisdiction presents a question of law for resolution by the\ncourt.” Id. Taken together, the MDLEA satisfied Arbaugh’s\n“readily administrable bright line” rule that Congress “clearly\nstate[] that a threshold limitation on a statute’s scope [would]\ncount as jurisdictional.” 546 U.S. at 5 15-16.\n\n Section 960a(b), in contrast, lacks any of these features,\nstating only that “{t]here is jurisdiction over an offense under\nthis section if’ one of several conditions is met. And each of\nthose conditions resembles the jurisdictional elements\ncommonly found to restrict the scope of congressional, rather\nthan judicial, power. Accordingly, “to ask what conduct\n [ 960a] reaches is to ask what conduct [the statute] prohibits,\nwhich is a merits question, not a question of subject matter\njurisdiction.” Miranda, 780 F.3d at 1191 (internal quotation\n\n 24\n PUBLIC COPY - SEALED INFORMATION DELETED\n\nmarks omitted); see also United States v. Harrington, 108 F.3d\n1460, 1465 n.2 (D.C. Cir. 1997) (describing jurisdictional\nelements that restrict the reach of statutes to Congress’s\nenumerated powers by requiring factual findings, such as an\neffect on interstate commerce). Appellant’s guilty plea thus\nwaives his claim that his conduct does not meet the\nrequirements of § 960a(b).\n\n\n Iv\n\n For these reasons, we affirm Appellant’s conviction.\n\n So ordered.\n\n PUBLIC COPY - SEALED ENFORMATION DELETED\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: In re: Sealed Case\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-30\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: Yes\\nAdditional notes: \\n\\n[CASE_TEXT]\\nPUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n tnitrb $Iati !Inurt nf ta1\\n FOR TIlE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued October 6, 2016 Decided June 25, 2019\\n Reissued August 30, 2019\\n\\n No. 14-3043\\n\\n IN RE: SEALED CASE\\n\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No:1 : 10-cr-00334-1)\\n\\n\\n Before: HENDERSON and GRIFFITH, Circuit Judges, and\\nSENTELLE, Senior Circuit Judge.\\n\\n Opinion for the Court filed by Circuit Judge GRIFFITH.\\n\\n Opinion concurring in part and concurring in the judgment\\nfiled by Circuit Judge HENDERSON.\\n\\n GRIFFITH, Circuit Judge: Appellant pleaded guilty to\\nviolating 21 U.S.C. § 960a, which prohibits using the proceeds\\nof drug trafficking to support foreign terrorist groups. He now\\nappeals his conviction. This case requires us to address how the\\nForeign Commerce and Due Process Clauses of the\\nConstitution limit the extraterritorial application of the statute,\\nand to consider the effect of Appellant’s guilty plea on his\\n\\n NOTE: Portions of this opinion contain Sealed information,\\nwhich has been redacted.\\n\\n 2\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nability to appeal his conviction. We conclude that Congress had\\nthe authority to criminalize Appellant’s conduct even though\\nhis actions occurred outside of the United States, and that his\\nplea agreement precludes his other arguments on appeal.\\nAccordingly, we affirm Appellant’s conviction.\\n\\n I\\n\\n Because of the threat they posed to the security of the\\n United States and its citizens, between Year and Year the State\\n Department designated terrorist ‘zation\\n 1 and Orc’”’ization 2\\n :h groups\\n I\\nI •\\n\\n controlled terrItory In Country A involved in the drug trade.\\nAppellant was a key leader of an extensive criminal enterprise\\nthat produced and transported drugs in those areas, flying large\\n amounts of drugs from Country A to Country B, Country C,\\n Country D, and Country E. For example, while in Country B,\\nAppellant supervised the delivery of large quantities of drugs,\\n reloading the planes used to drop off the drugs with millions of\\n U.S. dollars and other currency that were then flown to Country\\nA. He knew some of that money was being paid as a “tax” to\\n Organization 1 and Organization 2 in exchange for permission\\nto operate in their territory, and that those groups were actively\\n engaging in terrorism. At one point, he also personally\\nparticipated in the delivery of money and weapons to\\n Organization 1.\\n\\n A federal grand jury indicted Appellant on a single count\\nof violating 21 U.S.C. § 960a. Broadly speaking, the statute\\ncriminalizes supporting terrorist groups with the proceeds of\\nillegal drttg trafficking. The statute has three parts. First, it\\nidentifies a set of drug-related activities that would be punished\\nunder a separate statute if they were committed within the\\n\\n 3\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\njurisdiction of the United States. The statute then specifies that\\nit prohibits engaging in those drug-related activities “knowing\\nor intending to provide, directly or indirectly, anything of\\npecuniary value to any person or organization that has engaged\\nor engages in terrorist activity.” 21 U.S.C. § 960a(a).\\n\\n Finally, the statute contains five jurisdictional provisions.\\nThe government need prove only one, and three are relevant\\nhere. Jurisdiction exists when either the terrorist offense or\\ndrug activity involved in the charged conduct would\\nindependently violate U.S. criminal law; when “the offense, the\\nprohibited drug activity, or the terrorist offense occurs in or\\naffects interstate or foreign commerce”; or when a defendant,\\nafter violating the substantive provisions of § 960a, “is brought\\ninto or found in the United States, even if the conduct required\\nfor the offense occurs outside the United States.” Id.\\n§ 960a(b)(1), (2), (5).2\\n Here, the government initially alleged that Appellant had\\nconspired to manufacture, distribute, and possess with intent to\\ndistribute a quantity of drugs. The government charged that\\nAppellant then paid Organization 1 and Organization 2 for\\npermission to operate in their territory using proceeds of that\\nconspiracy. The government asserted that three of the statute’s\\njurisdictional elements were met because “the prohibited drug\\n\\n 2\\n Jurisdiction also exists when the support given is for “a\\nterrorist offense that causes or is designed to cause death or serious\\nbodily injury to a national of the United States while that national is\\noutside the United States, or substantial damage to the property of a\\nlegal entity organized under the laws of the United States . while\\n . .\\n\\n\\n\\n\\nthat property is outside of the United States,” or when the drug or\\nterrorist activity giving rise to the offense occurs outside of the\\nUnited States and the defendant is a U.S. national or organization. 21\\nU.S.C. § 960a(b)(3), (4).\\n\\n 4\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nactivity and terrorist offense violate[d] the criminal laws of the\\nUnited States, specifically” 18 U.S.C. § 23393, which\\ncriminalizes the material support of foreign terrorist\\norganizations; “the prohibited drug activity and the terrorist\\noffenses occur[edj in or affect[ed] interstate or foreign\\ncommerce”; and Appellant was brought into the United States\\nafter the conduct that violated the statute occurred abroad.\\n\\n Appellant was arrested on Date and extradited from\\nCountry A. He was arraigned in federal district court in the\\nDistrict of Columbia on Date. In Month, Appellant moved to\\ndismiss the indictment, arguing that (1) the indictment did not\\nallege all the elements of the offense charged; (2) the\\nindictment failed to put him on adequate notice of the conduct\\nthat was allegedly criminal; (3) no nexus between the charged\\nconduct and the United States existed, meaning that his\\nprosecution violated the Due Process Clause; and (4) the\\njurisdictional reach of 960a exceeded Congress’s enumerated\\npowers, both facially and as-applied in this case.\\n\\n 5\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nnever moved to dismiss the information to which he\\npleaded guilty.\\n\\n As part of the plea agreement, Appellant also stipulated to\\na statement of facts detailing his participation in the scheme\\nand agreed that those facts “satisfljedJ each of the essential\\nelements of the charge to which [he] [was] entering [Jhis plea.”\\nHe expressly waived the right to appeal his sentence in all but\\na few limited circumstances not applicable here.\\n the district court sentenced Appellant to time in\\nprison.\\n\\n Appellant did not expressly waive his right to appeal his\\nconviction, and we must first determine whether his guilty plea\\nbars his appeal. After we held oral argument, the Supreme\\nCourt granted certiorari in Class v. United States, 137 5. Ct.\\n1065 (2017), a case that raised the same issue. We held\\nAppellant’s appeal in abeyance pending a decision in Class,\\nwhich has been announced. We requested supplemental\\nbriefing regarding the effect of that decision on Appellant’s\\nappeal.\\n\\n II\\n\\n Appellant challenges his conviction on three grounds.\\nfirst, he argues that Article I of the Constitution does not\\nauthorize Congress to pass a criminal statute with the\\nextraterritorial reach of § 960a. Second, he claims that due\\nprocess requires a greater connection between the punished\\nconduct and the United States than existed here. finally,\\n\\n 6\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nAppellant contends that his behavior does not satisfy some of\\nthe statute’s elements.3 We reject each of these challenges.\\n\\n A\\n\\n Generally, “[u]nconditional guilty pleas” like Appellant’s\\nthat “are knowing and intelligent” and do not reserve some\\nground for appeal in writing serve to “waive the pleading\\ndefendant[’s] claims of error on appeal, even constitutional\\nclaims.” United States v. Delgado—Garcia, 374 F.3d 1337,\\n1341 (D.C. Cir. 2004); see also FED. R. CRIM. P. 1 1(a)(2).\\nThere are limited exceptions to this rule, and Appellant\\nmaintains this case falls within one: the right to argue on appeal\\nthat a prosecution was defective from its inception, because the\\ndefendant could not lawfully have been “haled into court at\\nall.” Blackledge v. Perry, 417 U.S. 21, 30-31 (1974).\\n\\n The Supreme Court first articulated this exception in\\nBlackledge v. Periy and Menna v. New York, 423 U.S. 61\\n(1975). The Blackledge-Menna doctrine recognizes that there\\nare some cases in which the Constitution prohibits a\\nprosecution even though the defendant has admitted in a plea\\nagreement that he did exactly what the government alleged.\\n\\n In his opening brief, Appellant also argued that “[t]he\\nsuperseding information in this case appears to equate the statutory\\nlanguage ‘has engaged or engages in terrorist activity . or\\n . .\\n\\n\\n\\n\\nterrorism’ with presence on the State Department list of [Foreign\\nTerrorist Organizations].” Appellant Br. 21. Appellant suggested\\nthat constitutes a bill of attainder and “a violation of separation of\\npowers.” Id. at 2 1-22. After the government clarified in its brief that\\nit does not, in fact, consider listing by the State Department to be\\nconclusive of that particular statutory element, Gov’t Br. 32,\\nAppellant withdrew this aspect of his appeal, Appellant Reply Br.\\n28.\\n\\n 7\\n PUBLIC COPY - SEALED INFOR1\\\\’IATION DELETED\\n\\nMenna, 423 U.S. at 62 n.2. In such a case, a guilty plea will not\\nbar a constitutional challenge to the prosecution. Id For\\nexample, a guilty plea does not preclude “a claim that the\\ncharged offense violates the double jeopardy clause,”\\nDelgado—Garcia, 374 F.3d at 1341, or bar the argument that\\nthe prosecution was vindictive, see Blackledge, 417 U.S. at 30.\\n\\n The boundaries of the Blackledge-Menna doctrine were\\nrecently clarified in Class v. United States, 13$ S. Ct. 79$\\n(2018). In that case, the Supreme Court held that a guilty plea\\nwill never, “by itself,” bar a defendant from challenging the\\nconstitutionality of his statute of conviction, Id. at $03, because\\nsuch an argument “call[s] into question the {g]overnment’s\\npower to ‘constitutionally prosecute’ him,” Id. at $05 (quoting\\nUnited States v. Broce, 48$ U.S. 563, 575 (1989)). A guilty\\nplea does, however, bar a defendant from making arguments\\non appeal that contradict admissions made in his plea\\nagreement. Id. at 805 (holding that a defendant who pleads\\nguilty “relinquishes any claim that would contradict the\\n‘admissions necessarily made upon entry of a voluntary plea of\\nguilty” (quoting Broce, 488 U.S. at 573-74)).\\n\\n As the government concedes, Appellant’s constitutional\\narguments are largely of the type that, as Class made clear,\\nsurvive a guilty plea under the Blackledge-Menna doctrine.\\nGov’t Suppi. Br. 3-4. Appellant’s challenges to the potential\\nreach of Congress’s power under Article I and the restraints on\\nthat power under the Due Process Clause “call into question the\\n[g]overnment’s power to constitutionally prosecute him.”\\nClass, 138 5. Ct. at 805 (internal quotation marks omitted).\\nAnd, critically, his challenge to Congress’s authority to enact\\n§ 960a and his due process claim do not contradict any of the\\nadmissions made in his plea agreement. See id. at 804. Instead,\\n“[t]hey are consistent with Appellant’s knowing, voluntary,\\n\\n 8\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nand intelligent admission that he did what the [charging\\ndocument] alleged,” and “can be resolved without any need to\\nventure beyond that record.” Id. (internal quotation marks\\nomitted). That means that Appellant may still raise at least\\nthese arguments despite pleading guilty. Id. at 807.\\n\\n Appellant argues that our review is de novo. The\\ngovernment counters that his failure to obtain a ruling on his\\nmotion to dismiss means that our review is for plain error.\\nBecause the challenges not barred by Appellant’s guilty plea\\nfail even under de novo review, we need not resolve which\\nstandard applies.\\n\\n B\\n\\n The parties agree that congressional authority to enact\\n§ 960a must be found among the powers enumerated in Article\\nI, section 8. Appellant Suppl. Br. 9; Gov’t Suppl. Br. 5. They\\ndisagree whether such authority exists, although Appellant’s\\nreasons for doubting Congress’s power are not always clear. At\\ntimes he characterizes his Article I claim as a facial challenge\\nto the statute, arguing that § 960a “applies only\\nextraterritorially” and “[i]n every case where the statute is\\napplied extraterritorially, it is invalid as beyond the power of\\nCongress to legislate.” Appellant Suppi. Br. 7. Elsewhere he\\nframes his argument as turning on “the facts of the instant\\ncase,” id. at 11, arguing that Congress lacked the power to\\ncriminalize the funding of terrorist organizations with the\\nproceeds of drug trafficking when both the trafficking and\\nfunding “occur[] solely on foreign soil” and “do[] not affect\\nAmerican citizens or property,” Id. at 5. This second version of\\nhis argument suggests that the statute may constitutionally be\\napplied to some defendants—just not to Appellant, given what\\n\\n 9\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nhe characterizes as the attenuated connection between his\\nconduct and the United States.\\n\\n In light of this uncertainty, we assume for purposes of this\\nappeal that Appellant intended to bring both a facial and an as-\\napplied challenge to § 960a.4 The government responds to the\\nformer by disputing that every application of the statute is\\nunconstitutional, and to the latter by arguing that, even\\naccepting as true Appellant’s characterization of his actions’\\nconnection to the United States, Congress had the power to\\ncriminalize his behavior.\\n\\n 1\\n\\n\\n ‘\\n Our colleague understands Appellant to bring only a facial\\nchallenge to § 960a. We agree that Appellant frequently describes\\nhis argument in those terms. But, for the reasons given in Parts\\n11.3.1-2, Appellant disputes only certain applications of § 960a:\\nspecifically, applications of the statute to conduct that occurs in\\ncommerce between foreign countries that affects commerce with the\\nUnited States. Put another way, although the statute contains\\nnumerous jurisdictional provisions, the substance of Appellant’s\\nbriefing addresses part of one of them. This sort of “separate attack\\non a defined subset of the statute’s applications” is better read as an\\nas-applied challenge. United States v. Stevens, 559 U.S. 460, 473 n.3\\n(2010). Contrary to our colleague’s suggestion, in reaching this\\nconclusion we have not forgotten “the cardinal principle ofjudicial\\nrestraint” and decide no more than is necessary to resolve the case.\\nPost, at 2-3 (Henderson, J., concurring in part and concurring in the\\njudgment). We have a simple disagreement about how best to\\ncharacterize Appellant’s arguments. That disagreement implicates\\nno grand constitutional values, only different approaches to reading\\nbriefs and arguments. We simply see an as-applied challenge amidst\\nthe sometimes confusing jumble of Appellant’s arguments when our\\ncolleague does not.\\n\\n 10\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n To succeed on a facial challenge, a defendant must\\ngenerally “establish[] that no set of circumstances exists under\\nwhich the [statute] would be valid,’ i.e., that the law is\\nunconstitutional in all of its applications.” Wash. State Grange\\n Wash. State Republican Party, 552 U.S. 442, 449 (2008)\\n(quoting United States v. Salerno, 481 U.S. 739, 745 (1987)).\\nAlthough there has been some debate whether every facial\\nchallenge must meet the Salerno standard, “all agree that a\\nfacial challenge must fail where the statute has a plainly\\nlegitimate sweep.” Id. (internal quotation marks omitted);\\nHodge v. Talkin, 799 f.3d 1145, 1156 (D.C. Cir. 2015).\\n\\n That means that Appellant’s facial challenge cannot\\nsucceed if § 960a has a “plainly legitimate sweep.” We\\nconclude that it does. Indeed, Appellant does not contest (nor\\ncould he) that Congress acted within its Article I power when\\nproviding for the prosecution of persons meeting many of the\\ncriteria set forth in § 960a(b). For example, courts have long\\nrecognized that Congress may criminalize many actions that\\ntarget U.S. nationals or interests abroad, or provide for the\\nprosecution of much conduct by U.S. nationals or organizations\\noverseas. 21 U.S.C. § 960a(b)(3), (4); see, e.g., United States\\nv. Bowman, 260 U.S. 94, 98-99 (1922). Prior prosecutions\\nunder § 960a involved defendants who had engaged in such\\nconduct. See, e.g., United States v. Mohammed, 693 F.3d 192,\\n195-96 (D.C. Cir. 2012) (defendant trafficked drugs to support\\nthe Taliban, which was engaged in a conflict with the U.S.\\nmilitary in Afghanistan); United States v. Viglakis, No. 12-cr-\\n585, 2013 WL 4477023, at *2 (S.D.N.Y. Aug. 14, 2013)\\n(defendant supplied weapons to FARC, which “has engaged in\\nviolence against U.S. persons, commercial [interests,] and\\nproperty interests” located abroad); United States v. Saade, No.\\n11-cr-Ill, 2012 WL 2878087, at *2 (S.D.N.Y. July 11,2012)\\n(denying a motion to dismiss an indictment on due process\\n\\n 11\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\ngrounds when the proceeds of the defendant’s drug trafficking\\n“would be used by the Taliban to buy weapons with which to\\nfight Americans”). Because Article I permits Congress to\\nauthorize these applications of § 960a, Appellant’s facial claim\\nfails.\\n\\n 2\\n\\n Appellant next argues that Article I does not authorize\\nCongress to criminalize the behavior in which he engaged.\\nUnder the Blackledge-Menna doctrine, that argument can only\\nsurvive Appellant’s guilty plea if it is consistent with the\\nfactual admissions Appellant made as part of his plea\\nagreement. Class, 138 5. Ct. at 804. Put another way, because\\nof what he already agreed to in his plea, Appellant may not\\nrelitigate before us what he did or did not do. His arguments on\\nappeal must be based on the facts established through his plea.\\nThe government and Appellant believe that to be the case here,\\nbut we disagree in part. Aspects of Appellant’s as-applied\\nchallenge ignore or revise facts to which he stipulated.\\n\\n As we noted in discussing Appellant’s facial challenge,\\nelements of § 960a(b) limit the extraterritorial reach of the\\nstatute. The government must find jurisdiction to prosecute\\nAppellant under at least one of those elements. Appellant\\nstipulated that the facts in this case fulfill three:\\n(1) Section 960a(b)(1), because the prohibited drug activity\\nand terrorist offense violated the criminal law of the United\\nStates that prohibits providing material support to foreign\\nterrorist groups, 1$ U.S.C. § 23393; (2) Section 960a(b)(2),\\nbecause the drug activity and terrorist offenses connected to\\nAppellant occurred in or affected interstate or foreign\\ncommerce; and (3) Section 960a(b)(5), because Appellant was\\nbrought to the United States for prosecution.\\n\\n 12\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\n\\n\\n Because parties may not stipulate to legal conclusions in\\nplea agreements, see United States v. Flores, 912 f.3d 613, 619\\n(D.C. Cir. 2019), Appellant’s plea then sets forth the specific\\nfacts that establish how his conduct fulfilled these elements.\\nRelevant here, the agreement describes how he provided\\nmoney and weapons to Organization 1 and Organization 2,\\n“which were designated Foreign Terrorist Organizations by the\\nUnited States Department of State during the period of the\\nconspiracy.” A Foreign Terrorist Organization is a “foreign\\norganization” that “engages in terrorist activity.. or terrorism\\n .\\n\\n\\n\\n\\n• . or retains the capability and intent to engage in terrorist\\n .\\n\\n\\n\\n\\nactivity or terrorism” that “threatens the security of United\\nStates nationals or the national security of the United States.”\\n8 U.S.C. § 1189. Indeed, Appellant stipulated that he knew that\\nOrganization 1 and Organization 2 were actively engaging in\\nterrorism while he was facilitating payments to them,\\nsometimes in weapons. Appellant also stipulated that the\\nmoney and weapons he supplied to Organization 1 and\\nOrganization 2 came from the large quantities of drugs he\\nmoved around the region. Taken together, Appellant admitted\\nthat he had trafficked millions of dollars’ worth of drugs and\\nweapons between multiple countries to fund groups engaging\\nin terrorism that threatened the security of U.S. nationals or the\\nnational security of the United States.\\n\\n Appellant now argues that Congress lacks the power under\\nArticle Ito authorize the application of § 960a to this admitted\\nconduct. In light of the admissions in Appellant’s plea\\nagreement, we disagree, and conclude that Congress had\\n\\n 13\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nauthority under the Foreign Commerce Clause to authorize\\nAppellant’s prosecution.5\\n\\n Appellant argues that his conduct was beyond the reach of\\nCongress’s power “[t]o regulate Commerce with foreign\\nNations” because it involved only “matters between two or\\nmore foreign countries not involving the United States,”\\nAppellant Suppl. Br. 9, and that his actions at most\\n“affect[ed]’ American commerce,” Appellant Reply Br. 18.\\nDoubt certainly exists as to Congress’s ability under the\\nForeign Commerce Clause to criminalize conduct with no\\neffect on the United States. See, e.g., United States v. Baston,\\n81$ f.3d 651, 667 (11th Cir. 2016) (observing that the Supreme\\nCourt has never “thoroughly explored the scope of the Foreign\\nCommerce Clause”); Baston v. United States, 137 5. Ct. $50,\\n$51 (2017) (Thomas, J., dissenting from denial of certiorari)\\n(observing that the Supreme Court has only evaluated laws\\ninvoking the foreign Commerce Clause to “regulat[e] conduct\\nwith a significant connection to the United States”). But see\\nHartford fire Ins. Co. v. Catfornia, 509 U.S. 764, 813-14\\n(1993) (Scalia, J., dissenting) (“Congress has broad power\\nunder [the Foreign Commerce Clause], and this Court has\\nrepeatedly upheld its power to make laws applicable to persons\\nor activities beyond our territorial boundaries where United\\nStates interests are affected.”); 3d. of Trs. of Univ. of Ill. v.\\nUnited States, 289 U.S. 48, 56 (1933) (describing as an\\n“essential attribute of the power” granted to Congress under the\\n\\n\\n Accordingly, we need not reach the government’s contention\\nthat Congress’s power under Article I, section 8 “[tb define and\\npunish Piracies and Felonies committed on the high Seas, and\\nOffenses against the Law of Nations” and “[t]o make all Laws which\\nshall be necessary and proper” to execute treaties justified\\napplication of 96Oato Appellant’s conduct.\\n\\n 14\\n PUBLIC COPY - SEALED INFORT’IATION DELETED\\n\\nForeign Commerce Clause “that it is exclusive and plenary”).\\nAfter all, the clause refers to “Commerce with foreign\\nNations,” lending support to the notion that some connection\\nto the United States must be established for Congress to\\nlegislate pursuant to this power. U.S. C0NsT. art. I, § 8, cI. 3\\n(emphasis added); Gibbons v. Ogden, 22 U.S. (9 Wheat) 1, 3\\n(1824) (“The power to regulate commerce extends to every\\nspecies of commercial intercourse between the United States\\nand foreign nations.” (emphasis added)); Anthony I.\\nColangelo, The foreign Commerce Clause, 96 VA. L. REv. 949\\n(2010) (arguing that the Foreign Commerce Clause authorizes\\nonly regulation of commerce with some nexus to the United\\nStates and that it may never authorize greater regulation of\\nactivities within foreign nations than would be authorized by\\nthe Interstate Commerce Clause within states).\\n\\n But Appellant’s argument goes further, asking us to hold\\nthat Congress can only make criminal a foreign commercial\\nactivity that takes place in the course of commerce with the\\nUnited States. Such a narrow construction would bar Congress\\nfrom outlawing drug trafficking that lends financial support to\\nterrorist organizations in foreign countries when that support\\nmerely “affects” commerce with the United States. Although\\nour sister circuits disagree on the precise test that should be\\nused to determine when overseas actions may be proscribed\\nunder the Foreign Commerce Clause, every Court of Appeals\\nto have considered the question in the context of commercial\\nactivity such as Appellant’s has disavowed his cramped\\nreading. Some circuits have used the familiar Interstate\\nCommerce Clause framework from United States v. Lopez, 514\\nU.S. 549, 558-59 (1995), and concluded that Congress has\\nbroad power to regulate overseas commercial conduct that has\\na “substantial effect” on commerce with the United States. See,\\ne.g., United States v. Durham, 902 F.3d 1180, 1192-93 (10th\\n\\n 15\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nCir. 201$); Baston, $18 f.3d at 667-68; United States v.\\nFendleton, 65$ F.3d 299, 308 (3d Cir. 2011). Some permit the\\nregulation of foreign conduct with less of an effect on the\\nUnited States. See, e.g., United States v. Bollinger, 79$ F.3d\\n201, 215-16 (4th Cir. 2015) (holding that Congress may\\nregulate conduct abroad that “demonstrably” rather than\\n“substantially” affects commerce with the United States);\\nUnited States v. Clark, 435 f.3d 1100, 1114 (9th Cir. 2006)\\n(requiring only a “constitutionally tenable nexus with foreign\\ncommerce”), partially superseded as noted in United States v.\\nFepe, 895 f.3d 679, 629 (9th Cir. 201$). But no court has\\nadopted the narrow view of congressional power Appellant\\nchampions. We find our sister circuits’ consistent adoption of\\nat least some sort of effects test to be persuasive. Cf RJR\\nNabisco, Inc. v. European Cmty., 136 S. Ct. 2090, 2105 (2016)\\n(holding that the term “foreign commerce” in the RICO statute\\nrestricts the reach of the law to enterprises that “engage in, or\\naffect in some significant way, commerce directly involving\\nthe United States—e.g., commerce between the United States\\nand a foreign country”).\\n\\n We need not, however, define today the precise level of\\n“effect” necessary, because the stipulations in Appellant’s plea\\nagreement place his conduct abroad within Congress’s reach\\nunder any version of the test. As already noted, his was not a\\nminor offense, the effects of which were only tangentially\\nrelated to the United States. Appellant was a leader of a large,\\ntransnational drug trafficking conspiracy that supported\\nestablished and powerful terrorist groups. Those groups\\nmurdered and kidnapped civilians, and the State Department\\nhad publicly concluded that they threatened U.S. nationals and\\ninterests. Appellant’s plea agreement does not specify where\\nthe drugs he trafficked were ultimately headed, but given the\\nquantity of drugs, the involvement of Organization 1, and the\\n\\n 16\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nnature of the Country A drug export markets in which\\nAppellant participated, his actions had a significant effect on\\nthe drug trade between the United States and the countries in\\n\\n\\n\\n I ant stipulated, moreover, that the terrorist\\norganizations he supported through his international drug\\ntrafficking threatened U.S. nationals or the national security of\\nthe United States. See Superseding Information at 2-3, No. 10-\\ncr-334 (D.D.C. Dec. 21, 2010), Dkt. No. 6 (stating that\\nOrganization 2 and Organization 1 “were designated Foreign\\nTerrorist Organizations by the United States Department of\\nState during the period of the conspiracy”); 2 U.S.C.\\n§ 1 189(a)(l)(C) (requiring that “the terrorist activity or\\nterrorism of’ an organization named as a foreign Terrorist\\nOrganization by the State Department “threaten[] the security\\nof United States nationals or the national security of the United\\nStates”). That magnifies the effect of his conduct on commerce\\nbetween the countries where he was operating and the United\\nStates. Cf $trassheim v. Daily, 221 U.S. 280, 225 (1911)\\n(“Acts done outside a jurisdiction, but intended to produce and\\nproducing detrimental effects within it, justify a state in\\npunishing the cause of the harm as if he had been present at the\\neffect, if the state should succeed in getting him within its\\npower.”).\\n\\n Appellant’s argument for why we should not apply any\\nversion of an effects test when evaluating Congress’s power\\nunder the foreign Commerce Clause is conclusory and without\\nmerit. He fails to address the reasoning given by those circuits\\nthat have considered the question or to explore the history of\\nthe clause. Instead, he simply relies on the bare assertion that a\\npower with such a weak limit would be too broad. Appellant\\n\\n 17\\n PUBLIC COPY - SEALED INFOR1vIATION DELETED\\n\\nReply Br. 18-19; Appellant Suppi. Br. 9. Given the Supreme\\nCourt’s longstanding approach to the Interstate Commerce\\nClause and the approach of our sister circuits to the Foreign\\nCommerce Clause, the need for the strict construction\\nAppellant urges is by no means self-evident. And once we\\nconclude that some version of an “effects test” applies,\\nAppellant’s remaining arguments are simply attempts to\\nrecharacterize the facts to which he stipulated, minimizing his\\nrole in the conspiracy and its magnitude to suggest that the\\neffect of his behavior on commerce with the United States was\\ntoo small for Congress to reach. Even after Class, we may not\\nconsider such arguments when, as here, they contradict the\\nadmissions made in the defendant’s plea agreement.\\n\\n The facts of Class illustrate why this aspect of Appellant’s\\nas-applied challenge is foreclosed by his plea agreement.\\nRodney Class pleaded guilty to possessing a firearm on the\\ngrounds of the U.S. Capitol. Class, 138 S. Ct. at 802; see also\\n40 U.S.C. § 5104(e)(1)(A)(i) (making it a crime to “carry on or\\nhave readily accessible to any individual on the Grounds or in\\nany of the Capitol Buildings a firearm”). The statute under\\nwhich he was convicted required proof of (1) carrying or\\nhaving readily accessible a firearm, dangerous weapon,\\nexplosives, or an incendiary device; and (2) presence “on the\\nGrounds or in any of the Capitol Buildings.” 40 U.S.C.\\n§ 5104(e)(1)(A)(i). As part of his plea, Class stipulated that\\nboth of these elements had been met because he had, in fact,\\npossessed a firearm on the Capitol Grounds when he drove his\\nvehicle into a parking lot near the Capitol. Class, 138 S. Ct. at\\n802, 804. On appeal, Class argued in relevant part “that he was\\ndenied fair notice that weapons were banned in the parking lot”\\nin violation of the Due Process Clause, both because no signs\\nexplained that the parking lot was an area in which firearms\\nwere prohibited and because the statute itself was so vague as\\n\\n 1$\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nto provide insufficient notice of what constituted the “Capitol\\nGrounds.” Id at $02; Id at 80$, $13 n.4 (Auto, J., dissenting).\\nCritically, neither his plea agreement nor the statute said\\nanything about his being aware that the parking lot was part of\\nthe Capitol Grounds.\\n\\n Because of that omission, the Supreme Court concluded\\nthat Class could press his due process claim on appeal because\\nit was “consistent with [his] knowing, voluntary, and intelligent\\nadmission that he did what the indictment alleged,” and could\\nbe resolved on appeal by reference to the existing record. Id. at\\n$04 (majority opinion) (internal quotation marks omitted).\\nAlthough Class had agreed that he was carrying weapons in a\\nprohibited place, he had not admitted to knowledge of the\\nnature of that place. Id. His argument that it violated the Due\\nProcess Clause to prosecute someone for unknowingly\\npossessing firearms in a prohibited location thus did not\\ncontradict any of the “admissions necessarily made” when he\\npleaded guilty. Id. at $05 (quoting Broce, 48$ U.S. at 573-74).\\n\\n This case more closely resembles a case that Class was\\nnot. Had Class admitted in his plea agreement that he knew he\\nwas in a place where possession of firearms was prohibited, he\\nwould not have been able to argue on appeal that he lacked\\nsufficient notice. Appellant made that kind of contradictory\\nadmission when he stipulated to the facts described above,\\nwhich demonstrate that his conduct affected commerce with\\nthe United States. As we concluded above, Appellant may\\nchallenge whether conduct that only “affects” foreign\\ncommerce with the United States may be prohibited by\\nCongress—just as Class, in our hypothetical, would remain\\nfree to claim that a statute that does not require notice violates\\nthe Due Process Clause. Appellant may not, however, argue\\n\\n 19\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nthat his conduct failed to affect foreign commerce without\\ncontradicting the terms of his plea agreement.6\\n\\n C\\n\\n Appellant next argues that his prosecution under § 960a,\\neven if authorized by Article I, violates the Due Process Clause\\nbecause the government has not established a sufficient nexus\\nbetween his conduct and the United States. We disagree.\\n\\n Appellant contends that Congress, subject to exceptions\\nnot applicable here, may only proscribe overseas conduct when\\na nexus exists between the conduct and the United States. We\\nhave repeatedly declined, however, to hold that the Due\\nProcess Clause demands such a nexus—or to even resolve\\n“whether the Due Process Clause constrains the extraterritorial\\n\\n\\n 6\\n Appellant’s claim reveals the difficulty faced by prosecutors\\nnegotiating plea agreements to resolve charges brought under\\nstatutes with jurisdictional elements such as § 960a. By pleading\\nguilty, Appellant prevented the government from putting on\\nevidence of his crime’s effect on foreign commerce with the United\\nStates. His appeal then posits that the tack of such evidence renders\\nhis prosecution unconstitutional. Prosecutors would be well-served\\nby ensuring that stipulations of fact entered into by pleading\\ndefendants include the factual bases for elements that bring the\\ndefendants’ conduct within the scope of Congress’s power to\\nlegislate, rather than simply reciting that such elements have been\\nmet. Cf Weston v. Wash. Metro. Area Transit Auth., 78 F.3d 682,\\n625 (D.C. Cir. 1996) (“While parties may enter into stipulations of\\nfact that are binding upon them unless they can show manifest\\ninjustice, parties may not stipulate to the legal conclusions to be\\nreached by the court.” (internal quotation marks and citations\\nomitted)). The stipulations of fact here, however, contained such\\nfacts.\\n\\n 20\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\napplication of federal criminal laws” at all. United States v.\\nBallestas, 795 F.3d 13$, 148 (D.C. Cir. 2015); see also United\\nStates v. Au, 718 f.3d 929, 943-44 & n.7 (D.C. Cir. 2013)\\n(declining to reach the question and observing that the court\\nhad “not found.. any case in which extraterritorial application\\n .\\n\\n\\n\\n\\nof a federal criminal statute was actually deemed a due process\\nviolation . .suggest[ing] [the defendant’s] burden is a heavy\\n .\\n\\n\\n\\n\\none”). Instead, “[t]he ultimate question under the Due Process\\nClause is not nexus, but is whether application of the statute to\\nthe defendant would be arbitrary or fundamentally unfair.”\\nBallestas, 795 F.3d at 148 (internal alterations and quotation\\nmarks omitted). “[T]he animating principle governing the due\\nprocess limits of extraterritorial jurisdiction,” to the extent such\\nlimits exist, “is the idea that ‘no man shall be held criminally\\nresponsible for conduct which he could not reasonably\\nunderstand to be proscribed.” Ali, 712 F.3d at 944 (quoting\\nBouie v. City of Columbia, 37$ U.S. 347, 351 (1964)).\\n\\n Appellant was a leader of an international drug trafficking\\norganization that funded and provided weapons to terrorist\\ngroups that openly murdered and kidnapped civilians. These\\ngroups, moreover, were publicly designated by the State\\nDepartment as directly threatening U.S. nationals or the\\nnational security of the United States. And various\\ninternational agreements specifically prohibit Appellant’s\\nconduct. See International Convention for the Suppression of\\nthe Financing of Terrorism, openedfor signature Dec. 9, 1999,\\n2178 U.N.T.S. 197 (entered into force Apr. 10, 2002); United\\nNations Convention Against Illicit Traffic in Narcotic Drugs\\nand Psychotropic Substances, opened for signature Dec. 20,\\n198$, 1582 U.N.T.S. 95 (entered into force Nov. 11, 1990).\\nAppellant can find no refuge in the Due Process Clause. He\\nadmitted, after all, that he engaged in conduct, proscribed by\\ninternational agreement, that supported groups that threaten\\n\\n 21\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nU.S. nationals or the national security of the United States. See\\nAll, 71$ F.3d at 944-45 (describing, on the basis of similar\\ninternational agreements, hostage taking as “an offense whose\\nproscription ‘is a result of universal condemnation of those\\nactivities and general interest in cooperating to suppress them”\\n(quoting RESTATEMENT (THIRD) OF FOREIGN RELATIONS LAW\\n§ 404 cmt. a (AM. LAW INST. 1987))). To apply § 960a to\\nAppellant “was neither arbitrary nor fundamentally unfair.”\\nBallestas, 795 f.3d at 148.\\n\\n III\\n\\n Appellant also argues that his conduct did not satisfy the\\njurisdictional elements of 21 U.S.C. § 960a(b), a defect in the\\ndistrict court’s subject mafterjurisdiction that cannot be waived\\nby his subsequent guilty plea. But Appellant misunderstands\\nthe work these elements of the statute perform.\\n\\n When used in statutory provisions like § 960a(b),\\njurisdiction can refer either to a “so-called ‘jurisdictional\\nelement’ addressing the reach of [Congress’s] legislative\\nauthority,” or “subject[ Jmatterj urisdiction—the jurisdiction of\\nthe ‘district courts.” United States v. Miranda, 780 F.3d 1185,\\n 1195 (D.C. Cir. 2015). The former, “consistent with the\\ndescription ‘jurisdictional element,’ treat[s] the relevant\\ncondition as an element of the offense to be found by a jury,”\\nand proof of such an element “is no different from proof of any\\nother element of a federal crime.” Id. (quoting Hugi v. United\\nStates, 164 F.3d 378, 381 (7th Cir. 1999)); see also Id. at 1191\\n(“[T]o ask ‘what conduct [a statute] reaches is to ask what\\nconduct [the statute] prohibits, which is a merits question,’ not\\na question of subject[ ]matter jurisdiction.” (quoting Morrison\\nv. Nat’l Austi. BankLtd., 561 U.S. 247, 254 (2010))); United\\nStates v. Gonzalez, 311 F.3d 440, 443 (1st Cir. 2002) (noting\\n\\n 22\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nthat Congress sometimes “use[s] the word ‘jurisdiction’\\nloosely to describe its own assertion of authority to regulate,”\\nsuch as when “it fixes an ‘affects interstate commerce’ . as a\\n . .\\n\\n\\n\\n\\ncondition of the crime,” and that these “conditions have\\nnothing whatever to do with the subject matter Uurisdiction] of\\nthe federal district court”). In contrast, subject matter\\njurisdiction “refers to a tribunal’s power to hear a case” and\\npresents a legal question. Miranda, 780 F.3d at 1191 (quoting\\nMorrison, 561 U.S. at 254).\\n\\n As discussed in Part 11.3, § 960a(a) refers to much conduct\\noccurring outside of the United States, and § 960a(b) puts\\nlimits on that extraterritorial sweep. “The extraterritorial reach\\nof a statute ordinarily presents a merits question, not a\\njurisdictional question.” Miranda, 780 F.3d at 1191. We\\ntherefore begin with the presumption that the provision does\\nnot speak to the subject matterjurisdiction of the federal courts.\\nWe will not, moreover, interpret “a threshold limitation on a\\nstatute’s scope” as referring to a court’s subject matter\\njurisdiction unless Congress “clearly states” as much. Arbaugh\\nv. Y&H Corp., 546 U.S. 500, 515 (2006). In arguing that\\nCongress has done so here, Appellant relies on Miranda, in\\nwhich this Court considered the Maritime Drug Law\\nEnforcement Act (MDLEA). The MDLEA prohibits drug\\nactivity on two categories of ships: “(1) a vessel of the United\\n States or a vessel subject to the jurisdiction of the United\\n States; or (2) any other vessel if the individual is a citizen of\\nthe United States or a resident alien of the United States.” 46\\nU.S.C. § 70503(a), (e). We concluded in Miranda that “subject\\nto the jurisdiction of the United States” represented a limit on\\nour subject matter jurisdiction, rather than an element of the\\noffense. 780 F.3d at 1191. Appellant argues that Miranda\\ncontrols the present case, and because § 960a(b) defines our\\n\\n 23\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nsubject matter jurisdiction, his challenge on this point cannot\\nbe waived by his plea.\\n\\n Miranda, however, is distinguishable based on the text of\\nthe statute at issue. The MDLEA provision defining\\n“[j]urisdiction of the United States with respect to a vessel\\nsubject to this chapter” explicitly states that such jurisdiction\\n“is not an element of an offense. Jurisdictional issues arising\\nunder this chapter are preliminary questions of law to be\\ndetermined solely by the trial judge.” 46 U.S.C. § 70504(a)\\n (emphasis added). Thus, the plain text prevented us from\\nfinding that “vessels subject to the jurisdiction of the United\\n States” represented ajurisdictional element because the inquiry\\nwas a “preliminary” or threshold issue—a key characteristic of\\nsubject matter jurisdiction. Miranda, 780 f.3d at 1193.\\nfurthermore, the text stated that the issue was “not an element\\nof the offense,” “fortifying its jurisdictional character.” Id. And\\nfinally, the text provided for the issue to be determined solely\\nby the trial judge as a question of law, just as “subject[ ]matter\\njurisdiction presents a question of law for resolution by the\\ncourt.” Id. Taken together, the MDLEA satisfied Arbaugh’s\\n“readily administrable bright line” rule that Congress “clearly\\nstate[] that a threshold limitation on a statute’s scope [would]\\ncount as jurisdictional.” 546 U.S. at 5 15-16.\\n\\n Section 960a(b), in contrast, lacks any of these features,\\nstating only that “{t]here is jurisdiction over an offense under\\nthis section if’ one of several conditions is met. And each of\\nthose conditions resembles the jurisdictional elements\\ncommonly found to restrict the scope of congressional, rather\\nthan judicial, power. Accordingly, “to ask what conduct\\n [ 960a] reaches is to ask what conduct [the statute] prohibits,\\nwhich is a merits question, not a question of subject matter\\njurisdiction.” Miranda, 780 F.3d at 1191 (internal quotation\\n\\n 24\\n PUBLIC COPY - SEALED INFORMATION DELETED\\n\\nmarks omitted); see also United States v. Harrington, 108 F.3d\\n1460, 1465 n.2 (D.C. Cir. 1997) (describing jurisdictional\\nelements that restrict the reach of statutes to Congress’s\\nenumerated powers by requiring factual findings, such as an\\neffect on interstate commerce). Appellant’s guilty plea thus\\nwaives his claim that his conduct does not meet the\\nrequirements of § 960a(b).\\n\\n\\n Iv\\n\\n For these reasons, we affirm Appellant’s conviction.\\n\\n So ordered.\\n\\n PUBLIC COPY - SEALED ENFORMATION DELETED\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4188,"cells":{"id":{"kind":"number","value":4657749,"string":"4,657,749"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"In re: Rail Freight Fuel Surcharge Antitrust Litigation"},"decision_date":{"kind":"string","value":"2019-08-30"},"docket_number":{"kind":"string","value":"18-7010"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before: GARLAND, Chief Judge, and ROGERS and KATSAS, Circuit Judges.\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the Court filed by Circuit Judge KATSAS\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued September 28, 2018 Decided August 16, 2019\\n Reissued August 30, 2019\\n\\n No. 18-7010\\n\\n IN RE: RAIL FREIGHT FUEL SURCHARGE ANTITRUST\\n LITIGATION - MDL NO. 1869,\\n\\n DAKOTA GRANITE COMPANY, ON BEHALF OF ITSELF AND ALL\\n OTHERS SIMILARLY SITUATED, ET AL.,\\n APPELLANTS\\n\\n v.\\n\\n BNSF RAILWAY COMPANY, ET AL.,\\n APPELLEES\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:07-mc-00489)\\n\\n\\n Kathleen M. Sullivan argued the cause for appellants.\\nWith her on the briefs were Stephen R. Neuwirth, Sami H.\\nRashid, Michael D. Hausfeld, and Michael P. Lehmann.\\n\\n Carter G. Phillips argued the cause for appellees. With\\nhim on the brief were Joseph R. Guerra, Kathleen Moriarty\\nMueller, Saul P. Morgenstern, Thomas A. Isaacson, John M.\\nNannes, Tara L. Reinhart, J. Scott Ballenger, Veronica S.\\n\\f 2\\nLewis, Samuel M. Sipe, Jr., Linda S. Stein, Andrew S.\\nTulumello, Lucas C. Townsend, and Kent A. Gardiner.\\n\\n Anton Metlitsky and Warren D. Postman were on the brief\\nfor amicus curiae Chamber of Commerce of the United States\\nof America in support of defendants-appellees.\\n\\n Before: GARLAND, Chief Judge, and ROGERS and KATSAS,\\nCircuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge KATSAS.\\n\\n KATSAS, Circuit Judge: This case involves a putative class\\nof over 16,000 shippers allegedly harmed by a price-fixing\\nconspiracy among the nation’s largest freight railroads. The\\ndistrict court denied class certification because the plaintiffs’\\nregression analysis—their evidence for proving causation,\\ninjury, and damages on a class-wide basis—measured negative\\ndamages for over 2,000 members of the proposed class. Based\\non that consideration, we affirm.\\n\\n I\\n\\n This appeal arises out of eighteen antitrust actions\\nconsolidated by the Multidistrict Litigation Panel. The\\ndefendants are the four largest freight railroads in the United\\nStates: BNSF Railway Company; CSX Transportation, Inc.;\\nNorfolk Southern Railway Company; and Union Pacific\\nRailroad Company. The plaintiffs, who are their customers,\\nallege that the railroads conspired to fix rate-based fuel\\nsurcharges. Railroads impose fuel surcharges—additional\\ncharges above the base shipping price—when the price of fuel\\nrises above a certain trigger price. Rate-based surcharges are\\ncalculated as a percentage of the base shipping price.\\n\\f 3\\n Following consolidation, the action was divided into one\\ncase involving direct purchasers and another involving indirect\\npurchasers. All plaintiffs alleged that the railroads violated\\nsection 1 of the Sherman Act, 15 U.S.C. § 1, by conspiring to\\nfix prices. The direct purchasers sought treble damages under\\nsection 4 of the Clayton Act, 15 U.S.C. § 15, and the district\\ncourt held that they stated a claim, In re Rail Freight Surcharge\\nAntitrust Litig., 587 F. Supp. 2d 27 (D.D.C. 2008). The indirect\\npurchasers sought injunctive relief under section 16 of the\\nClayton Act, 15 U.S.C. § 26, and raised various state-law\\nclaims. The district court held that the state claims were\\npreempted by federal law, but it declined to dismiss the federal\\nclaims. In re Rail Freight Surcharge Antitrust Litig., 593 F.\\nSupp. 2d 29 (D.D.C. 2008), aff’d, Fayus Enters. v. BNSF Ry.\\nCo., 602 F.3d 444 (D.C. Cir. 2010).\\n\\n The eight named plaintiffs in the direct-purchaser case—\\nCarter Distributing Company; Dakota Granite Company;\\nDonnelly Commodities, Inc.; Dust Pro, Inc.; Nyrstar Taylor\\nChemicals, Inc.; Olin Corporation; Strates Shows, Inc.; and US\\nMagnesium LLC—moved to certify a class under Federal Rule\\nof Civil Procedure 23(b)(3). The proposed class consisted of\\nall shippers who paid rate-based fuel surcharges for\\nunregulated services purchased from the defendants between\\nJuly 1, 2003 and December 31, 2008. To show that causation,\\ninjury, and damages could be proved on a class-wide basis, the\\nplaintiffs invoked two regression models constructed by their\\neconomist, Dr. Gordon Rausser. The “common factor model”\\nidentified seven variables said to determine the price of the\\ndefendants’ services, including fuel surcharges. The “damages\\nmodel,” controlling for those variables, sought to isolate price\\nincreases attributable to the alleged conspiracy. The railroads\\ncriticized these models on various grounds, including that they\\nmeasured damages for shipments made under legacy contracts\\nfixed before any conspiracy allegedly began.\\n\\f 4\\n The district court initially certified the class. It noted that\\nif individualized proof were necessary to establish causation\\nand injury, then the plaintiffs could not satisfy the Rule\\n23(b)(3) requirement that common questions predominate. See\\nIn re Rail Freight Fuel Surcharge Antitrust Litig., 287 F.R.D.\\n1, 43 (D.D.C. 2012). But the court found Dr. Rausser’s\\nregression analysis to be “plausible” and “workable,” so it\\nconcluded that causation, injury, and damages were\\n“susceptible to proof at trial through evidence common to the\\nclass.” Id. at 67. The court rejected many different criticisms\\nof the regression models, but it did not specifically address the\\nquestion of false positives for legacy contracts.\\n\\n On interlocutory review, we vacated the certification order\\nand remanded for reconsideration in light of Comcast Corp. v.\\nBehrend, 569 U.S. 27 (2013). In re Rail Freight Surcharge\\nAntitrust Litig.–MDL No. 1869, 725 F.3d 244 (D.C. Cir. 2013)\\n(Rail Freight I). We explained that, for an antitrust class\\naction, common questions “cannot predominate where there\\nexists no reliable means of proving classwide injury in fact.”\\nId. at 253. We expressed concern with the district court’s\\nfailure to address “the damages model’s propensity toward\\nfalse positives,” which left us with no way of knowing whether\\n“the overcharges the damages model calculates for class\\nmembers [are] any more accurate than the obviously false\\nestimates it produces for legacy shippers.” Id. at 254. Finally,\\nwe stressed that Rule 23, as construed in Comcast, requires a\\n“hard look at the soundness of statistical models that purport to\\nshow predominance.” Id. at 255.\\n\\n On remand, after permitting supplemental discovery and\\nexpert reports, the district court denied class certification. In\\nre Rail Freight Surcharge Antitrust Litig., 292 F. Supp. 3d 14\\n(D.D.C. 2017) (Rail Freight II). The court concluded that Dr.\\nRausser’s expert opinions were reliable enough to be\\n\\f 5\\nadmissible at trial. Id. at 49–63. But in assessing\\npredominance, the court identified three shortcomings in his\\ndamages model: first, it measured highly inflated damages for\\nintermodal traffic (i.e., shipments traveling by rail and another\\nmode of transportation such as trucks or airplanes), id. at 122–\\n26; second, as we had noted in the earlier appeal, the model\\nerroneously measured damages for shipments made under\\nlegacy contracts, id. at 126–31; and third, the model measured\\nnegative damages—and hence no injury—for over 2,000\\nmembers of the proposed class, id. at 132–41. The court\\nconcluded that any one of these problems was enough to defeat\\nthe plaintiffs’ argument for predominance. Id. at 122.\\n\\n The plaintiffs filed a petition for permission to appeal the\\nclass-certification decision under Federal Rule of Civil\\nProcedure 23(f). A motions panel of this Court granted the\\npetition without prejudice to reconsideration at the merits\\nstage. In re Rail Freight Fuel Surcharge Antitrust Litig.–MDL\\nNo. 1869, No. 17-8005 (D.C. Cir. Dec. 20, 2017).\\n\\n II\\n\\n We begin with the question of our jurisdiction. Orders\\ndenying class certification are neither final decisions under 28\\nU.S.C. § 1291, Coopers & Lybrand v. Livesay, 437 U.S. 463\\n(1978), nor injunctions immediately appealable under 28\\nU.S.C. § 1292(a)(1), Gardner v. Westinghouse Broad. Co., 437\\nU.S. 478 (1978). However, 28 U.S.C. § 1292(e) permits the\\nSupreme Court to promulgate rules creating new categories of\\ndecisions appealable before final judgment. Exercising that\\nauthority, the Court has provided that “[a] court of appeals may\\npermit an appeal from an order granting or denying class-action\\n\\f 6\\ncertification,” if a “petition for permission to appeal” is timely\\nfiled. Fed. R. Civ. P. 23(f).\\n\\n In this case, the plaintiffs filed a timely petition for\\npermission to appeal, which was enough under Rule 23(f) to\\nsecure our jurisdiction. That jurisdiction is “discretionary,”\\nRail Freight I, 725 F.3d at 250, and the railroads contest\\nwhether we should exercise it. But their argument on this point\\nis perfunctory—less than one page of briefing, with no case\\ncitations—and it almost entirely duplicates their merits\\narguments. According to the railroads, we should conclude\\nthat the denial of class certification was correct. Then, we\\nshould dismiss the appeal as raising neither a questionable\\ndecision nor an unsettled issue—considerations that bear on\\nwhether to permit the appeal in the first place, see id. at 250–\\n54. Because that disposition would make little sense at this\\njuncture, we decline to revisit the motions panel’s decision\\naccepting the appeal.\\n\\n III\\n\\n Federal Rule of Civil Procedure 23 sets forth various\\nrequirements for the certification of class actions. Rule 23(a)\\nprovides four “prerequisites” for any class certification,\\nincluding that there must be “questions of law or fact common\\nto the class.” If these prerequisites are met, Rule 23(b)(3)\\npermits certification if, among other things, “questions of law\\nor fact common to class members predominate over any\\nquestions affecting only individual members.” For purposes of\\nthese rules, a “common” question is one that is “capable of\\nclasswide resolution—which means that determination of its\\ntruth or falsity will resolve an issue that is central to the validity\\nof each one of the claims in one stroke.” Wal-Mart Stores, Inc.\\nv. Dukes, 564 U.S. 338, 350 (2011). In contrast, an\\n“individual” question is one for which “members of a proposed\\n\\f 7\\nclass will need to present evidence that varies from member to\\nmember.” Tyson Foods, Inc. v. Bouaphakeo, 136 S. Ct. 1036,\\n1045 (2016) (quotation marks omitted).\\n\\n The party seeking class certification “must affirmatively\\ndemonstrate” that the commonality and predominance\\nrequirements are satisfied. Wal-Mart, 564 U.S. at 350. This\\nrequires a “rigorous analysis” that often will “overlap with the\\nmerits.” Id. at 351 (quotation marks omitted). Three recent\\ncases address the contours of this analysis. In Wal-Mart, the\\nSupreme Court ordered decertification of a Title VII class\\nwhere the plaintiffs presented insufficient proof that the\\ndefendant had engaged in a general policy of sex\\ndiscrimination. Id. at 352–55. In Comcast, the Court ordered\\ndecertification where the regression analysis used to show\\ncommon injury did not track the underlying theory of liability.\\n569 U.S. at 36–38. The Court rejected a contention that, at the\\nclass-certification stage, “any method of measurement is\\nacceptable so long as it can be applied classwide, no matter\\nhow arbitrary the measurements may be.” Id. at 36. But in\\nTyson Foods, the Court held that concerns about the validity of\\na statistical sample used to prove class-wide averages—which\\nthe Court described as presenting a common objection to the\\nclaims of each class member—should be addressed “as a\\nmatter of summary judgment, not class certification.” 136 S.\\nCt. at 1047 (quotation marks omitted).\\n\\n IV\\n\\n The direct-purchaser plaintiffs raise claims under section\\n4 of the Clayton Act, which provides treble damages to any\\nperson “injured in his business or property by reason of\\nanything forbidden in the antitrust laws.” 15 U.S.C. § 15(a).\\nTo establish liability under section 4, each plaintiff must prove\\nnot only an antitrust violation, but also an injury to its business\\n\\f 8\\nor property and a causal relation between the two. Without\\ncommon proof of injury and causation, section 4 plaintiffs\\ncannot establish predominance. See, e.g., Comcast, 569 U.S.\\nat 36–38; Rail Freight I, 725 F.3d at 252–53.\\n\\n The parties dispute the extent to which a court, in\\nconducting the “hard look” required by Rule 23, should assess\\nthe reliability of common evidence. Specifically, they contest\\nwhether Rule 23 requires reliability above and beyond what is\\nnecessary to establish the admissibility of expert testimony\\nunder Federal Rule of Evidence 702 and Daubert v. Merrell\\nDow Pharmaceuticals, Inc., 509 U.S. 579 (1993). That\\nquestion arose because the district court found Dr. Rausser’s\\nregression models to be reliable for admissibility purposes,\\nRail Freight II, 292 F. Supp. 3d at 54–63; held that “reliability\\nunder Rule 23 is a higher standard than reliability under\\nDaubert,” id. at 91; and found no predominance based on\\nconcerns that “undermine[d] the reliability of Dr. Rausser’s\\ndamages model,” id. at 122. The defendants embrace this\\nreasoning; they argue that Wal-Mart and Comcast rejected\\ncertification based on concerns that common evidence was\\nunreliable, even though its admissibility was uncontested. The\\nplaintiffs disagree; they contend that predominance ultimately\\nturns on whether the relevant evidence is common or\\nindividualized, and they read Tyson Foods as holding that a\\ncourt may assess its reliability only in deciding admissibility or\\nsummary judgment.\\n\\n We need not resolve this dispute because Dr. Rausser’s\\ndamages model, even if sufficiently reliable, does not prove\\nclasswide injury. As the district court explained, his model\\nindicates that the proposed class consists of 16,065 shippers.\\nRail Freight II, 292 F. Supp. 3d at 136. The plaintiffs maintain\\nthat the alleged conspiracy injured every one of them. Yet the\\ndamages model also indicates that 2,037 members of the\\n\\f 9\\nproposed class—or 12.7 percent—suffered “only negative\\novercharges” and thus no injury from any conspiracy. Id. at\\n137. So even assuming the model can reliably show injury and\\ncausation for 87.3 percent of the class, that still leaves the\\nplaintiffs with no common proof of those essential elements of\\nliability for the remaining 12.7 percent. The district court held\\nthat the need for “individualized inquiries to determine which\\nof at least 2,037 (and possibly more) class members were\\nactually injured by the alleged conspiracy,” id. at 140,\\nprecluded a finding of predominance. Id. at 122. We review a\\ndistrict court’s assessment of predominance only for abuse of\\ndiscretion. Garcia v. Johanns, 444 F.3d 625, 631 (D.C. Cir.\\n2006). Here, we find no such abuse.\\n\\n The plaintiffs argue that their model measures these\\nnegative damages only because of normal prediction error. The\\ndistrict court found that prediction error could not “account for\\nall—or even a substantial portion of—the 2,037 shippers that\\nthe model shows to be uninjured.” Rail Freight II, 292 F. Supp.\\n3d at 139. The plaintiffs take issue with that finding, but it is\\nnot clearly erroneous. In any event, the plaintiffs’ argument\\nabout prediction error at most suggests that their damages\\nmodel might falsely have measured no injury for as many as\\n2,037 shippers. This line of reasoning describes a possible\\nproblem with their own evidence; it does not point to\\naffirmative evidence—much less common affirmative\\nevidence—that a conspiracy did in fact injure these shippers.\\n\\n The plaintiffs further argue that predominance does not\\nrequire common evidence extending to all class members.\\nThat contention appears inconsistent with our statement in Rail\\nFreight I that the plaintiffs, to establish predominance, must\\n“show that they can prove, through common evidence, that all\\nclass members were in fact injured by the alleged conspiracy.”\\n725 F.3d at 252; see also id. (“we do expect the common\\n\\f 10\\nevidence to show all class members suffered some injury”).\\nDespite these statements, the district court held that our opinion\\ndid not require common evidence of injury to all class\\nmembers. See Rail Freight II, 292 F. Supp. 3d at 132–34.\\nInstead, it agreed with the plaintiffs that common proof\\ncovering “virtually all” members of the proposed class, and\\nleaving only a “de minimis” number of cases requiring\\nindividualized proof of injury and causation, would be enough\\nto show predominance. Id. at 135. For the sake of argument,\\nwe assume that the district court correctly recognized a de\\nminimis exception to the general rule that, for claims under\\nsection 4 of the Clayton Act, causation and injury must be\\n“capable of classwide resolution,” Wal-Mart, 564 U.S. at 350.\\nThe court reasonably concluded that such a de minimis\\nexception would not encompass this case.\\n\\n In assessing how many individual adjudications are too\\nmany, both the district court and the parties invoke cases\\naddressing the question of when, if ever, a class may include\\nconcededly uninjured members. Strictly speaking, this case\\ndoes not present that question, for the plaintiffs here insist that\\neach member of the proposed class was injured. Nonetheless,\\nthe cited cases bear some similarity to this one: Uninjured class\\nmembers cannot prevail on the merits, so their claims must be\\nwinnowed away as part of the liability determination. And that\\nprospect raises the same kind of question at issue here—when\\ndoes the need for individualized proof of injury and causation\\ndestroy predominance? See, e.g., In re Asacol Antitrust Litig.,\\n907 F.3d 42, 51–58 (1st Cir. 2018); In re Nexium Antitrust\\nLitig., 777 F.3d 9, 18–22 (1st Cir. 2015).\\n\\n The plaintiffs complain that the district court arbitrarily\\nimposed a six-percent upper limit on the percentage of\\nuninjured parties who may be included in a certified class. In\\nfact, the court’s analysis was more nuanced. As the court\\n\\f 11\\nexplained, the “few reported decisions” involving uninjured\\nclass members “suggest that 5% to 6% constitutes the outer\\nlimits of a de minimis number.” Rail Freight II, 292 F. Supp.\\n3d at 137. The 12.7 percent figure in this case is more than\\ntwice that approximate upper bound reflected in analogous\\ncaselaw. Moreover, the district court considered raw numbers\\nas well as percentages: six percent of a “class totaling only\\nfifty-five” members might be de minimis, but 12.7 percent of\\nthis class yields “2,037 uninjured class members” (according\\nto the common proof), all of whom would need individualized\\nadjudications of causation and injury. Id. at 137–38. Finally,\\nthe district court stressed that the plaintiffs have proposed no\\n“further way”—short of full-blown, individual trials—“to\\nreduce this number and segregate the uninjured from the truly\\ninjured.” Id. at 138. None of this was an abuse of discretion.\\n\\n The absence of any winnowing mechanism sharply\\ndistinguishes Nexium, the plaintiffs’ best case. There, the class\\nincluded purchasers of a drug allegedly shielded from\\ncompetition by the unlawful suppression of a generic\\nalternative. 777 F.3d at 13–14. The problem of uninjured class\\nmembers arose because a small percentage of the class, due to\\nbrand loyalty, would have purchased the drug even if a less\\nexpensive generic alternative had been available. Id. at 19–20.\\nThe First Circuit held that the uninjured class members could\\nmanageably be winnowed by having individual consumers file\\nminimal, likely unrebutted affidavit testimony indicating\\nwhether, if given the choice, they would have purchased the\\nbranded drug or a generic alternative. Id. at 20–21.\\n\\n Nexium does not support class certification here. For one\\nthing, the First Circuit sharply limited that decision in Asacol.\\nThere, the Court explained that any winnowing mechanism\\nmust be truncated enough to ensure that the common issues\\npredominate, yet robust enough to preserve the defendants’\\n\\f 12\\nSeventh Amendment and due process rights to contest every\\nelement of liability and to present every colorable defense. 907\\nF.3d at 51–54. Moreover, the Court held that Nexium’s\\naffidavit mechanism could not satisfy both conditions where\\nthe defendant seeks to contest the question whether individual\\nclass members would have shifted from the branded drug to a\\nless expensive generic alternative. See id. That would require\\nindividual trials because genuinely contested affidavits do not\\nsupport summary judgment and are inadmissible. Id. Here, the\\ndefendants intend to contest whether any of the 2,037 shippers\\nsuffered injury as a result of any conspiracy. And the question\\npresented in these individual challenges—regarding impacts\\non shippers of different sizes, shipping different products in\\ndifferent geographic markets, with different transportation\\noptions and different degrees of leverage—would be far more\\ncomplex than the single unitary question (branded or generic?)\\nat issue in Nexium and Asacol.\\n\\n The plaintiffs also invoke the Supreme Court’s discussion\\nof uninjured class members in Tyson Foods. The defendants\\nthere sought review of the question whether a certified class\\nmay contain any uninjured members. But the Supreme Court\\nreserved that question, 136 S. Ct. at 1049, and the district court\\nhere decided it in the plaintiffs’ favor, Rail Freight II, 292 F.\\nSupp. 3d at 133–35. So, this aspect of Tyson Foods does not\\nadvance the plaintiffs’ appeal. The Supreme Court also\\naddressed a further “new argument” raised for the first time in\\nmerits briefing—that plaintiffs at the certification stage must\\nprove that all class members were injured or establish a\\nmanageable process for culling out uninjured class members.\\n136 S. Ct. at 1049. The Court held that this argument was\\n“premature” because the parties disagreed about what culling\\nmechanisms might be available and the district court had not\\nyet addressed the question. See 136 S. Ct. at 1050. This\\nholding rested on the inappropriateness of raising new issues\\n\\f 13\\nfor the first time in Supreme Court merits briefing. See id. It\\ndoes not, as the plaintiffs here contend, permit district courts\\nconsidering class certification to defer questions about the\\nnumber and nature of any individualized inquiries that might\\nbe necessary to establish liability. To the contrary, confronting\\nsuch questions is part-and-parcel of the “hard look” required\\nby Wal-Mart and Comcast, as recognized even by those courts\\npermitting a class to include some small number of concededly\\nuninjured individuals. See, e.g., In re Asacol, 907 F.3d at 51–\\n54; In re Nexium, 777 F.3d at 19–21.\\n\\n Finally, the plaintiffs argue that the 2,037 class members\\nfor whom their damages model shows no injury are a de\\nminimis portion of the class because their shipments make up\\nless than one percent of the railroads’ overall revenue from the\\nalleged conspiracy. But revenue is irrelevant to predominance,\\nwhich looks to whether elements such as causation and injury\\nmay be proved through common evidence, not how much the\\ndefendants benefited from any wrongdoing.\\n\\n V\\n\\n Looking beyond Dr. Rausser’s regression analysis, the\\nplaintiffs point to other evidence that they say can prove injury\\nand causation on a class-wide basis. In Rail Freight I, we\\nconcluded that Dr. Rausser’s analysis was “essential” to the\\nplaintiffs’ case for certification: “No damages model, no\\npredominance, no class certification.” 725 F.3d at 253. The\\ndistrict court reached the same conclusion on remand, after\\ncareful review of all the documentary and expert evidence.\\nThis was not an abuse of discretion.\\n\\n The plaintiffs invoke documentary evidence that the\\ndefendants enforced fuel surcharges “uniformly and with few\\nexceptions.” Rail Freight II, 292 F. Supp. 3d at 122; see id. at\\n\\f 14\\n103–07. But imposing fuel surcharges does not show injury\\ncaused by a conspiracy. The parties vigorously dispute\\nwhether higher overall prices during the class period were\\nattributable to causes besides any conspiracy—such as the\\nmarked increase in fuel prices that occurred around the\\nbeginning of the class period, see J.A. 6592–93. As the district\\ncourt explained, Dr. Rausser designed his regression models\\nprecisely to control for these kinds of potential alternative\\ncauses, see Rail Freight II, 292 F. Supp. 3d at 101, yet his\\ndamages model showed that 2,037 shippers were uninjured\\ndespite paying fuel surcharges. The plaintiffs’ evidence of\\nwidespread fuel surcharges helps explain why the proposed\\nclass is a large one, but it neither proves that the 2,037 shippers\\nwere injured by the alleged conspiracy nor otherwise compels\\na finding of predominance.\\n\\n Next, the plaintiffs invoke the expert testimony of Dr.\\nJames McClave, who argued that the 2,037 shippers must have\\nbeen harmed by the conspiracy. Dr. McClave reasoned that\\nbecause these shippers made fewer purchases on average, and\\nthus had less bargaining power than the rest of the class, they\\nmust have been more susceptible to injury. Moreover, Dr.\\nMcClave’s own study concluded that the defendants’ smallest\\ncustomers—roughly 10,000 shippers who made only one\\npurchase either before or during the class period—collectively\\npaid higher prices than did larger shippers.\\n\\n This analysis is not common proof of injury to the 2,037\\nshippers. For one thing, of the 10,000 shippers studied by Dr.\\nMcClave, only 3,700 were in the class, and Dr. McClave did\\nnot indicate how many of them were included among the 2,037\\nclass members for whom Dr. Rausser’s model showed no\\ninjury. Moreover, evidence that a group of one-time shippers\\nmay have paid higher prices collectively does not prove that all\\n(or almost all) of them were injured individually. And, as the\\n\\f 15\\ndistrict court explained, Dr. McClave did not attempt to\\nidentify how many of the one-time shippers did in fact pay\\nmore, let alone which of them were among the 2,037. See Rail\\nFreight II, 292 F. Supp. 3d at 139.\\n\\n The plaintiffs respond that Dr. McClave’s analysis, even if\\ninconclusive, at least suggests that reduced bargaining power\\nmade the 2,037 shippers more vulnerable to any conspiracy.\\nThis line of reasoning parallels one rejected by the Supreme\\nCourt in Wal-Mart. There, an expert opined that Wal-Mart’s\\n“strong corporate culture” made it “vulnerable to gender bias.”\\n564 U.S. at 354 (quotation marks omitted). The Court found\\nthis evidence insufficient to prove that injury could be\\nestablished on a class-wide basis, because the expert could not\\nsay what percentage of adverse employment decisions were in\\nfact caused by bias. Id. The McClave study is similarly\\nincomplete because, as explained above, it does not attempt to\\nidentify which of the small shippers, or what percentage of\\nthem, were in fact harmed by the alleged conspiracy.\\n\\n VI\\n\\n In Asacol, the First Circuit noted the absence of even a\\nsingle case “allowing, under Rule 23, a trial in which thousands\\nof class members testify.” 907 F.3d at 57–58. That Court\\ndeclined to create “the first such case.” Id. So do we. Given\\nthe need in this case for at least 2,037 individual determinations\\nof injury and causation, the district court did not abuse its\\ndiscretion in denying class certification on the ground that\\ncommon issues do not predominate.\\n\\n Affirmed.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435002/\", \"author_raw\": \"Opinion for the Court filed by Circuit Judge KATSAS\"}]}"},"judge1":{"kind":"string","value":"GARLAND"},"judge2":{"kind":"string","value":"ROGERS"},"judge3":{"kind":"string","value":"KATSAS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"GARLAND\": \", Chief\", \"ROGERS\": \", Circuit\", \"KATSAS\": \", Circuit\"}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4657749/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: In re: Rail Freight Fuel Surcharge Antitrust Litigation\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-30\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued September 28, 2018 Decided August 16, 2019\n Reissued August 30, 2019\n\n No. 18-7010\n\n IN RE: RAIL FREIGHT FUEL SURCHARGE ANTITRUST\n LITIGATION - MDL NO. 1869,\n\n DAKOTA GRANITE COMPANY, ON BEHALF OF ITSELF AND ALL\n OTHERS SIMILARLY SITUATED, ET AL.,\n APPELLANTS\n\n v.\n\n BNSF RAILWAY COMPANY, ET AL.,\n APPELLEES\n\n\n Appeal from the United States District Court\n for the District of Columbia\n (No. 1:07-mc-00489)\n\n\n Kathleen M. Sullivan argued the cause for appellants.\nWith her on the briefs were Stephen R. Neuwirth, Sami H.\nRashid, Michael D. Hausfeld, and Michael P. Lehmann.\n\n Carter G. Phillips argued the cause for appellees. With\nhim on the brief were Joseph R. Guerra, Kathleen Moriarty\nMueller, Saul P. Morgenstern, Thomas A. Isaacson, John M.\nNannes, Tara L. Reinhart, J. Scott Ballenger, Veronica S.\n\f 2\nLewis, Samuel M. Sipe, Jr., Linda S. Stein, Andrew S.\nTulumello, Lucas C. Townsend, and Kent A. Gardiner.\n\n Anton Metlitsky and Warren D. Postman were on the brief\nfor amicus curiae Chamber of Commerce of the United States\nof America in support of defendants-appellees.\n\n Before: GARLAND, Chief Judge, and ROGERS and KATSAS,\nCircuit Judges.\n\n Opinion for the Court filed by Circuit Judge KATSAS.\n\n KATSAS, Circuit Judge: This case involves a putative class\nof over 16,000 shippers allegedly harmed by a price-fixing\nconspiracy among the nation’s largest freight railroads. The\ndistrict court denied class certification because the plaintiffs’\nregression analysis—their evidence for proving causation,\ninjury, and damages on a class-wide basis—measured negative\ndamages for over 2,000 members of the proposed class. Based\non that consideration, we affirm.\n\n I\n\n This appeal arises out of eighteen antitrust actions\nconsolidated by the Multidistrict Litigation Panel. The\ndefendants are the four largest freight railroads in the United\nStates: BNSF Railway Company; CSX Transportation, Inc.;\nNorfolk Southern Railway Company; and Union Pacific\nRailroad Company. The plaintiffs, who are their customers,\nallege that the railroads conspired to fix rate-based fuel\nsurcharges. Railroads impose fuel surcharges—additional\ncharges above the base shipping price—when the price of fuel\nrises above a certain trigger price. Rate-based surcharges are\ncalculated as a percentage of the base shipping price.\n\f 3\n Following consolidation, the action was divided into one\ncase involving direct purchasers and another involving indirect\npurchasers. All plaintiffs alleged that the railroads violated\nsection 1 of the Sherman Act, 15 U.S.C. § 1, by conspiring to\nfix prices. The direct purchasers sought treble damages under\nsection 4 of the Clayton Act, 15 U.S.C. § 15, and the district\ncourt held that they stated a claim, In re Rail Freight Surcharge\nAntitrust Litig., 587 F. Supp. 2d 27 (D.D.C. 2008). The indirect\npurchasers sought injunctive relief under section 16 of the\nClayton Act, 15 U.S.C. § 26, and raised various state-law\nclaims. The district court held that the state claims were\npreempted by federal law, but it declined to dismiss the federal\nclaims. In re Rail Freight Surcharge Antitrust Litig., 593 F.\nSupp. 2d 29 (D.D.C. 2008), aff’d, Fayus Enters. v. BNSF Ry.\nCo., 602 F.3d 444 (D.C. Cir. 2010).\n\n The eight named plaintiffs in the direct-purchaser case—\nCarter Distributing Company; Dakota Granite Company;\nDonnelly Commodities, Inc.; Dust Pro, Inc.; Nyrstar Taylor\nChemicals, Inc.; Olin Corporation; Strates Shows, Inc.; and US\nMagnesium LLC—moved to certify a class under Federal Rule\nof Civil Procedure 23(b)(3). The proposed class consisted of\nall shippers who paid rate-based fuel surcharges for\nunregulated services purchased from the defendants between\nJuly 1, 2003 and December 31, 2008. To show that causation,\ninjury, and damages could be proved on a class-wide basis, the\nplaintiffs invoked two regression models constructed by their\neconomist, Dr. Gordon Rausser. The “common factor model”\nidentified seven variables said to determine the price of the\ndefendants’ services, including fuel surcharges. The “damages\nmodel,” controlling for those variables, sought to isolate price\nincreases attributable to the alleged conspiracy. The railroads\ncriticized these models on various grounds, including that they\nmeasured damages for shipments made under legacy contracts\nfixed before any conspiracy allegedly began.\n\f 4\n The district court initially certified the class. It noted that\nif individualized proof were necessary to establish causation\nand injury, then the plaintiffs could not satisfy the Rule\n23(b)(3) requirement that common questions predominate. See\nIn re Rail Freight Fuel Surcharge Antitrust Litig., 287 F.R.D.\n1, 43 (D.D.C. 2012). But the court found Dr. Rausser’s\nregression analysis to be “plausible” and “workable,” so it\nconcluded that causation, injury, and damages were\n“susceptible to proof at trial through evidence common to the\nclass.” Id. at 67. The court rejected many different criticisms\nof the regression models, but it did not specifically address the\nquestion of false positives for legacy contracts.\n\n On interlocutory review, we vacated the certification order\nand remanded for reconsideration in light of Comcast Corp. v.\nBehrend, 569 U.S. 27 (2013). In re Rail Freight Surcharge\nAntitrust Litig.–MDL No. 1869, 725 F.3d 244 (D.C. Cir. 2013)\n(Rail Freight I). We explained that, for an antitrust class\naction, common questions “cannot predominate where there\nexists no reliable means of proving classwide injury in fact.”\nId. at 253. We expressed concern with the district court’s\nfailure to address “the damages model’s propensity toward\nfalse positives,” which left us with no way of knowing whether\n“the overcharges the damages model calculates for class\nmembers [are] any more accurate than the obviously false\nestimates it produces for legacy shippers.” Id. at 254. Finally,\nwe stressed that Rule 23, as construed in Comcast, requires a\n“hard look at the soundness of statistical models that purport to\nshow predominance.” Id. at 255.\n\n On remand, after permitting supplemental discovery and\nexpert reports, the district court denied class certification. In\nre Rail Freight Surcharge Antitrust Litig., 292 F. Supp. 3d 14\n(D.D.C. 2017) (Rail Freight II). The court concluded that Dr.\nRausser’s expert opinions were reliable enough to be\n\f 5\nadmissible at trial. Id. at 49–63. But in assessing\npredominance, the court identified three shortcomings in his\ndamages model: first, it measured highly inflated damages for\nintermodal traffic (i.e., shipments traveling by rail and another\nmode of transportation such as trucks or airplanes), id. at 122–\n26; second, as we had noted in the earlier appeal, the model\nerroneously measured damages for shipments made under\nlegacy contracts, id. at 126–31; and third, the model measured\nnegative damages—and hence no injury—for over 2,000\nmembers of the proposed class, id. at 132–41. The court\nconcluded that any one of these problems was enough to defeat\nthe plaintiffs’ argument for predominance. Id. at 122.\n\n The plaintiffs filed a petition for permission to appeal the\nclass-certification decision under Federal Rule of Civil\nProcedure 23(f). A motions panel of this Court granted the\npetition without prejudice to reconsideration at the merits\nstage. In re Rail Freight Fuel Surcharge Antitrust Litig.–MDL\nNo. 1869, No. 17-8005 (D.C. Cir. Dec. 20, 2017).\n\n II\n\n We begin with the question of our jurisdiction. Orders\ndenying class certification are neither final decisions under 28\nU.S.C. § 1291, Coopers & Lybrand v. Livesay, 437 U.S. 463\n(1978), nor injunctions immediately appealable under 28\nU.S.C. § 1292(a)(1), Gardner v. Westinghouse Broad. Co., 437\nU.S. 478 (1978). However, 28 U.S.C. § 1292(e) permits the\nSupreme Court to promulgate rules creating new categories of\ndecisions appealable before final judgment. Exercising that\nauthority, the Court has provided that “[a] court of appeals may\npermit an appeal from an order granting or denying class-action\n\f 6\ncertification,” if a “petition for permission to appeal” is timely\nfiled. Fed. R. Civ. P. 23(f).\n\n In this case, the plaintiffs filed a timely petition for\npermission to appeal, which was enough under Rule 23(f) to\nsecure our jurisdiction. That jurisdiction is “discretionary,”\nRail Freight I, 725 F.3d at 250, and the railroads contest\nwhether we should exercise it. But their argument on this point\nis perfunctory—less than one page of briefing, with no case\ncitations—and it almost entirely duplicates their merits\narguments. According to the railroads, we should conclude\nthat the denial of class certification was correct. Then, we\nshould dismiss the appeal as raising neither a questionable\ndecision nor an unsettled issue—considerations that bear on\nwhether to permit the appeal in the first place, see id. at 250–\n54. Because that disposition would make little sense at this\njuncture, we decline to revisit the motions panel’s decision\naccepting the appeal.\n\n III\n\n Federal Rule of Civil Procedure 23 sets forth various\nrequirements for the certification of class actions. Rule 23(a)\nprovides four “prerequisites” for any class certification,\nincluding that there must be “questions of law or fact common\nto the class.” If these prerequisites are met, Rule 23(b)(3)\npermits certification if, among other things, “questions of law\nor fact common to class members predominate over any\nquestions affecting only individual members.” For purposes of\nthese rules, a “common” question is one that is “capable of\nclasswide resolution—which means that determination of its\ntruth or falsity will resolve an issue that is central to the validity\nof each one of the claims in one stroke.” Wal-Mart Stores, Inc.\nv. Dukes, 564 U.S. 338, 350 (2011). In contrast, an\n“individual” question is one for which “members of a proposed\n\f 7\nclass will need to present evidence that varies from member to\nmember.” Tyson Foods, Inc. v. Bouaphakeo, 136 S. Ct. 1036,\n1045 (2016) (quotation marks omitted).\n\n The party seeking class certification “must affirmatively\ndemonstrate” that the commonality and predominance\nrequirements are satisfied. Wal-Mart, 564 U.S. at 350. This\nrequires a “rigorous analysis” that often will “overlap with the\nmerits.” Id. at 351 (quotation marks omitted). Three recent\ncases address the contours of this analysis. In Wal-Mart, the\nSupreme Court ordered decertification of a Title VII class\nwhere the plaintiffs presented insufficient proof that the\ndefendant had engaged in a general policy of sex\ndiscrimination. Id. at 352–55. In Comcast, the Court ordered\ndecertification where the regression analysis used to show\ncommon injury did not track the underlying theory of liability.\n569 U.S. at 36–38. The Court rejected a contention that, at the\nclass-certification stage, “any method of measurement is\nacceptable so long as it can be applied classwide, no matter\nhow arbitrary the measurements may be.” Id. at 36. But in\nTyson Foods, the Court held that concerns about the validity of\na statistical sample used to prove class-wide averages—which\nthe Court described as presenting a common objection to the\nclaims of each class member—should be addressed “as a\nmatter of summary judgment, not class certification.” 136 S.\nCt. at 1047 (quotation marks omitted).\n\n IV\n\n The direct-purchaser plaintiffs raise claims under section\n4 of the Clayton Act, which provides treble damages to any\nperson “injured in his business or property by reason of\nanything forbidden in the antitrust laws.” 15 U.S.C. § 15(a).\nTo establish liability under section 4, each plaintiff must prove\nnot only an antitrust violation, but also an injury to its business\n\f 8\nor property and a causal relation between the two. Without\ncommon proof of injury and causation, section 4 plaintiffs\ncannot establish predominance. See, e.g., Comcast, 569 U.S.\nat 36–38; Rail Freight I, 725 F.3d at 252–53.\n\n The parties dispute the extent to which a court, in\nconducting the “hard look” required by Rule 23, should assess\nthe reliability of common evidence. Specifically, they contest\nwhether Rule 23 requires reliability above and beyond what is\nnecessary to establish the admissibility of expert testimony\nunder Federal Rule of Evidence 702 and Daubert v. Merrell\nDow Pharmaceuticals, Inc., 509 U.S. 579 (1993). That\nquestion arose because the district court found Dr. Rausser’s\nregression models to be reliable for admissibility purposes,\nRail Freight II, 292 F. Supp. 3d at 54–63; held that “reliability\nunder Rule 23 is a higher standard than reliability under\nDaubert,” id. at 91; and found no predominance based on\nconcerns that “undermine[d] the reliability of Dr. Rausser’s\ndamages model,” id. at 122. The defendants embrace this\nreasoning; they argue that Wal-Mart and Comcast rejected\ncertification based on concerns that common evidence was\nunreliable, even though its admissibility was uncontested. The\nplaintiffs disagree; they contend that predominance ultimately\nturns on whether the relevant evidence is common or\nindividualized, and they read Tyson Foods as holding that a\ncourt may assess its reliability only in deciding admissibility or\nsummary judgment.\n\n We need not resolve this dispute because Dr. Rausser’s\ndamages model, even if sufficiently reliable, does not prove\nclasswide injury. As the district court explained, his model\nindicates that the proposed class consists of 16,065 shippers.\nRail Freight II, 292 F. Supp. 3d at 136. The plaintiffs maintain\nthat the alleged conspiracy injured every one of them. Yet the\ndamages model also indicates that 2,037 members of the\n\f 9\nproposed class—or 12.7 percent—suffered “only negative\novercharges” and thus no injury from any conspiracy. Id. at\n137. So even assuming the model can reliably show injury and\ncausation for 87.3 percent of the class, that still leaves the\nplaintiffs with no common proof of those essential elements of\nliability for the remaining 12.7 percent. The district court held\nthat the need for “individualized inquiries to determine which\nof at least 2,037 (and possibly more) class members were\nactually injured by the alleged conspiracy,” id. at 140,\nprecluded a finding of predominance. Id. at 122. We review a\ndistrict court’s assessment of predominance only for abuse of\ndiscretion. Garcia v. Johanns, 444 F.3d 625, 631 (D.C. Cir.\n2006). Here, we find no such abuse.\n\n The plaintiffs argue that their model measures these\nnegative damages only because of normal prediction error. The\ndistrict court found that prediction error could not “account for\nall—or even a substantial portion of—the 2,037 shippers that\nthe model shows to be uninjured.” Rail Freight II, 292 F. Supp.\n3d at 139. The plaintiffs take issue with that finding, but it is\nnot clearly erroneous. In any event, the plaintiffs’ argument\nabout prediction error at most suggests that their damages\nmodel might falsely have measured no injury for as many as\n2,037 shippers. This line of reasoning describes a possible\nproblem with their own evidence; it does not point to\naffirmative evidence—much less common affirmative\nevidence—that a conspiracy did in fact injure these shippers.\n\n The plaintiffs further argue that predominance does not\nrequire common evidence extending to all class members.\nThat contention appears inconsistent with our statement in Rail\nFreight I that the plaintiffs, to establish predominance, must\n“show that they can prove, through common evidence, that all\nclass members were in fact injured by the alleged conspiracy.”\n725 F.3d at 252; see also id. (“we do expect the common\n\f 10\nevidence to show all class members suffered some injury”).\nDespite these statements, the district court held that our opinion\ndid not require common evidence of injury to all class\nmembers. See Rail Freight II, 292 F. Supp. 3d at 132–34.\nInstead, it agreed with the plaintiffs that common proof\ncovering “virtually all” members of the proposed class, and\nleaving only a “de minimis” number of cases requiring\nindividualized proof of injury and causation, would be enough\nto show predominance. Id. at 135. For the sake of argument,\nwe assume that the district court correctly recognized a de\nminimis exception to the general rule that, for claims under\nsection 4 of the Clayton Act, causation and injury must be\n“capable of classwide resolution,” Wal-Mart, 564 U.S. at 350.\nThe court reasonably concluded that such a de minimis\nexception would not encompass this case.\n\n In assessing how many individual adjudications are too\nmany, both the district court and the parties invoke cases\naddressing the question of when, if ever, a class may include\nconcededly uninjured members. Strictly speaking, this case\ndoes not present that question, for the plaintiffs here insist that\neach member of the proposed class was injured. Nonetheless,\nthe cited cases bear some similarity to this one: Uninjured class\nmembers cannot prevail on the merits, so their claims must be\nwinnowed away as part of the liability determination. And that\nprospect raises the same kind of question at issue here—when\ndoes the need for individualized proof of injury and causation\ndestroy predominance? See, e.g., In re Asacol Antitrust Litig.,\n907 F.3d 42, 51–58 (1st Cir. 2018); In re Nexium Antitrust\nLitig., 777 F.3d 9, 18–22 (1st Cir. 2015).\n\n The plaintiffs complain that the district court arbitrarily\nimposed a six-percent upper limit on the percentage of\nuninjured parties who may be included in a certified class. In\nfact, the court’s analysis was more nuanced. As the court\n\f 11\nexplained, the “few reported decisions” involving uninjured\nclass members “suggest that 5% to 6% constitutes the outer\nlimits of a de minimis number.” Rail Freight II, 292 F. Supp.\n3d at 137. The 12.7 percent figure in this case is more than\ntwice that approximate upper bound reflected in analogous\ncaselaw. Moreover, the district court considered raw numbers\nas well as percentages: six percent of a “class totaling only\nfifty-five” members might be de minimis, but 12.7 percent of\nthis class yields “2,037 uninjured class members” (according\nto the common proof), all of whom would need individualized\nadjudications of causation and injury. Id. at 137–38. Finally,\nthe district court stressed that the plaintiffs have proposed no\n“further way”—short of full-blown, individual trials—“to\nreduce this number and segregate the uninjured from the truly\ninjured.” Id. at 138. None of this was an abuse of discretion.\n\n The absence of any winnowing mechanism sharply\ndistinguishes Nexium, the plaintiffs’ best case. There, the class\nincluded purchasers of a drug allegedly shielded from\ncompetition by the unlawful suppression of a generic\nalternative. 777 F.3d at 13–14. The problem of uninjured class\nmembers arose because a small percentage of the class, due to\nbrand loyalty, would have purchased the drug even if a less\nexpensive generic alternative had been available. Id. at 19–20.\nThe First Circuit held that the uninjured class members could\nmanageably be winnowed by having individual consumers file\nminimal, likely unrebutted affidavit testimony indicating\nwhether, if given the choice, they would have purchased the\nbranded drug or a generic alternative. Id. at 20–21.\n\n Nexium does not support class certification here. For one\nthing, the First Circuit sharply limited that decision in Asacol.\nThere, the Court explained that any winnowing mechanism\nmust be truncated enough to ensure that the common issues\npredominate, yet robust enough to preserve the defendants’\n\f 12\nSeventh Amendment and due process rights to contest every\nelement of liability and to present every colorable defense. 907\nF.3d at 51–54. Moreover, the Court held that Nexium’s\naffidavit mechanism could not satisfy both conditions where\nthe defendant seeks to contest the question whether individual\nclass members would have shifted from the branded drug to a\nless expensive generic alternative. See id. That would require\nindividual trials because genuinely contested affidavits do not\nsupport summary judgment and are inadmissible. Id. Here, the\ndefendants intend to contest whether any of the 2,037 shippers\nsuffered injury as a result of any conspiracy. And the question\npresented in these individual challenges—regarding impacts\non shippers of different sizes, shipping different products in\ndifferent geographic markets, with different transportation\noptions and different degrees of leverage—would be far more\ncomplex than the single unitary question (branded or generic?)\nat issue in Nexium and Asacol.\n\n The plaintiffs also invoke the Supreme Court’s discussion\nof uninjured class members in Tyson Foods. The defendants\nthere sought review of the question whether a certified class\nmay contain any uninjured members. But the Supreme Court\nreserved that question, 136 S. Ct. at 1049, and the district court\nhere decided it in the plaintiffs’ favor, Rail Freight II, 292 F.\nSupp. 3d at 133–35. So, this aspect of Tyson Foods does not\nadvance the plaintiffs’ appeal. The Supreme Court also\naddressed a further “new argument” raised for the first time in\nmerits briefing—that plaintiffs at the certification stage must\nprove that all class members were injured or establish a\nmanageable process for culling out uninjured class members.\n136 S. Ct. at 1049. The Court held that this argument was\n“premature” because the parties disagreed about what culling\nmechanisms might be available and the district court had not\nyet addressed the question. See 136 S. Ct. at 1050. This\nholding rested on the inappropriateness of raising new issues\n\f 13\nfor the first time in Supreme Court merits briefing. See id. It\ndoes not, as the plaintiffs here contend, permit district courts\nconsidering class certification to defer questions about the\nnumber and nature of any individualized inquiries that might\nbe necessary to establish liability. To the contrary, confronting\nsuch questions is part-and-parcel of the “hard look” required\nby Wal-Mart and Comcast, as recognized even by those courts\npermitting a class to include some small number of concededly\nuninjured individuals. See, e.g., In re Asacol, 907 F.3d at 51–\n54; In re Nexium, 777 F.3d at 19–21.\n\n Finally, the plaintiffs argue that the 2,037 class members\nfor whom their damages model shows no injury are a de\nminimis portion of the class because their shipments make up\nless than one percent of the railroads’ overall revenue from the\nalleged conspiracy. But revenue is irrelevant to predominance,\nwhich looks to whether elements such as causation and injury\nmay be proved through common evidence, not how much the\ndefendants benefited from any wrongdoing.\n\n V\n\n Looking beyond Dr. Rausser’s regression analysis, the\nplaintiffs point to other evidence that they say can prove injury\nand causation on a class-wide basis. In Rail Freight I, we\nconcluded that Dr. Rausser’s analysis was “essential” to the\nplaintiffs’ case for certification: “No damages model, no\npredominance, no class certification.” 725 F.3d at 253. The\ndistrict court reached the same conclusion on remand, after\ncareful review of all the documentary and expert evidence.\nThis was not an abuse of discretion.\n\n The plaintiffs invoke documentary evidence that the\ndefendants enforced fuel surcharges “uniformly and with few\nexceptions.” Rail Freight II, 292 F. Supp. 3d at 122; see id. at\n\f 14\n103–07. But imposing fuel surcharges does not show injury\ncaused by a conspiracy. The parties vigorously dispute\nwhether higher overall prices during the class period were\nattributable to causes besides any conspiracy—such as the\nmarked increase in fuel prices that occurred around the\nbeginning of the class period, see J.A. 6592–93. As the district\ncourt explained, Dr. Rausser designed his regression models\nprecisely to control for these kinds of potential alternative\ncauses, see Rail Freight II, 292 F. Supp. 3d at 101, yet his\ndamages model showed that 2,037 shippers were uninjured\ndespite paying fuel surcharges. The plaintiffs’ evidence of\nwidespread fuel surcharges helps explain why the proposed\nclass is a large one, but it neither proves that the 2,037 shippers\nwere injured by the alleged conspiracy nor otherwise compels\na finding of predominance.\n\n Next, the plaintiffs invoke the expert testimony of Dr.\nJames McClave, who argued that the 2,037 shippers must have\nbeen harmed by the conspiracy. Dr. McClave reasoned that\nbecause these shippers made fewer purchases on average, and\nthus had less bargaining power than the rest of the class, they\nmust have been more susceptible to injury. Moreover, Dr.\nMcClave’s own study concluded that the defendants’ smallest\ncustomers—roughly 10,000 shippers who made only one\npurchase either before or during the class period—collectively\npaid higher prices than did larger shippers.\n\n This analysis is not common proof of injury to the 2,037\nshippers. For one thing, of the 10,000 shippers studied by Dr.\nMcClave, only 3,700 were in the class, and Dr. McClave did\nnot indicate how many of them were included among the 2,037\nclass members for whom Dr. Rausser’s model showed no\ninjury. Moreover, evidence that a group of one-time shippers\nmay have paid higher prices collectively does not prove that all\n(or almost all) of them were injured individually. And, as the\n\f 15\ndistrict court explained, Dr. McClave did not attempt to\nidentify how many of the one-time shippers did in fact pay\nmore, let alone which of them were among the 2,037. See Rail\nFreight II, 292 F. Supp. 3d at 139.\n\n The plaintiffs respond that Dr. McClave’s analysis, even if\ninconclusive, at least suggests that reduced bargaining power\nmade the 2,037 shippers more vulnerable to any conspiracy.\nThis line of reasoning parallels one rejected by the Supreme\nCourt in Wal-Mart. There, an expert opined that Wal-Mart’s\n“strong corporate culture” made it “vulnerable to gender bias.”\n564 U.S. at 354 (quotation marks omitted). The Court found\nthis evidence insufficient to prove that injury could be\nestablished on a class-wide basis, because the expert could not\nsay what percentage of adverse employment decisions were in\nfact caused by bias. Id. The McClave study is similarly\nincomplete because, as explained above, it does not attempt to\nidentify which of the small shippers, or what percentage of\nthem, were in fact harmed by the alleged conspiracy.\n\n VI\n\n In Asacol, the First Circuit noted the absence of even a\nsingle case “allowing, under Rule 23, a trial in which thousands\nof class members testify.” 907 F.3d at 57–58. That Court\ndeclined to create “the first such case.” Id. So do we. Given\nthe need in this case for at least 2,037 individual determinations\nof injury and causation, the district court did not abuse its\ndiscretion in denying class certification on the ground that\ncommon issues do not predominate.\n\n Affirmed.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: In re: Rail Freight Fuel Surcharge Antitrust Litigation\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-30\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued September 28, 2018 Decided August 16, 2019\\n Reissued August 30, 2019\\n\\n No. 18-7010\\n\\n IN RE: RAIL FREIGHT FUEL SURCHARGE ANTITRUST\\n LITIGATION - MDL NO. 1869,\\n\\n DAKOTA GRANITE COMPANY, ON BEHALF OF ITSELF AND ALL\\n OTHERS SIMILARLY SITUATED, ET AL.,\\n APPELLANTS\\n\\n v.\\n\\n BNSF RAILWAY COMPANY, ET AL.,\\n APPELLEES\\n\\n\\n Appeal from the United States District Court\\n for the District of Columbia\\n (No. 1:07-mc-00489)\\n\\n\\n Kathleen M. Sullivan argued the cause for appellants.\\nWith her on the briefs were Stephen R. Neuwirth, Sami H.\\nRashid, Michael D. Hausfeld, and Michael P. Lehmann.\\n\\n Carter G. Phillips argued the cause for appellees. With\\nhim on the brief were Joseph R. Guerra, Kathleen Moriarty\\nMueller, Saul P. Morgenstern, Thomas A. Isaacson, John M.\\nNannes, Tara L. Reinhart, J. Scott Ballenger, Veronica S.\\n\\f 2\\nLewis, Samuel M. Sipe, Jr., Linda S. Stein, Andrew S.\\nTulumello, Lucas C. Townsend, and Kent A. Gardiner.\\n\\n Anton Metlitsky and Warren D. Postman were on the brief\\nfor amicus curiae Chamber of Commerce of the United States\\nof America in support of defendants-appellees.\\n\\n Before: GARLAND, Chief Judge, and ROGERS and KATSAS,\\nCircuit Judges.\\n\\n Opinion for the Court filed by Circuit Judge KATSAS.\\n\\n KATSAS, Circuit Judge: This case involves a putative class\\nof over 16,000 shippers allegedly harmed by a price-fixing\\nconspiracy among the nation’s largest freight railroads. The\\ndistrict court denied class certification because the plaintiffs’\\nregression analysis—their evidence for proving causation,\\ninjury, and damages on a class-wide basis—measured negative\\ndamages for over 2,000 members of the proposed class. Based\\non that consideration, we affirm.\\n\\n I\\n\\n This appeal arises out of eighteen antitrust actions\\nconsolidated by the Multidistrict Litigation Panel. The\\ndefendants are the four largest freight railroads in the United\\nStates: BNSF Railway Company; CSX Transportation, Inc.;\\nNorfolk Southern Railway Company; and Union Pacific\\nRailroad Company. The plaintiffs, who are their customers,\\nallege that the railroads conspired to fix rate-based fuel\\nsurcharges. Railroads impose fuel surcharges—additional\\ncharges above the base shipping price—when the price of fuel\\nrises above a certain trigger price. Rate-based surcharges are\\ncalculated as a percentage of the base shipping price.\\n\\f 3\\n Following consolidation, the action was divided into one\\ncase involving direct purchasers and another involving indirect\\npurchasers. All plaintiffs alleged that the railroads violated\\nsection 1 of the Sherman Act, 15 U.S.C. § 1, by conspiring to\\nfix prices. The direct purchasers sought treble damages under\\nsection 4 of the Clayton Act, 15 U.S.C. § 15, and the district\\ncourt held that they stated a claim, In re Rail Freight Surcharge\\nAntitrust Litig., 587 F. Supp. 2d 27 (D.D.C. 2008). The indirect\\npurchasers sought injunctive relief under section 16 of the\\nClayton Act, 15 U.S.C. § 26, and raised various state-law\\nclaims. The district court held that the state claims were\\npreempted by federal law, but it declined to dismiss the federal\\nclaims. In re Rail Freight Surcharge Antitrust Litig., 593 F.\\nSupp. 2d 29 (D.D.C. 2008), aff’d, Fayus Enters. v. BNSF Ry.\\nCo., 602 F.3d 444 (D.C. Cir. 2010).\\n\\n The eight named plaintiffs in the direct-purchaser case—\\nCarter Distributing Company; Dakota Granite Company;\\nDonnelly Commodities, Inc.; Dust Pro, Inc.; Nyrstar Taylor\\nChemicals, Inc.; Olin Corporation; Strates Shows, Inc.; and US\\nMagnesium LLC—moved to certify a class under Federal Rule\\nof Civil Procedure 23(b)(3). The proposed class consisted of\\nall shippers who paid rate-based fuel surcharges for\\nunregulated services purchased from the defendants between\\nJuly 1, 2003 and December 31, 2008. To show that causation,\\ninjury, and damages could be proved on a class-wide basis, the\\nplaintiffs invoked two regression models constructed by their\\neconomist, Dr. Gordon Rausser. The “common factor model”\\nidentified seven variables said to determine the price of the\\ndefendants’ services, including fuel surcharges. The “damages\\nmodel,” controlling for those variables, sought to isolate price\\nincreases attributable to the alleged conspiracy. The railroads\\ncriticized these models on various grounds, including that they\\nmeasured damages for shipments made under legacy contracts\\nfixed before any conspiracy allegedly began.\\n\\f 4\\n The district court initially certified the class. It noted that\\nif individualized proof were necessary to establish causation\\nand injury, then the plaintiffs could not satisfy the Rule\\n23(b)(3) requirement that common questions predominate. See\\nIn re Rail Freight Fuel Surcharge Antitrust Litig., 287 F.R.D.\\n1, 43 (D.D.C. 2012). But the court found Dr. Rausser’s\\nregression analysis to be “plausible” and “workable,” so it\\nconcluded that causation, injury, and damages were\\n“susceptible to proof at trial through evidence common to the\\nclass.” Id. at 67. The court rejected many different criticisms\\nof the regression models, but it did not specifically address the\\nquestion of false positives for legacy contracts.\\n\\n On interlocutory review, we vacated the certification order\\nand remanded for reconsideration in light of Comcast Corp. v.\\nBehrend, 569 U.S. 27 (2013). In re Rail Freight Surcharge\\nAntitrust Litig.–MDL No. 1869, 725 F.3d 244 (D.C. Cir. 2013)\\n(Rail Freight I). We explained that, for an antitrust class\\naction, common questions “cannot predominate where there\\nexists no reliable means of proving classwide injury in fact.”\\nId. at 253. We expressed concern with the district court’s\\nfailure to address “the damages model’s propensity toward\\nfalse positives,” which left us with no way of knowing whether\\n“the overcharges the damages model calculates for class\\nmembers [are] any more accurate than the obviously false\\nestimates it produces for legacy shippers.” Id. at 254. Finally,\\nwe stressed that Rule 23, as construed in Comcast, requires a\\n“hard look at the soundness of statistical models that purport to\\nshow predominance.” Id. at 255.\\n\\n On remand, after permitting supplemental discovery and\\nexpert reports, the district court denied class certification. In\\nre Rail Freight Surcharge Antitrust Litig., 292 F. Supp. 3d 14\\n(D.D.C. 2017) (Rail Freight II). The court concluded that Dr.\\nRausser’s expert opinions were reliable enough to be\\n\\f 5\\nadmissible at trial. Id. at 49–63. But in assessing\\npredominance, the court identified three shortcomings in his\\ndamages model: first, it measured highly inflated damages for\\nintermodal traffic (i.e., shipments traveling by rail and another\\nmode of transportation such as trucks or airplanes), id. at 122–\\n26; second, as we had noted in the earlier appeal, the model\\nerroneously measured damages for shipments made under\\nlegacy contracts, id. at 126–31; and third, the model measured\\nnegative damages—and hence no injury—for over 2,000\\nmembers of the proposed class, id. at 132–41. The court\\nconcluded that any one of these problems was enough to defeat\\nthe plaintiffs’ argument for predominance. Id. at 122.\\n\\n The plaintiffs filed a petition for permission to appeal the\\nclass-certification decision under Federal Rule of Civil\\nProcedure 23(f). A motions panel of this Court granted the\\npetition without prejudice to reconsideration at the merits\\nstage. In re Rail Freight Fuel Surcharge Antitrust Litig.–MDL\\nNo. 1869, No. 17-8005 (D.C. Cir. Dec. 20, 2017).\\n\\n II\\n\\n We begin with the question of our jurisdiction. Orders\\ndenying class certification are neither final decisions under 28\\nU.S.C. § 1291, Coopers & Lybrand v. Livesay, 437 U.S. 463\\n(1978), nor injunctions immediately appealable under 28\\nU.S.C. § 1292(a)(1), Gardner v. Westinghouse Broad. Co., 437\\nU.S. 478 (1978). However, 28 U.S.C. § 1292(e) permits the\\nSupreme Court to promulgate rules creating new categories of\\ndecisions appealable before final judgment. Exercising that\\nauthority, the Court has provided that “[a] court of appeals may\\npermit an appeal from an order granting or denying class-action\\n\\f 6\\ncertification,” if a “petition for permission to appeal” is timely\\nfiled. Fed. R. Civ. P. 23(f).\\n\\n In this case, the plaintiffs filed a timely petition for\\npermission to appeal, which was enough under Rule 23(f) to\\nsecure our jurisdiction. That jurisdiction is “discretionary,”\\nRail Freight I, 725 F.3d at 250, and the railroads contest\\nwhether we should exercise it. But their argument on this point\\nis perfunctory—less than one page of briefing, with no case\\ncitations—and it almost entirely duplicates their merits\\narguments. According to the railroads, we should conclude\\nthat the denial of class certification was correct. Then, we\\nshould dismiss the appeal as raising neither a questionable\\ndecision nor an unsettled issue—considerations that bear on\\nwhether to permit the appeal in the first place, see id. at 250–\\n54. Because that disposition would make little sense at this\\njuncture, we decline to revisit the motions panel’s decision\\naccepting the appeal.\\n\\n III\\n\\n Federal Rule of Civil Procedure 23 sets forth various\\nrequirements for the certification of class actions. Rule 23(a)\\nprovides four “prerequisites” for any class certification,\\nincluding that there must be “questions of law or fact common\\nto the class.” If these prerequisites are met, Rule 23(b)(3)\\npermits certification if, among other things, “questions of law\\nor fact common to class members predominate over any\\nquestions affecting only individual members.” For purposes of\\nthese rules, a “common” question is one that is “capable of\\nclasswide resolution—which means that determination of its\\ntruth or falsity will resolve an issue that is central to the validity\\nof each one of the claims in one stroke.” Wal-Mart Stores, Inc.\\nv. Dukes, 564 U.S. 338, 350 (2011). In contrast, an\\n“individual” question is one for which “members of a proposed\\n\\f 7\\nclass will need to present evidence that varies from member to\\nmember.” Tyson Foods, Inc. v. Bouaphakeo, 136 S. Ct. 1036,\\n1045 (2016) (quotation marks omitted).\\n\\n The party seeking class certification “must affirmatively\\ndemonstrate” that the commonality and predominance\\nrequirements are satisfied. Wal-Mart, 564 U.S. at 350. This\\nrequires a “rigorous analysis” that often will “overlap with the\\nmerits.” Id. at 351 (quotation marks omitted). Three recent\\ncases address the contours of this analysis. In Wal-Mart, the\\nSupreme Court ordered decertification of a Title VII class\\nwhere the plaintiffs presented insufficient proof that the\\ndefendant had engaged in a general policy of sex\\ndiscrimination. Id. at 352–55. In Comcast, the Court ordered\\ndecertification where the regression analysis used to show\\ncommon injury did not track the underlying theory of liability.\\n569 U.S. at 36–38. The Court rejected a contention that, at the\\nclass-certification stage, “any method of measurement is\\nacceptable so long as it can be applied classwide, no matter\\nhow arbitrary the measurements may be.” Id. at 36. But in\\nTyson Foods, the Court held that concerns about the validity of\\na statistical sample used to prove class-wide averages—which\\nthe Court described as presenting a common objection to the\\nclaims of each class member—should be addressed “as a\\nmatter of summary judgment, not class certification.” 136 S.\\nCt. at 1047 (quotation marks omitted).\\n\\n IV\\n\\n The direct-purchaser plaintiffs raise claims under section\\n4 of the Clayton Act, which provides treble damages to any\\nperson “injured in his business or property by reason of\\nanything forbidden in the antitrust laws.” 15 U.S.C. § 15(a).\\nTo establish liability under section 4, each plaintiff must prove\\nnot only an antitrust violation, but also an injury to its business\\n\\f 8\\nor property and a causal relation between the two. Without\\ncommon proof of injury and causation, section 4 plaintiffs\\ncannot establish predominance. See, e.g., Comcast, 569 U.S.\\nat 36–38; Rail Freight I, 725 F.3d at 252–53.\\n\\n The parties dispute the extent to which a court, in\\nconducting the “hard look” required by Rule 23, should assess\\nthe reliability of common evidence. Specifically, they contest\\nwhether Rule 23 requires reliability above and beyond what is\\nnecessary to establish the admissibility of expert testimony\\nunder Federal Rule of Evidence 702 and Daubert v. Merrell\\nDow Pharmaceuticals, Inc., 509 U.S. 579 (1993). That\\nquestion arose because the district court found Dr. Rausser’s\\nregression models to be reliable for admissibility purposes,\\nRail Freight II, 292 F. Supp. 3d at 54–63; held that “reliability\\nunder Rule 23 is a higher standard than reliability under\\nDaubert,” id. at 91; and found no predominance based on\\nconcerns that “undermine[d] the reliability of Dr. Rausser’s\\ndamages model,” id. at 122. The defendants embrace this\\nreasoning; they argue that Wal-Mart and Comcast rejected\\ncertification based on concerns that common evidence was\\nunreliable, even though its admissibility was uncontested. The\\nplaintiffs disagree; they contend that predominance ultimately\\nturns on whether the relevant evidence is common or\\nindividualized, and they read Tyson Foods as holding that a\\ncourt may assess its reliability only in deciding admissibility or\\nsummary judgment.\\n\\n We need not resolve this dispute because Dr. Rausser’s\\ndamages model, even if sufficiently reliable, does not prove\\nclasswide injury. As the district court explained, his model\\nindicates that the proposed class consists of 16,065 shippers.\\nRail Freight II, 292 F. Supp. 3d at 136. The plaintiffs maintain\\nthat the alleged conspiracy injured every one of them. Yet the\\ndamages model also indicates that 2,037 members of the\\n\\f 9\\nproposed class—or 12.7 percent—suffered “only negative\\novercharges” and thus no injury from any conspiracy. Id. at\\n137. So even assuming the model can reliably show injury and\\ncausation for 87.3 percent of the class, that still leaves the\\nplaintiffs with no common proof of those essential elements of\\nliability for the remaining 12.7 percent. The district court held\\nthat the need for “individualized inquiries to determine which\\nof at least 2,037 (and possibly more) class members were\\nactually injured by the alleged conspiracy,” id. at 140,\\nprecluded a finding of predominance. Id. at 122. We review a\\ndistrict court’s assessment of predominance only for abuse of\\ndiscretion. Garcia v. Johanns, 444 F.3d 625, 631 (D.C. Cir.\\n2006). Here, we find no such abuse.\\n\\n The plaintiffs argue that their model measures these\\nnegative damages only because of normal prediction error. The\\ndistrict court found that prediction error could not “account for\\nall—or even a substantial portion of—the 2,037 shippers that\\nthe model shows to be uninjured.” Rail Freight II, 292 F. Supp.\\n3d at 139. The plaintiffs take issue with that finding, but it is\\nnot clearly erroneous. In any event, the plaintiffs’ argument\\nabout prediction error at most suggests that their damages\\nmodel might falsely have measured no injury for as many as\\n2,037 shippers. This line of reasoning describes a possible\\nproblem with their own evidence; it does not point to\\naffirmative evidence—much less common affirmative\\nevidence—that a conspiracy did in fact injure these shippers.\\n\\n The plaintiffs further argue that predominance does not\\nrequire common evidence extending to all class members.\\nThat contention appears inconsistent with our statement in Rail\\nFreight I that the plaintiffs, to establish predominance, must\\n“show that they can prove, through common evidence, that all\\nclass members were in fact injured by the alleged conspiracy.”\\n725 F.3d at 252; see also id. (“we do expect the common\\n\\f 10\\nevidence to show all class members suffered some injury”).\\nDespite these statements, the district court held that our opinion\\ndid not require common evidence of injury to all class\\nmembers. See Rail Freight II, 292 F. Supp. 3d at 132–34.\\nInstead, it agreed with the plaintiffs that common proof\\ncovering “virtually all” members of the proposed class, and\\nleaving only a “de minimis” number of cases requiring\\nindividualized proof of injury and causation, would be enough\\nto show predominance. Id. at 135. For the sake of argument,\\nwe assume that the district court correctly recognized a de\\nminimis exception to the general rule that, for claims under\\nsection 4 of the Clayton Act, causation and injury must be\\n“capable of classwide resolution,” Wal-Mart, 564 U.S. at 350.\\nThe court reasonably concluded that such a de minimis\\nexception would not encompass this case.\\n\\n In assessing how many individual adjudications are too\\nmany, both the district court and the parties invoke cases\\naddressing the question of when, if ever, a class may include\\nconcededly uninjured members. Strictly speaking, this case\\ndoes not present that question, for the plaintiffs here insist that\\neach member of the proposed class was injured. Nonetheless,\\nthe cited cases bear some similarity to this one: Uninjured class\\nmembers cannot prevail on the merits, so their claims must be\\nwinnowed away as part of the liability determination. And that\\nprospect raises the same kind of question at issue here—when\\ndoes the need for individualized proof of injury and causation\\ndestroy predominance? See, e.g., In re Asacol Antitrust Litig.,\\n907 F.3d 42, 51–58 (1st Cir. 2018); In re Nexium Antitrust\\nLitig., 777 F.3d 9, 18–22 (1st Cir. 2015).\\n\\n The plaintiffs complain that the district court arbitrarily\\nimposed a six-percent upper limit on the percentage of\\nuninjured parties who may be included in a certified class. In\\nfact, the court’s analysis was more nuanced. As the court\\n\\f 11\\nexplained, the “few reported decisions” involving uninjured\\nclass members “suggest that 5% to 6% constitutes the outer\\nlimits of a de minimis number.” Rail Freight II, 292 F. Supp.\\n3d at 137. The 12.7 percent figure in this case is more than\\ntwice that approximate upper bound reflected in analogous\\ncaselaw. Moreover, the district court considered raw numbers\\nas well as percentages: six percent of a “class totaling only\\nfifty-five” members might be de minimis, but 12.7 percent of\\nthis class yields “2,037 uninjured class members” (according\\nto the common proof), all of whom would need individualized\\nadjudications of causation and injury. Id. at 137–38. Finally,\\nthe district court stressed that the plaintiffs have proposed no\\n“further way”—short of full-blown, individual trials—“to\\nreduce this number and segregate the uninjured from the truly\\ninjured.” Id. at 138. None of this was an abuse of discretion.\\n\\n The absence of any winnowing mechanism sharply\\ndistinguishes Nexium, the plaintiffs’ best case. There, the class\\nincluded purchasers of a drug allegedly shielded from\\ncompetition by the unlawful suppression of a generic\\nalternative. 777 F.3d at 13–14. The problem of uninjured class\\nmembers arose because a small percentage of the class, due to\\nbrand loyalty, would have purchased the drug even if a less\\nexpensive generic alternative had been available. Id. at 19–20.\\nThe First Circuit held that the uninjured class members could\\nmanageably be winnowed by having individual consumers file\\nminimal, likely unrebutted affidavit testimony indicating\\nwhether, if given the choice, they would have purchased the\\nbranded drug or a generic alternative. Id. at 20–21.\\n\\n Nexium does not support class certification here. For one\\nthing, the First Circuit sharply limited that decision in Asacol.\\nThere, the Court explained that any winnowing mechanism\\nmust be truncated enough to ensure that the common issues\\npredominate, yet robust enough to preserve the defendants’\\n\\f 12\\nSeventh Amendment and due process rights to contest every\\nelement of liability and to present every colorable defense. 907\\nF.3d at 51–54. Moreover, the Court held that Nexium’s\\naffidavit mechanism could not satisfy both conditions where\\nthe defendant seeks to contest the question whether individual\\nclass members would have shifted from the branded drug to a\\nless expensive generic alternative. See id. That would require\\nindividual trials because genuinely contested affidavits do not\\nsupport summary judgment and are inadmissible. Id. Here, the\\ndefendants intend to contest whether any of the 2,037 shippers\\nsuffered injury as a result of any conspiracy. And the question\\npresented in these individual challenges—regarding impacts\\non shippers of different sizes, shipping different products in\\ndifferent geographic markets, with different transportation\\noptions and different degrees of leverage—would be far more\\ncomplex than the single unitary question (branded or generic?)\\nat issue in Nexium and Asacol.\\n\\n The plaintiffs also invoke the Supreme Court’s discussion\\nof uninjured class members in Tyson Foods. The defendants\\nthere sought review of the question whether a certified class\\nmay contain any uninjured members. But the Supreme Court\\nreserved that question, 136 S. Ct. at 1049, and the district court\\nhere decided it in the plaintiffs’ favor, Rail Freight II, 292 F.\\nSupp. 3d at 133–35. So, this aspect of Tyson Foods does not\\nadvance the plaintiffs’ appeal. The Supreme Court also\\naddressed a further “new argument” raised for the first time in\\nmerits briefing—that plaintiffs at the certification stage must\\nprove that all class members were injured or establish a\\nmanageable process for culling out uninjured class members.\\n136 S. Ct. at 1049. The Court held that this argument was\\n“premature” because the parties disagreed about what culling\\nmechanisms might be available and the district court had not\\nyet addressed the question. See 136 S. Ct. at 1050. This\\nholding rested on the inappropriateness of raising new issues\\n\\f 13\\nfor the first time in Supreme Court merits briefing. See id. It\\ndoes not, as the plaintiffs here contend, permit district courts\\nconsidering class certification to defer questions about the\\nnumber and nature of any individualized inquiries that might\\nbe necessary to establish liability. To the contrary, confronting\\nsuch questions is part-and-parcel of the “hard look” required\\nby Wal-Mart and Comcast, as recognized even by those courts\\npermitting a class to include some small number of concededly\\nuninjured individuals. See, e.g., In re Asacol, 907 F.3d at 51–\\n54; In re Nexium, 777 F.3d at 19–21.\\n\\n Finally, the plaintiffs argue that the 2,037 class members\\nfor whom their damages model shows no injury are a de\\nminimis portion of the class because their shipments make up\\nless than one percent of the railroads’ overall revenue from the\\nalleged conspiracy. But revenue is irrelevant to predominance,\\nwhich looks to whether elements such as causation and injury\\nmay be proved through common evidence, not how much the\\ndefendants benefited from any wrongdoing.\\n\\n V\\n\\n Looking beyond Dr. Rausser’s regression analysis, the\\nplaintiffs point to other evidence that they say can prove injury\\nand causation on a class-wide basis. In Rail Freight I, we\\nconcluded that Dr. Rausser’s analysis was “essential” to the\\nplaintiffs’ case for certification: “No damages model, no\\npredominance, no class certification.” 725 F.3d at 253. The\\ndistrict court reached the same conclusion on remand, after\\ncareful review of all the documentary and expert evidence.\\nThis was not an abuse of discretion.\\n\\n The plaintiffs invoke documentary evidence that the\\ndefendants enforced fuel surcharges “uniformly and with few\\nexceptions.” Rail Freight II, 292 F. Supp. 3d at 122; see id. at\\n\\f 14\\n103–07. But imposing fuel surcharges does not show injury\\ncaused by a conspiracy. The parties vigorously dispute\\nwhether higher overall prices during the class period were\\nattributable to causes besides any conspiracy—such as the\\nmarked increase in fuel prices that occurred around the\\nbeginning of the class period, see J.A. 6592–93. As the district\\ncourt explained, Dr. Rausser designed his regression models\\nprecisely to control for these kinds of potential alternative\\ncauses, see Rail Freight II, 292 F. Supp. 3d at 101, yet his\\ndamages model showed that 2,037 shippers were uninjured\\ndespite paying fuel surcharges. The plaintiffs’ evidence of\\nwidespread fuel surcharges helps explain why the proposed\\nclass is a large one, but it neither proves that the 2,037 shippers\\nwere injured by the alleged conspiracy nor otherwise compels\\na finding of predominance.\\n\\n Next, the plaintiffs invoke the expert testimony of Dr.\\nJames McClave, who argued that the 2,037 shippers must have\\nbeen harmed by the conspiracy. Dr. McClave reasoned that\\nbecause these shippers made fewer purchases on average, and\\nthus had less bargaining power than the rest of the class, they\\nmust have been more susceptible to injury. Moreover, Dr.\\nMcClave’s own study concluded that the defendants’ smallest\\ncustomers—roughly 10,000 shippers who made only one\\npurchase either before or during the class period—collectively\\npaid higher prices than did larger shippers.\\n\\n This analysis is not common proof of injury to the 2,037\\nshippers. For one thing, of the 10,000 shippers studied by Dr.\\nMcClave, only 3,700 were in the class, and Dr. McClave did\\nnot indicate how many of them were included among the 2,037\\nclass members for whom Dr. Rausser’s model showed no\\ninjury. Moreover, evidence that a group of one-time shippers\\nmay have paid higher prices collectively does not prove that all\\n(or almost all) of them were injured individually. And, as the\\n\\f 15\\ndistrict court explained, Dr. McClave did not attempt to\\nidentify how many of the one-time shippers did in fact pay\\nmore, let alone which of them were among the 2,037. See Rail\\nFreight II, 292 F. Supp. 3d at 139.\\n\\n The plaintiffs respond that Dr. McClave’s analysis, even if\\ninconclusive, at least suggests that reduced bargaining power\\nmade the 2,037 shippers more vulnerable to any conspiracy.\\nThis line of reasoning parallels one rejected by the Supreme\\nCourt in Wal-Mart. There, an expert opined that Wal-Mart’s\\n“strong corporate culture” made it “vulnerable to gender bias.”\\n564 U.S. at 354 (quotation marks omitted). The Court found\\nthis evidence insufficient to prove that injury could be\\nestablished on a class-wide basis, because the expert could not\\nsay what percentage of adverse employment decisions were in\\nfact caused by bias. Id. The McClave study is similarly\\nincomplete because, as explained above, it does not attempt to\\nidentify which of the small shippers, or what percentage of\\nthem, were in fact harmed by the alleged conspiracy.\\n\\n VI\\n\\n In Asacol, the First Circuit noted the absence of even a\\nsingle case “allowing, under Rule 23, a trial in which thousands\\nof class members testify.” 907 F.3d at 57–58. That Court\\ndeclined to create “the first such case.” Id. So do we. Given\\nthe need in this case for at least 2,037 individual determinations\\nof injury and causation, the district court did not abuse its\\ndiscretion in denying class certification on the ground that\\ncommon issues do not predominate.\\n\\n Affirmed.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4189,"cells":{"id":{"kind":"number","value":4657750,"string":"4,657,750"},"name":{"kind":"string","value":""},"name_abbreviation":{"kind":"string","value":"Coffeyville Resources Refining & Marketing, LLC v. EPA"},"decision_date":{"kind":"string","value":"2019-08-30"},"docket_number":{"kind":"string","value":"17-1044"},"court":{"kind":"string","value":"U.S. Court of Appeals for the District of Columbia Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Before: PILLARD and KATSAS, Circuit Judges, and WILLIAMS, Senior Circuit Judge.\", \"parties\": \"\", \"opinions\": [{\"author\": \"PER CURIAM\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued October 5, 2018 Decided August 30, 2019\\n\\n No. 16-1052\\n\\n ALON REFINING KROTZ SPRINGS, INC.,\\n PETITIONER\\n\\n v.\\n\\n ENVIRONMENTAL PROTECTION AGENCY,\\n RESPONDENT\\n\\n MONROE ENERGY, LLC, ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 16-1055, 17-1255, 17-1259, 18-1021,\\n 18-1024, 18-1025, 18-1029\\n\\n\\n On Petitions for Review of Agency Action of the\\n United States Environmental Protection Agency\\n\\n\\n Samara L. Kline argued the cause for petitioners. With her\\non the briefs were Evan A. Young, Megan H. Berge, Lisa M.\\nJaeger, Brittany M. Pemberton, Clara Poffenberger, Richard\\nS. Moskowitz, Robert J. Meyers, Thomas A. Lorenzen,\\nElizabeth B. Dawson, Warren R. Neufeld, LeAnn M. Johnson,\\nand Jonathan G. Hardin. Albert M. Ferlo Jr. and Krista\\nHughes entered appearances.\\n\\n 2\\n Meghan E. Greenfield, Trial Attorney, U.S. Department of\\nJustice, argued the cause for respondent. With her on the brief\\nwas Jeffrey H. Wood, Acting Assistant Attorney General.\\nDaniel R. Dertke, Attorney, entered an appearance.\\n\\n Robert A. Long Jr. argued the cause for intervenors\\nAmerican Petroleum Institute and Growth Energy. With him\\non the brief were Kevin King, Seth P. Waxman, David M. Lehn,\\nSaurabh Sanghvi, and Claire Chung. Stacy R. Linden entered\\nan appearance.\\n\\n Shannen W. Coffin and Linda C. Bailey were on the brief\\nfor amici curiae NACS, et al. in support of respondent EPA.\\n\\n 3\\n\\n No. 17-1044\\n\\nCOFFEYVILLE RESOURCES REFINING & MARKETING, LLC AND\\n WYNNEWOOD REFINING COMPANY, LLC,\\n PETITIONERS\\n\\n v.\\n\\n ENVIRONMENTAL PROTECTION AGENCY,\\n RESPONDENT\\n\\n ALON REFINING KROTZ SPRINGS, INC., ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 17-1045, 17-1047, 17-1049, 17-1051,\\n 17-1052\\n\\n\\n On Petitions for Review of Action of the\\n United States Environmental Protection Agency\\n\\n\\n Brian Killian argued the cause for petitioner The National\\nBiodiesel Board. With him on the briefs was Douglas A.\\nHastings.\\n\\n Samara L. Kline and Thomas Allen Lorenzen, argued the\\ncauses for Obligated Party Petitioners. With them on the briefs\\nwere Evan A. Young, Lisa M. Jaeger, Brittany M. Pemberton,\\nClara Poffenberger, Richard S. Moskowtiz, Robert J. Meyers,\\nElizabeth B. Dawson, David W. DeBruin, Thomas J. Perrelli,\\nMatthew E. Price, LeAnn M. Johnson, and Jonathan G.\\n\\n 4\\nHardin. David Y. Chung, Eric D. Miller, and Albert M. Ferlo\\nJr. entered appearances.\\n\\n Patrick R. Jacobi and Samara M. Spence, Attorneys, U.S.\\nDepartment of Justice, argued the causes for respondent. With\\nthem on the brief was Jeffrey H. Wood, Acting Assistant\\nAttorney General.\\n\\n Thomas Allen Lorenzen argued the cause for intervenors\\nAmerican Fuel & Petrochemical Manufacturers and American\\nPetroleum Institute in support of respondent regarding\\nBiomass-Based Diesel Issues. With him on the brief were\\nRobert J. Meyers, Elizabeth B. Dawson, Richard S. Moskowitz,\\nRobert A. Long, Jr., and Kevin King. Stacy R. Linden entered\\nan appearance.\\n\\n Robert A. Long, Jr., Kevin King, Bryan M. Killian,\\nDouglas A. Hastings, Seth P. Waxman, David M. Lehn,\\nSaurabh Sanghvi, and Claire H. Chung were on the brief for\\nintervenors Growth Energy, et al. in support of respondent.\\nEric D. Miller entered an appearance.\\n\\n Before: PILLARD and KATSAS, Circuit Judges, and\\nWILLIAMS, Senior Circuit Judge.\\n\\n Opinion for the Court filed PER CURIAM.\\n\\n Opinion concurring in part and concurring in the judgment\\nfiled by Senior Circuit Judge WILLIAMS.\\n\\n TABLE OF CONTENTS\\n\\nI. Introduction ..................................................................... 6\\nII. Background ..................................................................... 7\\n\\n 5\\n A. Legal Background ........................................................ 7\\n B. Procedural Background ............................................. 12\\n 1. 2007, 2010, and 2017 Point of Obligation\\n Proceedings ............................................................ 12\\n 2. 2017 Annual Volumetric Proceedings ................... 15\\nIII. Standard of Review .................................................... 17\\nIV. 2010 Point of Obligation Rule ................................... 17\\n A. Jurisdiction................................................................. 17\\n 1. Final Agency Action Under Section 7607(b)(1).... 19\\n 2. After-Arising Grounds Under Section\\n 7607(b)(1) .............................................................. 28\\n 3. Mandatory Reconsideration Under Section\\n 7607(d)(7)(B) ......................................................... 29\\n B. Merits of Challenges to EPA’s Refusal to Revise\\n the 2010 Point of Obligation Rule ............................. 32\\nV. 2017 Annual Volumetric Rule ...................................... 41\\n A. Point of Obligation .................................................... 42\\n 1. Jurisdiction ............................................................. 42\\n 2. Merits ..................................................................... 43\\n B. Cellulosic Biofuel Projection..................................... 53\\n C. Cellulosic Waiver ...................................................... 58\\nVI. 2018 Volume for Biomass-Based Diesel ................... 62\\n A. NBB’s Standing ......................................................... 63\\n B. Merits of NBB’s Challenges ...................................... 65\\nVII. Conclusion ................................................................. 70\\n\\n 6\\n PER CURIAM:\\n\\n\\n I. Introduction\\n\\n The Clean Air Act requires EPA to publish “renewable\\nfuel standards,” ultimately expressed as “applicable\\npercentages,” each year to ensure that the total supply of\\ntransportation fuel sold or imported into the United States\\ncontains specified proportions of each of four categories of\\nrenewable fuels. Congress intended the Renewable Fuel\\nStandards (RFS) program to “move the United States toward\\ngreater energy independence and security” and “increase the\\nproduction of clean renewable fuels.” See Energy\\nIndependence and Security Act of 2007 (EISA), Pub. L. No.\\n110-140, preamble, 121 Stat. 1492 (2007) (codified at 42\\nU.S.C. § 7545(o)).\\n\\n In these related cases, Alon Refining Krotz Springs,\\ntogether with other petroleum refineries and their trade\\nassociations—the “Alon Petitioners”—seek review of EPA’s\\ndecision not to revise its 2010 point of obligation regulation\\nrequiring refineries and importers, but not blenders, to bear the\\ndirect compliance obligation of ensuring that transportation\\nfuels sold or introduced into the U.S. market include the\\nrequisite percentages of renewables. Coffeyville Resources\\nRefining & Marketing and another group of refineries and trade\\nassociations—the “Coffeyville Petitioners”—challenge EPA’s\\nrefusal to reassess the appropriateness of the point of obligation\\nin the context of its 2017 annual volumetric rule, which set the\\n2017 applicable percentages for all four categories of\\nrenewable fuel and the 2018 applicable volume for one subset\\nof such fuel, biomass-based diesel. See 81 Fed. Reg. 89,746\\n(Dec. 12, 2016) (2017 Rule). The Coffeyville Petitioners also\\ncontend that EPA arbitrarily set the 2017 percentage standards\\n\\n 7\\ntoo high. The National Biodiesel Board (NBB)—a biomass-\\nbased diesel industry trade association—separately contends\\nthat EPA set the 2018 applicable volume for biomass-based\\ndiesel too low. Various trade associations representing\\nrefineries and producers of renewable fuels have intervened in\\nsupport of EPA. For the reasons that follow, we deny each of\\nthe petitions for review, many of which recycle arguments\\nraised and rejected in prior challenges.\\n\\n II. Background\\n\\nA. Legal Background\\n\\n Congress established the RFS program in 2005 as part of\\nthe Energy Policy Act, Pub. L. No. 109-58, 119 Stat. 594\\n(2005) (as amended at 42 U.S.C. § 7545(o)). The statute\\nmandates the gradual introduction of four nested categories of\\nrenewable fuels into the United States’ supply of gasoline,\\ndiesel, and other transportation fuels. See 42 U.S.C.\\n§ 7545(o)(2)(B). These categories include: (1) total renewable\\nfuel; (2) advanced biofuel; (3) cellulosic biofuel; and (4)\\nbiomass-based diesel. Id. § 7545(o)(2)(A)(i), (B). The\\numbrella category, total renewable fuel, covers the three other\\ncategories plus any conventional renewable fuels, such as corn-\\nbased ethanol. See id. § 7545(o)(1)(F), (J), (2)(A)(i). The\\nadvanced biofuel subset includes any renewable fuel (except\\nethanol from cornstarch) that has at least 50% lower lifecycle\\ngreenhouse gas emissions than fossil fuels. Id.\\n§ 7545(o)(1)(B). The statute further specifies two\\nnonexclusive subsets of advanced biofuels: cellulosic biofuel\\n(a renewable fuel derived from cellulose materials such as corn\\nstalks and husks) and biomass-based diesel (a diesel fuel\\nsubstitute made from feedstocks such as animal fats). Id.\\n§ 7545(o)(1)(B), (D), (E); EPA Coffeyville Br. 4–5. The\\n\\n 8\\nfollowing figure depicts the nested nature of the four fuel\\ncategories.\\n\\n\\n\\n\\n Source: Coffeyville Br. 11.\\n Four tables in the statute set forth gradually increasing\\nannual “applicable volume” requirements for each category of\\nrenewable fuel. See 42 U.S.C. § 7545(o)(2)(B)(i). The statute\\nsets applicable volumes for biomass-based diesel through\\n2012, id. § 7545(o)(2)(B)(i)(IV), and applicable volumes for\\nthe other three categories through 2022, id.\\n§ 7545(o)(2)(B)(i)(I)–(III). Under those tables, as the total\\nquantities of renewable fuel rise over time, the ratio of\\nadvanced biofuels relative to conventional renewable fuel\\ngradually increases. Id. For compliance years (which match\\ncalendar years) after those specified in the tables, the statute\\nrequires EPA, in coordination with the Secretaries of Energy\\nand Agriculture, to set the annual applicable volumes based on\\n\\n 9\\na review of the implementation of the program plus an analysis\\nof six listed factors. Id. § 7545(o)(2)(B)(ii). For years not\\nspecified in the table, EPA must publish the applicable volumes\\nfourteen months before the year in which they will apply—\\nvolumes that, shortly before the start of the compliance year,\\nEPA translates into percentage standards. Id.\\n Various “waiver” provisions require or permit EPA to\\nlower the annual applicable volumes. Two are relevant for the\\npurposes of this case. First, under the “cellulosic waiver\\nprovision,” EPA must make its own projection of the volume\\nof cellulosic biofuel that will be produced in the following year.\\nId. § 7545(o)(7)(D)(i). If that projection is less than the\\nstatutory figure, the agency must use its own projection as the\\napplicable volume of cellulosic biofuel. Id.; see Am. Petroleum\\nInst. v. EPA, 706 F.3d 474, 477–80 (D.C. Cir. 2013) (API). The\\nsame cellulosic waiver provision authorizes (but does not\\nrequire) EPA to also reduce the advanced biofuel and total\\nrenewable biofuel volume requirements “by the same or a\\nlesser volume” as the cellulosic biofuel reduction, 42 U.S.C.\\n§ 7545(o)(7)(D)(i), and EPA has “broad discretion” regarding\\nwhether and how to do that, Monroe Energy, LLC v. EPA, 750\\nF.3d 909, 915 (D.C. Cir. 2014). Separately, under the “general\\nwaiver provision,” EPA may reduce any of the statutory\\napplicable volumes if it determines “that implementation . . .\\nwould severely harm the economy or environment,” or “that\\nthere is an inadequate domestic supply.” 42 U.S.C.\\n§ 7545(o)(7)(A); see Ams. for Clean Energy v. EPA, 864 F.3d\\n691, 707–13 (D.C. Cir. 2017) (ACE).\\n After EPA determines the waiver-adjusted applicable\\nvolumes, it must translate those volumes into “renewable\\nvolume obligation[s]” for each category of renewable fuel for\\nthe upcoming compliance year. 42 U.S.C. § 7545(o)(3)(B)(i).\\nThe volume obligation for each category of renewable fuel is\\nexpressed as an “applicable percentage,” also known as a\\n\\n 10\\n“percentage standard,” calculated by dividing the adjusted\\napplicable volume for that category of fuel by the total\\nanticipated volume of non-renewable transportation fuel that\\nwill be introduced into commerce (which EPA derives based\\non an estimate provided by the Energy Information\\nAdministration) in the coming compliance year. Id.\\n§ 7545(o)(3)(A), (B)(ii)(II); 40 C.F.R. § 80.1405(c). The\\nstatute calls on EPA to publish the percentage standards not\\nlater than November 30—a month before the start of the\\ncompliance year. 42 U.S.C. § 7545(o)(3)(B)(i).\\n EPA must place the renewable volume obligations on\\n“refineries, blenders, and importers, as appropriate.” 42 U.S.C.\\n§ 7545(o)(3)(B)(ii)(I); see also id. § 7545(o)(2)(A) (requiring\\nEPA to promulgate implementing regulations, including\\n“compliance provisions applicable to refineries, blenders,\\ndistributors, and importers, as appropriate,” designed to ensure\\nthat transportation fuel sold or introduced into the United States\\n“contains at least” the required annual applicable volumes).\\nThe entities that EPA designates to meet the volume\\nobligations are known as “obligated parties.” Monroe Energy,\\n750 F.3d at 912. Each obligated party must ensure that the\\nvolume of non-renewable fuel it sells or introduces into U.S.\\ncommerce is matched by selling or introducing a corresponding\\nvolume of each category of renewable fuel at the level EPA’s\\npercentage standard requires for that category. See ACE, 864\\nF.3d at 699. The percentage standards are set in the\\nanticipation that, if each obligated party meets them and EPA’s\\nprojection regarding the country’s total transportation fuel\\nsupply bears out, the amount of each category of renewable\\nfuel introduced into the economy in the upcoming compliance\\nyear will equal the applicable volumes for that year. Id.\\nObligated parties bear no direct responsibility for any shortfalls\\nin the applicable volumes so long as they comply with the\\npercentage standards.\\n\\n 11\\n EPA assigns a set of “renewable identification numbers”\\n(RINs) to each batch of renewable fuel that is produced or\\nimported for use in the United States. 40 C.F.R. § 80.1426; see\\n42 U.S.C. § 7545(o)(5); Monroe Energy, 750 F.3d at 913. The\\nnumber of RINs assigned to each batch corresponds to the\\namount of ethanol-equivalent energy per gallon in that batch.\\nSee 40 C.F.R. § 80.1415; Monroe Energy, 750 F.3d at 913.\\nRINs remain attached to the renewable fuel until that fuel is\\npurchased by an obligated party or blended into fossil fuels to\\nbe used for transportation fuel. See ACE, 864 F.3d at 699\\n(citing 40 C.F.R. § 80.1429(b)(1)–(2)). At that point the RINs\\nbecome “separated,” meaning they are, in effect, a form of\\ncompliance credit. Id. Obligated parties demonstrate their\\ncompliance with their renewable fuel obligations by “retiring”\\nRINs in annual compliance demonstrations to EPA. 40 C.F.R.\\n§§ 80.1427(a), 80.1451(a)(1).\\n Because the four categories of renewable fuel are nested,\\nobligated parties can comply with their obligations for a type\\nof fuel by retiring any combination of RINs corresponding to\\nthat category of fuels or any subset thereof. See 40 C.F.R.\\n§ 80.1427(a)(3)(i). For instance, retiring a cellulosic biofuel or\\nbiomass-based diesel RIN counts not only toward the volume\\nobligation for that fuel, but also toward both the advanced\\nbiofuel and total renewable fuel obligations. Thus, “if one\\nmillion gallons of cellulosic biofuel are blended into the fuel\\nsupply, the statute allows those one million gallons to be\\ncredited toward the advanced biofuel and total renewable fuel\\nobligations in addition to the cellulosic biofuel obligation.”\\nACE, 864 F.3d at 698.\\n Obligated parties who have more RINs than they need may\\nsell or trade their excess, 40 C.F.R. § 80.1428(b), or they may\\n“bank” those RINs for use to meet up to 20 percent of their\\nobligations for the following compliance year, Monroe Energy,\\n750 F.3d at 913; see 40 C.F.R. § 80.1427(a)(1), (5); Regulation\\n\\n 12\\nof Fuels and Fuel Additives: Changes to Renewable Fuel\\nStandard Program, 75 Fed. Reg. 14,670, 14,734–35 (Mar. 26,\\n2010). Obligated parties without enough RINs to meet their\\ncompliance obligations may purchase RINs, use banked RINs\\nfrom the prior year, or carry a deficit forward to the following\\nyear to be satisfied together with the following year’s\\nobligations. See ACE, 864 F.3d at 699–700; see also 42 U.S.C.\\n§ 7545(o)(5)(D); 40 C.F.R. § 80.1427(b).\\n\\nB. Procedural Background\\n\\n The procedural history of these cases follows two paths:\\nfirst, the proceedings relevant to the challenge that EPA\\narbitrarily declined to initiate a rulemaking to modify the 2010\\nregulation designating refineries and importers, but not\\nblenders, as obligated parties; and second, the proceedings\\nchallenging the 2017 Rule.\\n\\n 1. 2007, 2010, and 2017 Point of Obligation\\n Proceedings\\n\\n In its 2007 regulations implementing the RFS program,\\nEPA designated refiners and importers, but not blenders, as the\\n“appropriate” parties to meet the renewable fuel obligation. 72\\nFed. Reg. 23,900, 23,923–24 (May 1, 2007). At the time, those\\ndesignations were not challenged in court. EPA reaffirmed its\\ndesignations in a 2010 regulation now commonly known as the\\n“point of obligation rule.” 75 Fed. Reg. at 14,721–22 (codified\\nat 40 C.F.R. § 80.1406(a)(1)). During the 2010 rulemaking,\\nseveral refiners—including petitioner Valero Energy\\nCorporation—argued that failing to obligate blenders, who\\ncombine renewable fuel with fossil fuels, would make the RFS\\nprogram unworkable. EPA concluded that the program was\\nfunctioning adequately and that the burdens and disruption\\nfrom changing the point of obligation would outweigh any\\n\\n 13\\nbenefits. See Summary and Analysis of Comments 3.9.2, Alon\\nJ.A. 287–90. Although other aspects of the 2010 regulations\\nwere challenged in court, see, e.g., Nat’l Chicken Council v.\\nEPA, 687 F.3d 393 (D.C. Cir. 2012); Nat’l Petrochemical &\\nRefiners Ass’n v. EPA, 630 F.3d 145 (D.C. Cir. 2010), the point\\nof obligation rule was not.\\n\\n On December 14, 2015, EPA promulgated the volume\\nrequirements for 2014, 2015, and 2016. Renewable Fuel\\nStandard Program, 80 Fed. Reg. 77,420 (Dec. 14, 2015). In so\\ndoing, EPA exercised its general waiver authority to lower the\\ntotal renewable fuel volumes based on a finding of inadequate\\ndomestic supply due to market factors “affecting the ability to\\ndistribute, blend, dispense, and consume . . . renewable fuels”\\nat the levels required by statute. Id. at 77,435/2. Among those\\nfactors was “the slower than expected development of the\\ncellulosic biofuel industry.” Id. at 77,422. The agency thought\\nan additional “real world constraint[]” was the “E10\\nblendwall”—the difficulty for most American vehicle engines\\nto run on blends containing more than 10% ethanol. Id. at\\n77,423. EPA explained that those factors made the statutory\\nrequirements “impossible to achieve.” Id. at 77,422/2. This\\nCourt later vacated the general waiver on the ground that EPA\\nhad misinterpreted the statutory term “inadequate domestic\\nsupply” to include demand-side constraints such as the E10\\nblendwall. See ACE, 864 F.3d at 704–13.\\n\\n On February 12, 2016, sixty days after EPA promulgated\\nthe volume requirements for 2014–16, the Alon Petitioners\\npetitioned this Court for review of the 2010 point of obligation\\nrule. These petitions contend that the rule was arbitrary and\\ncapricious insofar as it failed to impose the obligation on\\ndownstream blenders—the parties petitioners think are best\\nable to comply with it. The petitions assert jurisdiction under\\nthe after-arising provision in 42 U.S.C. § 7607(b)(1), which\\n\\n 14\\npermits otherwise-untimely challenges to a rule if the\\nchallenges are “based solely on grounds arising after” the sixty-\\nday deadline for seeking judicial review. The petitioners assert\\nthat EPA’s exercise of its general waiver authority in the 2014–\\n16 volume regulations, and its acknowledgment of the RFS\\nprogram’s shortcomings as of that time, provided such an after-\\narising ground.\\n\\n The Alon Petitioners simultaneously petitioned EPA to\\nrevise the point of obligation rule. Some of their requests were\\nstyled as petitions for a rulemaking. Others were styled as\\npetitions for mandatory reconsideration under 42 U.S.C.\\n§ 7607(d)(7)(B), which requires EPA to reconsider a rule if\\ncentrally important objections were impracticable to raise\\nduring the comment period or “arose after” that period “but\\nwithin the time specified for judicial review.” The petitions\\ncited the waiver in the 2014–16 volume regulations and EPA’s\\nacknowledgment of program difficulties as grounds supporting\\nmandatory reconsideration. This Court held in abeyance the\\npetitions for review of the point of obligation rule pending\\nresolution of the petitions to revise it.\\n\\n On November 10, 2016, EPA published a proposed denial\\nof the petitions to revise the point of obligation rule. On\\nNovember 22, 2017, after reviewing more than 18,000\\ncomments on the proposal, EPA denied the petitions. It\\nconcluded that the statutory requirements for mandatory\\nreconsideration were not met, so it treated all the filings as\\npetitions for a rulemaking. Denial of Petitions for Rulemaking\\nto Change the RFS Point of Obligation, EPA-HQ-OAR-2016-\\n0544-0525, at 7 (Nov. 22, 2017) (EPA Denial), Alon J.A. 61.\\nEPA then denied the petitions on the ground that “changing the\\npoint of obligation would . . . likely result in a decrease in the\\nproduction, distribution, and use of [renewable] fuels” and\\nwould “do nothing to incentivize the research, development,\\n\\n 15\\nand commercialization of cellulosic biofuel technologies\\ncritical for the growth of the RFS program in future years.”\\nEPA Denial at 8–9, Alon J.A. 62–63.\\n\\n Within sixty days (in December 2017 and January 2018),\\nthe Alon Petitioners sought judicial review of that denial,\\nwhich it cast as a final agency action under section 7607(b)(1).\\nThe two sets of petitions—the February 2016 petitions for\\nreview of the 2010 point of obligation rule and the 2017–18\\npetitions for review of EPA’s refusal to reconsider the rule—\\nwere consolidated and are now before us.\\n\\n 2. 2017 Annual Volumetric Proceedings\\n\\n EPA issued its 2017 annual volumetric rule on December\\n12, 2016. The 2017 Rule establishes: (1) the applicable volume\\nfor biomass-based diesel for 2018, 81 Fed. Reg. at 89,751/2;\\n(2) the waiver-adjusted applicable volumes for cellulosic\\nbiofuel, advanced biofuel, and total renewable fuel for 2017,\\nid. at 89,747 tbl. I-1; and (3) percentage standards for all four\\nfuel types for 2017, id. at 89,751, tbl. I.B.6-1.\\n EPA exercised its mandatory cellulosic waiver authority to\\ndecrease the 2017 applicable volume for cellulosic biofuel by\\nmore than 94 percent, dropping 5.189 billion gallons from the\\nstatutory target of 5.5 billion gallons, to 311 million gallons.\\nId. at 89,750/2; 42 U.S.C. § 7545(o)(2)(B)(i)(III). EPA then\\nhad discretion under that same waiver authority to cut as much\\nas 5.189 billion gallons off the statutory volumes for advanced\\nbiofuel and total renewable fuel. See 42 U.S.C.\\n§ 7545(o)(7)(D)(i); 81 Fed. Reg. at 89,762 & tbl. IV.A-1. EPA\\npartially exercised that authority, reducing the 9-billion-gallon\\nstatutory target for advanced biofuel by 4.72 billion gallons,\\nresulting in an adjusted applicable volume of 4.28 billion\\ngallons—a greater than 50% decrease. 81 Fed. Reg. at 89,750–\\n51; 42 U.S.C. § 7545(o)(2)(B)(i)(II). EPA reduced the total\\n\\n 16\\nrenewable fuel volume requirements by the same amount,\\nlowering the statutory target of 24 billion gallons to 19.28\\nbillion gallons. 81 Fed. Reg. at 89,751/1; 42 U.S.C.\\n§ 7545(o)(2)(B)(i)(I). EPA considered but decided against also\\nusing its general waiver authority to further lower the\\napplicable volume of total renewable fuel. 81 Fed. Reg. at\\n89,751/1.\\n Using the waiver-adjusted applicable volumes, EPA set\\nthe 2017 percentage standards for each of the four renewable\\nfuel categories. See id. at 89,751, 89,799–801. Finally, EPA\\nset the biomass-based diesel applicable volume for 2018 at 2.1\\nbillion gallons. Id. at 89,751/2. EPA received comments\\nurging it to reassess the point of obligation in the 2017 Rule,\\nbut declined to address them on the grounds that the comments\\nwere “beyond the scope” of the 2017 rulemaking. Response to\\nComments at 542, Coffeyville J.A. 761.\\n After EPA published the 2017 Rule, various parties\\npetitioned for judicial review. The Coffeyville Petitioners\\ncontend that EPA erred by refusing to reconsider which types\\nof parties would bear the direct compliance obligation under\\nthe 2017 Rule. They also argue that EPA arbitrarily calculated\\nthe 2017 production of cellulosic biofuel and arbitrarily\\nexercised its discretionary cellulosic waiver authority, resulting\\nin percentage standards that are too high. NBB argues that\\nEPA set the 2018 applicable volume for biomass-based diesel\\ntoo low by considering factors it should not have and omitting\\nor incorrectly assessing others. Two trade associations\\nrepresenting refineries have intervened in defense of EPA’s\\nbiomass-based diesel decision, and a coalition of trade\\nassociations representing renewable fuel producers and\\nrefineries have intervened to oppose the Coffeyville\\n\\n 17\\nPetitioners’ claims. None of the petitioners’ challenges\\nsucceeds.\\n\\n III. Standard of Review\\n\\n “This court applies the familiar, deferential standard\\nannounced in Chevron, U.S.A., Inc. v. Natural Resources\\nDefense Council, Inc., to sustain any reasonable agency\\ninterpretation of ambiguity in the Clean Air Act.” Nat’l Ass’n\\nfor Surface Finishing v. EPA, 795 F.3d 1, 7 (D.C. Cir. 2015).\\n“We employ the deferential State Farm standard of review\\nwhen reviewing arguments based on allegedly arbitrary or\\nunreasoned agency action.” ACE, 864 F.3d at 726 (citing\\nMotor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut.\\nAuto. Ins. Co., 463 U.S. 29, 43 (1983)). Under that rubric,\\nEPA’s actions are “presumptively valid provided [they] meet[]\\na minimum rationality standard.” Nat. Res. Def. Council, Inc.\\nv. EPA, 194 F.3d 130, 136 (D.C. Cir. 1999). We uphold EPA’s\\nactions so long as they are “reasonable and reasonably\\nexplained.” Jackson v. Mabus, 808 F.3d 933, 936 (D.C. Cir.\\n2015).\\n IV. 2010 Point of Obligation Rule\\n\\n We start with the Alon Petitioners and their challenges to\\nthe 2010 point of obligation rule.\\n\\nA. Jurisdiction\\n\\n We begin, as we must, with our jurisdiction. In general\\nterms, the question presented involves our review of rules\\npromulgated under the Clean Air Act, or EPA’s failure to\\namend them, after the initial window for seeking judicial\\n\\n 18\\nreview has passed. Various statutory provisions frame this\\ninquiry.\\n\\n Section 7607(b)(1) of Title 42 provides for judicial review\\nof regulations promulgated by the Administrator of EPA under\\nthe Clean Air Act. The first sentence of section 7607(b)(1)\\nvests this Court with exclusive jurisdiction to review\\n“nationally applicable regulations promulgated, or final action\\ntaken, by the Administrator” under the Act. The fourth\\nsentence of section 7607(b)(1) specifies the time for seeking\\njudicial review. It imposes a sixty-day time limit, but provides\\nan exception for petitions based on grounds arising after the\\nlimit: “Any petition for review under this subsection shall be\\nfiled within sixty days from the date notice of such\\npromulgation, approval, or action appears in the Federal\\nRegister, except that if such petition is based solely on grounds\\narising after such sixtieth day, then any petition for review\\nunder this subsection shall be filed within sixty days after such\\ngrounds arise.”\\n\\n Section 7607(d) of Title 42 sets forth provisions for\\nvarious rulemakings under the Clean Air Act, including for the\\n“promulgation or revision of any regulation” involving the RFS\\nprogram. Id. § 7607(d)(1)(E). Section 7607(d)(7)(B)\\naddresses various issues regarding exhaustion, agency\\nreconsideration, and judicial review. The first sentence of that\\nprovision imposes a conventional exhaustion requirement,\\nlimiting judicial review to objections “raised with reasonable\\nspecificity during the period for public comment.” The second\\nsentence requires EPA to reconsider regulations in certain\\nnarrow circumstances: “If the person raising an objection can\\ndemonstrate to the Administrator that it was impracticable to\\nraise such objection within such time or if the grounds for such\\nobjection arose after the period for public comment (but within\\nthe time specified for judicial review) and if such objection is\\n\\n 19\\nof central relevance to the outcome of the rule, the\\nAdministrator shall convene a proceeding for reconsideration\\nof the rule and provide the same procedural rights as would\\nhave been afforded had the information been available at the\\ntime the rule was promulgated.” The third sentence of section\\n7607(d)(7)(B) makes any “refusal” to provide such mandatory\\nreconsideration judicially reviewable.\\n\\n At various times in this litigation, the petitioners have\\nasserted three different jurisdictional theories. First, EPA’s\\nrefusal to revise the point of obligation rule in 2017 was a final\\naction reviewable under the first sentence of section\\n7607(b)(1), regardless of the after-arising provision. Second,\\nEPA’s statements and actions in its 2014–16 volume regulation\\nconstitute after-arising grounds permitting a challenge to the\\npoint of obligation rule as promulgated in 2010. Third, these\\nsame EPA statements and actions triggered a right to\\nmandatory reconsideration under section 7607(d)(7)(B), which\\nin turn makes the denial of reconsideration judicially\\nreviewable. As explained below, we conclude that the first\\ncontention is correct, the second has been abandoned, and the\\nthird lacks merit.\\n\\n 1. Final Agency Action Under Section 7607(b)(1)\\n\\n In 2016, various refiners petitioned EPA for a rulemaking\\nto modify the point of obligation rule. The petitions urged that\\nthe need for a modification became evident in 2015, when EPA\\nwaived certain statutory volume requirements and concluded\\nthat changing economic conditions had made the requirements\\n“impossible to achieve.” 2014–16 Rule, 80 Fed. Reg. at\\n77,422/2. In November 2017, EPA denied the rulemaking\\npetitions on the ground that any current problems with the RFS\\nprogram were manageable and that changing the point of\\nobligation at this juncture would be disruptive. EPA Denial at\\n\\n 20\\n8–9, Alon J.A. 62–63. The refiners sought review of the denial\\nin December 2017 and January 2018. As petitioners in this\\nCourt, they contend that the November 2017 denial constituted\\nfinal agency action reviewable under section 7607(b)(1). We\\nagree.\\n\\n As noted above, the first sentence of section 7607(b)(1)\\ngives this Court exclusive jurisdiction to review any nationally\\napplicable “final action” taken by EPA under the Clean Air\\nAct. The parties agree that the denial of the rulemaking\\npetitions was nationally applicable, final, and taken under the\\nClean Air Act. It was also agency “action” within the meaning\\nof the statute. That word “bears the same meaning in [section\\n7607(b)(1)] that it does under the Administrative Procedure\\nAct,” Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 478\\n(2001), which defines “agency” to include EPA, 5 U.S.C.\\n§ 551(1), and “agency action” to include “the whole or a part\\nof an agency rule, order, license, sanction, relief, or the\\nequivalent or denial thereof, or failure to act,” id. § 551(13)\\n(emphases added). So, EPA’s denial of the petitions for\\nrulemaking was a reviewable “action.”\\n\\n The petitions for review were timely. As a general matter,\\nsection 7607(b)(1) requires a petition for review to be filed\\nwithin sixty days of when “notice of such promulgation,\\napproval, or action appears in the Federal Register.” Here, the\\n“action” at issue—denial of the petitions for rulemaking—was\\npublished in the Federal Register on November 22, 2017, and\\nthe petitions for review of that action were filed within sixty\\ndays of that date. Moreover, this conclusion does not depend\\non the after-arising provision. To the contrary, because the\\npetitions for review were filed within sixty days of the “action”\\nunder review, the exception for “grounds arising after such\\nsixtieth day” was not triggered.\\n\\n 21\\n Our caselaw confirms this framing of the jurisdictional\\nissue. In Massachusetts v. EPA, 415 F.3d 50 (D.C. Cir. 2005),\\nthis Court held that EPA’s denial of a petition to regulate\\ngreenhouse gas emissions as air pollutants was itself “final\\naction” reviewable under section 7607(b)(1). See id. at 53–54\\n(opinion of Randolph, J.); id. at 61 (Sentelle, J., concurring in\\nthe judgment). The Supreme Court reversed our judgment on\\nthe merits, but agreed that we had jurisdiction. Massachusetts\\nv. EPA, 549 U.S. 497 (2007). In particular, the Court noted that\\nsection 7607(b)(1) “expressly permits review” of EPA’s\\n“rejection of [a] rulemaking petition.” Id. at 520, 528; see also\\nid. at 517 (section 7607(b)(1) affords “the right to challenge\\n[this] agency action unlawfully withheld”). Likewise, in\\nNatural Resources Defense Council v. EPA, 824 F.2d 1146\\n(D.C. Cir. 1987) (en banc) (NRDC), we held that a 1985\\ndecision to withdraw proposed amendments to certain 1976\\nregulations—which we described as a “decision not to amend”\\nthe regulations—was “reviewable agency action” under\\nsection 7607(b)(1). Id. at 1150 (quotation marks omitted). We\\nconcluded that the petition for review at issue, which on its face\\nchallenged the 1985 withdrawal decision, was not, in\\nsubstance, an untimely “‘back-door’ challenge to the 1976\\nregulations.” Id. On that point, we reasoned that the petitioners\\nclaimed legal errors in the 1985 withdrawal and sought vacatur\\nonly of that order. Likewise, in this case, the petitions for\\nreview filed in 2017 and 2018 raise no back-door challenge to\\nthe 2010 regulation: the petitions contend that EPA in 2017\\narbitrarily refused to take account of changing economic\\nconditions, and they seek vacatur only of the 2017 order\\ndenying a new rulemaking going forward.\\n\\n Background principles of administrative law reinforce our\\nconclusion that the denial of a petition to modify a rule based\\non changed circumstances is itself a reviewable order.\\nOrdinarily, the denial of a petition to amend or rescind a\\n\\n 22\\nregulation is judicially reviewable. See, e.g., NLRB Union v.\\nFLRA, 834 F.2d 191, 195–96 (D.C. Cir. 1987). When the\\npetition to amend attacks defects in the regulation as originally\\npromulgated, and when the time limit for seeking review of the\\nregulation has passed, questions can arise about whether the\\ntime limit is being improperly circumvented. In these\\ncircumstances, we have held that the petitioner cannot raise\\nprocedural challenges to the regulation, but can raise\\nsubstantive arguments that the regulation is unauthorized by or\\nconflicts with a statute. See id. at 196–97. But the\\ncircumvention concern does not even arise when the petitioner\\nraises arguments about changed circumstances or new\\ninformation. Those kinds of arguments—that recent\\ndevelopments compel the amendment of an older regulation—\\nare always cognizable through review of the denial of a petition\\nto amend, though they trigger an “extremely limited” review\\non the merits. Id. at 196. Our decision today harmonizes the\\njudicial-review provisions of the Clean Air Act with this\\ngeneral background principle.\\n\\n EPA recognizes the general rule that, under the NLRB\\nUnion line of cases, the denial of a petition to amend a rule is a\\nreviewable order, which supports both challenges based on\\nrecent developments and substantive challenges to the original\\nregulation. Nonetheless, EPA urges a different rule where the\\napplicable judicial-review provision contains a time limit for\\nseeking review and an exception for grounds arising after the\\ntime limit, as in the Clean Air Act. In those circumstances,\\naccording to EPA, a challenge to the denial of a petition to\\namend is untimely—and the denial is thus entirely\\nunreviewable—unless the after-arising provision is satisfied.\\nEPA rests this conclusion on National Mining Ass’n v.\\nDepartment of Interior, 70 F.3d 1345 (D.C. Cir. 1995), and\\nAmerican Road & Transportation Builders Ass’n v. EPA, 588\\n\\n 23\\nF.3d 1109 (D.C. Cir. 2009) (ARTBA), but neither decision\\nsupports its position.\\n\\n National Mining involved judicial review under the\\nSurface Mining Control and Reclamation Act (SMCRA),\\nwhich requires petitions for review to be filed “within sixty\\ndays” of the agency action at issue “or after such date if the\\npetition is based solely on grounds arising after the sixtieth\\nday.” See 70 F.3d at 1350. In 1986, parties petitioned the\\nDepartment of Interior to rescind a 1979 SMCRA regulation\\non two grounds. First, the petitioners argued that the rule was\\ninconsistent with the statute—an argument attacking the\\nregulation itself and “available” when the regulation was\\noriginally promulgated. See id. After the agency declined to\\nrescind the rule, the challengers sought judicial review. We\\nheld that the after-arising provision made this first challenge\\nuntimely, by explicitly requiring challenges to a regulation\\neither to be filed “within the statutory period” or to “meet the\\nafter-arising test.” Id. at 1350–51. At the same time, however,\\nwe held that the petitioners’ second challenge—to the agency’s\\n1986 decision refusing to “repeal” the regulation based on\\nchanged circumstances—was timely and reviewable. Id. at\\n1352. To be sure, we described the latter challenge as resting\\non “grounds that arose after the sixtieth day.” Id. But we failed\\nto explain what jurisdictional theory that observation\\nsupported: a challenge to the 1979 regulation rendered timely\\nby the after-arising exception; or a challenge, in substance as\\nwell as form, to the 1986 refusal to repeal, akin to the challenge\\nthat we held reviewable in NRDC. Instead, we simply\\nconcluded that, under the “limited scope of our review,” the\\nagency did not “act[] unreasonably in denying the petition for\\nrulemaking.” Id. at 1352–53. Thus, while National Mining\\nblessed jurisdiction to review agency refusals to amend\\nregulations based on changed factual circumstances, it did not\\n\\n 24\\nultimately address what we clarify today—the precise statutory\\nbasis for that jurisdiction.\\n\\n ARTBA applied the reasoning of National Mining to the\\nClean Air Act, which also contains a time limit for judicial\\nreview and an exception for after-arising grounds. ARTBA\\ninvolved a 2002 petition to amend 1997 regulations on the\\nground that they allowed states “to adopt precisely the kinds of\\nregulations that the statute forbids.” 588 F.3d at 1110. As in\\nNational Mining, the challenge was thus a substantive attack\\non a regulation as originally promulgated. We held that, under\\nNational Mining, EPA’s “denial of a revision-seeking petition\\ndoes not allow review of alleged substantive defects in the\\noriginal rule even under the deferential standards applicable to\\nreview of such denials, outside the statutory limitations period\\nrunning from the rule[’s] original promulgation.” Id. at 1113\\n(emphasis added). In other words, National Mining “require[d]\\nus to treat ARTBA’s petition to EPA as a challenge to the\\nregulations it sought revised.” Id. at 1110. As a result,\\ndismissal was necessary unless the petition satisfied the\\nexception for after-arising grounds, which it did not. See id.\\n\\n Neither decision controls here, where the 2017 and 2018\\npetitions for review challenge not the point of obligation\\nregulation as originally promulgated in 2010, but the failure to\\namend the regulation in light of changed circumstances flagged\\nby EPA in 2015. EPA rejected that argument in 2017, and the\\npetitioners sought review of the rejection within sixty days. In\\nsubstance as well as form, the challenge was to the 2017 refusal\\nto amend, not to the underlying 2010 regulation. Under these\\ncircumstances, there was no risk of circumventing the original\\ntime limit. Therefore, there was also no reason to treat the 2010\\npromulgation and the 2017 refusal to amend as one-and-the-\\nsame agency action, despite binding APA definitions treating\\nthem as separate.\\n\\n 25\\n Precedent aside, EPA’s proposed approach—making the\\nafter-arising provision the exclusive vehicle for challenging\\nrefusals to amend regulations based on new information or\\nchanged circumstances—creates various difficulties. For one\\nthing, there is a conceptual mismatch between that provision\\nand these kinds of challenges. Though the after-arising\\nexception and the opportunity to seek rule revision based on\\npost-rulemaking events may seem similar, the first allows an\\nintervening event to secure judicial review on the basis of\\ndefects extant at the time of the rulemaking, whereas the\\nsecond allows review on the question whether intervening\\nevents have fatally undermined the original justification for the\\nrule. The arrow of time runs forward, not backward, so it is at\\nbest awkward—and at worst incoherent—to speak of a later\\ndevelopment rendering unlawful an earlier promulgation.\\nEconomic developments in 2015 may have made it arbitrary\\nfor EPA to adhere to the point of obligation rule in 2017, but\\nthey cannot have retroactively made arbitrary its promulgation\\nin 2010.\\n\\n Even worse, the Clean Air Act’s after-arising provision, if\\nused to judge the timeliness of challenges based on new\\ninformation, would be difficult to apply, capriciously narrow,\\nor both. To satisfy that provision, a petition for review must be\\nboth (i) based “solely” on after-arising grounds and (ii) filed\\n“within sixty days after such grounds arise.” 42 U.S.C.\\n§ 7607(b)(1). Yet the case for changing an environmental\\nregulation will almost never manifest itself at one discrete\\nmoment. Instead, it will accumulate progressively over time,\\nas scientific knowledge advances or economic conditions\\nchange. And so, under EPA’s approach, the relevant filing\\ndeadlines would become practically unknowable. When did\\nsome environmental risk become serious and obvious enough\\nto compel a rulemaking to strengthen an existing regulation?\\nThat will usually be a hard question, and it would be little short\\n\\n 26\\nof miraculous if the answer turned out to be on a date certain\\nwithin sixty days of the filing of a petition for review, as\\nrequired to satisfy the after-arising provision.\\n\\n In contrast, the approach we have sketched out produces\\nsimple questions and discernible deadlines: ask when EPA\\ndenied the rulemaking petition, then add sixty days. If\\npossible, we should avoid trying to fix arbitrarily precise\\naccrual dates for events that develop incrementally over time.\\nSee Nat’l R.R. Passenger Corp. v. Morgan, 536 U.S. 101, 115–\\n21 (2002). And we should avoid jurisdictional tests that are\\ncomplex as opposed to straightforward. See, e.g., Hertz Corp.\\nv. Friend, 559 U.S. 77, 94 (2010). Treating the denial of a\\npetition to amend a rule based on changed circumstances as\\nreviewable agency action honors both principles, while\\nattempting to shoehorn such denials into the after-arising\\nprovision does the opposite.\\n\\n We acknowledge that, in Oljato Chapter of Navajo Tribe\\nv. Train, 515 F.2d 654 (D.C. Cir. 1975), we assumed that the\\nafter-arising provision would govern review of orders denying\\npetitions to modify EPA rules based on changed circumstances.\\nBut our only holdings were that such petitions must first be\\npresented to EPA, id. at 666, and then are reviewable directly\\nin the courts of appeals, id. at 657–65. Moreover, our\\nassumption was understandable in context; the petitioners had\\nmissed the statutory deadline to file a petition for review (then\\nthirty days), see id. at 657, so they pressed alternative\\njurisdictional theories (involving either district-court review or\\nthe after-arising provision) that would avoid that deadline.\\nFurthermore, we expressly reserved the precise bounds of what\\nconstituted a petition based solely on after-arising grounds. See\\nid. at 666–68. Finally, we stressed that review should generally\\nbe available “when new information casts doubt upon the\\nvalidity of a [regulatory] standard.” Id. at 665. At the time, the\\n\\n 27\\nafter-arising provision contained no separate deadline\\nrequiring a petition for review to be filed within sixty days of\\nthe after-arising ground. See id. at 657 n.3. So the difference\\nbetween “direct review” of a regulation through the after-\\narising provision and review of a “refusal to revise” the\\nregulation appeared largely semantic. See id. at 666. But, two\\nyears after Oljato was decided, Congress amended section\\n7607(b)(1) to include the separate filing deadline, see Pub. L.\\nNo. 95-95 § 305(c)(3), 91 Stat. 685, 776 (1977), which, as\\nexplained above, made the after-arising provision a singularly\\npoor vehicle for securing the review that Oljato assumed would\\nbe readily available.\\n\\n After Oljato, we held in Group Against Smog & Pollution\\nv. EPA, 665 F.2d 1284 (D.C. Cir. 1981), that the failure to\\nchallenge a 1974 regulation within the original sixty-day\\ndeadline did not bar “judicial review of the agency’s\\nsubsequent refusal to revise the standard on the basis of new\\ninformation.” Id. at 1291. Quoting liberally from Oljato, we\\nsuggested that such review would proceed through the after-\\narising provision. Id. at 1289. But we permitted review even\\nthough the after-arising provision, as amended, could not have\\napplied. The case involved comments filed with EPA in April\\n1977, which we treated as a petition for a rulemaking. Id. at\\n1290 & n.47. On April 13, 1978, EPA declined to amend the\\nrule as requested. Id. at 1288. The ensuing petition for review\\nwas filed on June 12, 1978—more than a year after the\\ntechnological changes discussed in the comments, but precisely\\nsixty days after the refusal to amend. See id. Because the\\npetition for review was not filed within sixty days of the\\nasserted after-arising grounds, the after-arising provision\\nplainly did not apply. So, review must have rested on the\\ntheory that the refusal to amend was itself a reviewable action\\ntriggering its own sixty-day filing window.\\n\\n 28\\n The approach we follow here—treating denials of\\nrulemakings based on new facts as independently reviewable\\ndecisions—does not reduce the after-arising provision to\\nsurplusage. Our cases have recognized other circumstances\\ntriggering the after-arising provision, including judicial\\ndecisions that significantly “changed the legal landscape”\\nfaced by petitioners, Honeywell Int’l, Inc. v. EPA, 705 F.3d\\n470, 473 (D.C. Cir. 2013), and “the occurrence of an event that\\nripens a claim,” Coalition for Responsible Regulation v. EPA,\\n684 F.3d 102, 129 (D.C. Cir. 2012) (per curiam). In cases like\\nthese—where, for example, an intervening statute, regulation,\\nor judicial decision extends old regulations to new parties—the\\nafter-arising ground is easily dated, the relevant filing\\ndeadlines are clear, and so the provision functions predictably.\\nMoreover, where the after-arising provision does apply, it\\npermits the petitioner to contend not only that changed\\ncircumstances warrant amending an existing regulation but\\nalso that the regulation was unlawful as originally\\npromulgated. See, e.g., Honeywell, 705 F.3d at 473.\\n\\n For these reasons, we conclude that we have jurisdiction\\nto consider the petitioners’ argument that EPA arbitrarily\\nrefused to amend the point of obligation rule based on the\\nchanged circumstances cited by the petitioners.\\n\\n 2. After-Arising Grounds Under Section 7607(b)(1)\\n\\n The February 2016 petitions for review, filed before EPA\\nresolved any of the rulemaking petitions, rested on the\\nalternative jurisdictional theory that EPA’s publication of the\\n2014–16 volume requirements constituted an after-arising\\nground within the meaning of section 7607(b)(1). Our\\nconclusion above does not moot this question, because if this\\nalternative theory were valid, then the petitioners could directly\\nattack the point of obligation rule as originally promulgated.\\n\\n 29\\nNonetheless, the petitioners have abandoned this theory of\\njurisdiction. In their merits briefs, they never actually attack\\nthe 2010 rule as originally promulgated; instead, they challenge\\nonly the 2017 denial of their rulemaking petitions. Moreover,\\nin requesting relief, they do not ask us to set aside the 2010\\nrule, but only to “vacate the [2017] Denial[] and remand to\\nEPA” for a rulemaking to change the point of obligation rule\\ngoing forward. Alon Br. 58. And at oral argument, they\\ndisclaimed any challenge to the 2010 rule itself and confirmed\\nthat their only challenge was to EPA’s 2017 refusal to revise\\nthe point of obligation rule. See Rec. of Oral Argument at\\n2:18:50–2:19:15.\\n\\n 3. Mandatory Reconsideration Under Section\\n 7607(d)(7)(B)\\n\\n Finally, the petitioners assert jurisdiction under section\\n7607(d)(7)(B), on the ground that EPA erroneously denied\\npetitions for mandatory reconsideration of the point of\\nobligation rule. To review, section 7607(d)(7)(B) provides in\\nrelevant part that only objections “raised . . . during the period\\nfor public comment” may be “raised during judicial review.”\\nBut if the objector “can demonstrate to the Administrator that\\nit was impracticable to raise such objection within such time or\\nif the grounds for such objection arose after the period for\\npublic comment (but within the time specified for judicial\\nreview) and if such objection is of central relevance to the\\noutcome of the rule,” then EPA must “convene a proceeding\\nfor reconsideration of the rule,” and any refusal to do so is\\njudicially reviewable. Here, the petitioners assert that the\\nground for their objections—EPA’s statements and actions in\\nits 2014–16 volume regulation—arose “after the period for\\npublic comment” on the 2010 point of obligation rule, and that\\n\\n 30\\nthe objections are “of central relevance to the outcome of the\\nrule.”\\n\\n The petitioners misapprehend the statutory text and\\nstructure. Section 7607(d)(7)(B) does not extend the\\njurisdictional deadline to seek judicial review imposed in\\nsection 7607(b)(1); instead, it specifies a non-jurisdictional\\nexhaustion requirement. See EPA v. EME Homer City\\nGeneration, L.P., 572 U.S. 489, 512 (2014). Its first sentence\\ngenerally requires parties to raise objections “during the period\\nfor public comment” in order to later present them in court. Its\\nsecond sentence allows a narrow exception when a centrally\\nimportant objection cannot feasibly be raised during the\\ncomment period—either because “the grounds for such\\nobjection arose after the period for public comment,” or\\nbecause commenting was otherwise “impracticable.” If an\\nobjection fits within this exception, the consequences are\\nweighty: EPA must grant reconsideration and conduct a new,\\nfull-dress, notice-and-comment rulemaking. And if EPA\\ndenies reconsideration, the objector may seek judicial review.\\n\\n This “limited exception” to the normal exhaustion\\ndeadline, Util. Air Regulatory Grp. v. EPA, 744 F.3d 741, 746\\n(D.C. Cir. 2014), does not come with a free pass from the\\nsubsequent deadline to seek judicial review. To the contrary,\\nthe second sentence of section 7607(d)(7)(B) covers only\\nobjections that arise after the close of the comment period, yet\\nwithin the time specified for judicial review. As noted above,\\nthat time for judicial review—sixty days from the promulgation\\nof the final rule—is specified in section 7607(b)(1). Section\\n7607(d)(7)(B) does not enlarge that filing period, but merely\\nfills a narrow gap within it: allowing orderly exhaustion of\\nimportant objections that “first became known to the petitioner\\nafter the comment period ended, but before the period for\\npetitioning for review expired.” Am. Petroleum Inst. v. Costle,\\n\\n 31\\n665 F.2d 1176, 1192 (D.C. Cir. 1981). We recognize that, as a\\ntextual matter, the statutory phrase “but within the time\\nspecified for review” qualifies the requirement that the grounds\\nmust have arisen “after the period for public comment,” but not\\nthe alternative requirement that “it was impracticable” to raise\\nthe objection “during the period for public comment.”\\nHowever, the petitioners do not invoke the impracticability\\nprong of section 7607(d)(7)(B). Moreover, we have construed\\nthat prong to cover instances when the final rule was not a\\nlogical outgrowth of the proposed rule, Clean Air Council v.\\nPruitt, 862 F.3d 1, 10 (D.C. Cir. 2017) (per curiam), which\\nlikewise involve problems during the period for public\\ncomment on or petitioning for review of the regulation itself—\\nnot problems that arise when circumstances change years or\\ndecades later.\\n\\n The petitioners argued in briefing that the “time specified\\nfor judicial review” referenced in section 7607(d)(7)(B)\\nencompasses not only the initial sixty-day window after a\\nrule’s promulgation, but also the secondary sixty-day limit\\nfrom after-arising grounds in the fourth sentence of section\\n7607(b)(1). But as noted above, the petitioners abandoned at\\noral argument any reliance on the latter after-arising provision.\\nMoreover, their theory would transform what we have\\ndescribed as a “limited” gap-filling provision, Util. Air\\nRegulatory Grp., 744 F.3d at 746, into a perpetually looming\\nthreat of mandatory notice-and-comment reconsideration.\\nTellingly, the petitioners can cite no case employing section\\n7607(d)(7)(B)’s mandatory reconsideration procedure for\\nobjections that arose after the close of the initial window for\\njudicial review. Their interpretation would “make a mockery\\nof Congress’[s] careful effort to force potential litigants to\\nbring challenges to a rule issued under this statute at the\\n\\n 32\\noutset.” Am. Rd. & Transp. Builders Ass’n v. EPA, 705 F.3d\\n453, 458 (D.C. Cir. 2013) (quotation marks omitted).\\n\\n Because the petitioners’ objections to the point of\\nobligation rule did not arise within the initial window for\\njudicial review of the 2010 point of obligation rule, but only\\nsome five years later, EPA properly denied mandatory\\nreconsideration.\\n\\nB. Merits of Challenges to EPA’s Refusal to Revise the\\n 2010 Point of Obligation Rule\\n\\n The Alon Petitioners offer an array of arguments to\\nchallenge the denial. None, however, is persuasive.\\n\\n We are reviewing EPA’s denial of a petition for\\nrulemaking to amend the agency’s point of obligation rule. See\\nsupra Part IV.A.1. Accordingly, our review is “‘extremely\\nlimited’ and ‘highly deferential.’” New York v. EPA, 921 F.3d\\n257, 261 (D.C. Cir. 2019) (quoting Massachusetts, 549 U.S. at\\n527–28). “To set aside the agency’s judgment, [we] must\\nconclude that EPA had not ‘adequately explained the facts and\\npolicy concerns relied on’ or that those facts did not ‘have some\\nbasis in the record.’” Id. (quoting WildEarth Guardians v.\\nEPA, 751 F.3d 649, 653 (D.C. Cir. 2014)). We have no basis\\nfor such a conclusion. In denying the petition for rulemaking,\\nEPA considered the “information currently before” it and\\ndetermined “that the point of obligation is appropriately\\nplaced,” wrestling with the petitioners’ claims to the contrary.\\nEPA Denial at 8, Alon J.A. 62. As is evident from our\\ndiscussion below, EPA did so with enough thoroughness and\\nreasonableness to satisfy our limited, deferential review.\\n\\n We start with EPA’s reasoning, which petitioners say is\\narbitrary and capricious. See 42 U.S.C. § 7607(d)(9)(A).\\n\\n 33\\n At the root of petitioners’ claim is a single premise: that\\nthe current point of obligation misaligns incentives by\\nrequiring those who refine fossil fuel, but not those who blend\\nit, to meet the RFS program’s annual standards. In petitioners’\\nview, this misalignment forces refiners to purchase RINs to\\nsatisfy their RFS obligations, jacking up their costs, while\\ngiving windfall profits to blenders, who produce (but don’t\\nconsume) RINs. From this cycle of “RINsanity,” petitioners\\nsay, flow harms galore. Alon Reply Br. 25. Higher RIN prices\\nnot only threaten the financial viability of refiners (putting our\\neconomy and energy security in jeopardy), see, e.g., Alon Br.\\n46, but they incentivize RIN hoarding, which feeds market\\nvolatility, and gives some market participants an unfair leg up,\\nsee, e.g., id. at 40.\\n\\n The problem with this argument, however, is that EPA\\nreasonably explained why, in its view, there is no misalignment\\nin the RFS program. According to EPA, refiners “recover the\\ncost of the RINs they purchase” by passing that cost along in\\nthe form of “higher prices for the petroleum based fuels they\\nproduce.” EPA Denial at 25, Alon J.A. 79. It grounded that\\nconclusion in studies and data in the record. EPA and the\\nauthors of the pertinent studies took advantage of the fact that\\nthere are pairs of petroleum products in which one variant is\\nsubject to the RIN obligation (such variants being awkwardly\\ncalled “obligated fuels”), whereas its not-quite-identical twin is\\nnot. For example, gasoline and diesel sold for use in the United\\nStates are “obligated,” whereas the same fuels sold for export\\nare not. EPA Denial at 23, Alon J.A. 77. The same goes for\\ndomestic diesel fuel, which is “obligated,” and jet fuel, which\\nis not. Christopher R. Knittel et al., The Pass-Through of RIN\\nPrices to Wholesale and Retail Fuels Under the Renewable\\nFuel Standard 4 (July 2015) (Knittel), Alon J.A. 534.\\nComparing these pairs, the agency found that as RIN prices\\nincreased, a gap “open[ed] up between” the price for obligated\\n\\n 34\\nand unobligated fuels, a gap rather precisely matching the\\ncontemporaneous increase in RIN price—a strong indication\\nthat refiners were “recoup[ing] the costs associated with RIN\\nprices.” EPA Denial at 23, Alon J.A. 77.\\n\\n Further confirming the price relationship, Professor\\nKnittel and his colleagues found that 73% of a change in RIN\\nprice was passed through in the form of higher petroleum prices\\nin the same day, 98% within two business days. Knittel 26,\\nAlon J.A. 536.\\n\\n Reviewing the findings, EPA (accurately) reported that the\\npapers by Knittel and his colleagues, and by Argus Consulting\\nServices, “concluded that the RIN cost was generally included\\nin the sale prices of obligated fuels.” EPA Denial at 25, Alon\\nJ.A. 79; see Knittel 26, Alon J.A. 536; Argus Consulting\\nServices, Do Obligated Parties Include RINs Costs in Product\\nPrices? 15 (Feb. 2017), Alon J.A. 564 (“There are very specific\\ncorrelating price data for diesel that indicate that refiners . . .\\npass along the RINs cost . . . .”).\\n\\n A similar analysis, EPA concluded, reveals that just as\\n(obligated) refiners do not pay excess costs, neither do blenders\\n(who are not obligated under the program) nor integrated\\nrefiners (who perform their own in-house blending) reap\\nwindfall profits. True, both earn RINs, without purchasing\\nthem on the open market, by blending renewable fuel into\\npetroleum blendstock. And true, as well, both can sell those\\nRINs, enjoying whatever revenues market conditions and their\\nown efficiencies permit. But as EPA quite accurately\\nexplained, this is only half the equation. In a competitive\\nmarket there’s no such thing as a free lunch, and blenders and\\nintegrated refiners pay their tab just as others do; they just do\\nso indirectly. To offer finished fuel without attached RINs at a\\ncompetitive price, these entities must discount their blended\\n\\n 35\\nfuel by roughly the value of the RINs that they detached and\\nkept for themselves. EPA Denial at 29, Alon J.A. 83. In other\\nwords, they “sell the finished transportation fuel at a loss,” but\\n“maintain[] profitability through RIN sales.” Id. at 27–28, 29,\\nAlon J.A. 81–82, 83.\\n\\n To be sure, in response to EPA’s proposed denial,\\ncommenters criticized the studies relied on by the agency.\\nThey contended, for example, that Professor Knittel and his\\ncolleagues erred by removing certain spreads from the analysis,\\nby including others, and by pooling the results of various\\ncomparisons. See EPA Denial at 25, Alon J.A. 79. But\\npetitioners have not raised these arguments here, and for that\\nreason we do not consider them. While petitioners do complain\\nthat EPA relied on a “preliminary” analysis, see Alon Br. 54;\\nAlon Reply Br. 23, that objection—whatever its persuasive\\nforce—says nothing about the other studies in the record (for\\nexample, by Professor Knittel et al.).\\n\\n Petitioners try, instead, to trace various refiner problems to\\nEPA’s refusal to obligate blenders. They suggest that the\\nalleged misplacement of the point of obligation causes\\nbankruptcies, see, e.g., Alon Br. 3, 47, and inflicts economic\\nhardship on small refineries, see, e.g., id. at 49, especially in\\nthe form of inflicting wildly disproportionate RIN acquisition\\ncosts on them, see, e.g., Alon Reply Br. 26. But some of these\\nevents occurred after EPA issued its denial, see, e.g., Alon Br.\\n49 (“following the Denial”); Alon Reply Br. 26 (“just after the\\nDenial”), and are therefore not properly before us, see\\nEnvironmental Defense Fund, Inc. v. Costle, 657 F.2d 275, 284\\n(D.C. Cir. 1981). More importantly, the claims presuppose that\\nrefiners cannot recover their RIN costs and that blenders reap\\nwindfall profits—suppositions that, as discussed above, EPA\\nreasonably rejected.\\n\\n 36\\n Petitioners respond by plucking snippets from the denial,\\nstringing them together with contrasting (bolded) conjunctions,\\nand asserting that EPA’s “discussion of RIN prices” is\\n“irreconcilably inconsistent.” Alon Br. 53. But we find no\\ninconsistency on EPA’s part.\\n\\n Take one of petitioners’ examples:\\n\\n . . . RIN prices had no “significant impact on retail\\n gasoline (E10) prices,” JA75;\\n\\n although “RINs . . . provide a price signal to\\n consumers to help achieve . . . greater renewable fuel\\n production and use,” JA75.\\n\\nAlon Br. 53 (second and third alterations in original) (quoting\\nEPA Denial at 21, Alon J.A. 75). At first blush, the two\\ncomments, located on the very same page, seem inconsistent.\\nHow could RIN prices have no “significant impact” on retail\\nprices, while, at the same time, provide “a price signal to\\nconsumers”?\\n\\n They can do so for the simple reason that the remarks refer\\nto different things, a detail omitted from petitioners’ brief. This\\nbecomes apparent when the passage from which petitioners\\nplucked their quotes (bolded and underscored below) is viewed\\nin full:\\n\\n External, non-EPA assessments similarly concluded\\n that increased RIN prices had not had a significant\\n impact on retail gasoline (E10) prices. When RIN\\n prices rise, the market price of the petroleum\\n blendstocks produced by refineries also rise to cover\\n the increased RIN costs, in much the same way as they\\n would rise in response to higher crude oil prices. The\\n effective price of renewable fuels (the price of the\\n\\n 37\\n renewable fuel with attached RIN minus the RIN\\n price), however, decreases as RIN prices increase.\\n When renewable fuels are blended into petroleum\\n fuels these two price impacts generally offset one\\n another for fuel blends such as E10 with a renewable\\n content approximately equal to the required\\n renewable fuel percentage standard. Higher RIN\\n prices also generally result in higher prices for fuels\\n with lower renewable content (such as E0 or\\n petroleum diesel) and lower prices for fuels with\\n higher renewable content (such as E85 or B20). The\\n cost of the RIN therefore serves as a cross-subsidy,\\n reducing the price of renewable fuels and increasing\\n the price of petroleum based fuels in transportation\\n fuel blends, thus incentivizing increased blending of\\n renewable fuels into the transportation fuel pool. In\\n this way the RINs also help provide a price signal to\\n consumers to help achieve the Congressional goals\\n of greater renewable fuel production and use.\\n Fuels with higher renewable content are relatively\\n cheaper to consumers than they would be absent high\\n RIN prices, while fuels with lower renewable content\\n are relatively more expensive when RIN prices are\\n high.\\n\\nEPA Denial at 20–21, Alon J.A. 74–75.\\n\\n As we can see, the first statement (no significant price\\nimpact) is referring to the price of E10—a blend of 90%\\ngasoline, 10% ethanol. As EPA explained, RINs work as a\\n“cross-subsidy,” effectively taxing the use of petroleum-based\\nfuels (e.g., gasoline) and subsidizing the use of renewables\\n(e.g., ethanol) in making a blended transportation fuel like E10.\\nEPA Denial at 21, Alon J.A. 75.\\n\\n 38\\n Before we dig in further, let’s take a step back. We must\\nfirst recognize that EPA assumes that we are talking of a market\\nwhere the RFS program, in effect, mandates minimum levels\\nof renewables in marketed fuels, a mandate that necessarily\\nimpacts fuel prices. EPA is not making a claim that the\\nmandatory inclusion of renewables in transportation fuel\\nrenders a gallon of gasoline lawfully purchased at the pump\\ncheaper than it would have been absent the RFS program.\\n\\n To see what this means, start on the subsidy side: Suppose\\na blender can realize $2.25 on a gallon of ethanol with an\\nattached RIN. If the blender can detach and sell that RIN for\\n$0.05, then the net ethanol value is only $2.20—a $0.05\\nsavings that in a competitive market should pass through to\\nconsumers. Now take the tax side: Because refiners must\\npurchase RINs to satisfy the RFS obligations that arise from\\nselling gasoline to blenders, a blender’s value for a gallon of\\ngasoline may rise, due to the RFS program, from, say, $2.75 to\\n$2.76. See Dallas Burkholder, Office of Transportation & Air\\nQuality, EPA, A Preliminary Assessment of RIN Market\\nDynamics, RIN Prices, and Their Effects 17, EPA-HQ-OAR-\\n2016-0544-0009 (May 14, 2015), Alon J.A. 337. As a result\\nof both price impacts, EPA described, blended fuels with a\\nhigher percentage of renewable content (e.g., 85% ethanol) will\\nbe cheaper than they would have been (absent the program),\\nwhereas fuels with a lower percentage of renewable content\\n(e.g., pure gasoline) will be more costly than they would have\\nbeen (absent the program). EPA Denial at 21, Alon J.A. 75.\\nFor E10, the “two price impacts generally offset one another,”\\nso (back to the first statement) any change in RIN price\\ngenerally has no “significant impact on” the E10 price. Id.\\n\\n But that’s just E10. There is an effect (of differing\\nmagnitude) on, say, E85 or E0. And that is where the second\\nstatement (“provide a price signal”) comes in: the signal arises\\n\\n 39\\nfrom a comparison of relative prices across the spectrum of\\ntransportation fuels. Again, as EPA explains, “[f]uels with\\nhigher renewable content are relatively cheaper to consumers\\nthan they would be absent high RIN prices, while fuels with\\nlower renewable content are relatively more expensive when\\nRIN prices are high.” Id. The two statements are consistent.\\n\\n Continuing the search for inconsistency, petitioners direct\\nour attention to “EPA’s past pronouncements.” Alon Br. 50.\\nIn them, they see an irrational “about-face”—with EPA saying,\\nat first, that “low RIN prices [were] a sign that the [RFS\\nprogram] was working,” but claiming, now, “that high RIN\\nprices are . . . desirable.” Id. at 51–52. Again, that’s not quite\\nright. All EPA originally said was that when it first adopted\\nthe point of obligation, it did so based, in part, on its\\n“expectation at that time that there would be an excess of RINs\\nat low cost.” Regulation of Fuels and Fuel Additives: Changes\\nto Renewable Fuel Standard Program, 74 Fed. Reg. 24,904,\\n24,963/2 (May 26, 2009) (proposed rule); see also Alon Br. 50\\n(citing EPA Denial at 13, Alon J.A. 67 (citing, in turn, 74 Fed.\\nReg. at 24,963)). EPA did not suggest that low RIN prices were\\na sign of market health—nor that high prices were a cause for\\nalarm.\\n\\n In any case, EPA addressed petitioners’ concern over high\\nRIN prices head on; the agency explicitly determined, on the\\ncurrent record, that “higher RIN prices” are not “indicative of\\na dysfunctional RIN market.” EPA Denial at 19, Alon J.A. 73.\\nRather, EPA explained, these prices accurately reflect the\\nincreasing cost associated with “getting ever-greater volumes\\nof renewable fuel into the transportation fuel pool—the explicit\\ngoal [of] the RFS program.” Id. Put more bluntly, the increases\\nin RIN prices are a completely understandable effect of the\\nprogram’s ever-increasing pressure to expand renewable\\nvolumes. Pushing out along the supply curve takes the raw\\n\\n 40\\nmarket price of the RIN-eligible fuel steadily into higher\\nrealms—except to the extent that production innovations or\\neconomies may tend to lower costs. So far as appears, it has\\nnothing to do with EPA’s allocation of the obligation.\\n\\n What about EPA’s concern, petitioners ask, that including\\nblenders in the point of obligation would expand the number of\\nobligated parties and, as a result, ratchet up the program’s\\ncomplexity? Isn’t that hard to square with EPA’s claim, made\\nyears earlier, that “essentially all downstream blenders . . . are\\n[already] regulated parties”? Alon Br. 42 (quoting 75 Fed.\\nReg. at 14,722/2). Again, not at all. Although the participation\\nof all (or nearly all) blenders in the RIN market subjects them\\nto RFS registration, recordkeeping, and reporting\\nrequirements, “the majority of these downstream [regulated]\\nparties are . . . currently not obligated parties.” EPA Denial at\\n69, Alon J.A. 123 (emphasis added). As EPA explained, there\\n“is a significant distinction between being a ‘regulated party’\\nand being an ‘obligated party.’” Id. “Obligated parties must\\nmeet all of [the requirements faced by regulated parties] and\\nalso calculate an annual renewable volume obligation, acquire\\nthe appropriate number of RINs in the market, practic[e] due\\ndiligence to ensure [the RINs’] validity, file annual compliance\\nreports demonstrating compliance, and maintain records to that\\neffect.” Id. at 69 n.205, Alon J.A. 123 (emphasis added); see,\\ne.g., 40 C.F.R. §§ 80.1427(a), 80.1450(a), 80.1451(a),\\n80.1454(a). It was not unreasonable for EPA to conclude that\\nimposing these burdens on additional entities would add to the\\nprogram’s complexity (and therefore be undesirable absent an\\nadequate offsetting benefit).\\n\\n Nor was it unreasonable to find that going down this\\nroute—overhauling a foundational element of the program—\\nwould create “uncertainty in the fuels marketplace.” EPA\\nDenial at 2, Alon J.A. 56. As EPA said, “all parties regulated\\n\\n 41\\nin the RFS program have made significant investments and\\ndecisions about their participation in the program and their\\nposition in the market on the basis of the existing regulations,\\nincluding the definition of obligated parties.” EPA Denial at\\n79, Alon J.A. 133. In these circumstances, it isn’t hard to\\nimagine how changing course could throw players off their\\ngame. Of course, as petitioners note, uncertainty may have\\n“plagued the RFS Program for years.” Alon Br. 37. But true\\nor not, EPA needn’t pile on; the cure for uncertainty isn’t\\nspawning more uncertainty.\\n\\n Taking a step back, petitioners launch a closing broadside\\nagainst the entire process. They assert that EPA “disregarded\\nthis Court’s remand” in ACE, 864 F.3d at 737, and arbitrarily\\ncredited some comments over others. Alon Br. 31–32, 55–56.\\nBut the ACE remand required, at most, that the agency “address\\nthe point of obligation issue.” 864 F.3d at 737. And, as\\ndetailed throughout this opinion, the agency has done so\\nreasonably, analyzing the data and explaining its decision.\\nNothing more was required.\\n\\n\\n * * *\\n\\n We have considered the Alon Petitioners’ other arguments\\nand have found them to be either without merit or, in the case\\nof the argument relying on 42 U.S.C. § 7545(o)(5)(A), see\\nAlon Br. 34, insufficiently developed, see, e.g., Masias v. EPA,\\n906 F.3d 1069, 1077 (D.C. Cir. 2018). For the foregoing\\nreasons, the petitions for review are denied.\\n\\n V. 2017 Annual Volumetric Rule\\n\\n We turn now to the Coffeyville Petitioners’ challenges.\\n\\n 42\\nA. Point of Obligation\\n\\n We first consider the Coffeyville Petitioners’ challenge to\\nEPA’s decision in the 2017 Rule not to reassess which\\ncategories of industry players are “obligated parties” under the\\nrenewable fuel program. As the Coffeyville Petitioners read it,\\nthe statutory provision requiring EPA to set annual renewable\\nfuel percentage standards also imposes on EPA a\\nnondiscretionary duty to reconsider—every year—which types\\nof entities are obligated to demonstrate to EPA compliance\\nwith the percentage standards. See 42 U.S.C. § 7545(o)(3)(B).\\nThey claim EPA shirked that duty when it treated the issue as\\nbeyond the scope of its 2017 annual rulemaking. EPA counters\\nthat it identified the obligated parties in 2007 pursuant to\\nCongress’s mandate to set “compliance provisions” for the new\\nrenewable fuel program, id. § 7545(o)(2), reaffirmed that\\ndecision in 2010, and that nothing in the mandate to calculate\\nthe annual percentage standards requires it to reconsider the\\npoint of obligation each year. EPA also asserts that it\\nappropriately addressed the Coffeyville Petitioners’ complaints\\nthat it obligated the wrong parties in a separate proceeding from\\nits annual volumetric rulemaking.\\n\\n 1. Jurisdiction\\n\\n EPA and a coalition of Respondent-Intervenors\\nrepresenting the renewable fuel and refinery industries assert\\nthat we lack jurisdiction because the Coffeyville Petitioners\\neffectively challenge the compliance rule that has been on the\\nbooks for a decade or so, see 40 C.F.R. § 80.1406(a)(1), and\\ndid not petition within 60 days of its publication or within 60\\ndays of any valid “grounds arising” thereafter. See 42 U.S.C.\\n§ 7607(b)(1); Med. Waste Inst. v. EPA, 645 F.3d 420, 427\\n(D.C. Cir. 2011). But petitioners are not challenging EPA’s\\ndecision to adopt the rule in 2007 or retain it in 2010. Rather,\\n\\n 43\\nthey contend that the provision calling on EPA to set annual\\nvolumes of biofuels “applicable to refineries, blenders, and\\nimporters, as appropriate,” requires EPA to reassess each year\\nwhether the point of obligation set when the agency established\\nthe program is still “appropriate,” or if EPA should re-assign it\\nand restructure the RIN market and other compliance\\ninfrastructure going forward. See 42 U.S.C.\\n§ 7545(o)(3)(B)(ii)(I). That challenge was timely filed within\\n60 days of the promulgation of the annual fuel standards. See\\nsupra Part IV.A.1.\\n\\n 2. Merits\\n\\n This dispute turns on the roles of two provisions of the\\nstatute directing the EPA to establish and run a Renewable Fuel\\nProgram, 42 U.S.C. § 7545(o)—paragraphs (2) and (3).\\n Paragraph (2) directs EPA to “promulgate regulations”\\nsetting up a program to “ensure that transportation fuel sold or\\nintroduced into commerce in the United States . . . contains at\\nleast the applicable volume[s] of renewable fuel,” as specified\\nin subparagraph (2)(B). Id. § 7545(o)(2)(A)(i). Among the\\nparameters Congress required EPA to include were\\n“compliance provisions applicable to refineries, blenders,\\ndistributors, and importers, as appropriate,” to ensure that the\\nrequirements of paragraph (2), including the applicable volume\\nrequirements specified in subparagraph (2)(B), are met. Id.\\n§ 7545(o)(2)(A)(iii)(I), (B). There is no question that EPA has\\nauthority to set those parameters, including the point of\\nobligation, and to adjust them if a change is needed.\\n Paragraph (3), in turn, requires EPA to determine and\\npublish annual renewable fuel obligations designed to\\n“ensure[] that” the applicable volumes specified in paragraph\\n(2) are met. Id. § 7545(o)(3)(B)(i). Those renewable fuel\\nobligations must:\\n\\n 44\\n (I) be applicable to refineries, blenders, and\\n importers, as appropriate;\\n (II) be expressed in terms of a volume percentage\\n of transportation fuel sold or introduced into\\n commerce in the United States; and\\n (III) . . . consist of a single applicable percentage\\n that applies to all categories of persons\\n specified in subclause (I).\\nId. § 7545(o)(3)(B)(ii)(I)–(III).\\n The parties’ dispute centers on the meaning of “as\\nappropriate” in subclause (3)(B)(ii)(I). The Coffeyville\\nPetitioners contend that the phrase unambiguously requires\\nEPA annually to reconsider which parties it is “appropriate” to\\nobligate to meet the renewable fuel obligations. EPA responds\\nthat the statute is, at most, ambiguous as to whether Congress\\nexpected EPA annually to revisit the obligated-parties\\ndesignation, or whether the agency may generally rely on the\\n“appropriate[ness]” finding it made pursuant to its paragraph\\n(2) authority. We begin by asking “whether Congress has\\ndirectly spoken to the precise question at issue,” and conclude\\nthat it has not. Chevron, 467 U.S. at 842.\\n EPA reads “as appropriate” in paragraph (3) to mean that\\nthe agency has “discretion” to decide whether, when, and how\\nto reassess which of three types of industry actors—refineries,\\nblenders, and importers—should continue to bear the point of\\nobligation, as originally designated in the compliance\\nprovisions. See EPA Denial, Coffeyville J.A. 779–80. At oral\\nargument, the agency conceded that its exercise of this\\ndiscretion is reviewable, so that the exclusion of the point of\\nobligation issue from an annual rulemaking could, under other\\ncircumstances, constitute an abuse of discretion. Rec. of Oral\\nArg. 1:22:22–1:24:00; 1:37:45–1:40:11.\\n\\n 45\\n The Coffeyville Petitioners object that the phase\\n“applicable . . . as appropriate” means applicable as\\ncontemporaneously determined to be appropriate in the annual\\nvolumetric rulemakings. Coffeyville Br. 30–33; see also Conc.\\nOp. 4–5, 7. But paragraph (3) does not specify when or in what\\ncontext EPA must make its appropriateness determination, nor\\ndoes the phrase “as appropriate” itself specify a particular\\ntemporal dimension—as between, for example, parties\\nappropriately designated in the past (as EPA interprets it) and\\nparties now appropriately selected (as Coffeyville insists). The\\nterm “appropriate” “naturally and traditionally includes\\nconsideration of all the relevant factors,” Michigan v. EPA,\\n135 S. Ct. 2699, 2707 (2015) (quoting White Stallion Energy\\nCenter, LLC v. EPA, 748 F.3d 1222, 1266 (D.C. Cir. 2014)\\n(Kavanaugh, J., concurring in part and dissenting in part)), but\\nit does not dictate when that consideration must be made. In\\nother words, the requirement that the point of obligation be\\n“appropriate” is at most grounds for assessing whether the\\nagency adequately explained its policy choices regarding the\\nappropriateness determination, not for imposing our own gloss\\non that broad term as a matter of law. See Kisor v. Wilkie, 139\\nS. Ct. 2400, 2448–49 (2019) (Kavanaugh, J., concurring in the\\njudgment) (“[S]ome cases involve regulations that employ\\nbroad and open-ended terms like ‘reasonable,’ ‘appropriate,’\\n‘feasible,’ or ‘practicable.’ Those kinds of terms afford\\nagencies broad policy discretion, and courts allow an agency to\\nreasonably exercise its discretion to choose among the options\\nallowed by the text of the rule. But that is more State Farm\\nthan Auer” or Chevron (emphasis added)). Here, as explained\\nbelow, EPA reasonably exercised its discretion, and explained\\nits decision, to address the point of obligation issue in a\\nseparate proceeding from its annual volumetric rulemaking.\\n The fact that paragraphs (2) and (3) both include the phrase\\n“as appropriate” does not make the Coffeyville Petitioners’ the\\n\\n 46\\nonly permissible interpretation. Even if, as our colleague\\ncontends, Conc. Op. 7, the two phrases bear the exact same\\nmeaning, but see Coffeyville Br. 30–31 (arguing that\\nparagraphs (2) and (3) are “worded differently” and have\\n“different contexts”), paragraph (3) simply does not dictate\\nwhen or in what context EPA must make the appropriateness\\ndetermination. It is the surrounding context, not the phrases\\nthemselves, that suggests when EPA might make that choice.\\n Unable to point to any express textual requirement that\\nEPA annually reconsider the point of obligation, the\\nCoffeyville Petitioners contend that, had Congress intended to\\nallow EPA in annual volumetric rulemakings to rest on its\\nparagraph (2) appropriateness determination, subclause\\n(3)(B)(ii)(I) could have more simply cross-referenced\\nparagraph (2). But replacing subclause (3)(B)(ii)(I) with a\\nsimpler cross-reference would not have achieved quite the\\nsame effect. While refineries, blenders, importers, and\\ndistributors may all be subject to compliance provisions under\\nparagraph (2), paragraph (3)’s applicability provision points to\\na more limited universe of potential obligated parties—to\\nrefineries, blenders, and importers, but not distributors. That\\nsupports EPA’s understanding of subclause (3)(B)(ii)(I) as a\\ncross-reference that also clarifies a limit on EPA’s options in\\nsetting the point of obligation. See 72 Fed. Reg. at 23,923/2\\n(preamble for compliance rule referencing paragraph (3)’s\\napplicability provision).\\n We are unpersuaded by the suggestion that such limitation\\nwas so clear even without subclause (3)(B)(ii)(I) that EPA’s\\nreading of that applicability provision renders it superfluous.\\nThe suggestion is that, because distributors do not “introduce”\\nfuel into commerce, they could not be obligated parties in any\\nevent, with or without the applicability provision. Our\\ncolleague posits that blenders—but not distributors—can in\\nfact “introduce” transportation fuel into commerce by blending\\n\\n 47\\ngasoline and diesel fuel with other fuels that have not already\\nbeen introduced by someone else. Conc. Op. 11. But that\\nargument rests on a complicated series of inferences from spare\\nstatutory text, as well as post-enactment regulations that do not\\nnecessarily show what the statute must have meant. For\\nexample, our colleague reasonably infers that the national\\n“volume[] of transportation fuel,” 42 U.S.C. § 7545(o)(3)(A),\\nmust be counted at the first moment each gallon of fuel enters\\ncommerce, in order to avoid double-counting. But, the statute\\nnowhere states the point directly. Still less clearly does the\\nstatute state our colleague’s corollary—necessary to the\\nsurplusage argument—that the point of obligation must also be\\nplaced, if at all, at the first moment a gallon of fuel enters\\ncommerce. That corollary is less obvious, because the\\nstatutory link between the point of obligation and entry into\\ncommerce is not ironclad: All parties agree that not every\\ngallon of transportation fuel must be subject to the point of\\nobligation upon entry into commerce. The statute plainly\\nallows EPA to obligate an “appropriate” subset of the three\\ncategories of parties. And obligating blenders would involve\\ndouble counting unless the transportation fuel they use to create\\nblends were not already counted upon its importation or sale to\\nthem. Under the circumstances, it is reasonable to read\\nsubclause (3)(B)(ii)(I) as clarifying what is at best a non-\\nobvious inference that distributors cannot be subjected to the\\npoint of obligation.\\n So subclause (3)(B)(ii)(I) as EPA reads it is not a\\nsuperfluity, but makes clear that EPA may have permissibly\\nplaced the point of obligation on refineries, blenders, and\\nimporters, but not distributors. For the same reason, subclause\\n(3)(B)(ii)(III), which requires that the annual standards apply\\n“to all categories of persons specified in subclause (I),” id.\\n§ 7545(o)(3)(B)(ii)(III), does not contain what our colleague\\nviews as an unnecessary double cross-reference to paragraph\\n\\n 48\\n(2), because it, too, is operative in not just cross-referencing,\\nbut also clarifying a limit on the three permissible targets of its\\n“single applicable percentage.”\\n The thrust of the Coffeyville Petitioners’ retort—that if\\nCongress had wanted to confer discretion or provide a limiting\\ncross-reference to paragraph (2), it would have said so more\\nplainly—applies with greater force against their own reading.\\nHad Congress intended EPA to consider on an annual basis\\nwhether to redo the point of obligation designation—a\\ndesignation that no-one disputes is a necessary cornerstone of\\nthe paragraph (2) compliance provisions—it knew how to\\nimpose such a requirement. The Clean Air Act’s provisions on\\nambient air quality, for instance, require EPA to “complete a\\nthorough review” of the air quality standards “at five-year\\nintervals” and “promulgate such new standards as may be\\nappropriate.” Id. § 7409(d)(1). The Act’s provisions\\ncontrolling hazardous air pollutants emitted from major and\\narea sources require EPA to “review, and revise as necessary”\\nthe applicable emission standards “no less often than every 8\\nyears.” Id. § 7412(d)(6). Paragraph (3) of the RFS program,\\nin contrast, does not tell EPA to “complete a thorough review,”\\nor “review, and revise as necessary” its point of obligation\\ndecision—or anything even close.\\n To be sure, EPA’s reading is not ineluctable. We do not\\ndoubt that Congress could have more directly provided that the\\nrenewable fuel obligations do not apply to distributors. See\\nConc. Op. 8. But, for the reasons discussed, we are\\nunconvinced that paragraph (3) plainly requires EPA to\\nconsider adjusting the point of obligation each year. See\\nValero Energy Corp. v. EPA, No. 7:17-cv-00004-O, 2017 WL\\n8780888, at *4 (N.D. Tex. Nov. 28, 2017) (holding that “there\\nis no clear statutory mandate . . . obligating [EPA] to evaluate\\nor adjust . . . what entities are ‘appropriate[ly]’ forced to\\ncomply with” the annual renewable fuel obligations\\n\\n 49\\n(alterations in original) (quoting 42 U.S.C.\\n§ 7545(o)(3)(B)(ii)(I))). Accordingly, we conclude that the\\nmeaning of “as appropriate” in paragraph (3) is ambiguous and\\nturn now to whether EPA’s construction is “based on a\\npermissible construction of the statute.” Chevron, 467 U.S. at\\n843.\\n The difficulty of squaring the Coffeyville Petitioners’\\nreading of “as appropriate” with the structure and purpose of\\nthe statute convinces us of the reasonableness of EPA’s\\ninterpretation. As a structural matter, the RFS program\\ncontains not only “annual” volumetric determinations, Conc.\\nOp. 1, but also a slew of compliance provisions that are not\\nannually re-determined. As a practical matter, the point of\\nobligation is the foundational “compliance provision” of the\\nentire renewable fuels program; EPA could not “ensure” that\\napplicable volumes of renewable fuels are introduced into the\\nnation’s transportation fuel supply without designating the\\nparties responsible for carrying the renewable fuel standards\\ninto operation. Id. § 7545(o)(2)(A)(i). To that end, in writing\\nthe compliance provisions, EPA placed the renewable fuel\\nobligation on the entities at the head of the United States supply\\nchain for nonrenewable fuels—domestic refiners, and\\nimporters of fuel refined elsewhere. See 72 Fed. Reg. at\\n23,923–24. After additional consideration, EPA in 2010\\nadhered to that decision. See 75 Fed. Reg. at 14,721–22\\n(codified at 40 C.F.R. § 80.1406(a)(1)); see also Monroe\\nEnergy, 750 F.3d at 912. No one challenged EPA’s decision\\nin 2007 or 2010, and EPA declined to revisit the issue in\\nresponse to comments in the 2017 annual rulemaking urging it\\nto shift the 2017 point of obligation to blenders. See Response\\nto Comments at 542, Coffeyville J.A. 761.\\n The focus of the annual rulemakings, in contrast, is to\\ntranslate the applicable volumes—as specified in paragraph\\n(2), or set according to the process there described—into\\n\\n 50\\npercentage requirements for each renewable fuel. 42 U.S.C.\\n§ 7545(o)(3)(B)(ii). It would be strange indeed if Congress\\nrequired EPA, as it went about its annual quantitative standard-\\nsetting duties, also to rethink a choice so basic to the RFS\\nprogram’s architecture. This implausibility is illuminated by\\nthe fact that Congress required EPA to facilitate statutory\\ncompliance through a credit trading program, which of\\nnecessity requires some year-to-year stability. See id.\\n§ 7545(o)(5). EPA responded by setting up the RIN system,\\nwith flexibility anchored to a fixed baseline—the point of\\nobligation. The compliance system is flexible in that RINs may\\nbe retired in compliance demonstrations not only in the\\ncompliance year during which they were generated, but also\\nthroughout the ensuing compliance year, 40 C.F.R.\\n§ 80.1427(a)(6), and obligated parties may carry over excess\\nRINs or RIN deficits from year to year, id. § 80.1427(a)(1),\\n(5)–(6); see Monroe Energy, 750 F.3d at 913.\\n Annual changes to the point of obligation could cause\\n“disparities in RIN-holdings,” leaving formerly obligated\\nparties with “significantly more RINs, including carryover\\nRINs, than they desire or can use” and newly obligated parties\\nwith “lower balances than they would desire to protect\\nthemselves against shortfalls in RIN availability or RIN price\\nvolatility.” EPA Denial at 78, Coffeyville J.A. 850. “[A]\\nchange to the point of obligation could also cause volatility in\\nthe [RIN] market,” inhibiting the “ability [of] parties that\\npossess excess carryover RINs to recover the cost of the RINs\\nthey hold by selling them to other parties.” Id. It is not\\nplausible that Congress meant EPA to consider uprooting the\\nbaseline of the RFS program every year. The real stretch is that\\nCongress would have imposed such an onerous and potentially\\ndisruptive duty merely by use of the phrase “as appropriate.”\\n The Coffeyville Petitioners’ reading is not made any more\\nplausible by highlighting the likelihood that, on annual\\n\\n 51\\nconsideration of the point of obligation, EPA would only need\\nto consider recent information, and likely would stay its course.\\nEven if the point of obligation in fact rarely changed, the mere\\n“reconsider[ation]” of the framework would “likely cause\\ndelays to the investments necessary to expand the supply of\\nrenewable fuels in the United States.” See EPA Denial at 2,\\nCoffeyville J.A. 774. EPA reasoned that “fuel[] industry\\nparticipants [would] withhold significant investment decisions\\nuntil the EPA’s final decision and the fallout from the decision\\nare known.” Id. at 81–82, Coffeyville J.A. 853–54. Insisting\\nthat the issue be on the regulatory agenda every year would sow\\n“significant market uncertainty and potential turmoil” into the\\nRFS program without offsetting benefit. Id.\\n Furthermore, any requirement that an agency repeatedly\\ngo through a regulatory process on an issue that promises to\\ndraw a regular parade of criticism from interest groups with\\nample resources is itself burdensome. See AT&T Corp. v. FCC,\\n220 F.3d 607, 630–31 (D.C. Cir. 2000). This issue is no\\nexception. As discussed above, EPA in 2016 and 2017\\nconsidered and decided against reopening its point of\\nobligation rule. In so doing, it received upwards of 18,000\\ncomments and published an exhaustive, 85-page decision. See\\nEPA Denial at 1–85, Coffeyville J.A. 771–857. “Given the\\ntime pressure associated with its annual standards rulemaking,”\\nEPA believes it would not be feasible or worthwhile to\\nundertake such reconsideration annually. Id. at 7 n.10,\\nCoffeyville J.A. 779. Indeed, as EPA acknowledged at oral\\nargument, the agency “has been late on [its annual rules]\\nbefore,” even “when [it hasn’t] taken up the point of\\nobligation.” Rec. of Oral Arg. 1:25:44–52. “[A]dd[ing] on\\nthe” duty to reassess the point of obligation annually, EPA tells\\nus, “would be a significant burden.” Id. at 1:25:55–1:26:05.\\nOur colleague doubts that EPA’s year-to-year burden would be\\nappreciable, but we see no ground to question EPA’s judgment\\n\\n 52\\nto the contrary. It seems unlikely that Congress wrote the\\napplicability provision in order to heap that annual duty onto\\nEPA’s plate. It seems even less likely given the absence of\\nreason to think that yearly second-guessing of program\\nfundamentals makes sense, or that, when and if the need for a\\nprogram restructuring arises, EPA would fail to act. Indeed,\\nthe statute elsewhere explicitly requires EPA to conduct\\n“periodic reviews of . . . the feasibility of achieving compliance\\nwith the [applicable volume] requirements.” 42 U.S.C. §\\n7545(o)(11). That provision has not been briefed, but would\\nappear to require EPA to reconsider the point of obligation if it\\nconcluded that its placement was obstructing compliance.\\n Finally, EPA’s approach coheres with basic principles of\\nadministrative law. In general, the choice between various\\nprocedural channels lies within the “informed discretion of the\\nadministrative agency.” SEC v. Chenery Corp., 332 U.S. 194,\\n203 (1947). That discretion properly includes judgments about\\nthe scope of rulemakings and when to relegate ancillary issues\\nto separate proceedings: “Agencies, like legislatures, do not\\ngenerally resolve massive problems in one fell regulatory\\nswoop.” Massachusetts, 549 U.S. at 524; see, e.g., Grp.\\nAgainst Smog & Pollution, 665 F.2d at 1292 (“. . . EPA cannot\\nsoundly be charged with arbitrariness merely because it chose\\na separate rulemaking proceeding as the process for proposing\\na revised standard in lieu of an undertaking to do so in the\\nnarrower context of the opacity standard proceedings as\\npetitioners requested.”). Once the agency has resolved an issue\\nin a separate proceeding, it may defend against related criticism\\nby “simply refer[ing]” to the other proceeding, so long as the\\n“reasoning remains applicable and adequately refutes the\\nchallenge.” Bechtel v. FCC, 10 F.3d 875, 878 (D.C. Cir. 1993).\\nEPA reasonably reads “as appropriate,” in paragraph (3)(B), to\\nleave undisturbed these background norms of broad but\\nreviewable procedural discretion.\\n\\n 53\\n Our holding today does not give EPA the limitless and\\nunreviewable discretion feared by our colleague. As we have\\nsaid, EPA’s determination as to whether it is “appropriate” to\\nreconsider the point of obligation in the context of an annual\\nvolumetric rulemaking is reviewable for abuse of discretion.\\nEPA did not abuse its discretion in refusing to do so here.\\nIndeed, it considered whether to change the point of obligation\\nrule in a separate, contemporaneous proceeding that yielded a\\nfinal order that we also have reviewed and found to be\\nadequately justified. See supra Part IV.B. We do not address\\nwhether it would be an abuse of discretion for EPA to refuse to\\nreconsider the point of obligation—in an annual volumetric\\nrulemaking or otherwise—in extreme circumstances akin to\\nthose posited by our colleague’s hypothetical about the\\ncontinuing study of an abolished tort. See Conc. Op. 5.\\n In sum, we hold that EPA permissibly rejected the claim\\nthat paragraph (3) requires the agency annually to reassess the\\npoint of obligation in the renewable fuel program. Because\\nEPA has no duty to reconsider the appropriateness of its point\\nof obligation regulation as part of its yearly determination of\\nvolumetric requirements, it was not arbitrary for EPA to treat\\ncomments complaining that it obligated the wrong parties as\\nappropriately assessed in a separate proceeding, and beyond\\nthe scope of proceedings for the 2017 volumetric rulemaking.\\nFor these reasons, we deny the Coffeyville Petitioners’ petition.\\nB. Cellulosic Biofuel Projection\\n\\n The Coffeyville Petitioners lob a variety of challenges at\\nEPA’s cellulosic biofuel projection for 2017. Many of these\\npetitioners, however, raised many of the same arguments\\nbefore. See ACE, 864 F.3d at 727–29 (addressing challenges\\nto EPA’s 2014–16 projection). We rejected those arguments\\nonce—and do so again.\\n\\n 54\\n First, the Coffeyville Petitioners contend that “EPA’s\\n[m]ethodology” for projecting cellulosic biofuel production is\\ninvalid because it “[c]hronically [o]verestimates [a]ctual\\n[p]roduction.” Coffeyville Br. 40. But that argument—that\\nEPA has “repeatedly . . . overshot the mark,” id. at 41—doesn’t\\napply to the methodology EPA actually used here, as we found\\nin ACE, 864 F.3d at 727–28. As we explained when petitioners\\ndeployed this same argument in challenging the 2014–16\\nprojection, “the majority of EPA’s prior overestimations”\\nutilized a different methodology—one that we rejected in API,\\n706 F.3d at 478–81, and that the EPA accordingly abandoned.\\nACE, 864 F.3d at 727. The new methodology—the one EPA\\nused here—has been applied only twice before. At the time\\nEPA made its final evaluation for 2017, that methodology had\\n(as detailed in the table below) undershot for 2015 and overshot\\nfor 2016. See Assessment of the Accuracy of Cellulosic\\nBiofuel Production Projections in 2015 and 2016, EPA-HQ-\\nOAR-2016-0004-3687, at 1–4 (Dec. 12, 2016), Coffeyville\\nJ.A. 515–18. This is hardly a pattern of chronic overestimation.\\n\\n EPA Actual\\n RFS\\n Estimate Production EPA Record\\n Compliance\\n (millions of (millions of Error ** Citation\\n Year\\n RINs) RINs)\\n Q1\\n Q2 [No Data in the Record]\\n 2015\\n Q3\\n Q4 35.00 53.36 - 34.4% J.A. 515\\n Q1\\n Q2 J.A.\\n 2016 230.00 198.39* + 15.9%\\n Q3 516–17\\n\\n Q4\\n\\n 55\\n * At the time of EPA’s assessment, the agency had actual RIN\\nproduction data for only the first nine months of 2016 (123.99 million\\ngallons). To calculate actual production for the year, EPA extrapolated the\\nlikely RIN generation for the last three months of the year based on the\\nhistorical relation (a multiple of 1.8) between the average quarterly\\ngeneration in the first three quarters and that of the last quarter, yielding a\\nfigure of 74.39 million RINs for the last quarter. See Coffeyville J.A. 516–\\n17.\\n\\n ** The EPA error has been calculated as the difference between the\\nEPA estimate and actual production, divided by the actual production.\\n\\n Second, the Coffeyville Petitioners claim that EPA failed\\nto generate a projection “based on” the cellulosic biofuel\\nestimate provided by the Energy Information Administration\\n(EIA), as required by statute. 42 U.S.C. § 7545(o)(7)(D)(i).\\nThat is so, they say, because 96% of EPA’s projected volume\\nwas biogas, a type of cellulosic biofuel that EIA did not include\\nin its estimate. See Coffeyville Br. 43–44. The problem for\\npetitioners, however, is that this closely parallels an argument\\nwe rejected in ACE: “[W]e do not agree that EPA failed to\\ngenerate projections ‘based on’ the [EIA’s] estimates,” even\\nthough those “estimates did not contain figures for [biogas]\\nproduction—production that accounts for the vast majority of\\ncellulosic biofuel” (“around 90 percent”). ACE, 864 F.3d at\\n724, 729. Here, as there, EPA showed sufficient “respect” for\\nEIA’s estimates. Id. at 729. When limited to fuels actually\\nanalyzed by EIA, EPA’s estimates were “very similar” to\\nEIA’s, id.; see 2017 Rule, 81 Fed. Reg. at 89,758/1, a fact that\\nthe Coffeyville Petitioners do not contest.\\n\\n Congress demanded no more. Nothing in the statute\\nrequired EPA to, as the Coffeyville Petitioners insist, “work[]\\nwith the EIA to develop information” about biogas.\\nCoffeyville Br. 44. “[T]he Administrator of the Energy\\nInformation Administration shall provide . . . an estimate,” 42\\nU.S.C. § 7545(o)(3)(A), and EPA shall “respect” it, API, 706\\n\\n 56\\nF.3d at 478. That’s it. In showing such respect, EPA, of\\ncourse, must “understand how EIA derived” its estimate.\\nCoffeyville Reply Br. 24. But, contrary to the Coffeyville\\nPetitioners’ contention, EPA did just that. The agency\\nidentified the types of cellulosic biofuels that EIA considered\\nand then, to test the integrity of its projection, conducted an\\napples-to-apples comparison, “limiting the scope of [its]\\nprojection to the companies assessed by EIA.” See 2017 Rule,\\n81 Fed. Reg. at 89,757–58. Nothing more was required.\\n\\n Third, the Coffeyville Petitioners object that EPA relied on\\n“information from the [biogas] industry”—an industry “with a\\ndirect financial interest in the outcome of the rule.” Coffeyville\\nBr. 45. The Petitioners characterize this as “reliance on\\nundisclosed information.” Id. But EPA did disclose the\\ninformation from the biogas industry—and that the information\\ncame from that industry; it just did so in the aggregate. See\\nOctober 2016 Assessment of Cellulosic Biofuel Production\\nfrom Biogas (2017), EPA-HQ-OAR-2016-0004-3711, at 2–6\\n(Dec. 12, 2016), Coffeyville J.A. 536–40. All the agency\\nwithheld was company-specific information, claiming that it\\nhad to withhold such data as confidential business information,\\nsee id. at 7, Coffeyville J.A. 541; see also 40 C.F.R. § 2.211(b),\\na claim that petitioners never even attempt to rebut, see Masias\\nv. EPA, 906 F.3d 1069, 1077 (D.C. Cir. 2018) (“It is not enough\\nmerely to mention a possible argument in the most skeletal\\nway, leaving the court to do counsel’s work . . . .” (quoting\\nSchneider v. Kissinger, 412 F.3d 190, 200 n.1 (D.C. Cir.\\n2005))).\\n\\n As for the implication of bias, we have previously upheld\\nEPA’s reliance on “biofuel producers’ own forecasts.” ACE,\\n864 F.3d at 728; see also API, 706 F.3d at 478 (recognizing\\nthat producers are an “almost inevitable source of\\ninformation”). Here, as in ACE, EPA did not “blindly adopt[]\\n\\n 57\\nthe facilities’ own forecasts”; it “performed its own\\ninvestigation.” 864 F.3d at 728; see October 2016 Assessment\\nof Cellulosic Biofuel Production from Biogas, supra, at 4,\\nCoffeyville J.A. 538 (“To verify the reasonableness of these\\nprojections, EPA compared the projected volume from each\\nregistered facility to the registered capacity of that facility.”).\\nPetitioners point to no unreasonable step by EPA in its efforts\\nto address “the uncertainty and unreliability identified by the\\n[Coffeyville] Petitioners.” ACE, 864 F.3d at 728.\\n\\n Fourth, the Coffeyville Petitioners protest EPA’s reliance\\non “facilities’ actual production in prior years” as a floor for\\nprojecting future cellulosic biofuel production. Coffeyville Br.\\n46. This was error, they say, because some companies might\\ncease production. Perhaps so. But, as we said in ACE,\\nalthough unforeseen issues “could prevent a producer from\\nmeeting” its prior year’s production, it was “reasonable” for\\nEPA to expect, as a general matter, “that a company’s output\\nwould grow year-over-year as the company gained\\nexperience.” 864 F.3d at 728. This seems especially true in an\\nindustry with the government’s wind surging at its back. And\\neven were EPA’s assumption not true for each company, any\\none facility’s shortfall, EPA explained, could be “off-set” by\\nnew facilities coming online or existing facilities exceeding the\\nhigh end of their projected production range. See Renewable\\nFuel Standard Program—Standards for 2017 and Biomass-\\nBased Diesel Volume for 2018: Response to Comments, EPA-\\nHQ-OAR-2016-0004-3753, at 444 (Dec. 12, 2016),\\nCoffeyville J.A. 707. This explanation fulfills EPA’s “duty to\\narticulate a ‘reasonable and reasonably explained’ approach to\\nsetting the low end of the production ranges.” ACE, 864 F.3d\\nat 729 (quoting Comtys. for a Better Env’t v. EPA, 748 F.3d\\n333, 335 (D.C. Cir. 2014)).\\n\\n 58\\n Fifth, the Coffeyville Petitioners complain that EPA\\nshould have based its cellulosic biofuel projections on “actual”\\nprior production. Coffeyville Br. 47. But this backward\\nlooking approach would have, in EPA’s view, “ignore[d] the\\npotential for facilities . . . to increase their fuel production\\nrates” and would have been “inappropriately conservative” in\\nlight of the “year-over-year increases” that EPA had observed\\n“in recent years.” 2017 Rule, 81 Fed. Reg. at 89,761/1. We\\ncannot say that in rejecting such an approach EPA violated “its\\nduty to take a ‘neutral aim at accuracy.’” ACE, 864 F.3d at 727\\n(quoting API, 706 F.3d at 476).\\n\\n For these reasons, we reject the Coffeyville Petitioners’\\nchallenges to EPA’s cellulosic biofuel projection for 2017. See\\nACE, 864 F.3d at 729.\\n\\nC. Cellulosic Waiver\\n\\n The Coffeyville Petitioners also challenge EPA’s decision\\nto use less than all of its discretionary cellulosic waiver\\nauthority to lower the 2017 requirements for advanced biofuel\\nand total renewable fuel. Having reduced the 2017 cellulosic\\nbiofuel requirement by 5.189 billion gallons, EPA had\\nauthority to reduce the advanced biofuel and total renewable\\nfuel requirements “by the same or a lesser volume.” 42 U.S.C.\\n§ 7545(o)(7)(D)(i). To decide by how much to reduce these\\nstatutory requirements, EPA first determines what reduction in\\nthe advanced biofuel requirement will yield a “reasonably\\nattainable” volume, and it then mechanically applies an\\nequivalent reduction to the total renewable fuel volume. 2017\\nRule, 81 Fed. Reg. at 89,752–53. Petitioners do not directly\\nchallenge this methodology. Instead, they argue that EPA\\napplied it arbitrarily in deciding to waive only 4.719 billion\\ngallons of the advanced biofuel volume for 2017, rather than\\nthe maximum available waiver of 5.189 billion gallons. We\\n\\n 59\\nagain reject some of their arguments as foreclosed by precedent\\nand others on their own terms.\\n\\n First, the Coffeyville Petitioners argue that EPA sought to\\njustify its 2017 advanced biofuel volume in part by making an\\nimpermissible comparison to the statutory volume set by\\nCongress for 2022. In response to a comment expressing\\nconcerns about utilization of non-cellulosic advanced biofuels\\n(which could be food-based) and possible adverse effects on\\nfood availability, EPA noted that its “reasonably attainable”\\nnon-cellulosic advanced biofuel volume for 2017\\n(approximately 4 billion gallons) was “somewhat higher than\\nthe level envisioned in the statute for 2017” (3.5 billion), “but\\nwell below the level of such fuels Congress expected would be\\nused by 2022” (5 billion). Response to Comments at 214,\\nCoffeyville J.A. 689. According to petitioners, “by comparing\\n2017 volumes with 2022 statutory targets, EPA departed from\\nCongress’s intent.” Coffeyville Br. 50.\\n\\n However, nothing in the statute forbids EPA from taking\\naccount of future statutory volumes in this way. Although\\nCongress specified presumptively applicable volumes for\\ncertain years, it also provided waiver authority to depart from\\nthose volumes. Indeed, the discretionary waiver provision\\nnecessarily empowers EPA to depart upward from the statutory\\nlevel of non-cellulosic advanced biofuel for a given year:\\nreducing the advanced biofuel volume by less than the\\nreduction in cellulosic biofuel, as section 7545(o)(7)(D)(i)\\npermits, is mathematically equivalent to increasing the volume\\nof non-cellulosic advanced biofuels, to “partially backfill for\\nmissing cellulosic biofuel.” 2017 Rule, 81 Fed. Reg. at\\n89,763/1. As we have noted, the cellulosic waiver provision\\n“grants EPA ‘broad discretion’ to consider a variety of factors”\\nin exercising this authority to depart from the presumptive\\nstatutory volumes. ACE, 864 F.3d at 733 (quoting Monroe\\n\\n 60\\nEnergy, 750 F.3d at 915). In this case, while deflecting a\\ncomment about food availability, EPA observed that its non-\\ncellulosic advanced biofuel volume for 2017—while higher\\nthan the statutory volume envisioned for that year—was lower\\nthan the presumptive statutory volume for the near future. And\\nit then reasonably concluded that this “somewhat higher\\ninterim volume reflect[ed] [its] assessment that it is appropriate\\nto allow non-cellulosic advanced biofuels to partially backfill\\nfor missing cellulosic volumes in light of the associated\\n[greenhouse gas] and energy security benefits.” Response to\\nComments at 214, Coffeyville J.A. 689.\\n\\n Second, the Coffeyville Petitioners argue that EPA failed\\nto explain its estimate of reasonably attainable 2017 imports of\\nsugarcane ethanol, a type of non-cellulosic advanced biofuel.\\nSugarcane ethanol imports have varied greatly from year to\\nyear, reaching a high of 681 million gallons in 2006 but falling\\nto 64 million gallons in 2014 and 89 million gallons in 2015.\\nSee 2017 Rule, 81 Fed. Reg. at 89,764. At the time of the 2017\\nRule, EPA expected only 76 million gallons to be imported in\\n2016, but it nonetheless adhered to its proposed estimate of 200\\nmillion gallons for 2017—an estimate originally based on\\nEPA’s judgment that circumstances in 2017 were “not . . .\\nsignificantly different” from circumstances in 2016, for which\\nEPA had also projected 200 million gallons. Id. at 89,763/3.\\nEPA acknowledged the “recent low import levels,” but also\\ncited “the difficulty in precisely identifying the reasons” for the\\nhistorical “high variability,” given “uncertainty” as to market\\nfactors including “ongoing growth in gasoline demand in\\nBrazil, and competing world demand for sugar.” Id. at 89,764–\\n65. The agency accordingly reaffirmed that 200 million\\ngallons “reflects a reasonable intermediate point between the\\nlower levels imported recently and the considerably higher\\nlevels that have been achieved in earlier years.” Id. at 89,765/2.\\n\\n 61\\n There is some force to petitioners’ objection to EPA’s\\nadherence to an estimate well over double the actual imports in\\nthe three preceding years. However, our review is “particularly\\ndeferential in matters implicating predictive judgments.”\\nMusic Choice v. Copyright Royalty Bd., 774 F.3d 1000, 1015\\n(D.C. Cir. 2014) (quoting Rural Cellular Ass’n v. FCC, 588\\nF.3d 1095, 1105 (D.C. Cir. 2009)). We accordingly upheld\\nEPA’s identical 2016 sugarcane ethanol estimate as\\n“reasonable and reasonably explained” in ACE, 864 F.3d at 736\\n(quotation marks omitted). In that case, we held that EPA\\nreasonably “concluded that ‘a somewhat lower level of imports\\nwill occur than the historic average’ of 300 million,” based on\\na similar analysis of market factors. Id. (quoting 2014–16\\nRule, 80 Fed. Reg. at 77,478/2). Here, we cannot say that one\\nmore year of low imports made it arbitrary for EPA to adhere\\nto that same projection for 2017.\\n\\n Third, the Coffeyville Petitioners object to EPA’s analysis\\nof supply and demand for regular gasoline (E0) and gasoline\\nwith added ethanol (E15 and E85). However, this analysis\\nplayed no role in EPA’s exercise of its discretionary cellulosic\\nwaiver authority under section 7545(o)(7)(D)(i). As noted\\nabove, EPA’s exercise of that authority rested entirely on its\\ndetermination of reasonably attainable advanced biofuel\\nvolumes. See 2017 Rule, 81 Fed. Reg. at 89,773–74. The\\ndisputed analysis of E0, E15, and E85 supported EPA’s\\nseparate decision not to invoke its “general waiver” authority,\\nunder section 7545(o)(7)(A)(ii), based on “inadequate\\ndomestic supply.” See generally ACE, 864 F.3d at 705–13.\\nBut in their opening brief, petitioners failed to challenge EPA’s\\ndecision not to invoke that separate waiver provision for 2017.\\nAlthough their reply brief gestures at this point, “an argument\\nfirst made in a reply brief is forfeited.” Bartko v. SEC, 845\\nF.3d 1217, 1225 n.7 (D.C. Cir. 2017).\\n\\n 62\\n Finally, the Coffeyville Petitioners take issue with EPA’s\\nresponse to various comments. We have considered these\\narguments and find them to be without merit.\\n\\n For these reasons, we reject the Coffeyville Petitioners’\\nchallenges to EPA’s exercise of its discretionary cellulosic\\nwaiver authority to reduce advanced biofuel and total\\nrenewable fuel volumes for 2017.\\n\\n VI. 2018 Volume for Biomass-Based Diesel\\n\\n Since 2012, EPA, acting in coordination with the\\nSecretaries of Energy and Agriculture, has calculated the\\nannual applicable volume (also known as the “volume\\nrequirement”) for biomass-based diesel based on a holistic,\\nbackward- and forward-looking consideration of relevant\\nfactors. In particular, it has set the volume requirement “based\\non a review of the implementation of the program during\\ncalendar years specified in the tables, and an analysis of” six\\nstatutorily enumerated factors: (1) “the impact of the\\nproduction and use of renewable fuels on the environment”; (2)\\n“the impact of renewable fuels on the energy security of the\\nUnited States”; (3) “the expected annual rate of future\\ncommercial production of renewable fuels, including advanced\\nbiofuels in each category (cellulosic biofuel and biomass-based\\ndiesel)”; (4) “the impact of renewable fuels on the\\ninfrastructure of the United States”; (5) “the impact of the use\\nof renewable fuels on the cost to consumers of transportation\\nfuel and on the cost to transport goods”; and (6) “the impact of\\nthe use of renewable fuels on other factors, including job\\ncreation, the price and supply of agricultural commodities,\\nrural economic development, and food prices.” 42 U.S.C.\\n§ 7545(o)(2)(B)(ii)(I)–(VI).\\n\\n EPA set the 2018 applicable volume for biomass-based\\ndiesel at 2.1 billion gallons, up from 2.0 billion gallons in 2017,\\n\\n 63\\nand 1.1 billion gallons above a statutory minimum that\\nCongress set to plateau at 1 billion gallons as of 2012. 2017\\nRule, 81 Fed. Reg. at 89,798/1; see 42 U.S.C.\\n§ 7545(o)(2)(B)(i)(IV), (v). NBB had asked EPA to set the\\nbiomass-based diesel volume at 2.5 billion gallons, and now\\nchallenges the volume EPA set as arbitrary and capricious and\\ncontrary to the Clean Air Act.\\n\\nA. NBB’s Standing\\n\\n Before considering the merits of NBB’s claims, we must\\nsatisfy ourselves that NBB has standing to assert them.\\nRespondent-Intervenors, the American Fuel & Petrochemical\\nManufacturers and the American Petroleum Institute, contend\\nthat NBB lacks standing because, they say, it has not shown\\nthat the 2017 Rule inflicted a cognizable injury on any of its\\nmembers.\\n\\n NBB has associational standing here for the same reasons\\nwe held it did in National Biodiesel Board v. EPA, 843 F.3d\\n1010, 1015 (D.C. Cir. 2016) (NBB v. EPA), where EPA’s\\nactions “incentivize[d] . . . compet[ition] with [NBB’s\\nmembers’] domestic production.” Here, too, NBB’s members\\n“compete with” the other industry players EPA’s rule is\\ndesigned to affect. Id. at 1016. Recall that biomass-based\\ndiesel is a nested subset of advanced and total renewable fuels,\\nsuch that NBB’s members get (1) a market for compelled\\nbuyers of the specified volume of biomass-based diesel, for\\nwhich they are the exclusive suppliers, plus (2) a market for\\ncompelled buyers of advanced and other renewable fuels\\nalongside a broad array of competing suppliers. See supra at\\n6–8, 11. The 2017 Rule preamble explains that biomass-based\\ndiesel “compet[es] for research and development dollars with\\nother types of advanced biofuels,” and that, “[b]y establishing\\n[the biomass-based diesel] volume requirement[] at [a] level[]\\n\\n 64\\nlower than . . . the expected production of [biomass-based\\ndiesel],” EPA was “creating the potential for some competition\\nbetween [biomass-based diesel] and other advanced biofuels to\\nsatisfy the advanced biofuel” applicable volume and providing\\n“incentives for the continued development of” those\\ncompetitors’ fuels. 81 Fed. Reg. at 89,797; see also EPA\\nCoffeyville Br. 24 (“Above 2.1 billion gallons, biomass-based\\ndiesel will have to compete with other types of advanced\\nbiofuel.”). Such competition will likely “temper to some extent\\n[biomass-based diesel] prices.” Final Statutory Factors\\nAssessment for the 2018 Biomass Based Diesel (BBD)\\nApplicable Volume, EPA-HQ-OAR-2016-0004-3708, at 10\\n(Dec. 12, 2016) (Supplemental Assessment), Coffeyville J.A.\\n533. That is a cognizable injury to NBB’s members. See NBB\\nv. EPA, 843 F.3d at 1015–16; see also Delta Constr. Co. v.\\nEPA, 783 F.3d 1291, 1299 (D.C. Cir. 2015) (per curiam).\\n\\n Though NBB failed to identify any of its members—\\nordinarily a prerequisite for organizations alleging\\nassociational standing, see Summers v. Earth Island Inst., 555\\nU.S. 488, 497–98 (2009)—that omission is not fatal here\\nbecause NBB’s members comprise “the entire biomass-based\\ndiesel category of the Renewable Fuel Standard[s]” and\\nrepresent no other interests. Coffeyville J.A. 134. Consistent\\nwith “the real purpose of the [standing] inquiry—that is, for the\\ncourt to be satisfied that the requisite injury really has occurred\\nor will occur in the future to members of the organization[],”\\nPub. Citizen v. FTC, 869 F.2d 1541, 1552 (D.C. Cir. 1989),\\nthere is no need to identify injured members when “all the\\nmembers of the organization are affected by the challenged\\nactivity,” Summers, 555 U.S. at 499 (citing NAACP v. Ala. ex\\nrel. Patterson, 357 U.S. 449, 459 (1958)). Because EPA’s rule\\nsubjects the biomass-based diesel industry to increased\\ncompetition, with anticipated pricing effects, NBB “meet[s] the\\n\\n 65\\nconstitutional prerequisites of injury, causation, and\\nredressability.” NBB v. EPA, 843 F.3d at 1015.\\n\\nB. Merits of NBB’s Challenges\\n\\n NBB advances two challenges to the applicable volume\\nEPA set for biomass-based diesel: First, that EPA erred in\\nconsidering the interaction of biomass-based diesel with the\\nyet-to-be established 2018 advanced biofuel applicable\\nvolume, and second, that EPA’s consideration of the six\\nstatutory factors was arbitrary and capricious and contrary to\\nlaw. We reject both claims.\\n\\n First, EPA reasonably chose a 2018 biomass-based diesel\\napplicable volume that would “maintain[] support for growth\\nin [biomass-based diesel] volumes” while also encouraging the\\n“development of other advanced biofuels.” 2017 Rule, 81 Fed.\\nReg. at 89,798/1. Congress directed EPA to consider the\\nlessons learned from its retrospective “review” of the program,\\napply them in its prospective “analysis of” the six statutory\\nfactors, and set a biomass-based diesel volume that will apply\\nfourteen months in the future. See 42 U.S.C.\\n§ 7545(o)(2)(B)(ii).\\n\\n EPA’s approach is consistent with the structure and\\npurposes of the statute. Congress set a minimum applicable\\nvolume for biomass-based diesel of one billion gallons for each\\nyear from 2012 forward, id. § 7545(o)(2)(B)(i)(IV), (v), while\\nspecifying statutory minimum volumes for the advanced\\nbiofuel category containing biomass-based diesel that grow\\nyear by year to 21 billion gallons by 2022, id.\\n§ 7545(o)(2)(B)(i)(II), (iii). EPA reasonably concluded that,\\nby nesting biomass-based diesel together with cellulosic (and\\nother unspecified) biofuels within the advanced biofuel\\ncategory, and specifically charting a higher, steeper, and longer\\ninitial growth curve for advanced biofuel, Congress anticipated\\n\\n 66\\nthat production of other types of advanced biofuels could step\\nup to help meet the advanced biofuel volume requirement. See\\n2017 Rule, 81 Fed. Reg. at 89,797/1. EPA also reasonably\\nconcluded that increasing fuel diversity serves one of\\nCongress’s primary goals in establishing the Renewable Fuel\\nStandards program: improving the nation’s “energy\\nindependence and security.” See Pub. L. No. 110-140,\\npreamble; see also 2017 Rule, 81 Fed. Reg. at 89,798/3. EPA\\nalso reasonably anticipated that enhanced competition in the\\nadvanced biofuels market would help “temper to some extent\\n[biomass-based diesel] prices,” Supplemental Assessment 10,\\nCoffeyville J.A. 533, thereby ameliorating Congress’s concern\\nthat, with a too-high target volume, the “price of biomass-based\\ndiesel fuel” would “increase significantly,” 42 U.S.C.\\n§ 7545(o)(7)(E)(ii). And fuel diversity may produce\\nenvironmental benefits insofar as certain advanced biofuels,\\nsuch as ethanol from food waste, will “likely have significantly\\nlower impacts on wetlands, ecosystems, and wildlife habitats”\\nthan would greater reliance on biomass-based diesel.\\nSupplemental Assessment 6, Coffeyville J.A. 529.\\n\\n NBB’s arguments to the contrary turn on reading the\\nstatutory directive that EPA “review . . . the implementation of\\nthe program during calendar years specified in the tables,” id.\\n§ 7545(o)(2)(B)(ii), to confine EPA’s consideration to\\nbiomass-based diesel’s statutory volumes and actual\\nperformance, and to prevent EPA from considering other fuel\\ncategories or future years. In particular, NBB takes issue with\\nEPA’s consideration of the not-yet-finalized 2018 advanced\\nbiofuel applicable volume, which NBB contends led EPA to\\nset the biomass-based diesel volume too low.\\n\\n NBB’s objections are not supported by the text or purpose\\nof the statute. Assuming NBB is right that EPA’s “review of\\nthe implementation of the program” consists of a retrospective\\n\\n 67\\nassessment, the agency must also conduct “an analysis of” six\\nstatutory factors. Id. § 7545(o)(2)(B)(ii). And those factors\\nplainly require a prospective assessment—an assessment that\\nwould likely miss “important aspects of the problem,” State\\nFarm, 463 U.S. at 43, if it ignored the interaction, now and in\\nthe future, of the requirements for all the categories of\\nrenewable fuels. See, e.g., 42 U.S.C. § 7545(o)(2)(B)(ii)(I)\\n(requiring an “analysis of” the “impact of the production and\\nuse of renewable fuels on,” among other things, “the\\nenvironment”). Though EPA set the biomass-based diesel\\nrequirement lower than NBB wished, Congress did not intend\\nto incentivize growth of biomass-based diesel “at all costs.”\\nACE, 864 F.3d at 714 (quoting Am. Express Co. v. Italian\\nColors Rest., 570 U.S. 228, 234 (2013)).\\n\\n NBB objects that setting the 2018 biomass-based diesel\\napplicable volume below expected production might lead to a\\ndepressed advanced biofuel volume for 2018. But the agency\\nspecifically anticipated “that the 2018 advanced biofuel\\nrequirement will be larger than the 2017 advanced biofuel\\nvolume requirement.” 2017 Rule, 81 Fed. Reg. at 89,798/1.\\nEPA has never set the biomass-based diesel applicable volume\\nat the “maximum potential production” level, id. at 89,799/3,\\nyet the “growing supply of” biomass-based diesel has\\nconsistently “allowed EPA to establish higher advanced\\nbiofuel” applicable volumes, id. at 89,797/3. EPA opted for\\n“allowing room within the advanced biofuel volume\\nrequirement for the participation of non-[biomass-based diesel]\\nadvanced fuels” as a reasonable way “to encourage the\\ndevelopment and production of a variety of advanced biofuels\\nover the long term without reducing the incentive for [biomass-\\nbased diesel] beyond the [biomass-based diesel applicable\\nvolume] in 2018.” Id. at 89,797–98.\\n\\n 68\\n Second, in setting the 2018 biomass-based diesel\\napplicable volume, EPA reasonably compared the advantages\\nand disadvantages of biomass-based diesel to those of other\\nfuels. NBB contends that the statute confines EPA to assessing\\nadvantages of biomass-based diesel over petroleum, not\\nconsidering other renewable fuels, and that the agency failed to\\n“meaningfully” consider the six factors. NBB Br. 9–10. Both\\narguments miss the mark.\\n\\n NBB suggests that, because the statute “was intended to\\n‘increase the production of clean renewable fuels’ as a\\nsubstitute for petroleum fuel,” id. at 21 (quoting Pub. L. No.\\n110-140, preamble), the only relevant comparison is to\\npetroleum, not to other categories of renewable fuel. But NBB\\nidentifies nothing in section 7545(o)(2)(B)(ii) or any other\\nsection that requires EPA to assess the performance of a\\nparticular renewable solely by reference to petroleum fuel. Its\\nanalysis would require us to read the term “renewable fuels”\\nused throughout section 7545(o)(2)(B)(ii) to refer to the single\\nrenewable fuel being analyzed, even though the statutory\\ndefinition of “renewable fuel” includes all types of renewables.\\nSee 42 U.S.C. § 7545(o)(1)(J). And if EPA could compare the\\nbenefits of each specific fuel only to petroleum, it might be\\nunable to set rational applicable volumes for each specified\\ncategory of renewable fuel after 2022, when the statute no\\nlonger sets any specific volumes. See id. § 7545(o)(2)(B)(iii)–\\n(v). EPA could easily conclude, for example, that each\\nrenewable fuel had a lower “impact . . . on the environment”\\nthan petroleum fuel, see id. § 7545(o)(2)(B)(ii)(I), but, no\\nmatter their differing merits in serving the statute’s goals, the\\nagency would be barred from making relative judgments\\namong renewable fuel categories.\\n\\n NBB also argues that EPA failed to give meaningful\\nconsideration to the six statutory factors, and instead “pre-\\n\\n 69\\ndetermined the outcome,” NBB Br. 20, but the record shows\\notherwise. EPA considered in detail how setting the biomass-\\nbased diesel applicable volume at a level higher or lower than\\n2.1 billion gallons would affect the six statutory factors. See\\n2017 Rule, 81 Fed. Reg. at 89,798–99. EPA further elaborated\\nits analysis of the factors in an 11-page supplemental\\nmemorandum evaluating effects of its proposed biomass-based\\ndiesel volume on renewable fuel production rates, the\\nenvironment, and the economy. See Supplemental Assessment\\n1–11, Coffeyville J.A. 524–34. EPA concluded that, over the\\nlong term, “[a] variety of different types of advanced biofuels,\\nrather than a single type such as [biomass-based diesel], would\\npositively impact energy security . . . and increase the\\nlikelihood of the development of lower cost advanced biofuels\\nthat meet the same [greenhouse gas] reduction threshold as\\n[biomass-based diesel].” Supplemental Assessment 3,\\nCoffeyville J.A. 526. EPA thus concluded that the statutory\\nfactors supported its biomass-based diesel applicable volume.\\nSee 2017 Rule, 81 Fed. Reg. at 89,798/3.\\n\\n At bottom, NBB’s objections rest on a policy\\ndisagreement: NBB urges that, instead of setting a level that\\nwould support continued investment in the biomass-based\\ndiesel industry while also encouraging producers of other types\\nof advanced biofuel to compete to satisfy the 2018 advanced\\nbiofuel applicable volume at lower cost, EPA should have\\nreserved to biomass-based diesel alone a volume nearer to that\\nindustry’s maximum production potential. But NBB’s\\nproposed “simple solution”—that EPA should have “set[] a\\nmeaningful [biomass-based diesel] volume” while planning to\\n“increas[e] the 2018 advanced-biofuel volume to provide room\\nfor the production of other advanced biofuels when it set that\\nvolume a year later,” NBB Br. 23—describes what EPA\\nactually did. A mere disagreement with the particular\\ncalibration of a line drawn in the exercise of an agency’s\\n\\n 70\\nreasonable judgment is no basis to invalidate a rule. Therefore,\\nwe deny NBB’s petition.\\n\\n VII. Conclusion\\n\\n For these reasons, the petitions for review are denied.\\n\\n So ordered.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435003/\", \"author_raw\": \"PER CURIAM\"}, {\"author\": \"WILLIAMS, Senior Circuit Judge, concurring\", \"type\": \"concurrence\", \"text\": \"WILLIAMS, Senior Circuit Judge, concurring in part and\\nconcurring in the judgment:\\n\\n The Clean Air Act’s Renewable Fuel Program operates on\\nan annual cycle. It provides annual credits, authorizes annual\\nwaivers, and calls for annual reviews, see, e.g., 42 U.S.C.\\n§ 7545(o)(5), (7), (10)—all to implement Congress’s annual\\ngoals, see id. § 7545(o)(2)(B).\\n\\n Each year, as part of this annual affair, the Environmental\\nProtection Agency embarks on an elaborate rulemaking. Id.\\n§ 7545(o)(3)(B). In doing so, it receives an annual estimate of\\nthe total volume of fuel to be sold to inform it in setting the\\nannual “renewable fuel obligation,” id. § 7545(o)(B)(i), which\\n“shall . . . be applicable to refineries, blenders, and importers,\\nas appropriate,” id. § 7545(o)(3)(B)(ii)(I). So EPA is to specify\\nappropriateness among those three categories. But\\n“appropriate” as of when?\\n\\n Coffeyville Petitioners say appropriate as of the annual\\nrulemaking.\\n\\n But EPA says appropriate as of the last time EPA happened\\nto consider the issue, no matter how many years earlier that\\nwas. The initial determination sticks for “all years,” EPA says,\\n“unless and until” EPA chooses, in its “discretion,” to\\n“undertake [an] annual reevaluation[].” Denial of Petitions for\\nRulemaking to Change the RFS Point of Obligation, EPA-HQ-\\nOAR-2016-0544-0525, at 7 (Mar. 13, 2018) (“EPA Denial”),\\nJ.A. 779; see also EPA Br. 66 (claiming “discretion” to decide\\n“whether, how, and when” it will “reconsider its initial\\ndesignation”). Even when affected parties point to a series of\\nmarket “disparities” that they say have developed and render\\nthe earlier determination “not appropriate,” and point to\\n§ 7545(o)(3)(B)(ii)(I) as entitling them to a fresh\\ndetermination, see, e.g., Valero Energy Corporation\\n\\n 2\\n\\nComments, EPA-HQ-OAR-2016-0004-1746, at 1, 14 (July 11,\\n2016), J.A. 138, 151, EPA claims that that section does nothing\\nof the sort, see EPA Denial at 8, J.A. 780 (asserting full\\n“discretion” to decide “when” and “under what circumstances”\\nit will consider the issue). In Part V.A of the court’s opinion,\\nmy colleagues accept EPA’s theory. I, however, disagree. So\\nwhile I otherwise join the court’s opinion in full, I cannot joint\\nPart V.A—though I do, in the end, concur in the judgment.\\n\\n\\n * * *\\n\\n At the risk of oversimplifying, we can boil this annual\\nprocess down to three steps.\\n\\n First, the annual goal. Congress sets annual (steadily\\nincreasing) goals for the volume of renewable transportation\\nfuel to be sold or introduced into commerce in the United\\nStates, with special targets for some subsets of renewable fuel.\\n42 U.S.C. § 7545(o)(2)(B)(i).\\n\\n Next, the annual estimate. The Energy Information\\nAdministration projects the total volume of transportation fuel\\nthat will be sold into commerce in a given year (as well as\\nvolumes of biomass-based diesel and cellulosic biofuel). 42\\nU.S.C. § 7545(o)(3)(A); see, e.g., Letter from Adam Sieminski,\\nAdministrator, U.S. Energy Information Administration, to\\nGina McCarthy, Administrator, U.S. Environmental Protection\\nAgency, EPA-HQ-OAR-2016-0004-3646 (Oct. 19, 2016), J.A.\\n494.\\n\\n Finally, the annual obligation. This is set by EPA during\\nthe agency’s annual rulemaking. And it is expressed in terms\\nof a single percentage of transportation fuel sold into commerce\\n(the “renewable fuel obligation”) by any obligated party\\n(regardless of category). 42 U.S.C. § 7545(o)(3)(B)(ii); see,\\ne.g., Renewable Fuel Standard Program: Standards for 2017\\n\\n 3\\n\\nand Biomass-Based Diesel Volume for 2018, 81 Fed. Reg.\\n89,746, 89,751/3 (Dec. 12, 2016) (“2017 Rule”). The basic\\nidea is this: If EPA knows (i) the annual goal for the volume of\\nrenewable fuel introduced into commerce (see step one above),\\nand (ii) the annual estimate for the total volume of fuel to be\\nintroduced into commerce (see step two above), then EPA—\\nafter filling in any gaps in the goals left by Congress, see 42\\nU.S.C. § 7545(o)(2)(B)(ii), and making any necessary\\nadjustments to the estimates provided by the Energy\\nInformation Administration, see id. § 7545(o)(3)(A)—can set\\nthe minimum percentage of renewable fuel that must be\\nintroduced into commerce by “obligated parties.” If everything\\nworks out well, Congress’s annual goal should, more or less,\\nbe met.\\n\\n But who are these “obligated parties”? Under the Act,\\nEPA must tell us. The first among the three “Required\\nelements” of the annual determination is that it “be applicable\\nto refineries, blenders, and importers, as appropriate.” 42\\nU.S.C. § 7545(o)(3)(B)(ii)(I) (emphasis added).\\n\\n Even though EPA “determine[s] and publish[es]” the\\nannual obligation anew “[e]ach [] calendar year[],” 42 U.S.C.\\n§ 7545(o)(3)(B)(i), since 2010 it hasn’t considered what parties\\nare “appropriate” to obligate. Regulation of Rules and Fuel\\nAdditives, 75 Fed. Reg. 14,670, 14,722 (Mar. 26, 2010); see\\nalso Regulation of Fuels and Fuel Additives, 72 Fed. Reg.\\n23,900, 23,923/2 (May 1, 2007). Rather, year in and year out,\\nthe agency has simply “indicated,” “in passing,” that the\\nrenewable fuel obligation “would apply to ‘. . . producers and\\nimporters,’” “consistent with [its] preexisting” determination.\\nEPA Br. 69–70 (quoting 2017 Rule, 81 Fed. Reg. at 89,746/2).\\nThat’s it. In my view, however, the language of the statute\\nrequires more. EPA’s contrary reading seems to me to go\\nunreasonably “beyond the meaning that the statute can bear.”\\nU.S. Postal Serv. v. Postal Regulatory Comm’n, 886 F.3d 1253,\\n\\n 4\\n\\n1255 (D.C. Cir. 2018) (quoting MCI Telecomm. Corp. v. AT&T\\nCo., 512 U.S. 218, 229 (1994)).\\n\\n\\n * * *\\n\\n The key provision says, “[n]ot later than November 30 of\\neach [] calendar year[],” EPA “shall determine and publish in\\nthe Federal Register . . . the renewable fuel obligation.” 42\\nU.S.C. § 7545(o)(3)(B)(i). The first of the “Required\\nelements” of that annual obligation is that it shall “be applicable\\nto refineries, blenders, and importers, as appropriate.” Id.\\n§ 7545(o)(3)(B)(ii)(I).\\n\\n This much tells us a few things. First, Congress required\\nEPA to set the renewable fuel obligation annually. That feature\\nof the requirement pretty clearly indicates a congressional\\nexpectation of possible year-to-year variation in all the\\nmandatory elements—not merely in the percentage chosen\\n(which is addressed in subclauses (II) and (III)). Second, one\\nexplicitly required element of this annual determination is a\\nselection among “refineries, blenders and importers,” a\\nselection that must be “appropriate.” Taken together, the Act\\nseems inevitably to require EPA to apply (at least) some\\nthought to the issue of what market sectors should be\\nobligated—thought that the agency must apply each time it sets\\nthe annual obligation. After all, the term “appropriate”\\n“naturally and traditionally includes consideration of all the\\nrelevant factors,” not just a recitation that some time ago the\\nagency considered the factors that it then thought relevant.\\nMichigan v. EPA, 135 S. Ct. 2699, 2707 (2015) (emphasis\\nadded) (quoting White Stallion Energy Center, LLC v. EPA,\\n748 F.3d 1222, 1266 (D.C. Cir. 2014) (Kavanaugh, J.,\\nconcurring in part and dissenting in part)). The agency, in other\\nwords, must “exercise its discretion to choose among the\\noptions” that Congress has given it, Maj. op. 45 (quoting Kisor\\n\\n 5\\n\\nv. Wilkie, 139 S. Ct. 2400, 2449 (2019) (Kavanaugh, J.,\\nconcurring in the judgment)), not “explain[]” why, in the\\nagency’s opinion, it’s “appropriate” not to choose among the\\noptions that Congress has given it, id.; see Response to\\nComments, EPA-HQ-OAR-2016-0004-3753, at 542 (Dec. 12,\\n2016), J.A. 761 (declaring the point-of-obligation “issue”\\n“beyond the scope of this rulemaking”).\\n\\n Suppose a law school charter—adopted at the school’s\\nfounding in 1920—calls on the dean to annually set a “tort\\ncredits obligation,” consisting of a minimum number of credit\\nhours students must devote to certain tort subjects; the dean is\\nto make the obligation “applicable to negligence, defamation,\\nbattery, and alienation of affections, as appropriate.” The first\\ndean, in 1921, sets the obligation at three credit hours per\\nsubject—and applies it to all the subjects. For the 2020–21\\nacademic year, the tenth dean likewise duly requires students\\nto devote at least three credits hours to those same subjects—\\nincluding alienation of affections. Students understandably\\nprotest, since that tort is now a bygone relic. See Fitch v.\\nValentine, 2005-CA-01800-SCT (¶¶ 79–81) (Miss. 2007)\\n(Dickinson, J., concurring), 959 So. 2d 1012, 1036 (noting 31\\nstates have “completely abolished” it). But the dean adamantly\\nrefuses even to consider their entreaties, “explain[ing]” (Maj.\\nop. 45) they’re “beyond the scope” (J.A. 761) of topics relevant\\nto the annual credit determination, which, after all, is perfectly\\n“consistent with [a] preexisting” 1921 determination that that\\napplication was “appropriate” (EPA Br. 70). EPA’s reasoning\\n(on the procedural point—whether or not the phrase\\n“applicable . . . as appropriate” requires it to consider the issue)\\nis, in essence, as startling as the dean’s. Never mind whether,\\nas a substantive matter, studying the tort—or exempting\\nblenders—is actually “appropriate.” Cf. Maj. op. 52. EPA tells\\nus it need not even address the point—ever again.\\n\\n 6\\n\\n EPA’s response does more to hurt than to help its cause.\\nThe agency points us to similarities between the provision\\nwe’ve been discussing, § 7545(o)(3)(B)(ii)(I), and\\n§ 7545(o)(2)(A)(iii)(I), which I’ll call the “compliance\\nprovision.” The two echo each other, see Oral Arg. Tr. 70:19–\\n25, both using the “applicable . . . as appropriate” formulation.\\n\\nAnnual determination, 42 U.S.C. § 7545(o)(3)(B)(i), (ii)(I):\\n\\n [E]ach . . . calendar year[] . . . , the Administrator of the\\n Environmental Protection Agency shall determine and\\n publish in the Federal Register . . . the renewable fuel\\n obligation . . . . The renewable fuel obligation . . . shall . . .\\n be applicable to refineries, blenders, and importers, as\\n appropriate.\\n\\nCompliance provision, 42 U.S.C. § 7545(o)(2)(A)(i), (iii)(I):\\n\\n Not later than [August 8, 2006], the Administrator shall\\n promulgate regulations . . . . [T]he regulations . . . shall\\n contain compliance provisions applicable to refineries,\\n blenders, distributors, and importers, as appropriate . . . .\\n\\n As EPA reads the two, the agency may define the point of\\nobligation once—while announcing the compliance provisions\\nat the outset of the program. See EPA Br. 66. Congress’s\\ncommand to make the annual renewable fuel obligation\\n“applicable . . as appropriate” is simply, in the agency’s view,\\na cross-reference back to the “applicable . . . as appropriate”\\ndetermination made by EPA at the outset in its adoption of\\ncompliance regulations. See, e.g., id. at 69–70; Oral Arg. Tr.\\n70:19–71:15, 72:13–24, 73:16–74:13.\\n\\n The agency’s reading, however, seems utterly implausible.\\nWhen Congress uses “identical words” in “different parts of the\\nsame statute,” we normally infer that those words carry “the\\nsame meaning.” Henson v. Santander Consumer USA Inc., 137\\n\\n 7\\n\\nS. Ct. 1718, 1723 (2017) (quoting IBP, Inc. v. Alvarez, 546 U.S.\\n21, 34 (2005)). So if “applicable . . . as appropriate,” in the\\ncontext of setting the compliance regulations, means (as\\neveryone agrees it means) that EPA is to contemporaneously\\nassess the appropriateness of its decision, then the same phrase,\\nin the context of setting the annual renewable fuel obligation,\\nmust mean the same thing: EPA is to make a contemporaneous\\nassessment of appropriateness—rather than, as the agency\\nimplausibly claims, treat a decision made long ago as\\ndispositive for the present.\\n\\n The majority responds—somewhat bafflingly—that\\nnothing in the phrase “applicable . . . as appropriate” indicates\\n“when or in what context EPA must make the appropriateness\\ndetermination.” Maj. op. 46 (emphasis added). But that can’t\\nbe right. Imagine a daycare advertises that it will dress kids for\\nrecess, “as appropriate.” Would any reasonable speaker of\\nEnglish really harbor any doubt as to whether there existed a\\n“particular temporal” connection between the selection made\\nand the selection’s appropriateness? Id. at 45. Surely parents\\nwould be surprised to learn that the school’s clothing selection\\nfor a snowy, December day was not “appropriate” in light of\\nthe then-pounding blizzard, but, rather, was “appropriate” in\\nlight of the sunshine from six months earlier, when the daycare\\nfirst opened.\\n\\n In fact, had Congress wanted EPA to readopt a prior\\ndetermination, without any contemporaneous analysis as to\\nappropriateness, “it could easily have chosen clearer language”\\nto do just that. NLRB v. SW General, Inc., 137 S. Ct. 929, 939\\n(2017). Related provisions of the same statute provide\\nexamples of such straightforward wording. An obvious\\npossibility would be to replace “applicable to refineries,\\nblenders, and importers, as appropriate,” with “applicable to\\nObligated Parties (as defined by the Administrator under 42\\nU.S.C. § 7545(o)(2)),” thus using the pattern adopted in\\n\\n 8\\n\\n§ 7545(h)(1), (k)(3)(B)(i). Another obvious way of expressing\\nwhat EPA says Congress meant would have been to modify\\n“refineries, blenders, and importers” with the phrase, “in\\nconformity with the compliance provisions established by the\\nAdministrator,” thus paralleling the approach of § 7545(b)(2).\\nBoth formulations, relying on a past participle, easily invite the\\nconstruction that EPA prefers—allowing the administrator to\\nrely on a decision made at some unspecified time in the past.\\n“The fact that [Congress] did not adopt [any of these] readily\\navailable and apparent alternative[s] strongly supports rejecting\\n[EPA’s] reading.” Knight v. Commissioner, 552 U.S. 181, 188\\n(2008).\\n\\n Further, rather than using such easy alternatives, Congress\\nchose language that, as read by EPA, makes a mess of virtually\\nall of § 7545(o)(3)(B)(ii). Again, subclause (I) requires the\\n“renewable fuel obligation” to “be applicable to refineries,\\nblenders, and importers, as appropriate.” 42 U.S.C.\\n§ 7545(o)(3)(B)(ii)(I). If Congress had envisioned EPA\\n“identif[ying] the ‘appropriate’ obligated parties” in its exercise\\nof the compliance provision (§ 7545(o)(2)(A)(iii)(I)), rather\\nthan of this clause, as EPA says it did, see EPA Br. 7, then\\nsubclause (I) would be doing no work at all—contrary to the\\n“principle of statutory construction that we must ‘give effect, if\\npossible, to every clause and word of a statute,’” Williams v.\\nTaylor, 529 U.S. 362, 404 (2000) (quoting United States v.\\nMenasche, 348 U.S. 528, 538–39 (1955)).\\n\\n EPA and the majority respond that subclause (I) is needed\\nto “clarify[]” that distributors—who can be subjected to the\\ncompliance provisions—“cannot be” subjected to the\\nrenewable fuel obligation. Oral Arg. Tr. 75:9–12 (emphasis\\nadded); see also Maj. op. 47. Compare 42 U.S.C.\\n§ 7545(o)(2)(A)(iii)(I) (providing that the compliance\\nprovisions shall be “applicable to refineries, blenders,\\ndistributors, and importers, as appropriate” (emphasis added)),\\n\\n 9\\n\\nwith id. § 7545(o)(3)(B)(ii)(I) (providing that the renewable\\nfuel obligation shall be “applicable to refineries, blenders, and\\nimporters, as appropriate”). But the need for clarity could be\\nattributed to “most superfluous language.” SW General, 137 S.\\nCt. at 941. And if clarity were actually Congress’s goal, if all\\nCongress wanted to do in subclause (I) was exclude\\n“distributors” from the universe of potential obligated parties,\\nMaj. op. 47, it chose an exceedingly odd way of getting there:\\ninserting into an annual exercise the task of indicating what\\nentities are “appropriate” targets for the renewable fuel\\nobligation. Wouldn’t it have been more straightforward to just\\nreference EPA’s prior determination, and then directly state—\\nfor the purpose of clarity—that the renewable fuel obligation\\nmay not apply to “distributors”?\\n\\n In any case, it’s hard to see what distributor-based\\nobscurity EPA sees a need for subclause (I) to correct. Because\\nthe renewable fuel obligation concerns only fuel that is “sold or\\nintroduced into commerce in the United States,” 42 U.S.C.\\n§ 7545(o)(2)(A)(i), (o)(3)(B)(ii)(II) (emphasis added), the\\nobligation applies, for any gallon of fuel, only once—i.e., when\\nthe fuel enters the American economy upstream, not when\\ndistributors transport the same fuel downstream.\\n\\n Once the sale or introduction “into” commerce is\\ncomplete—once a given unit of fuel is already flowing through\\nAmerican commerce—that same unit of fuel cannot be sold or\\nintroduced “into” American commerce again; it’s already there.\\nWhile one, for example, might say that a fuel line, which carries\\nfuel from a car’s tank to its engine, carries fuel “in” the car, no\\none would say that it carries fuel “into” the car. So too, while\\none might say that a distributor, which transports fuel from the\\neconomy’s refineries to its retailers, see 40 C.F.R. § 80.2(l);\\nEPA Denial at 9, J.A. 781, transports fuel “in” the economy, no\\none would say that it transports (or sells or introduces) fuel\\n“into” the economy; again, the fuel is already in the relevant\\n\\n 10\\n\\nprocess. Congress itself recognizes the distinction, referring to\\nfuel that is “sold or introduced into commerce,” 42 U.S.C.\\n§ 7545(o)(2)(A)(i), (o)(3)(B)(ii)(II) (emphasis added), and fuel\\nthat is “sold or distributed in . . . commerce,” id. § 7545(u)(4)\\n(emphasis added). Because distributors do only the latter—\\nthey move fuel “in,” not “into,” commerce—there is nothing\\nfor subclause (I) to clarify. These downstream intermediaries\\ncan never fall within the universe of potentially obligated\\nparties.\\n\\n My colleagues don’t claim to disagree; at most, they\\ndeclare it “non-obvious” that “distributors cannot be subjected\\nto the point of obligation.” Maj. op. 47. But what’s “non-\\nobvious” about it, even if we put the plain meaning of “into\\ncommerce in the United States” aside? That phrase appears\\nthroughout the statute—and can’t possibly include\\ndownstream, distributor transactions. Take the statutory\\nprovision concerning the Energy Information Administration,\\nwhich says that the agency must provide EPA with an estimate\\nof the “volume[] of transportation fuel . . . projected to be sold\\nor introduced into commerce in the United States.” 42 U.S.C.\\n§ 7545(o)(3)(A). Does Congress really expect that estimate—\\nand the regulatory burdens “based on” that estimate, id.\\n§ 7545(o)(3)(B)(i), (o)(7)(D)(i)—to radically fluctuate based\\non the frequency of transactions among the distributors that\\nhappen to line the distribution network? So if every distributor\\nstarts selling to another distributor, or several of them, the\\ncalculated volume of fuel “sold or introduced into commerce in\\nthe United States” would balloon overnight? I doubt it.\\n\\n EPA, it seems, shares my skepticism. The agency itself\\ndescribes the renewable fuel obligation, not in terms of\\ndownstream intermediaries, like distributors, but in terms of the\\ninitial, upstream players—those “responsible for introducing\\n[fuel] into the domestic gasoline pool.” 72 Fed. Reg. at\\n23,904/1 (emphasis added). Indeed, when defining the\\n\\n 11\\n\\nrenewable fuel obligation, EPA speaks not of sales that happen\\nto occur, distributor-to-distributor, along the supply chain, but\\nonly of initial injections into U.S. commerce as a consequence\\nof the upstream “produc[tion]” or “import[ation]” of\\ntransportation fuel. 40 C.F.R. § 80.1407(a), (b).\\n\\n What about blenders, asks the majority? Aren’t they\\npotentially obligated parties, even though they, like\\ndistributors, handle fuels that have already been “introduce[d]”\\ninto U.S. commerce by other upstream entities, like refineries?\\nMaj. op. 46–47. Yes, of course, they are. But that’s because\\nblenders—unlike distributors—are the ones who initially sell\\nor introduce various types of finished transportation fuel “into\\ncommerce in the United States.” E15, for instance, a blend of\\n85% gasoline, 15% ethanol, generally enters “into” American\\ncommerce at the hands of a blender—the entity that actually\\nblends the various components. Just ask EPA, which\\nreferences the “ethanol blenders that introduce E15 into\\ncommerce.” 76 Fed. Reg. 44,406, 44,410/3 (July 25, 2011). A\\ndistributor, in contrast—and by definition, whether that’s a\\n“post-enactment regulat[ory]” definition, Maj. op. 47, or a pre-\\nenactment dictionary definition—never introduces anything\\n“into” commerce. It only distributes (i.e., “transports” or\\n“deliver[s]”) finished transportation fuel, such as E15, from one\\npoint to another. See 40 C.F.R. § 80.2(l); Webster’s Third New\\nInternational Dictionary 660 (1961) (defining “distribute”). So\\nsubclause (I), as EPA reads it, is, in fact, a superfluity, because\\nthe agency could not place the point of obligation on\\ndistributors whether that clause existed or not.\\n\\n The muddle generated by EPA’s reading doesn’t end there.\\nConsider the effect on subclause (III). That provision provides\\nthat the “renewable fuel obligation . . . shall . . . consist of a\\nsingle applicable percentage that applies to all categories of\\npersons specified in subclause (I).” 42 U.S.C.\\n§ 7545(o)(3)(B)(ii)(III). But if EPA is right, and the point of\\n\\n 12\\n\\nobligation is determined, not under subclause (I), but under the\\ncompliance provision, why does Congress take such a\\ncircuitous route to get there—a reference in subclause (III) to\\nsubclause (I), which, in turn, in EPA’s reasoning (but without\\nlinguistic underpinning), refers back to the compliance\\nprovision? Couldn’t Congress in subclause (III) have just\\nalluded to decisions made by EPA under the compliance\\nprovision directly? Cf., e.g., 42 U.S.C. § 7545(o)(4)(A). EPA\\ndoesn’t say.\\n\\n Instead, the agency puts essentially all its eggs in the\\ncompliance provision basket. EPA argues, first and foremost,\\nthat its power to promulgate compliance provisions is broad\\nand includes the power to set the point of obligation. And\\n“nothing,” it says, requires it to “reconsider” that\\ndetermination. See, e.g., EPA Br. 67–68. My colleagues offer\\na similar thought, claiming that Congress knew how to call for\\na “redo” if that is what it really wanted. Maj. op. 48. Both\\narguments, however, miss the point. When Congress mandates\\nan annual “determin[ation]” in 42 § 7545(o)(3)(B)(ii), there is\\nnothing to be redone or reviewed. The determination must\\nhappen anew each year, and the specific instruction to apply\\nthat determination “to refineries, blenders, and importers, as\\nappropriate,” controls, id. § 7545(o)(3)(B)(ii)(I); any general\\nauthorization to promulgate compliance provisions (including,\\nI’ll assume, license to not “reconsider” them) must yield to that\\nspecific instruction. See SW General, 137 S. Ct. at 941 (“[I]t is\\na commonplace of statutory construction that the specific\\ngoverns the general.” (alteration in original) (quoting RadLAX\\nGateway Hotel, LLC v. Amalgamated Bank, 566 U.S. 639, 645\\n(2012)).\\n\\n “[B]asic principles of administrative law,” unfortunately\\nfor the majority, only further erode EPA’s position. Maj. op.\\n52. We “generally ‘presume[] that Congress expects it statutes\\nto be read in conformity with the[] [Supreme] Court’s\\n\\n 13\\n\\nprecedents.’” Porter v. Nussle, 534 U.S. 516, 528 (2002)\\n(second alteration in original) (quoting United States v. Wells,\\n519 U.S. 482, 495 (1997)). And those precedents make clear\\nthat an agency, when exercising its congressionally delegated\\nauthority, must “consider [every] important aspect of the\\nproblem.” Motor Vehicle Mfrs. Ass’n of U.S. v. State Farm\\nMutual Ins. Co., 463 U.S. 29, 43 (1983). Failure to do so\\n“would be arbitrary and capricious.” Id. With that background\\nin mind, it “would be strange indeed” if Congress really\\nexpected EPA, year in and year out, to set the renewable fuel\\nstandards for the entire economy, yet allowed the agency—sub\\nsilentio—to do so without considering ever again whether a\\n“foundational” element of the regulatory program was\\n“appropriate.” Maj. op. 49.\\n\\n Retreating from the statutory language, EPA claims that\\nreading the Act to require it to appropriately identify the point\\nof obligation each year would be inconsistent with Congress’s\\n“purpose.” Specifically, the agency says, it would “reduce the\\nregulatory certainty required for private parties to plan for\\ngrowth.” EPA Br. 72. But EPA’s fears are vastly overblown.\\nIts concern about upsetting investment-backed expectations is\\na reason to not change the point of obligation; it is not a reason\\nto not consider doing so. The same goes for my colleagues’\\nconcerns about the credit trading program, see Maj. op. 50,\\neven if that program really does require (as my colleagues seem\\nto assume it does) rock solid stability in the point of\\nobligation—a dubious proposition, given that credits are held\\nindividual-entity-by-individual-entity, so that shrinkage or\\nswelling of the number of covered entities has no impact on the\\nneeded computations. EPA’s duty is to “articulate a\\nsatisfactory explanation for its action,” State Farm, 463 U.S. at\\n43—an explanation that must consider the industry’s\\n(including the credit traders’) reliance on a prior determination,\\nsee, e.g., FCC v. Fox Television Stations, Inc., 556 U.S. 502,\\n515 (2009) (explaining that it “would be arbitrary and\\n\\n 14\\n\\ncapricious to ignore” the fact that a “prior policy has\\nengendered serious reliance interests”); Nat’l Cable &\\nTelecomms. Ass’n v. FCC, 567 F.3d 659, 670 (D.C. Cir. 2009)\\n(similar). In fact, given the substantial reliance interests at\\nstake, along with the agency’s prior findings, it seems likely\\nthat (in the absence of significantly changed circumstances or\\na compelling new analysis) EPA would be able to make rather\\nshort work of the annual analysis. In most years, the prior\\nanalyses and the reliance interests would probably dictate the\\nconclusion.\\n\\n In any event, especially when the alleged downside of\\npetitioners’ claim is so chimerical, our “role is not to ‘correct’\\nthe [statutory] text so that it better serves [Congress’s]\\npurposes.” Va. Dep’t of Medical Assistance Servs. v. U.S.\\nDep’t of Health & Human Servs., 678 F.3d 918, 926 (D.C. Cir.\\n2012) (some internal quotation marks omitted) (quoting Engine\\nManufacturers Ass’n v. EPA, 88 F.3d 1075, 1089 (D.C. Cir.\\n1996)). That is a job for Congress.\\n\\n For these reasons, I respectfully disagree with the panel’s\\nconclusion, which grants EPA essentially unfettered discretion\\nas to when—or even if—it will consider the appropriateness of\\nthe point of obligation.\\n\\n Indeed, the panel, it seems to me, arrived at its conclusion\\nonly by extending to EPA the type of “reflexive” deference that\\nthe Supreme Court has recently criticized. Kisor, 139 S. Ct. at\\n2415 (quoting Pereira v. Sessions, 138 S. Ct. 2105, 2120 (2018)\\n(Kennedy, J., concurring)). The Court has made clear that\\nbefore we may declare a statute genuinely ambiguous—and,\\nthus, before we, an Article III court, may surrender to an\\nexecutive agency’s (often self-serving) declaration of what the\\nlaw means—we must exhaust all the “traditional tools” of\\nstatutory construction. Kisor, 139 S. Ct. at 2415 (quoting\\nChevron U.S.A. Inc. v. Natural Resources Defense Council,\\n\\n 15\\n\\nInc., 467 U.S. 837, 843 n.9 (1984)). Then and only then—\\n“when that legal toolkit is empty”—may we “wave the\\nambiguity flag.” Id.\\n\\n The majority, however, in apparent haste to bow to EPA’s\\nadmittedly self-serving declaration of what the law means, see\\nMaj. op. 51 (describing the “burden[s]” that EPA would rather\\navoid), doesn’t actually use any of the tools of statutory\\nconstruction in an attempt to discern Congress’s meaning. For\\nexample, besides acknowledging that EPA’s reading of the\\nphrase “applicable . . . as appropriate” “is not ineluctable,” Maj.\\nop. 48, the majority has almost nothing to say about that\\nphrase’s ordinary meaning. Although the majority declares it\\n“ambiguous,” id. at 49, my colleagues do not offer a single\\nexample of the phrase being used in the way EPA desires—\\nwhere the duty to make a selection, “as appropriate,”\\n(somehow) permits the decisionmaker wholly to ignore the\\ncontemporaneous context of his selection. But see supra pp. 5,\\n7 (offering examples where EPA’s interpretation makes no\\nsense). The majority’s treatment of the presumption of\\nconsistent usage isn’t much better. It says that there are\\nmultiple “permissible” ways to ascribe the same meaning to the\\nsame words, but doesn’t offer any, see Maj. op. 46—all the\\nwhile overlooking an obvious interpretation that satisfies the\\npresumption (i.e., EPA must consider the factors that are\\nrelevant at the time of its decision), see supra pp. 6–7. Finally,\\nthe majority writes off the canon against surplusage without\\nactually finding that the language at issue isn’t superfluous.\\nThe majority avers that a finding of superfluity “rests on a\\ncomplicated series of inferences,” Maj. op. 47, but that’s not\\nunusual, or reason to shy away from wading through the\\nmuddle. Complex regulatory schemes “can sometimes make\\nthe eyes glaze over. But hard interpretive conundrums, even\\nrelating to complex rules, can often be solved.” Kisor, 139 S.\\nCt. at 2415. To solve such conundrums, however, we must\\nembrace the canons of interpretation as the useful tools that\\n\\n 16\\n\\nthey are for discerning Congress’s meaning, not as pests to be\\ndodged and swatted away in our rush to deference. Here, when\\nthose tools are properly applied, we can discern Congress’s\\nmeaning—which “is the law and must be given effect.”\\nChevron, 467 U.S. at 843 n.9.\\n\\n Nonetheless, I concur in the judgment. As we explain\\ntoday with regard to claims brought by the Alon Petitioners,\\nEPA adequately explained, at around the time it set the annual\\nobligation for 2017, why it was not “appropriate” (in light of\\nthe facts as they then existed) to change the point of obligation.\\nSee Maj. op., Part IV.B. Although that explanation arose in the\\ncontext of a petition for rulemaking—and was thus subject to a\\nmore deferential form of arbitrary and capricious review—I\\nwould hold here (for the same reasons that we give in Part IV.B\\nof the majority opinion) that EPA’s reasoning was sufficient\\neven under the deference level that demands more of the\\nagency.\\n\\n The difference in our standard of review between an appeal\\nfrom the agency’s annual determination under\\n§ 7545(o)(3)(B)(i), (ii)(I), on the one hand, and an agency’s\\nconventional duty to entertain a petition for a rulemaking to\\nrevise an existing regulation, on the other, is in practice fairly\\nslight. Under both understandings, the agency is bound to give\\nsuitable weight to reliance interests, and indeed to the general\\nadvantage of regulators’ not rocking too many boats. A party\\nchallenging the status quo faces some sort of burden in either\\ncontext—to point to new facts, or to new discoveries of facts,\\nor to previously unnoticed flaws in the agency’s analysis, etc.\\nThere is, to be sure, a subtle difference in the deference level,\\nbut deference levels themselves build in a good deal of\\nsubjectivity. I nonetheless write separately because I see\\nCongress as having imposed a specific, if modest, duty, on the\\nagency, and having thereby provided an explicit avenue for\\nreview. That explicitness seems to me designed to, and likely\\n\\n 17\\n\\nto, concentrate the mind of the administrator—a congressional\\nchoice that we should honor.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4435003/\", \"author_raw\": \"WILLIAMS, Senior Circuit Judge, concurring\"}]}"},"judge1":{"kind":"string","value":"PILLARD"},"judge2":{"kind":"string","value":"KATSAS"},"judge3":{"kind":"string","value":"WILLIAMS"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{\"PILLARD\": \", Circuit\", \"KATSAS\": \", Circuit\", \"WILLIAMS\": \", Senior Circuit\"}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":1,"string":"1"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4657750/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Coffeyville Resources Refining & Marketing, LLC v. EPA\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\nDate decided: 2019-08-30\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: Yes\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\n\n\n\nArgued October 5, 2018 Decided August 30, 2019\n\n No. 16-1052\n\n ALON REFINING KROTZ SPRINGS, INC.,\n PETITIONER\n\n v.\n\n ENVIRONMENTAL PROTECTION AGENCY,\n RESPONDENT\n\n MONROE ENERGY, LLC, ET AL.,\n INTERVENORS\n\n\n Consolidated with 16-1055, 17-1255, 17-1259, 18-1021,\n 18-1024, 18-1025, 18-1029\n\n\n On Petitions for Review of Agency Action of the\n United States Environmental Protection Agency\n\n\n Samara L. Kline argued the cause for petitioners. With her\non the briefs were Evan A. Young, Megan H. Berge, Lisa M.\nJaeger, Brittany M. Pemberton, Clara Poffenberger, Richard\nS. Moskowitz, Robert J. Meyers, Thomas A. Lorenzen,\nElizabeth B. Dawson, Warren R. Neufeld, LeAnn M. Johnson,\nand Jonathan G. Hardin. Albert M. Ferlo Jr. and Krista\nHughes entered appearances.\n\n 2\n Meghan E. Greenfield, Trial Attorney, U.S. Department of\nJustice, argued the cause for respondent. With her on the brief\nwas Jeffrey H. Wood, Acting Assistant Attorney General.\nDaniel R. Dertke, Attorney, entered an appearance.\n\n Robert A. Long Jr. argued the cause for intervenors\nAmerican Petroleum Institute and Growth Energy. With him\non the brief were Kevin King, Seth P. Waxman, David M. Lehn,\nSaurabh Sanghvi, and Claire Chung. Stacy R. Linden entered\nan appearance.\n\n Shannen W. Coffin and Linda C. Bailey were on the brief\nfor amici curiae NACS, et al. in support of respondent EPA.\n\n 3\n\n No. 17-1044\n\nCOFFEYVILLE RESOURCES REFINING & MARKETING, LLC AND\n WYNNEWOOD REFINING COMPANY, LLC,\n PETITIONERS\n\n v.\n\n ENVIRONMENTAL PROTECTION AGENCY,\n RESPONDENT\n\n ALON REFINING KROTZ SPRINGS, INC., ET AL.,\n INTERVENORS\n\n\n Consolidated with 17-1045, 17-1047, 17-1049, 17-1051,\n 17-1052\n\n\n On Petitions for Review of Action of the\n United States Environmental Protection Agency\n\n\n Brian Killian argued the cause for petitioner The National\nBiodiesel Board. With him on the briefs was Douglas A.\nHastings.\n\n Samara L. Kline and Thomas Allen Lorenzen, argued the\ncauses for Obligated Party Petitioners. With them on the briefs\nwere Evan A. Young, Lisa M. Jaeger, Brittany M. Pemberton,\nClara Poffenberger, Richard S. Moskowtiz, Robert J. Meyers,\nElizabeth B. Dawson, David W. DeBruin, Thomas J. Perrelli,\nMatthew E. Price, LeAnn M. Johnson, and Jonathan G.\n\n 4\nHardin. David Y. Chung, Eric D. Miller, and Albert M. Ferlo\nJr. entered appearances.\n\n Patrick R. Jacobi and Samara M. Spence, Attorneys, U.S.\nDepartment of Justice, argued the causes for respondent. With\nthem on the brief was Jeffrey H. Wood, Acting Assistant\nAttorney General.\n\n Thomas Allen Lorenzen argued the cause for intervenors\nAmerican Fuel & Petrochemical Manufacturers and American\nPetroleum Institute in support of respondent regarding\nBiomass-Based Diesel Issues. With him on the brief were\nRobert J. Meyers, Elizabeth B. Dawson, Richard S. Moskowitz,\nRobert A. Long, Jr., and Kevin King. Stacy R. Linden entered\nan appearance.\n\n Robert A. Long, Jr., Kevin King, Bryan M. Killian,\nDouglas A. Hastings, Seth P. Waxman, David M. Lehn,\nSaurabh Sanghvi, and Claire H. Chung were on the brief for\nintervenors Growth Energy, et al. in support of respondent.\nEric D. Miller entered an appearance.\n\n Before: PILLARD and KATSAS, Circuit Judges, and\nWILLIAMS, Senior Circuit Judge.\n\n Opinion for the Court filed PER CURIAM.\n\n Opinion concurring in part and concurring in the judgment\nfiled by Senior Circuit Judge WILLIAMS.\n\n TABLE OF CONTENTS\n\nI. Introduction ..................................................................... 6\nII. Background ..................................................................... 7\n\n 5\n A. Legal Background ........................................................ 7\n B. Procedural Background ............................................. 12\n 1. 2007, 2010, and 2017 Point of Obligation\n Proceedings ............................................................ 12\n 2. 2017 Annual Volumetric Proceedings ................... 15\nIII. Standard of Review .................................................... 17\nIV. 2010 Point of Obligation Rule ................................... 17\n A. Jurisdiction................................................................. 17\n 1. Final Agency Action Under Section 7607(b)(1).... 19\n 2. After-Arising Grounds Under Section\n 7607(b)(1) .............................................................. 28\n 3. Mandatory Reconsideration Under Section\n 7607(d)(7)(B) ......................................................... 29\n B. Merits of Challenges to EPA’s Refusal to Revise\n the 2010 Point of Obligation Rule ............................. 32\nV. 2017 Annual Volumetric Rule ...................................... 41\n A. Point of Obligation .................................................... 42\n 1. Jurisdiction ............................................................. 42\n 2. Merits ..................................................................... 43\n B. Cellulosic Biofuel Projection..................................... 53\n C. Cellulosic Waiver ...................................................... 58\nVI. 2018 Volume for Biomass-Based Diesel ................... 62\n A. NBB’s Standing ......................................................... 63\n B. Merits of NBB’s Challenges ...................................... 65\nVII. Conclusion ................................................................. 70\n\n 6\n PER CURIAM:\n\n\n I. Introduction\n\n The Clean Air Act requires EPA to publish “renewable\nfuel standards,” ultimately expressed as “applicable\npercentages,” each year to ensure that the total supply of\ntransportation fuel sold or imported into the United States\ncontains specified proportions of each of four categories of\nrenewable fuels. Congress intended the Renewable Fuel\nStandards (RFS) program to “move the United States toward\ngreater energy independence and security” and “increase the\nproduction of clean renewable fuels.” See Energy\nIndependence and Security Act of 2007 (EISA), Pub. L. No.\n110-140, preamble, 121 Stat. 1492 (2007) (codified at 42\nU.S.C. § 7545(o)).\n\n In these related cases, Alon Refining Krotz Springs,\ntogether with other petroleum refineries and their trade\nassociations—the “Alon Petitioners”—seek review of EPA’s\ndecision not to revise its 2010 point of obligation regulation\nrequiring refineries and importers, but not blenders, to bear the\ndirect compliance obligation of ensuring that transportation\nfuels sold or introduced into the U.S. market include the\nrequisite percentages of renewables. Coffeyville Resources\nRefining & Marketing and another group of refineries and trade\nassociations—the “Coffeyville Petitioners”—challenge EPA’s\nrefusal to reassess the appropriateness of the point of obligation\nin the context of its 2017 annual volumetric rule, which set the\n2017 applicable percentages for all four categories of\nrenewable fuel and the 2018 applicable volume for one subset\nof such fuel, biomass-based diesel. See 81 Fed. Reg. 89,746\n(Dec. 12, 2016) (2017 Rule). The Coffeyville Petitioners also\ncontend that EPA arbitrarily set the 2017 percentage standards\n\n 7\ntoo high. The National Biodiesel Board (NBB)—a biomass-\nbased diesel industry trade association—separately contends\nthat EPA set the 2018 applicable volume for biomass-based\ndiesel too low. Various trade associations representing\nrefineries and producers of renewable fuels have intervened in\nsupport of EPA. For the reasons that follow, we deny each of\nthe petitions for review, many of which recycle arguments\nraised and rejected in prior challenges.\n\n II. Background\n\nA. Legal Background\n\n Congress established the RFS program in 2005 as part of\nthe Energy Policy Act, Pub. L. No. 109-58, 119 Stat. 594\n(2005) (as amended at 42 U.S.C. § 7545(o)). The statute\nmandates the gradual introduction of four nested categories of\nrenewable fuels into the United States’ supply of gasoline,\ndiesel, and other transportation fuels. See 42 U.S.C.\n§ 7545(o)(2)(B). These categories include: (1) total renewable\nfuel; (2) advanced biofuel; (3) cellulosic biofuel; and (4)\nbiomass-based diesel. Id. § 7545(o)(2)(A)(i), (B). The\numbrella category, total renewable fuel, covers the three other\ncategories plus any conventional renewable fuels, such as corn-\nbased ethanol. See id. § 7545(o)(1)(F), (J), (2)(A)(i). The\nadvanced biofuel subset includes any renewable fuel (except\nethanol from cornstarch) that has at least 50% lower lifecycle\ngreenhouse gas emissions than fossil fuels. Id.\n§ 7545(o)(1)(B). The statute further specifies two\nnonexclusive subsets of advanced biofuels: cellulosic biofuel\n(a renewable fuel derived from cellulose materials such as corn\nstalks and husks) and biomass-based diesel (a diesel fuel\nsubstitute made from feedstocks such as animal fats). Id.\n§ 7545(o)(1)(B), (D), (E); EPA Coffeyville Br. 4–5. The\n\n 8\nfollowing figure depicts the nested nature of the four fuel\ncategories.\n\n\n\n\n Source: Coffeyville Br. 11.\n Four tables in the statute set forth gradually increasing\nannual “applicable volume” requirements for each category of\nrenewable fuel. See 42 U.S.C. § 7545(o)(2)(B)(i). The statute\nsets applicable volumes for biomass-based diesel through\n2012, id. § 7545(o)(2)(B)(i)(IV), and applicable volumes for\nthe other three categories through 2022, id.\n§ 7545(o)(2)(B)(i)(I)–(III). Under those tables, as the total\nquantities of renewable fuel rise over time, the ratio of\nadvanced biofuels relative to conventional renewable fuel\ngradually increases. Id. For compliance years (which match\ncalendar years) after those specified in the tables, the statute\nrequires EPA, in coordination with the Secretaries of Energy\nand Agriculture, to set the annual applicable volumes based on\n\n 9\na review of the implementation of the program plus an analysis\nof six listed factors. Id. § 7545(o)(2)(B)(ii). For years not\nspecified in the table, EPA must publish the applicable volumes\nfourteen months before the year in which they will apply—\nvolumes that, shortly before the start of the compliance year,\nEPA translates into percentage standards. Id.\n Various “waiver” provisions require or permit EPA to\nlower the annual applicable volumes. Two are relevant for the\npurposes of this case. First, under the “cellulosic waiver\nprovision,” EPA must make its own projection of the volume\nof cellulosic biofuel that will be produced in the following year.\nId. § 7545(o)(7)(D)(i). If that projection is less than the\nstatutory figure, the agency must use its own projection as the\napplicable volume of cellulosic biofuel. Id.; see Am. Petroleum\nInst. v. EPA, 706 F.3d 474, 477–80 (D.C. Cir. 2013) (API). The\nsame cellulosic waiver provision authorizes (but does not\nrequire) EPA to also reduce the advanced biofuel and total\nrenewable biofuel volume requirements “by the same or a\nlesser volume” as the cellulosic biofuel reduction, 42 U.S.C.\n§ 7545(o)(7)(D)(i), and EPA has “broad discretion” regarding\nwhether and how to do that, Monroe Energy, LLC v. EPA, 750\nF.3d 909, 915 (D.C. Cir. 2014). Separately, under the “general\nwaiver provision,” EPA may reduce any of the statutory\napplicable volumes if it determines “that implementation . . .\nwould severely harm the economy or environment,” or “that\nthere is an inadequate domestic supply.” 42 U.S.C.\n§ 7545(o)(7)(A); see Ams. for Clean Energy v. EPA, 864 F.3d\n691, 707–13 (D.C. Cir. 2017) (ACE).\n After EPA determines the waiver-adjusted applicable\nvolumes, it must translate those volumes into “renewable\nvolume obligation[s]” for each category of renewable fuel for\nthe upcoming compliance year. 42 U.S.C. § 7545(o)(3)(B)(i).\nThe volume obligation for each category of renewable fuel is\nexpressed as an “applicable percentage,” also known as a\n\n 10\n“percentage standard,” calculated by dividing the adjusted\napplicable volume for that category of fuel by the total\nanticipated volume of non-renewable transportation fuel that\nwill be introduced into commerce (which EPA derives based\non an estimate provided by the Energy Information\nAdministration) in the coming compliance year. Id.\n§ 7545(o)(3)(A), (B)(ii)(II); 40 C.F.R. § 80.1405(c). The\nstatute calls on EPA to publish the percentage standards not\nlater than November 30—a month before the start of the\ncompliance year. 42 U.S.C. § 7545(o)(3)(B)(i).\n EPA must place the renewable volume obligations on\n“refineries, blenders, and importers, as appropriate.” 42 U.S.C.\n§ 7545(o)(3)(B)(ii)(I); see also id. § 7545(o)(2)(A) (requiring\nEPA to promulgate implementing regulations, including\n“compliance provisions applicable to refineries, blenders,\ndistributors, and importers, as appropriate,” designed to ensure\nthat transportation fuel sold or introduced into the United States\n“contains at least” the required annual applicable volumes).\nThe entities that EPA designates to meet the volume\nobligations are known as “obligated parties.” Monroe Energy,\n750 F.3d at 912. Each obligated party must ensure that the\nvolume of non-renewable fuel it sells or introduces into U.S.\ncommerce is matched by selling or introducing a corresponding\nvolume of each category of renewable fuel at the level EPA’s\npercentage standard requires for that category. See ACE, 864\nF.3d at 699. The percentage standards are set in the\nanticipation that, if each obligated party meets them and EPA’s\nprojection regarding the country’s total transportation fuel\nsupply bears out, the amount of each category of renewable\nfuel introduced into the economy in the upcoming compliance\nyear will equal the applicable volumes for that year. Id.\nObligated parties bear no direct responsibility for any shortfalls\nin the applicable volumes so long as they comply with the\npercentage standards.\n\n 11\n EPA assigns a set of “renewable identification numbers”\n(RINs) to each batch of renewable fuel that is produced or\nimported for use in the United States. 40 C.F.R. § 80.1426; see\n42 U.S.C. § 7545(o)(5); Monroe Energy, 750 F.3d at 913. The\nnumber of RINs assigned to each batch corresponds to the\namount of ethanol-equivalent energy per gallon in that batch.\nSee 40 C.F.R. § 80.1415; Monroe Energy, 750 F.3d at 913.\nRINs remain attached to the renewable fuel until that fuel is\npurchased by an obligated party or blended into fossil fuels to\nbe used for transportation fuel. See ACE, 864 F.3d at 699\n(citing 40 C.F.R. § 80.1429(b)(1)–(2)). At that point the RINs\nbecome “separated,” meaning they are, in effect, a form of\ncompliance credit. Id. Obligated parties demonstrate their\ncompliance with their renewable fuel obligations by “retiring”\nRINs in annual compliance demonstrations to EPA. 40 C.F.R.\n§§ 80.1427(a), 80.1451(a)(1).\n Because the four categories of renewable fuel are nested,\nobligated parties can comply with their obligations for a type\nof fuel by retiring any combination of RINs corresponding to\nthat category of fuels or any subset thereof. See 40 C.F.R.\n§ 80.1427(a)(3)(i). For instance, retiring a cellulosic biofuel or\nbiomass-based diesel RIN counts not only toward the volume\nobligation for that fuel, but also toward both the advanced\nbiofuel and total renewable fuel obligations. Thus, “if one\nmillion gallons of cellulosic biofuel are blended into the fuel\nsupply, the statute allows those one million gallons to be\ncredited toward the advanced biofuel and total renewable fuel\nobligations in addition to the cellulosic biofuel obligation.”\nACE, 864 F.3d at 698.\n Obligated parties who have more RINs than they need may\nsell or trade their excess, 40 C.F.R. § 80.1428(b), or they may\n“bank” those RINs for use to meet up to 20 percent of their\nobligations for the following compliance year, Monroe Energy,\n750 F.3d at 913; see 40 C.F.R. § 80.1427(a)(1), (5); Regulation\n\n 12\nof Fuels and Fuel Additives: Changes to Renewable Fuel\nStandard Program, 75 Fed. Reg. 14,670, 14,734–35 (Mar. 26,\n2010). Obligated parties without enough RINs to meet their\ncompliance obligations may purchase RINs, use banked RINs\nfrom the prior year, or carry a deficit forward to the following\nyear to be satisfied together with the following year’s\nobligations. See ACE, 864 F.3d at 699–700; see also 42 U.S.C.\n§ 7545(o)(5)(D); 40 C.F.R. § 80.1427(b).\n\nB. Procedural Background\n\n The procedural history of these cases follows two paths:\nfirst, the proceedings relevant to the challenge that EPA\narbitrarily declined to initiate a rulemaking to modify the 2010\nregulation designating refineries and importers, but not\nblenders, as obligated parties; and second, the proceedings\nchallenging the 2017 Rule.\n\n 1. 2007, 2010, and 2017 Point of Obligation\n Proceedings\n\n In its 2007 regulations implementing the RFS program,\nEPA designated refiners and importers, but not blenders, as the\n“appropriate” parties to meet the renewable fuel obligation. 72\nFed. Reg. 23,900, 23,923–24 (May 1, 2007). At the time, those\ndesignations were not challenged in court. EPA reaffirmed its\ndesignations in a 2010 regulation now commonly known as the\n“point of obligation rule.” 75 Fed. Reg. at 14,721–22 (codified\nat 40 C.F.R. § 80.1406(a)(1)). During the 2010 rulemaking,\nseveral refiners—including petitioner Valero Energy\nCorporation—argued that failing to obligate blenders, who\ncombine renewable fuel with fossil fuels, would make the RFS\nprogram unworkable. EPA concluded that the program was\nfunctioning adequately and that the burdens and disruption\nfrom changing the point of obligation would outweigh any\n\n 13\nbenefits. See Summary and Analysis of Comments 3.9.2, Alon\nJ.A. 287–90. Although other aspects of the 2010 regulations\nwere challenged in court, see, e.g., Nat’l Chicken Council v.\nEPA, 687 F.3d 393 (D.C. Cir. 2012); Nat’l Petrochemical &\nRefiners Ass’n v. EPA, 630 F.3d 145 (D.C. Cir. 2010), the point\nof obligation rule was not.\n\n On December 14, 2015, EPA promulgated the volume\nrequirements for 2014, 2015, and 2016. Renewable Fuel\nStandard Program, 80 Fed. Reg. 77,420 (Dec. 14, 2015). In so\ndoing, EPA exercised its general waiver authority to lower the\ntotal renewable fuel volumes based on a finding of inadequate\ndomestic supply due to market factors “affecting the ability to\ndistribute, blend, dispense, and consume . . . renewable fuels”\nat the levels required by statute. Id. at 77,435/2. Among those\nfactors was “the slower than expected development of the\ncellulosic biofuel industry.” Id. at 77,422. The agency thought\nan additional “real world constraint[]” was the “E10\nblendwall”—the difficulty for most American vehicle engines\nto run on blends containing more than 10% ethanol. Id. at\n77,423. EPA explained that those factors made the statutory\nrequirements “impossible to achieve.” Id. at 77,422/2. This\nCourt later vacated the general waiver on the ground that EPA\nhad misinterpreted the statutory term “inadequate domestic\nsupply” to include demand-side constraints such as the E10\nblendwall. See ACE, 864 F.3d at 704–13.\n\n On February 12, 2016, sixty days after EPA promulgated\nthe volume requirements for 2014–16, the Alon Petitioners\npetitioned this Court for review of the 2010 point of obligation\nrule. These petitions contend that the rule was arbitrary and\ncapricious insofar as it failed to impose the obligation on\ndownstream blenders—the parties petitioners think are best\nable to comply with it. The petitions assert jurisdiction under\nthe after-arising provision in 42 U.S.C. § 7607(b)(1), which\n\n 14\npermits otherwise-untimely challenges to a rule if the\nchallenges are “based solely on grounds arising after” the sixty-\nday deadline for seeking judicial review. The petitioners assert\nthat EPA’s exercise of its general waiver authority in the 2014–\n16 volume regulations, and its acknowledgment of the RFS\nprogram’s shortcomings as of that time, provided such an after-\narising ground.\n\n The Alon Petitioners simultaneously petitioned EPA to\nrevise the point of obligation rule. Some of their requests were\nstyled as petitions for a rulemaking. Others were styled as\npetitions for mandatory reconsideration under 42 U.S.C.\n§ 7607(d)(7)(B), which requires EPA to reconsider a rule if\ncentrally important objections were impracticable to raise\nduring the comment period or “arose after” that period “but\nwithin the time specified for judicial review.” The petitions\ncited the waiver in the 2014–16 volume regulations and EPA’s\nacknowledgment of program difficulties as grounds supporting\nmandatory reconsideration. This Court held in abeyance the\npetitions for review of the point of obligation rule pending\nresolution of the petitions to revise it.\n\n On November 10, 2016, EPA published a proposed denial\nof the petitions to revise the point of obligation rule. On\nNovember 22, 2017, after reviewing more than 18,000\ncomments on the proposal, EPA denied the petitions. It\nconcluded that the statutory requirements for mandatory\nreconsideration were not met, so it treated all the filings as\npetitions for a rulemaking. Denial of Petitions for Rulemaking\nto Change the RFS Point of Obligation, EPA-HQ-OAR-2016-\n0544-0525, at 7 (Nov. 22, 2017) (EPA Denial), Alon J.A. 61.\nEPA then denied the petitions on the ground that “changing the\npoint of obligation would . . . likely result in a decrease in the\nproduction, distribution, and use of [renewable] fuels” and\nwould “do nothing to incentivize the research, development,\n\n 15\nand commercialization of cellulosic biofuel technologies\ncritical for the growth of the RFS program in future years.”\nEPA Denial at 8–9, Alon J.A. 62–63.\n\n Within sixty days (in December 2017 and January 2018),\nthe Alon Petitioners sought judicial review of that denial,\nwhich it cast as a final agency action under section 7607(b)(1).\nThe two sets of petitions—the February 2016 petitions for\nreview of the 2010 point of obligation rule and the 2017–18\npetitions for review of EPA’s refusal to reconsider the rule—\nwere consolidated and are now before us.\n\n 2. 2017 Annual Volumetric Proceedings\n\n EPA issued its 2017 annual volumetric rule on December\n12, 2016. The 2017 Rule establishes: (1) the applicable volume\nfor biomass-based diesel for 2018, 81 Fed. Reg. at 89,751/2;\n(2) the waiver-adjusted applicable volumes for cellulosic\nbiofuel, advanced biofuel, and total renewable fuel for 2017,\nid. at 89,747 tbl. I-1; and (3) percentage standards for all four\nfuel types for 2017, id. at 89,751, tbl. I.B.6-1.\n EPA exercised its mandatory cellulosic waiver authority to\ndecrease the 2017 applicable volume for cellulosic biofuel by\nmore than 94 percent, dropping 5.189 billion gallons from the\nstatutory target of 5.5 billion gallons, to 311 million gallons.\nId. at 89,750/2; 42 U.S.C. § 7545(o)(2)(B)(i)(III). EPA then\nhad discretion under that same waiver authority to cut as much\nas 5.189 billion gallons off the statutory volumes for advanced\nbiofuel and total renewable fuel. See 42 U.S.C.\n§ 7545(o)(7)(D)(i); 81 Fed. Reg. at 89,762 & tbl. IV.A-1. EPA\npartially exercised that authority, reducing the 9-billion-gallon\nstatutory target for advanced biofuel by 4.72 billion gallons,\nresulting in an adjusted applicable volume of 4.28 billion\ngallons—a greater than 50% decrease. 81 Fed. Reg. at 89,750–\n51; 42 U.S.C. § 7545(o)(2)(B)(i)(II). EPA reduced the total\n\n 16\nrenewable fuel volume requirements by the same amount,\nlowering the statutory target of 24 billion gallons to 19.28\nbillion gallons. 81 Fed. Reg. at 89,751/1; 42 U.S.C.\n§ 7545(o)(2)(B)(i)(I). EPA considered but decided against also\nusing its general waiver authority to further lower the\napplicable volume of total renewable fuel. 81 Fed. Reg. at\n89,751/1.\n Using the waiver-adjusted applicable volumes, EPA set\nthe 2017 percentage standards for each of the four renewable\nfuel categories. See id. at 89,751, 89,799–801. Finally, EPA\nset the biomass-based diesel applicable volume for 2018 at 2.1\nbillion gallons. Id. at 89,751/2. EPA received comments\nurging it to reassess the point of obligation in the 2017 Rule,\nbut declined to address them on the grounds that the comments\nwere “beyond the scope” of the 2017 rulemaking. Response to\nComments at 542, Coffeyville J.A. 761.\n After EPA published the 2017 Rule, various parties\npetitioned for judicial review. The Coffeyville Petitioners\ncontend that EPA erred by refusing to reconsider which types\nof parties would bear the direct compliance obligation under\nthe 2017 Rule. They also argue that EPA arbitrarily calculated\nthe 2017 production of cellulosic biofuel and arbitrarily\nexercised its discretionary cellulosic waiver authority, resulting\nin percentage standards that are too high. NBB argues that\nEPA set the 2018 applicable volume for biomass-based diesel\ntoo low by considering factors it should not have and omitting\nor incorrectly assessing others. Two trade associations\nrepresenting refineries have intervened in defense of EPA’s\nbiomass-based diesel decision, and a coalition of trade\nassociations representing renewable fuel producers and\nrefineries have intervened to oppose the Coffeyville\n\n 17\nPetitioners’ claims. None of the petitioners’ challenges\nsucceeds.\n\n III. Standard of Review\n\n “This court applies the familiar, deferential standard\nannounced in Chevron, U.S.A., Inc. v. Natural Resources\nDefense Council, Inc., to sustain any reasonable agency\ninterpretation of ambiguity in the Clean Air Act.” Nat’l Ass’n\nfor Surface Finishing v. EPA, 795 F.3d 1, 7 (D.C. Cir. 2015).\n“We employ the deferential State Farm standard of review\nwhen reviewing arguments based on allegedly arbitrary or\nunreasoned agency action.” ACE, 864 F.3d at 726 (citing\nMotor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut.\nAuto. Ins. Co., 463 U.S. 29, 43 (1983)). Under that rubric,\nEPA’s actions are “presumptively valid provided [they] meet[]\na minimum rationality standard.” Nat. Res. Def. Council, Inc.\nv. EPA, 194 F.3d 130, 136 (D.C. Cir. 1999). We uphold EPA’s\nactions so long as they are “reasonable and reasonably\nexplained.” Jackson v. Mabus, 808 F.3d 933, 936 (D.C. Cir.\n2015).\n IV. 2010 Point of Obligation Rule\n\n We start with the Alon Petitioners and their challenges to\nthe 2010 point of obligation rule.\n\nA. Jurisdiction\n\n We begin, as we must, with our jurisdiction. In general\nterms, the question presented involves our review of rules\npromulgated under the Clean Air Act, or EPA’s failure to\namend them, after the initial window for seeking judicial\n\n 18\nreview has passed. Various statutory provisions frame this\ninquiry.\n\n Section 7607(b)(1) of Title 42 provides for judicial review\nof regulations promulgated by the Administrator of EPA under\nthe Clean Air Act. The first sentence of section 7607(b)(1)\nvests this Court with exclusive jurisdiction to review\n“nationally applicable regulations promulgated, or final action\ntaken, by the Administrator” under the Act. The fourth\nsentence of section 7607(b)(1) specifies the time for seeking\njudicial review. It imposes a sixty-day time limit, but provides\nan exception for petitions based on grounds arising after the\nlimit: “Any petition for review under this subsection shall be\nfiled within sixty days from the date notice of such\npromulgation, approval, or action appears in the Federal\nRegister, except that if such petition is based solely on grounds\narising after such sixtieth day, then any petition for review\nunder this subsection shall be filed within sixty days after such\ngrounds arise.”\n\n Section 7607(d) of Title 42 sets forth provisions for\nvarious rulemakings under the Clean Air Act, including for the\n“promulgation or revision of any regulation” involving the RFS\nprogram. Id. § 7607(d)(1)(E). Section 7607(d)(7)(B)\naddresses various issues regarding exhaustion, agency\nreconsideration, and judicial review. The first sentence of that\nprovision imposes a conventional exhaustion requirement,\nlimiting judicial review to objections “raised with reasonable\nspecificity during the period for public comment.” The second\nsentence requires EPA to reconsider regulations in certain\nnarrow circumstances: “If the person raising an objection can\ndemonstrate to the Administrator that it was impracticable to\nraise such objection within such time or if the grounds for such\nobjection arose after the period for public comment (but within\nthe time specified for judicial review) and if such objection is\n\n 19\nof central relevance to the outcome of the rule, the\nAdministrator shall convene a proceeding for reconsideration\nof the rule and provide the same procedural rights as would\nhave been afforded had the information been available at the\ntime the rule was promulgated.” The third sentence of section\n7607(d)(7)(B) makes any “refusal” to provide such mandatory\nreconsideration judicially reviewable.\n\n At various times in this litigation, the petitioners have\nasserted three different jurisdictional theories. First, EPA’s\nrefusal to revise the point of obligation rule in 2017 was a final\naction reviewable under the first sentence of section\n7607(b)(1), regardless of the after-arising provision. Second,\nEPA’s statements and actions in its 2014–16 volume regulation\nconstitute after-arising grounds permitting a challenge to the\npoint of obligation rule as promulgated in 2010. Third, these\nsame EPA statements and actions triggered a right to\nmandatory reconsideration under section 7607(d)(7)(B), which\nin turn makes the denial of reconsideration judicially\nreviewable. As explained below, we conclude that the first\ncontention is correct, the second has been abandoned, and the\nthird lacks merit.\n\n 1. Final Agency Action Under Section 7607(b)(1)\n\n In 2016, various refiners petitioned EPA for a rulemaking\nto modify the point of obligation rule. The petitions urged that\nthe need for a modification became evident in 2015, when EPA\nwaived certain statutory volume requirements and concluded\nthat changing economic conditions had made the requirements\n“impossible to achieve.” 2014–16 Rule, 80 Fed. Reg. at\n77,422/2. In November 2017, EPA denied the rulemaking\npetitions on the ground that any current problems with the RFS\nprogram were manageable and that changing the point of\nobligation at this juncture would be disruptive. EPA Denial at\n\n 20\n8–9, Alon J.A. 62–63. The refiners sought review of the denial\nin December 2017 and January 2018. As petitioners in this\nCourt, they contend that the November 2017 denial constituted\nfinal agency action reviewable under section 7607(b)(1). We\nagree.\n\n As noted above, the first sentence of section 7607(b)(1)\ngives this Court exclusive jurisdiction to review any nationally\napplicable “final action” taken by EPA under the Clean Air\nAct. The parties agree that the denial of the rulemaking\npetitions was nationally applicable, final, and taken under the\nClean Air Act. It was also agency “action” within the meaning\nof the statute. That word “bears the same meaning in [section\n7607(b)(1)] that it does under the Administrative Procedure\nAct,” Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 478\n(2001), which defines “agency” to include EPA, 5 U.S.C.\n§ 551(1), and “agency action” to include “the whole or a part\nof an agency rule, order, license, sanction, relief, or the\nequivalent or denial thereof, or failure to act,” id. § 551(13)\n(emphases added). So, EPA’s denial of the petitions for\nrulemaking was a reviewable “action.”\n\n The petitions for review were timely. As a general matter,\nsection 7607(b)(1) requires a petition for review to be filed\nwithin sixty days of when “notice of such promulgation,\napproval, or action appears in the Federal Register.” Here, the\n“action” at issue—denial of the petitions for rulemaking—was\npublished in the Federal Register on November 22, 2017, and\nthe petitions for review of that action were filed within sixty\ndays of that date. Moreover, this conclusion does not depend\non the after-arising provision. To the contrary, because the\npetitions for review were filed within sixty days of the “action”\nunder review, the exception for “grounds arising after such\nsixtieth day” was not triggered.\n\n 21\n Our caselaw confirms this framing of the jurisdictional\nissue. In Massachusetts v. EPA, 415 F.3d 50 (D.C. Cir. 2005),\nthis Court held that EPA’s denial of a petition to regulate\ngreenhouse gas emissions as air pollutants was itself “final\naction” reviewable under section 7607(b)(1). See id. at 53–54\n(opinion of Randolph, J.); id. at 61 (Sentelle, J., concurring in\nthe judgment). The Supreme Court reversed our judgment on\nthe merits, but agreed that we had jurisdiction. Massachusetts\nv. EPA, 549 U.S. 497 (2007). In particular, the Court noted that\nsection 7607(b)(1) “expressly permits review” of EPA’s\n“rejection of [a] rulemaking petition.” Id. at 520, 528; see also\nid. at 517 (section 7607(b)(1) affords “the right to challenge\n[this] agency action unlawfully withheld”). Likewise, in\nNatural Resources Defense Council v. EPA, 824 F.2d 1146\n(D.C. Cir. 1987) (en banc) (NRDC), we held that a 1985\ndecision to withdraw proposed amendments to certain 1976\nregulations—which we described as a “decision not to amend”\nthe regulations—was “reviewable agency action” under\nsection 7607(b)(1). Id. at 1150 (quotation marks omitted). We\nconcluded that the petition for review at issue, which on its face\nchallenged the 1985 withdrawal decision, was not, in\nsubstance, an untimely “‘back-door’ challenge to the 1976\nregulations.” Id. On that point, we reasoned that the petitioners\nclaimed legal errors in the 1985 withdrawal and sought vacatur\nonly of that order. Likewise, in this case, the petitions for\nreview filed in 2017 and 2018 raise no back-door challenge to\nthe 2010 regulation: the petitions contend that EPA in 2017\narbitrarily refused to take account of changing economic\nconditions, and they seek vacatur only of the 2017 order\ndenying a new rulemaking going forward.\n\n Background principles of administrative law reinforce our\nconclusion that the denial of a petition to modify a rule based\non changed circumstances is itself a reviewable order.\nOrdinarily, the denial of a petition to amend or rescind a\n\n 22\nregulation is judicially reviewable. See, e.g., NLRB Union v.\nFLRA, 834 F.2d 191, 195–96 (D.C. Cir. 1987). When the\npetition to amend attacks defects in the regulation as originally\npromulgated, and when the time limit for seeking review of the\nregulation has passed, questions can arise about whether the\ntime limit is being improperly circumvented. In these\ncircumstances, we have held that the petitioner cannot raise\nprocedural challenges to the regulation, but can raise\nsubstantive arguments that the regulation is unauthorized by or\nconflicts with a statute. See id. at 196–97. But the\ncircumvention concern does not even arise when the petitioner\nraises arguments about changed circumstances or new\ninformation. Those kinds of arguments—that recent\ndevelopments compel the amendment of an older regulation—\nare always cognizable through review of the denial of a petition\nto amend, though they trigger an “extremely limited” review\non the merits. Id. at 196. Our decision today harmonizes the\njudicial-review provisions of the Clean Air Act with this\ngeneral background principle.\n\n EPA recognizes the general rule that, under the NLRB\nUnion line of cases, the denial of a petition to amend a rule is a\nreviewable order, which supports both challenges based on\nrecent developments and substantive challenges to the original\nregulation. Nonetheless, EPA urges a different rule where the\napplicable judicial-review provision contains a time limit for\nseeking review and an exception for grounds arising after the\ntime limit, as in the Clean Air Act. In those circumstances,\naccording to EPA, a challenge to the denial of a petition to\namend is untimely—and the denial is thus entirely\nunreviewable—unless the after-arising provision is satisfied.\nEPA rests this conclusion on National Mining Ass’n v.\nDepartment of Interior, 70 F.3d 1345 (D.C. Cir. 1995), and\nAmerican Road & Transportation Builders Ass’n v. EPA, 588\n\n 23\nF.3d 1109 (D.C. Cir. 2009) (ARTBA), but neither decision\nsupports its position.\n\n National Mining involved judicial review under the\nSurface Mining Control and Reclamation Act (SMCRA),\nwhich requires petitions for review to be filed “within sixty\ndays” of the agency action at issue “or after such date if the\npetition is based solely on grounds arising after the sixtieth\nday.” See 70 F.3d at 1350. In 1986, parties petitioned the\nDepartment of Interior to rescind a 1979 SMCRA regulation\non two grounds. First, the petitioners argued that the rule was\ninconsistent with the statute—an argument attacking the\nregulation itself and “available” when the regulation was\noriginally promulgated. See id. After the agency declined to\nrescind the rule, the challengers sought judicial review. We\nheld that the after-arising provision made this first challenge\nuntimely, by explicitly requiring challenges to a regulation\neither to be filed “within the statutory period” or to “meet the\nafter-arising test.” Id. at 1350–51. At the same time, however,\nwe held that the petitioners’ second challenge—to the agency’s\n1986 decision refusing to “repeal” the regulation based on\nchanged circumstances—was timely and reviewable. Id. at\n1352. To be sure, we described the latter challenge as resting\non “grounds that arose after the sixtieth day.” Id. But we failed\nto explain what jurisdictional theory that observation\nsupported: a challenge to the 1979 regulation rendered timely\nby the after-arising exception; or a challenge, in substance as\nwell as form, to the 1986 refusal to repeal, akin to the challenge\nthat we held reviewable in NRDC. Instead, we simply\nconcluded that, under the “limited scope of our review,” the\nagency did not “act[] unreasonably in denying the petition for\nrulemaking.” Id. at 1352–53. Thus, while National Mining\nblessed jurisdiction to review agency refusals to amend\nregulations based on changed factual circumstances, it did not\n\n 24\nultimately address what we clarify today—the precise statutory\nbasis for that jurisdiction.\n\n ARTBA applied the reasoning of National Mining to the\nClean Air Act, which also contains a time limit for judicial\nreview and an exception for after-arising grounds. ARTBA\ninvolved a 2002 petition to amend 1997 regulations on the\nground that they allowed states “to adopt precisely the kinds of\nregulations that the statute forbids.” 588 F.3d at 1110. As in\nNational Mining, the challenge was thus a substantive attack\non a regulation as originally promulgated. We held that, under\nNational Mining, EPA’s “denial of a revision-seeking petition\ndoes not allow review of alleged substantive defects in the\noriginal rule even under the deferential standards applicable to\nreview of such denials, outside the statutory limitations period\nrunning from the rule[’s] original promulgation.” Id. at 1113\n(emphasis added). In other words, National Mining “require[d]\nus to treat ARTBA’s petition to EPA as a challenge to the\nregulations it sought revised.” Id. at 1110. As a result,\ndismissal was necessary unless the petition satisfied the\nexception for after-arising grounds, which it did not. See id.\n\n Neither decision controls here, where the 2017 and 2018\npetitions for review challenge not the point of obligation\nregulation as originally promulgated in 2010, but the failure to\namend the regulation in light of changed circumstances flagged\nby EPA in 2015. EPA rejected that argument in 2017, and the\npetitioners sought review of the rejection within sixty days. In\nsubstance as well as form, the challenge was to the 2017 refusal\nto amend, not to the underlying 2010 regulation. Under these\ncircumstances, there was no risk of circumventing the original\ntime limit. Therefore, there was also no reason to treat the 2010\npromulgation and the 2017 refusal to amend as one-and-the-\nsame agency action, despite binding APA definitions treating\nthem as separate.\n\n 25\n Precedent aside, EPA’s proposed approach—making the\nafter-arising provision the exclusive vehicle for challenging\nrefusals to amend regulations based on new information or\nchanged circumstances—creates various difficulties. For one\nthing, there is a conceptual mismatch between that provision\nand these kinds of challenges. Though the after-arising\nexception and the opportunity to seek rule revision based on\npost-rulemaking events may seem similar, the first allows an\nintervening event to secure judicial review on the basis of\ndefects extant at the time of the rulemaking, whereas the\nsecond allows review on the question whether intervening\nevents have fatally undermined the original justification for the\nrule. The arrow of time runs forward, not backward, so it is at\nbest awkward—and at worst incoherent—to speak of a later\ndevelopment rendering unlawful an earlier promulgation.\nEconomic developments in 2015 may have made it arbitrary\nfor EPA to adhere to the point of obligation rule in 2017, but\nthey cannot have retroactively made arbitrary its promulgation\nin 2010.\n\n Even worse, the Clean Air Act’s after-arising provision, if\nused to judge the timeliness of challenges based on new\ninformation, would be difficult to apply, capriciously narrow,\nor both. To satisfy that provision, a petition for review must be\nboth (i) based “solely” on after-arising grounds and (ii) filed\n“within sixty days after such grounds arise.” 42 U.S.C.\n§ 7607(b)(1). Yet the case for changing an environmental\nregulation will almost never manifest itself at one discrete\nmoment. Instead, it will accumulate progressively over time,\nas scientific knowledge advances or economic conditions\nchange. And so, under EPA’s approach, the relevant filing\ndeadlines would become practically unknowable. When did\nsome environmental risk become serious and obvious enough\nto compel a rulemaking to strengthen an existing regulation?\nThat will usually be a hard question, and it would be little short\n\n 26\nof miraculous if the answer turned out to be on a date certain\nwithin sixty days of the filing of a petition for review, as\nrequired to satisfy the after-arising provision.\n\n In contrast, the approach we have sketched out produces\nsimple questions and discernible deadlines: ask when EPA\ndenied the rulemaking petition, then add sixty days. If\npossible, we should avoid trying to fix arbitrarily precise\naccrual dates for events that develop incrementally over time.\nSee Nat’l R.R. Passenger Corp. v. Morgan, 536 U.S. 101, 115–\n21 (2002). And we should avoid jurisdictional tests that are\ncomplex as opposed to straightforward. See, e.g., Hertz Corp.\nv. Friend, 559 U.S. 77, 94 (2010). Treating the denial of a\npetition to amend a rule based on changed circumstances as\nreviewable agency action honors both principles, while\nattempting to shoehorn such denials into the after-arising\nprovision does the opposite.\n\n We acknowledge that, in Oljato Chapter of Navajo Tribe\nv. Train, 515 F.2d 654 (D.C. Cir. 1975), we assumed that the\nafter-arising provision would govern review of orders denying\npetitions to modify EPA rules based on changed circumstances.\nBut our only holdings were that such petitions must first be\npresented to EPA, id. at 666, and then are reviewable directly\nin the courts of appeals, id. at 657–65. Moreover, our\nassumption was understandable in context; the petitioners had\nmissed the statutory deadline to file a petition for review (then\nthirty days), see id. at 657, so they pressed alternative\njurisdictional theories (involving either district-court review or\nthe after-arising provision) that would avoid that deadline.\nFurthermore, we expressly reserved the precise bounds of what\nconstituted a petition based solely on after-arising grounds. See\nid. at 666–68. Finally, we stressed that review should generally\nbe available “when new information casts doubt upon the\nvalidity of a [regulatory] standard.” Id. at 665. At the time, the\n\n 27\nafter-arising provision contained no separate deadline\nrequiring a petition for review to be filed within sixty days of\nthe after-arising ground. See id. at 657 n.3. So the difference\nbetween “direct review” of a regulation through the after-\narising provision and review of a “refusal to revise” the\nregulation appeared largely semantic. See id. at 666. But, two\nyears after Oljato was decided, Congress amended section\n7607(b)(1) to include the separate filing deadline, see Pub. L.\nNo. 95-95 § 305(c)(3), 91 Stat. 685, 776 (1977), which, as\nexplained above, made the after-arising provision a singularly\npoor vehicle for securing the review that Oljato assumed would\nbe readily available.\n\n After Oljato, we held in Group Against Smog & Pollution\nv. EPA, 665 F.2d 1284 (D.C. Cir. 1981), that the failure to\nchallenge a 1974 regulation within the original sixty-day\ndeadline did not bar “judicial review of the agency’s\nsubsequent refusal to revise the standard on the basis of new\ninformation.” Id. at 1291. Quoting liberally from Oljato, we\nsuggested that such review would proceed through the after-\narising provision. Id. at 1289. But we permitted review even\nthough the after-arising provision, as amended, could not have\napplied. The case involved comments filed with EPA in April\n1977, which we treated as a petition for a rulemaking. Id. at\n1290 & n.47. On April 13, 1978, EPA declined to amend the\nrule as requested. Id. at 1288. The ensuing petition for review\nwas filed on June 12, 1978—more than a year after the\ntechnological changes discussed in the comments, but precisely\nsixty days after the refusal to amend. See id. Because the\npetition for review was not filed within sixty days of the\nasserted after-arising grounds, the after-arising provision\nplainly did not apply. So, review must have rested on the\ntheory that the refusal to amend was itself a reviewable action\ntriggering its own sixty-day filing window.\n\n 28\n The approach we follow here—treating denials of\nrulemakings based on new facts as independently reviewable\ndecisions—does not reduce the after-arising provision to\nsurplusage. Our cases have recognized other circumstances\ntriggering the after-arising provision, including judicial\ndecisions that significantly “changed the legal landscape”\nfaced by petitioners, Honeywell Int’l, Inc. v. EPA, 705 F.3d\n470, 473 (D.C. Cir. 2013), and “the occurrence of an event that\nripens a claim,” Coalition for Responsible Regulation v. EPA,\n684 F.3d 102, 129 (D.C. Cir. 2012) (per curiam). In cases like\nthese—where, for example, an intervening statute, regulation,\nor judicial decision extends old regulations to new parties—the\nafter-arising ground is easily dated, the relevant filing\ndeadlines are clear, and so the provision functions predictably.\nMoreover, where the after-arising provision does apply, it\npermits the petitioner to contend not only that changed\ncircumstances warrant amending an existing regulation but\nalso that the regulation was unlawful as originally\npromulgated. See, e.g., Honeywell, 705 F.3d at 473.\n\n For these reasons, we conclude that we have jurisdiction\nto consider the petitioners’ argument that EPA arbitrarily\nrefused to amend the point of obligation rule based on the\nchanged circumstances cited by the petitioners.\n\n 2. After-Arising Grounds Under Section 7607(b)(1)\n\n The February 2016 petitions for review, filed before EPA\nresolved any of the rulemaking petitions, rested on the\nalternative jurisdictional theory that EPA’s publication of the\n2014–16 volume requirements constituted an after-arising\nground within the meaning of section 7607(b)(1). Our\nconclusion above does not moot this question, because if this\nalternative theory were valid, then the petitioners could directly\nattack the point of obligation rule as originally promulgated.\n\n 29\nNonetheless, the petitioners have abandoned this theory of\njurisdiction. In their merits briefs, they never actually attack\nthe 2010 rule as originally promulgated; instead, they challenge\nonly the 2017 denial of their rulemaking petitions. Moreover,\nin requesting relief, they do not ask us to set aside the 2010\nrule, but only to “vacate the [2017] Denial[] and remand to\nEPA” for a rulemaking to change the point of obligation rule\ngoing forward. Alon Br. 58. And at oral argument, they\ndisclaimed any challenge to the 2010 rule itself and confirmed\nthat their only challenge was to EPA’s 2017 refusal to revise\nthe point of obligation rule. See Rec. of Oral Argument at\n2:18:50–2:19:15.\n\n 3. Mandatory Reconsideration Under Section\n 7607(d)(7)(B)\n\n Finally, the petitioners assert jurisdiction under section\n7607(d)(7)(B), on the ground that EPA erroneously denied\npetitions for mandatory reconsideration of the point of\nobligation rule. To review, section 7607(d)(7)(B) provides in\nrelevant part that only objections “raised . . . during the period\nfor public comment” may be “raised during judicial review.”\nBut if the objector “can demonstrate to the Administrator that\nit was impracticable to raise such objection within such time or\nif the grounds for such objection arose after the period for\npublic comment (but within the time specified for judicial\nreview) and if such objection is of central relevance to the\noutcome of the rule,” then EPA must “convene a proceeding\nfor reconsideration of the rule,” and any refusal to do so is\njudicially reviewable. Here, the petitioners assert that the\nground for their objections—EPA’s statements and actions in\nits 2014–16 volume regulation—arose “after the period for\npublic comment” on the 2010 point of obligation rule, and that\n\n 30\nthe objections are “of central relevance to the outcome of the\nrule.”\n\n The petitioners misapprehend the statutory text and\nstructure. Section 7607(d)(7)(B) does not extend the\njurisdictional deadline to seek judicial review imposed in\nsection 7607(b)(1); instead, it specifies a non-jurisdictional\nexhaustion requirement. See EPA v. EME Homer City\nGeneration, L.P., 572 U.S. 489, 512 (2014). Its first sentence\ngenerally requires parties to raise objections “during the period\nfor public comment” in order to later present them in court. Its\nsecond sentence allows a narrow exception when a centrally\nimportant objection cannot feasibly be raised during the\ncomment period—either because “the grounds for such\nobjection arose after the period for public comment,” or\nbecause commenting was otherwise “impracticable.” If an\nobjection fits within this exception, the consequences are\nweighty: EPA must grant reconsideration and conduct a new,\nfull-dress, notice-and-comment rulemaking. And if EPA\ndenies reconsideration, the objector may seek judicial review.\n\n This “limited exception” to the normal exhaustion\ndeadline, Util. Air Regulatory Grp. v. EPA, 744 F.3d 741, 746\n(D.C. Cir. 2014), does not come with a free pass from the\nsubsequent deadline to seek judicial review. To the contrary,\nthe second sentence of section 7607(d)(7)(B) covers only\nobjections that arise after the close of the comment period, yet\nwithin the time specified for judicial review. As noted above,\nthat time for judicial review—sixty days from the promulgation\nof the final rule—is specified in section 7607(b)(1). Section\n7607(d)(7)(B) does not enlarge that filing period, but merely\nfills a narrow gap within it: allowing orderly exhaustion of\nimportant objections that “first became known to the petitioner\nafter the comment period ended, but before the period for\npetitioning for review expired.” Am. Petroleum Inst. v. Costle,\n\n 31\n665 F.2d 1176, 1192 (D.C. Cir. 1981). We recognize that, as a\ntextual matter, the statutory phrase “but within the time\nspecified for review” qualifies the requirement that the grounds\nmust have arisen “after the period for public comment,” but not\nthe alternative requirement that “it was impracticable” to raise\nthe objection “during the period for public comment.”\nHowever, the petitioners do not invoke the impracticability\nprong of section 7607(d)(7)(B). Moreover, we have construed\nthat prong to cover instances when the final rule was not a\nlogical outgrowth of the proposed rule, Clean Air Council v.\nPruitt, 862 F.3d 1, 10 (D.C. Cir. 2017) (per curiam), which\nlikewise involve problems during the period for public\ncomment on or petitioning for review of the regulation itself—\nnot problems that arise when circumstances change years or\ndecades later.\n\n The petitioners argued in briefing that the “time specified\nfor judicial review” referenced in section 7607(d)(7)(B)\nencompasses not only the initial sixty-day window after a\nrule’s promulgation, but also the secondary sixty-day limit\nfrom after-arising grounds in the fourth sentence of section\n7607(b)(1). But as noted above, the petitioners abandoned at\noral argument any reliance on the latter after-arising provision.\nMoreover, their theory would transform what we have\ndescribed as a “limited” gap-filling provision, Util. Air\nRegulatory Grp., 744 F.3d at 746, into a perpetually looming\nthreat of mandatory notice-and-comment reconsideration.\nTellingly, the petitioners can cite no case employing section\n7607(d)(7)(B)’s mandatory reconsideration procedure for\nobjections that arose after the close of the initial window for\njudicial review. Their interpretation would “make a mockery\nof Congress’[s] careful effort to force potential litigants to\nbring challenges to a rule issued under this statute at the\n\n 32\noutset.” Am. Rd. & Transp. Builders Ass’n v. EPA, 705 F.3d\n453, 458 (D.C. Cir. 2013) (quotation marks omitted).\n\n Because the petitioners’ objections to the point of\nobligation rule did not arise within the initial window for\njudicial review of the 2010 point of obligation rule, but only\nsome five years later, EPA properly denied mandatory\nreconsideration.\n\nB. Merits of Challenges to EPA’s Refusal to Revise the\n 2010 Point of Obligation Rule\n\n The Alon Petitioners offer an array of arguments to\nchallenge the denial. None, however, is persuasive.\n\n We are reviewing EPA’s denial of a petition for\nrulemaking to amend the agency’s point of obligation rule. See\nsupra Part IV.A.1. Accordingly, our review is “‘extremely\nlimited’ and ‘highly deferential.’” New York v. EPA, 921 F.3d\n257, 261 (D.C. Cir. 2019) (quoting Massachusetts, 549 U.S. at\n527–28). “To set aside the agency’s judgment, [we] must\nconclude that EPA had not ‘adequately explained the facts and\npolicy concerns relied on’ or that those facts did not ‘have some\nbasis in the record.’” Id. (quoting WildEarth Guardians v.\nEPA, 751 F.3d 649, 653 (D.C. Cir. 2014)). We have no basis\nfor such a conclusion. In denying the petition for rulemaking,\nEPA considered the “information currently before” it and\ndetermined “that the point of obligation is appropriately\nplaced,” wrestling with the petitioners’ claims to the contrary.\nEPA Denial at 8, Alon J.A. 62. As is evident from our\ndiscussion below, EPA did so with enough thoroughness and\nreasonableness to satisfy our limited, deferential review.\n\n We start with EPA’s reasoning, which petitioners say is\narbitrary and capricious. See 42 U.S.C. § 7607(d)(9)(A).\n\n 33\n At the root of petitioners’ claim is a single premise: that\nthe current point of obligation misaligns incentives by\nrequiring those who refine fossil fuel, but not those who blend\nit, to meet the RFS program’s annual standards. In petitioners’\nview, this misalignment forces refiners to purchase RINs to\nsatisfy their RFS obligations, jacking up their costs, while\ngiving windfall profits to blenders, who produce (but don’t\nconsume) RINs. From this cycle of “RINsanity,” petitioners\nsay, flow harms galore. Alon Reply Br. 25. Higher RIN prices\nnot only threaten the financial viability of refiners (putting our\neconomy and energy security in jeopardy), see, e.g., Alon Br.\n46, but they incentivize RIN hoarding, which feeds market\nvolatility, and gives some market participants an unfair leg up,\nsee, e.g., id. at 40.\n\n The problem with this argument, however, is that EPA\nreasonably explained why, in its view, there is no misalignment\nin the RFS program. According to EPA, refiners “recover the\ncost of the RINs they purchase” by passing that cost along in\nthe form of “higher prices for the petroleum based fuels they\nproduce.” EPA Denial at 25, Alon J.A. 79. It grounded that\nconclusion in studies and data in the record. EPA and the\nauthors of the pertinent studies took advantage of the fact that\nthere are pairs of petroleum products in which one variant is\nsubject to the RIN obligation (such variants being awkwardly\ncalled “obligated fuels”), whereas its not-quite-identical twin is\nnot. For example, gasoline and diesel sold for use in the United\nStates are “obligated,” whereas the same fuels sold for export\nare not. EPA Denial at 23, Alon J.A. 77. The same goes for\ndomestic diesel fuel, which is “obligated,” and jet fuel, which\nis not. Christopher R. Knittel et al., The Pass-Through of RIN\nPrices to Wholesale and Retail Fuels Under the Renewable\nFuel Standard 4 (July 2015) (Knittel), Alon J.A. 534.\nComparing these pairs, the agency found that as RIN prices\nincreased, a gap “open[ed] up between” the price for obligated\n\n 34\nand unobligated fuels, a gap rather precisely matching the\ncontemporaneous increase in RIN price—a strong indication\nthat refiners were “recoup[ing] the costs associated with RIN\nprices.” EPA Denial at 23, Alon J.A. 77.\n\n Further confirming the price relationship, Professor\nKnittel and his colleagues found that 73% of a change in RIN\nprice was passed through in the form of higher petroleum prices\nin the same day, 98% within two business days. Knittel 26,\nAlon J.A. 536.\n\n Reviewing the findings, EPA (accurately) reported that the\npapers by Knittel and his colleagues, and by Argus Consulting\nServices, “concluded that the RIN cost was generally included\nin the sale prices of obligated fuels.” EPA Denial at 25, Alon\nJ.A. 79; see Knittel 26, Alon J.A. 536; Argus Consulting\nServices, Do Obligated Parties Include RINs Costs in Product\nPrices? 15 (Feb. 2017), Alon J.A. 564 (“There are very specific\ncorrelating price data for diesel that indicate that refiners . . .\npass along the RINs cost . . . .”).\n\n A similar analysis, EPA concluded, reveals that just as\n(obligated) refiners do not pay excess costs, neither do blenders\n(who are not obligated under the program) nor integrated\nrefiners (who perform their own in-house blending) reap\nwindfall profits. True, both earn RINs, without purchasing\nthem on the open market, by blending renewable fuel into\npetroleum blendstock. And true, as well, both can sell those\nRINs, enjoying whatever revenues market conditions and their\nown efficiencies permit. But as EPA quite accurately\nexplained, this is only half the equation. In a competitive\nmarket there’s no such thing as a free lunch, and blenders and\nintegrated refiners pay their tab just as others do; they just do\nso indirectly. To offer finished fuel without attached RINs at a\ncompetitive price, these entities must discount their blended\n\n 35\nfuel by roughly the value of the RINs that they detached and\nkept for themselves. EPA Denial at 29, Alon J.A. 83. In other\nwords, they “sell the finished transportation fuel at a loss,” but\n“maintain[] profitability through RIN sales.” Id. at 27–28, 29,\nAlon J.A. 81–82, 83.\n\n To be sure, in response to EPA’s proposed denial,\ncommenters criticized the studies relied on by the agency.\nThey contended, for example, that Professor Knittel and his\ncolleagues erred by removing certain spreads from the analysis,\nby including others, and by pooling the results of various\ncomparisons. See EPA Denial at 25, Alon J.A. 79. But\npetitioners have not raised these arguments here, and for that\nreason we do not consider them. While petitioners do complain\nthat EPA relied on a “preliminary” analysis, see Alon Br. 54;\nAlon Reply Br. 23, that objection—whatever its persuasive\nforce—says nothing about the other studies in the record (for\nexample, by Professor Knittel et al.).\n\n Petitioners try, instead, to trace various refiner problems to\nEPA’s refusal to obligate blenders. They suggest that the\nalleged misplacement of the point of obligation causes\nbankruptcies, see, e.g., Alon Br. 3, 47, and inflicts economic\nhardship on small refineries, see, e.g., id. at 49, especially in\nthe form of inflicting wildly disproportionate RIN acquisition\ncosts on them, see, e.g., Alon Reply Br. 26. But some of these\nevents occurred after EPA issued its denial, see, e.g., Alon Br.\n49 (“following the Denial”); Alon Reply Br. 26 (“just after the\nDenial”), and are therefore not properly before us, see\nEnvironmental Defense Fund, Inc. v. Costle, 657 F.2d 275, 284\n(D.C. Cir. 1981). More importantly, the claims presuppose that\nrefiners cannot recover their RIN costs and that blenders reap\nwindfall profits—suppositions that, as discussed above, EPA\nreasonably rejected.\n\n 36\n Petitioners respond by plucking snippets from the denial,\nstringing them together with contrasting (bolded) conjunctions,\nand asserting that EPA’s “discussion of RIN prices” is\n“irreconcilably inconsistent.” Alon Br. 53. But we find no\ninconsistency on EPA’s part.\n\n Take one of petitioners’ examples:\n\n . . . RIN prices had no “significant impact on retail\n gasoline (E10) prices,” JA75;\n\n although “RINs . . . provide a price signal to\n consumers to help achieve . . . greater renewable fuel\n production and use,” JA75.\n\nAlon Br. 53 (second and third alterations in original) (quoting\nEPA Denial at 21, Alon J.A. 75). At first blush, the two\ncomments, located on the very same page, seem inconsistent.\nHow could RIN prices have no “significant impact” on retail\nprices, while, at the same time, provide “a price signal to\nconsumers”?\n\n They can do so for the simple reason that the remarks refer\nto different things, a detail omitted from petitioners’ brief. This\nbecomes apparent when the passage from which petitioners\nplucked their quotes (bolded and underscored below) is viewed\nin full:\n\n External, non-EPA assessments similarly concluded\n that increased RIN prices had not had a significant\n impact on retail gasoline (E10) prices. When RIN\n prices rise, the market price of the petroleum\n blendstocks produced by refineries also rise to cover\n the increased RIN costs, in much the same way as they\n would rise in response to higher crude oil prices. The\n effective price of renewable fuels (the price of the\n\n 37\n renewable fuel with attached RIN minus the RIN\n price), however, decreases as RIN prices increase.\n When renewable fuels are blended into petroleum\n fuels these two price impacts generally offset one\n another for fuel blends such as E10 with a renewable\n content approximately equal to the required\n renewable fuel percentage standard. Higher RIN\n prices also generally result in higher prices for fuels\n with lower renewable content (such as E0 or\n petroleum diesel) and lower prices for fuels with\n higher renewable content (such as E85 or B20). The\n cost of the RIN therefore serves as a cross-subsidy,\n reducing the price of renewable fuels and increasing\n the price of petroleum based fuels in transportation\n fuel blends, thus incentivizing increased blending of\n renewable fuels into the transportation fuel pool. In\n this way the RINs also help provide a price signal to\n consumers to help achieve the Congressional goals\n of greater renewable fuel production and use.\n Fuels with higher renewable content are relatively\n cheaper to consumers than they would be absent high\n RIN prices, while fuels with lower renewable content\n are relatively more expensive when RIN prices are\n high.\n\nEPA Denial at 20–21, Alon J.A. 74–75.\n\n As we can see, the first statement (no significant price\nimpact) is referring to the price of E10—a blend of 90%\ngasoline, 10% ethanol. As EPA explained, RINs work as a\n“cross-subsidy,” effectively taxing the use of petroleum-based\nfuels (e.g., gasoline) and subsidizing the use of renewables\n(e.g., ethanol) in making a blended transportation fuel like E10.\nEPA Denial at 21, Alon J.A. 75.\n\n 38\n Before we dig in further, let’s take a step back. We must\nfirst recognize that EPA assumes that we are talking of a market\nwhere the RFS program, in effect, mandates minimum levels\nof renewables in marketed fuels, a mandate that necessarily\nimpacts fuel prices. EPA is not making a claim that the\nmandatory inclusion of renewables in transportation fuel\nrenders a gallon of gasoline lawfully purchased at the pump\ncheaper than it would have been absent the RFS program.\n\n To see what this means, start on the subsidy side: Suppose\na blender can realize $2.25 on a gallon of ethanol with an\nattached RIN. If the blender can detach and sell that RIN for\n$0.05, then the net ethanol value is only $2.20—a $0.05\nsavings that in a competitive market should pass through to\nconsumers. Now take the tax side: Because refiners must\npurchase RINs to satisfy the RFS obligations that arise from\nselling gasoline to blenders, a blender’s value for a gallon of\ngasoline may rise, due to the RFS program, from, say, $2.75 to\n$2.76. See Dallas Burkholder, Office of Transportation & Air\nQuality, EPA, A Preliminary Assessment of RIN Market\nDynamics, RIN Prices, and Their Effects 17, EPA-HQ-OAR-\n2016-0544-0009 (May 14, 2015), Alon J.A. 337. As a result\nof both price impacts, EPA described, blended fuels with a\nhigher percentage of renewable content (e.g., 85% ethanol) will\nbe cheaper than they would have been (absent the program),\nwhereas fuels with a lower percentage of renewable content\n(e.g., pure gasoline) will be more costly than they would have\nbeen (absent the program). EPA Denial at 21, Alon J.A. 75.\nFor E10, the “two price impacts generally offset one another,”\nso (back to the first statement) any change in RIN price\ngenerally has no “significant impact on” the E10 price. Id.\n\n But that’s just E10. There is an effect (of differing\nmagnitude) on, say, E85 or E0. And that is where the second\nstatement (“provide a price signal”) comes in: the signal arises\n\n 39\nfrom a comparison of relative prices across the spectrum of\ntransportation fuels. Again, as EPA explains, “[f]uels with\nhigher renewable content are relatively cheaper to consumers\nthan they would be absent high RIN prices, while fuels with\nlower renewable content are relatively more expensive when\nRIN prices are high.” Id. The two statements are consistent.\n\n Continuing the search for inconsistency, petitioners direct\nour attention to “EPA’s past pronouncements.” Alon Br. 50.\nIn them, they see an irrational “about-face”—with EPA saying,\nat first, that “low RIN prices [were] a sign that the [RFS\nprogram] was working,” but claiming, now, “that high RIN\nprices are . . . desirable.” Id. at 51–52. Again, that’s not quite\nright. All EPA originally said was that when it first adopted\nthe point of obligation, it did so based, in part, on its\n“expectation at that time that there would be an excess of RINs\nat low cost.” Regulation of Fuels and Fuel Additives: Changes\nto Renewable Fuel Standard Program, 74 Fed. Reg. 24,904,\n24,963/2 (May 26, 2009) (proposed rule); see also Alon Br. 50\n(citing EPA Denial at 13, Alon J.A. 67 (citing, in turn, 74 Fed.\nReg. at 24,963)). EPA did not suggest that low RIN prices were\na sign of market health—nor that high prices were a cause for\nalarm.\n\n In any case, EPA addressed petitioners’ concern over high\nRIN prices head on; the agency explicitly determined, on the\ncurrent record, that “higher RIN prices” are not “indicative of\na dysfunctional RIN market.” EPA Denial at 19, Alon J.A. 73.\nRather, EPA explained, these prices accurately reflect the\nincreasing cost associated with “getting ever-greater volumes\nof renewable fuel into the transportation fuel pool—the explicit\ngoal [of] the RFS program.” Id. Put more bluntly, the increases\nin RIN prices are a completely understandable effect of the\nprogram’s ever-increasing pressure to expand renewable\nvolumes. Pushing out along the supply curve takes the raw\n\n 40\nmarket price of the RIN-eligible fuel steadily into higher\nrealms—except to the extent that production innovations or\neconomies may tend to lower costs. So far as appears, it has\nnothing to do with EPA’s allocation of the obligation.\n\n What about EPA’s concern, petitioners ask, that including\nblenders in the point of obligation would expand the number of\nobligated parties and, as a result, ratchet up the program’s\ncomplexity? Isn’t that hard to square with EPA’s claim, made\nyears earlier, that “essentially all downstream blenders . . . are\n[already] regulated parties”? Alon Br. 42 (quoting 75 Fed.\nReg. at 14,722/2). Again, not at all. Although the participation\nof all (or nearly all) blenders in the RIN market subjects them\nto RFS registration, recordkeeping, and reporting\nrequirements, “the majority of these downstream [regulated]\nparties are . . . currently not obligated parties.” EPA Denial at\n69, Alon J.A. 123 (emphasis added). As EPA explained, there\n“is a significant distinction between being a ‘regulated party’\nand being an ‘obligated party.’” Id. “Obligated parties must\nmeet all of [the requirements faced by regulated parties] and\nalso calculate an annual renewable volume obligation, acquire\nthe appropriate number of RINs in the market, practic[e] due\ndiligence to ensure [the RINs’] validity, file annual compliance\nreports demonstrating compliance, and maintain records to that\neffect.” Id. at 69 n.205, Alon J.A. 123 (emphasis added); see,\ne.g., 40 C.F.R. §§ 80.1427(a), 80.1450(a), 80.1451(a),\n80.1454(a). It was not unreasonable for EPA to conclude that\nimposing these burdens on additional entities would add to the\nprogram’s complexity (and therefore be undesirable absent an\nadequate offsetting benefit).\n\n Nor was it unreasonable to find that going down this\nroute—overhauling a foundational element of the program—\nwould create “uncertainty in the fuels marketplace.” EPA\nDenial at 2, Alon J.A. 56. As EPA said, “all parties regulated\n\n 41\nin the RFS program have made significant investments and\ndecisions about their participation in the program and their\nposition in the market on the basis of the existing regulations,\nincluding the definition of obligated parties.” EPA Denial at\n79, Alon J.A. 133. In these circumstances, it isn’t hard to\nimagine how changing course could throw players off their\ngame. Of course, as petitioners note, uncertainty may have\n“plagued the RFS Program for years.” Alon Br. 37. But true\nor not, EPA needn’t pile on; the cure for uncertainty isn’t\nspawning more uncertainty.\n\n Taking a step back, petitioners launch a closing broadside\nagainst the entire process. They assert that EPA “disregarded\nthis Court’s remand” in ACE, 864 F.3d at 737, and arbitrarily\ncredited some comments over others. Alon Br. 31–32, 55–56.\nBut the ACE remand required, at most, that the agency “address\nthe point of obligation issue.” 864 F.3d at 737. And, as\ndetailed throughout this opinion, the agency has done so\nreasonably, analyzing the data and explaining its decision.\nNothing more was required.\n\n\n * * *\n\n We have considered the Alon Petitioners’ other arguments\nand have found them to be either without merit or, in the case\nof the argument relying on 42 U.S.C. § 7545(o)(5)(A), see\nAlon Br. 34, insufficiently developed, see, e.g., Masias v. EPA,\n906 F.3d 1069, 1077 (D.C. Cir. 2018). For the foregoing\nreasons, the petitions for review are denied.\n\n V. 2017 Annual Volumetric Rule\n\n We turn now to the Coffeyville Petitioners’ challenges.\n\n 42\nA. Point of Obligation\n\n We first consider the Coffeyville Petitioners’ challenge to\nEPA’s decision in the 2017 Rule not to reassess which\ncategories of industry players are “obligated parties” under the\nrenewable fuel program. As the Coffeyville Petitioners read it,\nthe statutory provision requiring EPA to set annual renewable\nfuel percentage standards also imposes on EPA a\nnondiscretionary duty to reconsider—every year—which types\nof entities are obligated to demonstrate to EPA compliance\nwith the percentage standards. See 42 U.S.C. § 7545(o)(3)(B).\nThey claim EPA shirked that duty when it treated the issue as\nbeyond the scope of its 2017 annual rulemaking. EPA counters\nthat it identified the obligated parties in 2007 pursuant to\nCongress’s mandate to set “compliance provisions” for the new\nrenewable fuel program, id. § 7545(o)(2), reaffirmed that\ndecision in 2010, and that nothing in the mandate to calculate\nthe annual percentage standards requires it to reconsider the\npoint of obligation each year. EPA also asserts that it\nappropriately addressed the Coffeyville Petitioners’ complaints\nthat it obligated the wrong parties in a separate proceeding from\nits annual volumetric rulemaking.\n\n 1. Jurisdiction\n\n EPA and a coalition of Respondent-Intervenors\nrepresenting the renewable fuel and refinery industries assert\nthat we lack jurisdiction because the Coffeyville Petitioners\neffectively challenge the compliance rule that has been on the\nbooks for a decade or so, see 40 C.F.R. § 80.1406(a)(1), and\ndid not petition within 60 days of its publication or within 60\ndays of any valid “grounds arising” thereafter. See 42 U.S.C.\n§ 7607(b)(1); Med. Waste Inst. v. EPA, 645 F.3d 420, 427\n(D.C. Cir. 2011). But petitioners are not challenging EPA’s\ndecision to adopt the rule in 2007 or retain it in 2010. Rather,\n\n 43\nthey contend that the provision calling on EPA to set annual\nvolumes of biofuels “applicable to refineries, blenders, and\nimporters, as appropriate,” requires EPA to reassess each year\nwhether the point of obligation set when the agency established\nthe program is still “appropriate,” or if EPA should re-assign it\nand restructure the RIN market and other compliance\ninfrastructure going forward. See 42 U.S.C.\n§ 7545(o)(3)(B)(ii)(I). That challenge was timely filed within\n60 days of the promulgation of the annual fuel standards. See\nsupra Part IV.A.1.\n\n 2. Merits\n\n This dispute turns on the roles of two provisions of the\nstatute directing the EPA to establish and run a Renewable Fuel\nProgram, 42 U.S.C. § 7545(o)—paragraphs (2) and (3).\n Paragraph (2) directs EPA to “promulgate regulations”\nsetting up a program to “ensure that transportation fuel sold or\nintroduced into commerce in the United States . . . contains at\nleast the applicable volume[s] of renewable fuel,” as specified\nin subparagraph (2)(B). Id. § 7545(o)(2)(A)(i). Among the\nparameters Congress required EPA to include were\n“compliance provisions applicable to refineries, blenders,\ndistributors, and importers, as appropriate,” to ensure that the\nrequirements of paragraph (2), including the applicable volume\nrequirements specified in subparagraph (2)(B), are met. Id.\n§ 7545(o)(2)(A)(iii)(I), (B). There is no question that EPA has\nauthority to set those parameters, including the point of\nobligation, and to adjust them if a change is needed.\n Paragraph (3), in turn, requires EPA to determine and\npublish annual renewable fuel obligations designed to\n“ensure[] that” the applicable volumes specified in paragraph\n(2) are met. Id. § 7545(o)(3)(B)(i). Those renewable fuel\nobligations must:\n\n 44\n (I) be applicable to refineries, blenders, and\n importers, as appropriate;\n (II) be expressed in terms of a volume percentage\n of transportation fuel sold or introduced into\n commerce in the United States; and\n (III) . . . consist of a single applicable percentage\n that applies to all categories of persons\n specified in subclause (I).\nId. § 7545(o)(3)(B)(ii)(I)–(III).\n The parties’ dispute centers on the meaning of “as\nappropriate” in subclause (3)(B)(ii)(I). The Coffeyville\nPetitioners contend that the phrase unambiguously requires\nEPA annually to reconsider which parties it is “appropriate” to\nobligate to meet the renewable fuel obligations. EPA responds\nthat the statute is, at most, ambiguous as to whether Congress\nexpected EPA annually to revisit the obligated-parties\ndesignation, or whether the agency may generally rely on the\n“appropriate[ness]” finding it made pursuant to its paragraph\n(2) authority. We begin by asking “whether Congress has\ndirectly spoken to the precise question at issue,” and conclude\nthat it has not. Chevron, 467 U.S. at 842.\n EPA reads “as appropriate” in paragraph (3) to mean that\nthe agency has “discretion” to decide whether, when, and how\nto reassess which of three types of industry actors—refineries,\nblenders, and importers—should continue to bear the point of\nobligation, as originally designated in the compliance\nprovisions. See EPA Denial, Coffeyville J.A. 779–80. At oral\nargument, the agency conceded that its exercise of this\ndiscretion is reviewable, so that the exclusion of the point of\nobligation issue from an annual rulemaking could, under other\ncircumstances, constitute an abuse of discretion. Rec. of Oral\nArg. 1:22:22–1:24:00; 1:37:45–1:40:11.\n\n 45\n The Coffeyville Petitioners object that the phase\n“applicable . . . as appropriate” means applicable as\ncontemporaneously determined to be appropriate in the annual\nvolumetric rulemakings. Coffeyville Br. 30–33; see also Conc.\nOp. 4–5, 7. But paragraph (3) does not specify when or in what\ncontext EPA must make its appropriateness determination, nor\ndoes the phrase “as appropriate” itself specify a particular\ntemporal dimension—as between, for example, parties\nappropriately designated in the past (as EPA interprets it) and\nparties now appropriately selected (as Coffeyville insists). The\nterm “appropriate” “naturally and traditionally includes\nconsideration of all the relevant factors,” Michigan v. EPA,\n135 S. Ct. 2699, 2707 (2015) (quoting White Stallion Energy\nCenter, LLC v. EPA, 748 F.3d 1222, 1266 (D.C. Cir. 2014)\n(Kavanaugh, J., concurring in part and dissenting in part)), but\nit does not dictate when that consideration must be made. In\nother words, the requirement that the point of obligation be\n“appropriate” is at most grounds for assessing whether the\nagency adequately explained its policy choices regarding the\nappropriateness determination, not for imposing our own gloss\non that broad term as a matter of law. See Kisor v. Wilkie, 139\nS. Ct. 2400, 2448–49 (2019) (Kavanaugh, J., concurring in the\njudgment) (“[S]ome cases involve regulations that employ\nbroad and open-ended terms like ‘reasonable,’ ‘appropriate,’\n‘feasible,’ or ‘practicable.’ Those kinds of terms afford\nagencies broad policy discretion, and courts allow an agency to\nreasonably exercise its discretion to choose among the options\nallowed by the text of the rule. But that is more State Farm\nthan Auer” or Chevron (emphasis added)). Here, as explained\nbelow, EPA reasonably exercised its discretion, and explained\nits decision, to address the point of obligation issue in a\nseparate proceeding from its annual volumetric rulemaking.\n The fact that paragraphs (2) and (3) both include the phrase\n“as appropriate” does not make the Coffeyville Petitioners’ the\n\n 46\nonly permissible interpretation. Even if, as our colleague\ncontends, Conc. Op. 7, the two phrases bear the exact same\nmeaning, but see Coffeyville Br. 30–31 (arguing that\nparagraphs (2) and (3) are “worded differently” and have\n“different contexts”), paragraph (3) simply does not dictate\nwhen or in what context EPA must make the appropriateness\ndetermination. It is the surrounding context, not the phrases\nthemselves, that suggests when EPA might make that choice.\n Unable to point to any express textual requirement that\nEPA annually reconsider the point of obligation, the\nCoffeyville Petitioners contend that, had Congress intended to\nallow EPA in annual volumetric rulemakings to rest on its\nparagraph (2) appropriateness determination, subclause\n(3)(B)(ii)(I) could have more simply cross-referenced\nparagraph (2). But replacing subclause (3)(B)(ii)(I) with a\nsimpler cross-reference would not have achieved quite the\nsame effect. While refineries, blenders, importers, and\ndistributors may all be subject to compliance provisions under\nparagraph (2), paragraph (3)’s applicability provision points to\na more limited universe of potential obligated parties—to\nrefineries, blenders, and importers, but not distributors. That\nsupports EPA’s understanding of subclause (3)(B)(ii)(I) as a\ncross-reference that also clarifies a limit on EPA’s options in\nsetting the point of obligation. See 72 Fed. Reg. at 23,923/2\n(preamble for compliance rule referencing paragraph (3)’s\napplicability provision).\n We are unpersuaded by the suggestion that such limitation\nwas so clear even without subclause (3)(B)(ii)(I) that EPA’s\nreading of that applicability provision renders it superfluous.\nThe suggestion is that, because distributors do not “introduce”\nfuel into commerce, they could not be obligated parties in any\nevent, with or without the applicability provision. Our\ncolleague posits that blenders—but not distributors—can in\nfact “introduce” transportation fuel into commerce by blending\n\n 47\ngasoline and diesel fuel with other fuels that have not already\nbeen introduced by someone else. Conc. Op. 11. But that\nargument rests on a complicated series of inferences from spare\nstatutory text, as well as post-enactment regulations that do not\nnecessarily show what the statute must have meant. For\nexample, our colleague reasonably infers that the national\n“volume[] of transportation fuel,” 42 U.S.C. § 7545(o)(3)(A),\nmust be counted at the first moment each gallon of fuel enters\ncommerce, in order to avoid double-counting. But, the statute\nnowhere states the point directly. Still less clearly does the\nstatute state our colleague’s corollary—necessary to the\nsurplusage argument—that the point of obligation must also be\nplaced, if at all, at the first moment a gallon of fuel enters\ncommerce. That corollary is less obvious, because the\nstatutory link between the point of obligation and entry into\ncommerce is not ironclad: All parties agree that not every\ngallon of transportation fuel must be subject to the point of\nobligation upon entry into commerce. The statute plainly\nallows EPA to obligate an “appropriate” subset of the three\ncategories of parties. And obligating blenders would involve\ndouble counting unless the transportation fuel they use to create\nblends were not already counted upon its importation or sale to\nthem. Under the circumstances, it is reasonable to read\nsubclause (3)(B)(ii)(I) as clarifying what is at best a non-\nobvious inference that distributors cannot be subjected to the\npoint of obligation.\n So subclause (3)(B)(ii)(I) as EPA reads it is not a\nsuperfluity, but makes clear that EPA may have permissibly\nplaced the point of obligation on refineries, blenders, and\nimporters, but not distributors. For the same reason, subclause\n(3)(B)(ii)(III), which requires that the annual standards apply\n“to all categories of persons specified in subclause (I),” id.\n§ 7545(o)(3)(B)(ii)(III), does not contain what our colleague\nviews as an unnecessary double cross-reference to paragraph\n\n 48\n(2), because it, too, is operative in not just cross-referencing,\nbut also clarifying a limit on the three permissible targets of its\n“single applicable percentage.”\n The thrust of the Coffeyville Petitioners’ retort—that if\nCongress had wanted to confer discretion or provide a limiting\ncross-reference to paragraph (2), it would have said so more\nplainly—applies with greater force against their own reading.\nHad Congress intended EPA to consider on an annual basis\nwhether to redo the point of obligation designation—a\ndesignation that no-one disputes is a necessary cornerstone of\nthe paragraph (2) compliance provisions—it knew how to\nimpose such a requirement. The Clean Air Act’s provisions on\nambient air quality, for instance, require EPA to “complete a\nthorough review” of the air quality standards “at five-year\nintervals” and “promulgate such new standards as may be\nappropriate.” Id. § 7409(d)(1). The Act’s provisions\ncontrolling hazardous air pollutants emitted from major and\narea sources require EPA to “review, and revise as necessary”\nthe applicable emission standards “no less often than every 8\nyears.” Id. § 7412(d)(6). Paragraph (3) of the RFS program,\nin contrast, does not tell EPA to “complete a thorough review,”\nor “review, and revise as necessary” its point of obligation\ndecision—or anything even close.\n To be sure, EPA’s reading is not ineluctable. We do not\ndoubt that Congress could have more directly provided that the\nrenewable fuel obligations do not apply to distributors. See\nConc. Op. 8. But, for the reasons discussed, we are\nunconvinced that paragraph (3) plainly requires EPA to\nconsider adjusting the point of obligation each year. See\nValero Energy Corp. v. EPA, No. 7:17-cv-00004-O, 2017 WL\n8780888, at *4 (N.D. Tex. Nov. 28, 2017) (holding that “there\nis no clear statutory mandate . . . obligating [EPA] to evaluate\nor adjust . . . what entities are ‘appropriate[ly]’ forced to\ncomply with” the annual renewable fuel obligations\n\n 49\n(alterations in original) (quoting 42 U.S.C.\n§ 7545(o)(3)(B)(ii)(I))). Accordingly, we conclude that the\nmeaning of “as appropriate” in paragraph (3) is ambiguous and\nturn now to whether EPA’s construction is “based on a\npermissible construction of the statute.” Chevron, 467 U.S. at\n843.\n The difficulty of squaring the Coffeyville Petitioners’\nreading of “as appropriate” with the structure and purpose of\nthe statute convinces us of the reasonableness of EPA’s\ninterpretation. As a structural matter, the RFS program\ncontains not only “annual” volumetric determinations, Conc.\nOp. 1, but also a slew of compliance provisions that are not\nannually re-determined. As a practical matter, the point of\nobligation is the foundational “compliance provision” of the\nentire renewable fuels program; EPA could not “ensure” that\napplicable volumes of renewable fuels are introduced into the\nnation’s transportation fuel supply without designating the\nparties responsible for carrying the renewable fuel standards\ninto operation. Id. § 7545(o)(2)(A)(i). To that end, in writing\nthe compliance provisions, EPA placed the renewable fuel\nobligation on the entities at the head of the United States supply\nchain for nonrenewable fuels—domestic refiners, and\nimporters of fuel refined elsewhere. See 72 Fed. Reg. at\n23,923–24. After additional consideration, EPA in 2010\nadhered to that decision. See 75 Fed. Reg. at 14,721–22\n(codified at 40 C.F.R. § 80.1406(a)(1)); see also Monroe\nEnergy, 750 F.3d at 912. No one challenged EPA’s decision\nin 2007 or 2010, and EPA declined to revisit the issue in\nresponse to comments in the 2017 annual rulemaking urging it\nto shift the 2017 point of obligation to blenders. See Response\nto Comments at 542, Coffeyville J.A. 761.\n The focus of the annual rulemakings, in contrast, is to\ntranslate the applicable volumes—as specified in paragraph\n(2), or set according to the process there described—into\n\n 50\npercentage requirements for each renewable fuel. 42 U.S.C.\n§ 7545(o)(3)(B)(ii). It would be strange indeed if Congress\nrequired EPA, as it went about its annual quantitative standard-\nsetting duties, also to rethink a choice so basic to the RFS\nprogram’s architecture. This implausibility is illuminated by\nthe fact that Congress required EPA to facilitate statutory\ncompliance through a credit trading program, which of\nnecessity requires some year-to-year stability. See id.\n§ 7545(o)(5). EPA responded by setting up the RIN system,\nwith flexibility anchored to a fixed baseline—the point of\nobligation. The compliance system is flexible in that RINs may\nbe retired in compliance demonstrations not only in the\ncompliance year during which they were generated, but also\nthroughout the ensuing compliance year, 40 C.F.R.\n§ 80.1427(a)(6), and obligated parties may carry over excess\nRINs or RIN deficits from year to year, id. § 80.1427(a)(1),\n(5)–(6); see Monroe Energy, 750 F.3d at 913.\n Annual changes to the point of obligation could cause\n“disparities in RIN-holdings,” leaving formerly obligated\nparties with “significantly more RINs, including carryover\nRINs, than they desire or can use” and newly obligated parties\nwith “lower balances than they would desire to protect\nthemselves against shortfalls in RIN availability or RIN price\nvolatility.” EPA Denial at 78, Coffeyville J.A. 850. “[A]\nchange to the point of obligation could also cause volatility in\nthe [RIN] market,” inhibiting the “ability [of] parties that\npossess excess carryover RINs to recover the cost of the RINs\nthey hold by selling them to other parties.” Id. It is not\nplausible that Congress meant EPA to consider uprooting the\nbaseline of the RFS program every year. The real stretch is that\nCongress would have imposed such an onerous and potentially\ndisruptive duty merely by use of the phrase “as appropriate.”\n The Coffeyville Petitioners’ reading is not made any more\nplausible by highlighting the likelihood that, on annual\n\n 51\nconsideration of the point of obligation, EPA would only need\nto consider recent information, and likely would stay its course.\nEven if the point of obligation in fact rarely changed, the mere\n“reconsider[ation]” of the framework would “likely cause\ndelays to the investments necessary to expand the supply of\nrenewable fuels in the United States.” See EPA Denial at 2,\nCoffeyville J.A. 774. EPA reasoned that “fuel[] industry\nparticipants [would] withhold significant investment decisions\nuntil the EPA’s final decision and the fallout from the decision\nare known.” Id. at 81–82, Coffeyville J.A. 853–54. Insisting\nthat the issue be on the regulatory agenda every year would sow\n“significant market uncertainty and potential turmoil” into the\nRFS program without offsetting benefit. Id.\n Furthermore, any requirement that an agency repeatedly\ngo through a regulatory process on an issue that promises to\ndraw a regular parade of criticism from interest groups with\nample resources is itself burdensome. See AT&T Corp. v. FCC,\n220 F.3d 607, 630–31 (D.C. Cir. 2000). This issue is no\nexception. As discussed above, EPA in 2016 and 2017\nconsidered and decided against reopening its point of\nobligation rule. In so doing, it received upwards of 18,000\ncomments and published an exhaustive, 85-page decision. See\nEPA Denial at 1–85, Coffeyville J.A. 771–857. “Given the\ntime pressure associated with its annual standards rulemaking,”\nEPA believes it would not be feasible or worthwhile to\nundertake such reconsideration annually. Id. at 7 n.10,\nCoffeyville J.A. 779. Indeed, as EPA acknowledged at oral\nargument, the agency “has been late on [its annual rules]\nbefore,” even “when [it hasn’t] taken up the point of\nobligation.” Rec. of Oral Arg. 1:25:44–52. “[A]dd[ing] on\nthe” duty to reassess the point of obligation annually, EPA tells\nus, “would be a significant burden.” Id. at 1:25:55–1:26:05.\nOur colleague doubts that EPA’s year-to-year burden would be\nappreciable, but we see no ground to question EPA’s judgment\n\n 52\nto the contrary. It seems unlikely that Congress wrote the\napplicability provision in order to heap that annual duty onto\nEPA’s plate. It seems even less likely given the absence of\nreason to think that yearly second-guessing of program\nfundamentals makes sense, or that, when and if the need for a\nprogram restructuring arises, EPA would fail to act. Indeed,\nthe statute elsewhere explicitly requires EPA to conduct\n“periodic reviews of . . . the feasibility of achieving compliance\nwith the [applicable volume] requirements.” 42 U.S.C. §\n7545(o)(11). That provision has not been briefed, but would\nappear to require EPA to reconsider the point of obligation if it\nconcluded that its placement was obstructing compliance.\n Finally, EPA’s approach coheres with basic principles of\nadministrative law. In general, the choice between various\nprocedural channels lies within the “informed discretion of the\nadministrative agency.” SEC v. Chenery Corp., 332 U.S. 194,\n203 (1947). That discretion properly includes judgments about\nthe scope of rulemakings and when to relegate ancillary issues\nto separate proceedings: “Agencies, like legislatures, do not\ngenerally resolve massive problems in one fell regulatory\nswoop.” Massachusetts, 549 U.S. at 524; see, e.g., Grp.\nAgainst Smog & Pollution, 665 F.2d at 1292 (“. . . EPA cannot\nsoundly be charged with arbitrariness merely because it chose\na separate rulemaking proceeding as the process for proposing\na revised standard in lieu of an undertaking to do so in the\nnarrower context of the opacity standard proceedings as\npetitioners requested.”). Once the agency has resolved an issue\nin a separate proceeding, it may defend against related criticism\nby “simply refer[ing]” to the other proceeding, so long as the\n“reasoning remains applicable and adequately refutes the\nchallenge.” Bechtel v. FCC, 10 F.3d 875, 878 (D.C. Cir. 1993).\nEPA reasonably reads “as appropriate,” in paragraph (3)(B), to\nleave undisturbed these background norms of broad but\nreviewable procedural discretion.\n\n 53\n Our holding today does not give EPA the limitless and\nunreviewable discretion feared by our colleague. As we have\nsaid, EPA’s determination as to whether it is “appropriate” to\nreconsider the point of obligation in the context of an annual\nvolumetric rulemaking is reviewable for abuse of discretion.\nEPA did not abuse its discretion in refusing to do so here.\nIndeed, it considered whether to change the point of obligation\nrule in a separate, contemporaneous proceeding that yielded a\nfinal order that we also have reviewed and found to be\nadequately justified. See supra Part IV.B. We do not address\nwhether it would be an abuse of discretion for EPA to refuse to\nreconsider the point of obligation—in an annual volumetric\nrulemaking or otherwise—in extreme circumstances akin to\nthose posited by our colleague’s hypothetical about the\ncontinuing study of an abolished tort. See Conc. Op. 5.\n In sum, we hold that EPA permissibly rejected the claim\nthat paragraph (3) requires the agency annually to reassess the\npoint of obligation in the renewable fuel program. Because\nEPA has no duty to reconsider the appropriateness of its point\nof obligation regulation as part of its yearly determination of\nvolumetric requirements, it was not arbitrary for EPA to treat\ncomments complaining that it obligated the wrong parties as\nappropriately assessed in a separate proceeding, and beyond\nthe scope of proceedings for the 2017 volumetric rulemaking.\nFor these reasons, we deny the Coffeyville Petitioners’ petition.\nB. Cellulosic Biofuel Projection\n\n The Coffeyville Petitioners lob a variety of challenges at\nEPA’s cellulosic biofuel projection for 2017. Many of these\npetitioners, however, raised many of the same arguments\nbefore. See ACE, 864 F.3d at 727–29 (addressing challenges\nto EPA’s 2014–16 projection). We rejected those arguments\nonce—and do so again.\n\n 54\n First, the Coffeyville Petitioners contend that “EPA’s\n[m]ethodology” for projecting cellulosic biofuel production is\ninvalid because it “[c]hronically [o]verestimates [a]ctual\n[p]roduction.” Coffeyville Br. 40. But that argument—that\nEPA has “repeatedly . . . overshot the mark,” id. at 41—doesn’t\napply to the methodology EPA actually used here, as we found\nin ACE, 864 F.3d at 727–28. As we explained when petitioners\ndeployed this same argument in challenging the 2014–16\nprojection, “the majority of EPA’s prior overestimations”\nutilized a different methodology—one that we rejected in API,\n706 F.3d at 478–81, and that the EPA accordingly abandoned.\nACE, 864 F.3d at 727. The new methodology—the one EPA\nused here—has been applied only twice before. At the time\nEPA made its final evaluation for 2017, that methodology had\n(as detailed in the table below) undershot for 2015 and overshot\nfor 2016. See Assessment of the Accuracy of Cellulosic\nBiofuel Production Projections in 2015 and 2016, EPA-HQ-\nOAR-2016-0004-3687, at 1–4 (Dec. 12, 2016), Coffeyville\nJ.A. 515–18. This is hardly a pattern of chronic overestimation.\n\n EPA Actual\n RFS\n Estimate Production EPA Record\n Compliance\n (millions of (millions of Error ** Citation\n Year\n RINs) RINs)\n Q1\n Q2 [No Data in the Record]\n 2015\n Q3\n Q4 35.00 53.36 - 34.4% J.A. 515\n Q1\n Q2 J.A.\n 2016 230.00 198.39* + 15.9%\n Q3 516–17\n\n Q4\n\n 55\n * At the time of EPA’s assessment, the agency had actual RIN\nproduction data for only the first nine months of 2016 (123.99 million\ngallons). To calculate actual production for the year, EPA extrapolated the\nlikely RIN generation for the last three months of the year based on the\nhistorical relation (a multiple of 1.8) between the average quarterly\ngeneration in the first three quarters and that of the last quarter, yielding a\nfigure of 74.39 million RINs for the last quarter. See Coffeyville J.A. 516–\n17.\n\n ** The EPA error has been calculated as the difference between the\nEPA estimate and actual production, divided by the actual production.\n\n Second, the Coffeyville Petitioners claim that EPA failed\nto generate a projection “based on” the cellulosic biofuel\nestimate provided by the Energy Information Administration\n(EIA), as required by statute. 42 U.S.C. § 7545(o)(7)(D)(i).\nThat is so, they say, because 96% of EPA’s projected volume\nwas biogas, a type of cellulosic biofuel that EIA did not include\nin its estimate. See Coffeyville Br. 43–44. The problem for\npetitioners, however, is that this closely parallels an argument\nwe rejected in ACE: “[W]e do not agree that EPA failed to\ngenerate projections ‘based on’ the [EIA’s] estimates,” even\nthough those “estimates did not contain figures for [biogas]\nproduction—production that accounts for the vast majority of\ncellulosic biofuel” (“around 90 percent”). ACE, 864 F.3d at\n724, 729. Here, as there, EPA showed sufficient “respect” for\nEIA’s estimates. Id. at 729. When limited to fuels actually\nanalyzed by EIA, EPA’s estimates were “very similar” to\nEIA’s, id.; see 2017 Rule, 81 Fed. Reg. at 89,758/1, a fact that\nthe Coffeyville Petitioners do not contest.\n\n Congress demanded no more. Nothing in the statute\nrequired EPA to, as the Coffeyville Petitioners insist, “work[]\nwith the EIA to develop information” about biogas.\nCoffeyville Br. 44. “[T]he Administrator of the Energy\nInformation Administration shall provide . . . an estimate,” 42\nU.S.C. § 7545(o)(3)(A), and EPA shall “respect” it, API, 706\n\n 56\nF.3d at 478. That’s it. In showing such respect, EPA, of\ncourse, must “understand how EIA derived” its estimate.\nCoffeyville Reply Br. 24. But, contrary to the Coffeyville\nPetitioners’ contention, EPA did just that. The agency\nidentified the types of cellulosic biofuels that EIA considered\nand then, to test the integrity of its projection, conducted an\napples-to-apples comparison, “limiting the scope of [its]\nprojection to the companies assessed by EIA.” See 2017 Rule,\n81 Fed. Reg. at 89,757–58. Nothing more was required.\n\n Third, the Coffeyville Petitioners object that EPA relied on\n“information from the [biogas] industry”—an industry “with a\ndirect financial interest in the outcome of the rule.” Coffeyville\nBr. 45. The Petitioners characterize this as “reliance on\nundisclosed information.” Id. But EPA did disclose the\ninformation from the biogas industry—and that the information\ncame from that industry; it just did so in the aggregate. See\nOctober 2016 Assessment of Cellulosic Biofuel Production\nfrom Biogas (2017), EPA-HQ-OAR-2016-0004-3711, at 2–6\n(Dec. 12, 2016), Coffeyville J.A. 536–40. All the agency\nwithheld was company-specific information, claiming that it\nhad to withhold such data as confidential business information,\nsee id. at 7, Coffeyville J.A. 541; see also 40 C.F.R. § 2.211(b),\na claim that petitioners never even attempt to rebut, see Masias\nv. EPA, 906 F.3d 1069, 1077 (D.C. Cir. 2018) (“It is not enough\nmerely to mention a possible argument in the most skeletal\nway, leaving the court to do counsel’s work . . . .” (quoting\nSchneider v. Kissinger, 412 F.3d 190, 200 n.1 (D.C. Cir.\n2005))).\n\n As for the implication of bias, we have previously upheld\nEPA’s reliance on “biofuel producers’ own forecasts.” ACE,\n864 F.3d at 728; see also API, 706 F.3d at 478 (recognizing\nthat producers are an “almost inevitable source of\ninformation”). Here, as in ACE, EPA did not “blindly adopt[]\n\n 57\nthe facilities’ own forecasts”; it “performed its own\ninvestigation.” 864 F.3d at 728; see October 2016 Assessment\nof Cellulosic Biofuel Production from Biogas, supra, at 4,\nCoffeyville J.A. 538 (“To verify the reasonableness of these\nprojections, EPA compared the projected volume from each\nregistered facility to the registered capacity of that facility.”).\nPetitioners point to no unreasonable step by EPA in its efforts\nto address “the uncertainty and unreliability identified by the\n[Coffeyville] Petitioners.” ACE, 864 F.3d at 728.\n\n Fourth, the Coffeyville Petitioners protest EPA’s reliance\non “facilities’ actual production in prior years” as a floor for\nprojecting future cellulosic biofuel production. Coffeyville Br.\n46. This was error, they say, because some companies might\ncease production. Perhaps so. But, as we said in ACE,\nalthough unforeseen issues “could prevent a producer from\nmeeting” its prior year’s production, it was “reasonable” for\nEPA to expect, as a general matter, “that a company’s output\nwould grow year-over-year as the company gained\nexperience.” 864 F.3d at 728. This seems especially true in an\nindustry with the government’s wind surging at its back. And\neven were EPA’s assumption not true for each company, any\none facility’s shortfall, EPA explained, could be “off-set” by\nnew facilities coming online or existing facilities exceeding the\nhigh end of their projected production range. See Renewable\nFuel Standard Program—Standards for 2017 and Biomass-\nBased Diesel Volume for 2018: Response to Comments, EPA-\nHQ-OAR-2016-0004-3753, at 444 (Dec. 12, 2016),\nCoffeyville J.A. 707. This explanation fulfills EPA’s “duty to\narticulate a ‘reasonable and reasonably explained’ approach to\nsetting the low end of the production ranges.” ACE, 864 F.3d\nat 729 (quoting Comtys. for a Better Env’t v. EPA, 748 F.3d\n333, 335 (D.C. Cir. 2014)).\n\n 58\n Fifth, the Coffeyville Petitioners complain that EPA\nshould have based its cellulosic biofuel projections on “actual”\nprior production. Coffeyville Br. 47. But this backward\nlooking approach would have, in EPA’s view, “ignore[d] the\npotential for facilities . . . to increase their fuel production\nrates” and would have been “inappropriately conservative” in\nlight of the “year-over-year increases” that EPA had observed\n“in recent years.” 2017 Rule, 81 Fed. Reg. at 89,761/1. We\ncannot say that in rejecting such an approach EPA violated “its\nduty to take a ‘neutral aim at accuracy.’” ACE, 864 F.3d at 727\n(quoting API, 706 F.3d at 476).\n\n For these reasons, we reject the Coffeyville Petitioners’\nchallenges to EPA’s cellulosic biofuel projection for 2017. See\nACE, 864 F.3d at 729.\n\nC. Cellulosic Waiver\n\n The Coffeyville Petitioners also challenge EPA’s decision\nto use less than all of its discretionary cellulosic waiver\nauthority to lower the 2017 requirements for advanced biofuel\nand total renewable fuel. Having reduced the 2017 cellulosic\nbiofuel requirement by 5.189 billion gallons, EPA had\nauthority to reduce the advanced biofuel and total renewable\nfuel requirements “by the same or a lesser volume.” 42 U.S.C.\n§ 7545(o)(7)(D)(i). To decide by how much to reduce these\nstatutory requirements, EPA first determines what reduction in\nthe advanced biofuel requirement will yield a “reasonably\nattainable” volume, and it then mechanically applies an\nequivalent reduction to the total renewable fuel volume. 2017\nRule, 81 Fed. Reg. at 89,752–53. Petitioners do not directly\nchallenge this methodology. Instead, they argue that EPA\napplied it arbitrarily in deciding to waive only 4.719 billion\ngallons of the advanced biofuel volume for 2017, rather than\nthe maximum available waiver of 5.189 billion gallons. We\n\n 59\nagain reject some of their arguments as foreclosed by precedent\nand others on their own terms.\n\n First, the Coffeyville Petitioners argue that EPA sought to\njustify its 2017 advanced biofuel volume in part by making an\nimpermissible comparison to the statutory volume set by\nCongress for 2022. In response to a comment expressing\nconcerns about utilization of non-cellulosic advanced biofuels\n(which could be food-based) and possible adverse effects on\nfood availability, EPA noted that its “reasonably attainable”\nnon-cellulosic advanced biofuel volume for 2017\n(approximately 4 billion gallons) was “somewhat higher than\nthe level envisioned in the statute for 2017” (3.5 billion), “but\nwell below the level of such fuels Congress expected would be\nused by 2022” (5 billion). Response to Comments at 214,\nCoffeyville J.A. 689. According to petitioners, “by comparing\n2017 volumes with 2022 statutory targets, EPA departed from\nCongress’s intent.” Coffeyville Br. 50.\n\n However, nothing in the statute forbids EPA from taking\naccount of future statutory volumes in this way. Although\nCongress specified presumptively applicable volumes for\ncertain years, it also provided waiver authority to depart from\nthose volumes. Indeed, the discretionary waiver provision\nnecessarily empowers EPA to depart upward from the statutory\nlevel of non-cellulosic advanced biofuel for a given year:\nreducing the advanced biofuel volume by less than the\nreduction in cellulosic biofuel, as section 7545(o)(7)(D)(i)\npermits, is mathematically equivalent to increasing the volume\nof non-cellulosic advanced biofuels, to “partially backfill for\nmissing cellulosic biofuel.” 2017 Rule, 81 Fed. Reg. at\n89,763/1. As we have noted, the cellulosic waiver provision\n“grants EPA ‘broad discretion’ to consider a variety of factors”\nin exercising this authority to depart from the presumptive\nstatutory volumes. ACE, 864 F.3d at 733 (quoting Monroe\n\n 60\nEnergy, 750 F.3d at 915). In this case, while deflecting a\ncomment about food availability, EPA observed that its non-\ncellulosic advanced biofuel volume for 2017—while higher\nthan the statutory volume envisioned for that year—was lower\nthan the presumptive statutory volume for the near future. And\nit then reasonably concluded that this “somewhat higher\ninterim volume reflect[ed] [its] assessment that it is appropriate\nto allow non-cellulosic advanced biofuels to partially backfill\nfor missing cellulosic volumes in light of the associated\n[greenhouse gas] and energy security benefits.” Response to\nComments at 214, Coffeyville J.A. 689.\n\n Second, the Coffeyville Petitioners argue that EPA failed\nto explain its estimate of reasonably attainable 2017 imports of\nsugarcane ethanol, a type of non-cellulosic advanced biofuel.\nSugarcane ethanol imports have varied greatly from year to\nyear, reaching a high of 681 million gallons in 2006 but falling\nto 64 million gallons in 2014 and 89 million gallons in 2015.\nSee 2017 Rule, 81 Fed. Reg. at 89,764. At the time of the 2017\nRule, EPA expected only 76 million gallons to be imported in\n2016, but it nonetheless adhered to its proposed estimate of 200\nmillion gallons for 2017—an estimate originally based on\nEPA’s judgment that circumstances in 2017 were “not . . .\nsignificantly different” from circumstances in 2016, for which\nEPA had also projected 200 million gallons. Id. at 89,763/3.\nEPA acknowledged the “recent low import levels,” but also\ncited “the difficulty in precisely identifying the reasons” for the\nhistorical “high variability,” given “uncertainty” as to market\nfactors including “ongoing growth in gasoline demand in\nBrazil, and competing world demand for sugar.” Id. at 89,764–\n65. The agency accordingly reaffirmed that 200 million\ngallons “reflects a reasonable intermediate point between the\nlower levels imported recently and the considerably higher\nlevels that have been achieved in earlier years.” Id. at 89,765/2.\n\n 61\n There is some force to petitioners’ objection to EPA’s\nadherence to an estimate well over double the actual imports in\nthe three preceding years. However, our review is “particularly\ndeferential in matters implicating predictive judgments.”\nMusic Choice v. Copyright Royalty Bd., 774 F.3d 1000, 1015\n(D.C. Cir. 2014) (quoting Rural Cellular Ass’n v. FCC, 588\nF.3d 1095, 1105 (D.C. Cir. 2009)). We accordingly upheld\nEPA’s identical 2016 sugarcane ethanol estimate as\n“reasonable and reasonably explained” in ACE, 864 F.3d at 736\n(quotation marks omitted). In that case, we held that EPA\nreasonably “concluded that ‘a somewhat lower level of imports\nwill occur than the historic average’ of 300 million,” based on\na similar analysis of market factors. Id. (quoting 2014–16\nRule, 80 Fed. Reg. at 77,478/2). Here, we cannot say that one\nmore year of low imports made it arbitrary for EPA to adhere\nto that same projection for 2017.\n\n Third, the Coffeyville Petitioners object to EPA’s analysis\nof supply and demand for regular gasoline (E0) and gasoline\nwith added ethanol (E15 and E85). However, this analysis\nplayed no role in EPA’s exercise of its discretionary cellulosic\nwaiver authority under section 7545(o)(7)(D)(i). As noted\nabove, EPA’s exercise of that authority rested entirely on its\ndetermination of reasonably attainable advanced biofuel\nvolumes. See 2017 Rule, 81 Fed. Reg. at 89,773–74. The\ndisputed analysis of E0, E15, and E85 supported EPA’s\nseparate decision not to invoke its “general waiver” authority,\nunder section 7545(o)(7)(A)(ii), based on “inadequate\ndomestic supply.” See generally ACE, 864 F.3d at 705–13.\nBut in their opening brief, petitioners failed to challenge EPA’s\ndecision not to invoke that separate waiver provision for 2017.\nAlthough their reply brief gestures at this point, “an argument\nfirst made in a reply brief is forfeited.” Bartko v. SEC, 845\nF.3d 1217, 1225 n.7 (D.C. Cir. 2017).\n\n 62\n Finally, the Coffeyville Petitioners take issue with EPA’s\nresponse to various comments. We have considered these\narguments and find them to be without merit.\n\n For these reasons, we reject the Coffeyville Petitioners’\nchallenges to EPA’s exercise of its discretionary cellulosic\nwaiver authority to reduce advanced biofuel and total\nrenewable fuel volumes for 2017.\n\n VI. 2018 Volume for Biomass-Based Diesel\n\n Since 2012, EPA, acting in coordination with the\nSecretaries of Energy and Agriculture, has calculated the\nannual applicable volume (also known as the “volume\nrequirement”) for biomass-based diesel based on a holistic,\nbackward- and forward-looking consideration of relevant\nfactors. In particular, it has set the volume requirement “based\non a review of the implementation of the program during\ncalendar years specified in the tables, and an analysis of” six\nstatutorily enumerated factors: (1) “the impact of the\nproduction and use of renewable fuels on the environment”; (2)\n“the impact of renewable fuels on the energy security of the\nUnited States”; (3) “the expected annual rate of future\ncommercial production of renewable fuels, including advanced\nbiofuels in each category (cellulosic biofuel and biomass-based\ndiesel)”; (4) “the impact of renewable fuels on the\ninfrastructure of the United States”; (5) “the impact of the use\nof renewable fuels on the cost to consumers of transportation\nfuel and on the cost to transport goods”; and (6) “the impact of\nthe use of renewable fuels on other factors, including job\ncreation, the price and supply of agricultural commodities,\nrural economic development, and food prices.” 42 U.S.C.\n§ 7545(o)(2)(B)(ii)(I)–(VI).\n\n EPA set the 2018 applicable volume for biomass-based\ndiesel at 2.1 billion gallons, up from 2.0 billion gallons in 2017,\n\n 63\nand 1.1 billion gallons above a statutory minimum that\nCongress set to plateau at 1 billion gallons as of 2012. 2017\nRule, 81 Fed. Reg. at 89,798/1; see 42 U.S.C.\n§ 7545(o)(2)(B)(i)(IV), (v). NBB had asked EPA to set the\nbiomass-based diesel volume at 2.5 billion gallons, and now\nchallenges the volume EPA set as arbitrary and capricious and\ncontrary to the Clean Air Act.\n\nA. NBB’s Standing\n\n Before considering the merits of NBB’s claims, we must\nsatisfy ourselves that NBB has standing to assert them.\nRespondent-Intervenors, the American Fuel & Petrochemical\nManufacturers and the American Petroleum Institute, contend\nthat NBB lacks standing because, they say, it has not shown\nthat the 2017 Rule inflicted a cognizable injury on any of its\nmembers.\n\n NBB has associational standing here for the same reasons\nwe held it did in National Biodiesel Board v. EPA, 843 F.3d\n1010, 1015 (D.C. Cir. 2016) (NBB v. EPA), where EPA’s\nactions “incentivize[d] . . . compet[ition] with [NBB’s\nmembers’] domestic production.” Here, too, NBB’s members\n“compete with” the other industry players EPA’s rule is\ndesigned to affect. Id. at 1016. Recall that biomass-based\ndiesel is a nested subset of advanced and total renewable fuels,\nsuch that NBB’s members get (1) a market for compelled\nbuyers of the specified volume of biomass-based diesel, for\nwhich they are the exclusive suppliers, plus (2) a market for\ncompelled buyers of advanced and other renewable fuels\nalongside a broad array of competing suppliers. See supra at\n6–8, 11. The 2017 Rule preamble explains that biomass-based\ndiesel “compet[es] for research and development dollars with\nother types of advanced biofuels,” and that, “[b]y establishing\n[the biomass-based diesel] volume requirement[] at [a] level[]\n\n 64\nlower than . . . the expected production of [biomass-based\ndiesel],” EPA was “creating the potential for some competition\nbetween [biomass-based diesel] and other advanced biofuels to\nsatisfy the advanced biofuel” applicable volume and providing\n“incentives for the continued development of” those\ncompetitors’ fuels. 81 Fed. Reg. at 89,797; see also EPA\nCoffeyville Br. 24 (“Above 2.1 billion gallons, biomass-based\ndiesel will have to compete with other types of advanced\nbiofuel.”). Such competition will likely “temper to some extent\n[biomass-based diesel] prices.” Final Statutory Factors\nAssessment for the 2018 Biomass Based Diesel (BBD)\nApplicable Volume, EPA-HQ-OAR-2016-0004-3708, at 10\n(Dec. 12, 2016) (Supplemental Assessment), Coffeyville J.A.\n533. That is a cognizable injury to NBB’s members. See NBB\nv. EPA, 843 F.3d at 1015–16; see also Delta Constr. Co. v.\nEPA, 783 F.3d 1291, 1299 (D.C. Cir. 2015) (per curiam).\n\n Though NBB failed to identify any of its members—\nordinarily a prerequisite for organizations alleging\nassociational standing, see Summers v. Earth Island Inst., 555\nU.S. 488, 497–98 (2009)—that omission is not fatal here\nbecause NBB’s members comprise “the entire biomass-based\ndiesel category of the Renewable Fuel Standard[s]” and\nrepresent no other interests. Coffeyville J.A. 134. Consistent\nwith “the real purpose of the [standing] inquiry—that is, for the\ncourt to be satisfied that the requisite injury really has occurred\nor will occur in the future to members of the organization[],”\nPub. Citizen v. FTC, 869 F.2d 1541, 1552 (D.C. Cir. 1989),\nthere is no need to identify injured members when “all the\nmembers of the organization are affected by the challenged\nactivity,” Summers, 555 U.S. at 499 (citing NAACP v. Ala. ex\nrel. Patterson, 357 U.S. 449, 459 (1958)). Because EPA’s rule\nsubjects the biomass-based diesel industry to increased\ncompetition, with anticipated pricing effects, NBB “meet[s] the\n\n 65\nconstitutional prerequisites of injury, causation, and\nredressability.” NBB v. EPA, 843 F.3d at 1015.\n\nB. Merits of NBB’s Challenges\n\n NBB advances two challenges to the applicable volume\nEPA set for biomass-based diesel: First, that EPA erred in\nconsidering the interaction of biomass-based diesel with the\nyet-to-be established 2018 advanced biofuel applicable\nvolume, and second, that EPA’s consideration of the six\nstatutory factors was arbitrary and capricious and contrary to\nlaw. We reject both claims.\n\n First, EPA reasonably chose a 2018 biomass-based diesel\napplicable volume that would “maintain[] support for growth\nin [biomass-based diesel] volumes” while also encouraging the\n“development of other advanced biofuels.” 2017 Rule, 81 Fed.\nReg. at 89,798/1. Congress directed EPA to consider the\nlessons learned from its retrospective “review” of the program,\napply them in its prospective “analysis of” the six statutory\nfactors, and set a biomass-based diesel volume that will apply\nfourteen months in the future. See 42 U.S.C.\n§ 7545(o)(2)(B)(ii).\n\n EPA’s approach is consistent with the structure and\npurposes of the statute. Congress set a minimum applicable\nvolume for biomass-based diesel of one billion gallons for each\nyear from 2012 forward, id. § 7545(o)(2)(B)(i)(IV), (v), while\nspecifying statutory minimum volumes for the advanced\nbiofuel category containing biomass-based diesel that grow\nyear by year to 21 billion gallons by 2022, id.\n§ 7545(o)(2)(B)(i)(II), (iii). EPA reasonably concluded that,\nby nesting biomass-based diesel together with cellulosic (and\nother unspecified) biofuels within the advanced biofuel\ncategory, and specifically charting a higher, steeper, and longer\ninitial growth curve for advanced biofuel, Congress anticipated\n\n 66\nthat production of other types of advanced biofuels could step\nup to help meet the advanced biofuel volume requirement. See\n2017 Rule, 81 Fed. Reg. at 89,797/1. EPA also reasonably\nconcluded that increasing fuel diversity serves one of\nCongress’s primary goals in establishing the Renewable Fuel\nStandards program: improving the nation’s “energy\nindependence and security.” See Pub. L. No. 110-140,\npreamble; see also 2017 Rule, 81 Fed. Reg. at 89,798/3. EPA\nalso reasonably anticipated that enhanced competition in the\nadvanced biofuels market would help “temper to some extent\n[biomass-based diesel] prices,” Supplemental Assessment 10,\nCoffeyville J.A. 533, thereby ameliorating Congress’s concern\nthat, with a too-high target volume, the “price of biomass-based\ndiesel fuel” would “increase significantly,” 42 U.S.C.\n§ 7545(o)(7)(E)(ii). And fuel diversity may produce\nenvironmental benefits insofar as certain advanced biofuels,\nsuch as ethanol from food waste, will “likely have significantly\nlower impacts on wetlands, ecosystems, and wildlife habitats”\nthan would greater reliance on biomass-based diesel.\nSupplemental Assessment 6, Coffeyville J.A. 529.\n\n NBB’s arguments to the contrary turn on reading the\nstatutory directive that EPA “review . . . the implementation of\nthe program during calendar years specified in the tables,” id.\n§ 7545(o)(2)(B)(ii), to confine EPA’s consideration to\nbiomass-based diesel’s statutory volumes and actual\nperformance, and to prevent EPA from considering other fuel\ncategories or future years. In particular, NBB takes issue with\nEPA’s consideration of the not-yet-finalized 2018 advanced\nbiofuel applicable volume, which NBB contends led EPA to\nset the biomass-based diesel volume too low.\n\n NBB’s objections are not supported by the text or purpose\nof the statute. Assuming NBB is right that EPA’s “review of\nthe implementation of the program” consists of a retrospective\n\n 67\nassessment, the agency must also conduct “an analysis of” six\nstatutory factors. Id. § 7545(o)(2)(B)(ii). And those factors\nplainly require a prospective assessment—an assessment that\nwould likely miss “important aspects of the problem,” State\nFarm, 463 U.S. at 43, if it ignored the interaction, now and in\nthe future, of the requirements for all the categories of\nrenewable fuels. See, e.g., 42 U.S.C. § 7545(o)(2)(B)(ii)(I)\n(requiring an “analysis of” the “impact of the production and\nuse of renewable fuels on,” among other things, “the\nenvironment”). Though EPA set the biomass-based diesel\nrequirement lower than NBB wished, Congress did not intend\nto incentivize growth of biomass-based diesel “at all costs.”\nACE, 864 F.3d at 714 (quoting Am. Express Co. v. Italian\nColors Rest., 570 U.S. 228, 234 (2013)).\n\n NBB objects that setting the 2018 biomass-based diesel\napplicable volume below expected production might lead to a\ndepressed advanced biofuel volume for 2018. But the agency\nspecifically anticipated “that the 2018 advanced biofuel\nrequirement will be larger than the 2017 advanced biofuel\nvolume requirement.” 2017 Rule, 81 Fed. Reg. at 89,798/1.\nEPA has never set the biomass-based diesel applicable volume\nat the “maximum potential production” level, id. at 89,799/3,\nyet the “growing supply of” biomass-based diesel has\nconsistently “allowed EPA to establish higher advanced\nbiofuel” applicable volumes, id. at 89,797/3. EPA opted for\n“allowing room within the advanced biofuel volume\nrequirement for the participation of non-[biomass-based diesel]\nadvanced fuels” as a reasonable way “to encourage the\ndevelopment and production of a variety of advanced biofuels\nover the long term without reducing the incentive for [biomass-\nbased diesel] beyond the [biomass-based diesel applicable\nvolume] in 2018.” Id. at 89,797–98.\n\n 68\n Second, in setting the 2018 biomass-based diesel\napplicable volume, EPA reasonably compared the advantages\nand disadvantages of biomass-based diesel to those of other\nfuels. NBB contends that the statute confines EPA to assessing\nadvantages of biomass-based diesel over petroleum, not\nconsidering other renewable fuels, and that the agency failed to\n“meaningfully” consider the six factors. NBB Br. 9–10. Both\narguments miss the mark.\n\n NBB suggests that, because the statute “was intended to\n‘increase the production of clean renewable fuels’ as a\nsubstitute for petroleum fuel,” id. at 21 (quoting Pub. L. No.\n110-140, preamble), the only relevant comparison is to\npetroleum, not to other categories of renewable fuel. But NBB\nidentifies nothing in section 7545(o)(2)(B)(ii) or any other\nsection that requires EPA to assess the performance of a\nparticular renewable solely by reference to petroleum fuel. Its\nanalysis would require us to read the term “renewable fuels”\nused throughout section 7545(o)(2)(B)(ii) to refer to the single\nrenewable fuel being analyzed, even though the statutory\ndefinition of “renewable fuel” includes all types of renewables.\nSee 42 U.S.C. § 7545(o)(1)(J). And if EPA could compare the\nbenefits of each specific fuel only to petroleum, it might be\nunable to set rational applicable volumes for each specified\ncategory of renewable fuel after 2022, when the statute no\nlonger sets any specific volumes. See id. § 7545(o)(2)(B)(iii)–\n(v). EPA could easily conclude, for example, that each\nrenewable fuel had a lower “impact . . . on the environment”\nthan petroleum fuel, see id. § 7545(o)(2)(B)(ii)(I), but, no\nmatter their differing merits in serving the statute’s goals, the\nagency would be barred from making relative judgments\namong renewable fuel categories.\n\n NBB also argues that EPA failed to give meaningful\nconsideration to the six statutory factors, and instead “pre-\n\n 69\ndetermined the outcome,” NBB Br. 20, but the record shows\notherwise. EPA considered in detail how setting the biomass-\nbased diesel applicable volume at a level higher or lower than\n2.1 billion gallons would affect the six statutory factors. See\n2017 Rule, 81 Fed. Reg. at 89,798–99. EPA further elaborated\nits analysis of the factors in an 11-page supplemental\nmemorandum evaluating effects of its proposed biomass-based\ndiesel volume on renewable fuel production rates, the\nenvironment, and the economy. See Supplemental Assessment\n1–11, Coffeyville J.A. 524–34. EPA concluded that, over the\nlong term, “[a] variety of different types of advanced biofuels,\nrather than a single type such as [biomass-based diesel], would\npositively impact energy security . . . and increase the\nlikelihood of the development of lower cost advanced biofuels\nthat meet the same [greenhouse gas] reduction threshold as\n[biomass-based diesel].” Supplemental Assessment 3,\nCoffeyville J.A. 526. EPA thus concluded that the statutory\nfactors supported its biomass-based diesel applicable volume.\nSee 2017 Rule, 81 Fed. Reg. at 89,798/3.\n\n At bottom, NBB’s objections rest on a policy\ndisagreement: NBB urges that, instead of setting a level that\nwould support continued investment in the biomass-based\ndiesel industry while also encouraging producers of other types\nof advanced biofuel to compete to satisfy the 2018 advanced\nbiofuel applicable volume at lower cost, EPA should have\nreserved to biomass-based diesel alone a volume nearer to that\nindustry’s maximum production potential. But NBB’s\nproposed “simple solution”—that EPA should have “set[] a\nmeaningful [biomass-based diesel] volume” while planning to\n“increas[e] the 2018 advanced-biofuel volume to provide room\nfor the production of other advanced biofuels when it set that\nvolume a year later,” NBB Br. 23—describes what EPA\nactually did. A mere disagreement with the particular\ncalibration of a line drawn in the exercise of an agency’s\n\n 70\nreasonable judgment is no basis to invalidate a rule. Therefore,\nwe deny NBB’s petition.\n\n VII. Conclusion\n\n For these reasons, the petitions for review are denied.\n\n So ordered.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Coffeyville Resources Refining & Marketing, LLC v. EPA\\nCourt: U.S. Court of Appeals for the District of Columbia Circuit\\nDate decided: 2019-08-30\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: Yes\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n FOR THE DISTRICT OF COLUMBIA CIRCUIT\\n\\n\\n\\nArgued October 5, 2018 Decided August 30, 2019\\n\\n No. 16-1052\\n\\n ALON REFINING KROTZ SPRINGS, INC.,\\n PETITIONER\\n\\n v.\\n\\n ENVIRONMENTAL PROTECTION AGENCY,\\n RESPONDENT\\n\\n MONROE ENERGY, LLC, ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 16-1055, 17-1255, 17-1259, 18-1021,\\n 18-1024, 18-1025, 18-1029\\n\\n\\n On Petitions for Review of Agency Action of the\\n United States Environmental Protection Agency\\n\\n\\n Samara L. Kline argued the cause for petitioners. With her\\non the briefs were Evan A. Young, Megan H. Berge, Lisa M.\\nJaeger, Brittany M. Pemberton, Clara Poffenberger, Richard\\nS. Moskowitz, Robert J. Meyers, Thomas A. Lorenzen,\\nElizabeth B. Dawson, Warren R. Neufeld, LeAnn M. Johnson,\\nand Jonathan G. Hardin. Albert M. Ferlo Jr. and Krista\\nHughes entered appearances.\\n\\n 2\\n Meghan E. Greenfield, Trial Attorney, U.S. Department of\\nJustice, argued the cause for respondent. With her on the brief\\nwas Jeffrey H. Wood, Acting Assistant Attorney General.\\nDaniel R. Dertke, Attorney, entered an appearance.\\n\\n Robert A. Long Jr. argued the cause for intervenors\\nAmerican Petroleum Institute and Growth Energy. With him\\non the brief were Kevin King, Seth P. Waxman, David M. Lehn,\\nSaurabh Sanghvi, and Claire Chung. Stacy R. Linden entered\\nan appearance.\\n\\n Shannen W. Coffin and Linda C. Bailey were on the brief\\nfor amici curiae NACS, et al. in support of respondent EPA.\\n\\n 3\\n\\n No. 17-1044\\n\\nCOFFEYVILLE RESOURCES REFINING & MARKETING, LLC AND\\n WYNNEWOOD REFINING COMPANY, LLC,\\n PETITIONERS\\n\\n v.\\n\\n ENVIRONMENTAL PROTECTION AGENCY,\\n RESPONDENT\\n\\n ALON REFINING KROTZ SPRINGS, INC., ET AL.,\\n INTERVENORS\\n\\n\\n Consolidated with 17-1045, 17-1047, 17-1049, 17-1051,\\n 17-1052\\n\\n\\n On Petitions for Review of Action of the\\n United States Environmental Protection Agency\\n\\n\\n Brian Killian argued the cause for petitioner The National\\nBiodiesel Board. With him on the briefs was Douglas A.\\nHastings.\\n\\n Samara L. Kline and Thomas Allen Lorenzen, argued the\\ncauses for Obligated Party Petitioners. With them on the briefs\\nwere Evan A. Young, Lisa M. Jaeger, Brittany M. Pemberton,\\nClara Poffenberger, Richard S. Moskowtiz, Robert J. Meyers,\\nElizabeth B. Dawson, David W. DeBruin, Thomas J. Perrelli,\\nMatthew E. Price, LeAnn M. Johnson, and Jonathan G.\\n\\n 4\\nHardin. David Y. Chung, Eric D. Miller, and Albert M. Ferlo\\nJr. entered appearances.\\n\\n Patrick R. Jacobi and Samara M. Spence, Attorneys, U.S.\\nDepartment of Justice, argued the causes for respondent. With\\nthem on the brief was Jeffrey H. Wood, Acting Assistant\\nAttorney General.\\n\\n Thomas Allen Lorenzen argued the cause for intervenors\\nAmerican Fuel & Petrochemical Manufacturers and American\\nPetroleum Institute in support of respondent regarding\\nBiomass-Based Diesel Issues. With him on the brief were\\nRobert J. Meyers, Elizabeth B. Dawson, Richard S. Moskowitz,\\nRobert A. Long, Jr., and Kevin King. Stacy R. Linden entered\\nan appearance.\\n\\n Robert A. Long, Jr., Kevin King, Bryan M. Killian,\\nDouglas A. Hastings, Seth P. Waxman, David M. Lehn,\\nSaurabh Sanghvi, and Claire H. Chung were on the brief for\\nintervenors Growth Energy, et al. in support of respondent.\\nEric D. Miller entered an appearance.\\n\\n Before: PILLARD and KATSAS, Circuit Judges, and\\nWILLIAMS, Senior Circuit Judge.\\n\\n Opinion for the Court filed PER CURIAM.\\n\\n Opinion concurring in part and concurring in the judgment\\nfiled by Senior Circuit Judge WILLIAMS.\\n\\n TABLE OF CONTENTS\\n\\nI. Introduction ..................................................................... 6\\nII. Background ..................................................................... 7\\n\\n 5\\n A. Legal Background ........................................................ 7\\n B. Procedural Background ............................................. 12\\n 1. 2007, 2010, and 2017 Point of Obligation\\n Proceedings ............................................................ 12\\n 2. 2017 Annual Volumetric Proceedings ................... 15\\nIII. Standard of Review .................................................... 17\\nIV. 2010 Point of Obligation Rule ................................... 17\\n A. Jurisdiction................................................................. 17\\n 1. Final Agency Action Under Section 7607(b)(1).... 19\\n 2. After-Arising Grounds Under Section\\n 7607(b)(1) .............................................................. 28\\n 3. Mandatory Reconsideration Under Section\\n 7607(d)(7)(B) ......................................................... 29\\n B. Merits of Challenges to EPA’s Refusal to Revise\\n the 2010 Point of Obligation Rule ............................. 32\\nV. 2017 Annual Volumetric Rule ...................................... 41\\n A. Point of Obligation .................................................... 42\\n 1. Jurisdiction ............................................................. 42\\n 2. Merits ..................................................................... 43\\n B. Cellulosic Biofuel Projection..................................... 53\\n C. Cellulosic Waiver ...................................................... 58\\nVI. 2018 Volume for Biomass-Based Diesel ................... 62\\n A. NBB’s Standing ......................................................... 63\\n B. Merits of NBB’s Challenges ...................................... 65\\nVII. Conclusion ................................................................. 70\\n\\n 6\\n PER CURIAM:\\n\\n\\n I. Introduction\\n\\n The Clean Air Act requires EPA to publish “renewable\\nfuel standards,” ultimately expressed as “applicable\\npercentages,” each year to ensure that the total supply of\\ntransportation fuel sold or imported into the United States\\ncontains specified proportions of each of four categories of\\nrenewable fuels. Congress intended the Renewable Fuel\\nStandards (RFS) program to “move the United States toward\\ngreater energy independence and security” and “increase the\\nproduction of clean renewable fuels.” See Energy\\nIndependence and Security Act of 2007 (EISA), Pub. L. No.\\n110-140, preamble, 121 Stat. 1492 (2007) (codified at 42\\nU.S.C. § 7545(o)).\\n\\n In these related cases, Alon Refining Krotz Springs,\\ntogether with other petroleum refineries and their trade\\nassociations—the “Alon Petitioners”—seek review of EPA’s\\ndecision not to revise its 2010 point of obligation regulation\\nrequiring refineries and importers, but not blenders, to bear the\\ndirect compliance obligation of ensuring that transportation\\nfuels sold or introduced into the U.S. market include the\\nrequisite percentages of renewables. Coffeyville Resources\\nRefining & Marketing and another group of refineries and trade\\nassociations—the “Coffeyville Petitioners”—challenge EPA’s\\nrefusal to reassess the appropriateness of the point of obligation\\nin the context of its 2017 annual volumetric rule, which set the\\n2017 applicable percentages for all four categories of\\nrenewable fuel and the 2018 applicable volume for one subset\\nof such fuel, biomass-based diesel. See 81 Fed. Reg. 89,746\\n(Dec. 12, 2016) (2017 Rule). The Coffeyville Petitioners also\\ncontend that EPA arbitrarily set the 2017 percentage standards\\n\\n 7\\ntoo high. The National Biodiesel Board (NBB)—a biomass-\\nbased diesel industry trade association—separately contends\\nthat EPA set the 2018 applicable volume for biomass-based\\ndiesel too low. Various trade associations representing\\nrefineries and producers of renewable fuels have intervened in\\nsupport of EPA. For the reasons that follow, we deny each of\\nthe petitions for review, many of which recycle arguments\\nraised and rejected in prior challenges.\\n\\n II. Background\\n\\nA. Legal Background\\n\\n Congress established the RFS program in 2005 as part of\\nthe Energy Policy Act, Pub. L. No. 109-58, 119 Stat. 594\\n(2005) (as amended at 42 U.S.C. § 7545(o)). The statute\\nmandates the gradual introduction of four nested categories of\\nrenewable fuels into the United States’ supply of gasoline,\\ndiesel, and other transportation fuels. See 42 U.S.C.\\n§ 7545(o)(2)(B). These categories include: (1) total renewable\\nfuel; (2) advanced biofuel; (3) cellulosic biofuel; and (4)\\nbiomass-based diesel. Id. § 7545(o)(2)(A)(i), (B). The\\numbrella category, total renewable fuel, covers the three other\\ncategories plus any conventional renewable fuels, such as corn-\\nbased ethanol. See id. § 7545(o)(1)(F), (J), (2)(A)(i). The\\nadvanced biofuel subset includes any renewable fuel (except\\nethanol from cornstarch) that has at least 50% lower lifecycle\\ngreenhouse gas emissions than fossil fuels. Id.\\n§ 7545(o)(1)(B). The statute further specifies two\\nnonexclusive subsets of advanced biofuels: cellulosic biofuel\\n(a renewable fuel derived from cellulose materials such as corn\\nstalks and husks) and biomass-based diesel (a diesel fuel\\nsubstitute made from feedstocks such as animal fats). Id.\\n§ 7545(o)(1)(B), (D), (E); EPA Coffeyville Br. 4–5. The\\n\\n 8\\nfollowing figure depicts the nested nature of the four fuel\\ncategories.\\n\\n\\n\\n\\n Source: Coffeyville Br. 11.\\n Four tables in the statute set forth gradually increasing\\nannual “applicable volume” requirements for each category of\\nrenewable fuel. See 42 U.S.C. § 7545(o)(2)(B)(i). The statute\\nsets applicable volumes for biomass-based diesel through\\n2012, id. § 7545(o)(2)(B)(i)(IV), and applicable volumes for\\nthe other three categories through 2022, id.\\n§ 7545(o)(2)(B)(i)(I)–(III). Under those tables, as the total\\nquantities of renewable fuel rise over time, the ratio of\\nadvanced biofuels relative to conventional renewable fuel\\ngradually increases. Id. For compliance years (which match\\ncalendar years) after those specified in the tables, the statute\\nrequires EPA, in coordination with the Secretaries of Energy\\nand Agriculture, to set the annual applicable volumes based on\\n\\n 9\\na review of the implementation of the program plus an analysis\\nof six listed factors. Id. § 7545(o)(2)(B)(ii). For years not\\nspecified in the table, EPA must publish the applicable volumes\\nfourteen months before the year in which they will apply—\\nvolumes that, shortly before the start of the compliance year,\\nEPA translates into percentage standards. Id.\\n Various “waiver” provisions require or permit EPA to\\nlower the annual applicable volumes. Two are relevant for the\\npurposes of this case. First, under the “cellulosic waiver\\nprovision,” EPA must make its own projection of the volume\\nof cellulosic biofuel that will be produced in the following year.\\nId. § 7545(o)(7)(D)(i). If that projection is less than the\\nstatutory figure, the agency must use its own projection as the\\napplicable volume of cellulosic biofuel. Id.; see Am. Petroleum\\nInst. v. EPA, 706 F.3d 474, 477–80 (D.C. Cir. 2013) (API). The\\nsame cellulosic waiver provision authorizes (but does not\\nrequire) EPA to also reduce the advanced biofuel and total\\nrenewable biofuel volume requirements “by the same or a\\nlesser volume” as the cellulosic biofuel reduction, 42 U.S.C.\\n§ 7545(o)(7)(D)(i), and EPA has “broad discretion” regarding\\nwhether and how to do that, Monroe Energy, LLC v. EPA, 750\\nF.3d 909, 915 (D.C. Cir. 2014). Separately, under the “general\\nwaiver provision,” EPA may reduce any of the statutory\\napplicable volumes if it determines “that implementation . . .\\nwould severely harm the economy or environment,” or “that\\nthere is an inadequate domestic supply.” 42 U.S.C.\\n§ 7545(o)(7)(A); see Ams. for Clean Energy v. EPA, 864 F.3d\\n691, 707–13 (D.C. Cir. 2017) (ACE).\\n After EPA determines the waiver-adjusted applicable\\nvolumes, it must translate those volumes into “renewable\\nvolume obligation[s]” for each category of renewable fuel for\\nthe upcoming compliance year. 42 U.S.C. § 7545(o)(3)(B)(i).\\nThe volume obligation for each category of renewable fuel is\\nexpressed as an “applicable percentage,” also known as a\\n\\n 10\\n“percentage standard,” calculated by dividing the adjusted\\napplicable volume for that category of fuel by the total\\nanticipated volume of non-renewable transportation fuel that\\nwill be introduced into commerce (which EPA derives based\\non an estimate provided by the Energy Information\\nAdministration) in the coming compliance year. Id.\\n§ 7545(o)(3)(A), (B)(ii)(II); 40 C.F.R. § 80.1405(c). The\\nstatute calls on EPA to publish the percentage standards not\\nlater than November 30—a month before the start of the\\ncompliance year. 42 U.S.C. § 7545(o)(3)(B)(i).\\n EPA must place the renewable volume obligations on\\n“refineries, blenders, and importers, as appropriate.” 42 U.S.C.\\n§ 7545(o)(3)(B)(ii)(I); see also id. § 7545(o)(2)(A) (requiring\\nEPA to promulgate implementing regulations, including\\n“compliance provisions applicable to refineries, blenders,\\ndistributors, and importers, as appropriate,” designed to ensure\\nthat transportation fuel sold or introduced into the United States\\n“contains at least” the required annual applicable volumes).\\nThe entities that EPA designates to meet the volume\\nobligations are known as “obligated parties.” Monroe Energy,\\n750 F.3d at 912. Each obligated party must ensure that the\\nvolume of non-renewable fuel it sells or introduces into U.S.\\ncommerce is matched by selling or introducing a corresponding\\nvolume of each category of renewable fuel at the level EPA’s\\npercentage standard requires for that category. See ACE, 864\\nF.3d at 699. The percentage standards are set in the\\nanticipation that, if each obligated party meets them and EPA’s\\nprojection regarding the country’s total transportation fuel\\nsupply bears out, the amount of each category of renewable\\nfuel introduced into the economy in the upcoming compliance\\nyear will equal the applicable volumes for that year. Id.\\nObligated parties bear no direct responsibility for any shortfalls\\nin the applicable volumes so long as they comply with the\\npercentage standards.\\n\\n 11\\n EPA assigns a set of “renewable identification numbers”\\n(RINs) to each batch of renewable fuel that is produced or\\nimported for use in the United States. 40 C.F.R. § 80.1426; see\\n42 U.S.C. § 7545(o)(5); Monroe Energy, 750 F.3d at 913. The\\nnumber of RINs assigned to each batch corresponds to the\\namount of ethanol-equivalent energy per gallon in that batch.\\nSee 40 C.F.R. § 80.1415; Monroe Energy, 750 F.3d at 913.\\nRINs remain attached to the renewable fuel until that fuel is\\npurchased by an obligated party or blended into fossil fuels to\\nbe used for transportation fuel. See ACE, 864 F.3d at 699\\n(citing 40 C.F.R. § 80.1429(b)(1)–(2)). At that point the RINs\\nbecome “separated,” meaning they are, in effect, a form of\\ncompliance credit. Id. Obligated parties demonstrate their\\ncompliance with their renewable fuel obligations by “retiring”\\nRINs in annual compliance demonstrations to EPA. 40 C.F.R.\\n§§ 80.1427(a), 80.1451(a)(1).\\n Because the four categories of renewable fuel are nested,\\nobligated parties can comply with their obligations for a type\\nof fuel by retiring any combination of RINs corresponding to\\nthat category of fuels or any subset thereof. See 40 C.F.R.\\n§ 80.1427(a)(3)(i). For instance, retiring a cellulosic biofuel or\\nbiomass-based diesel RIN counts not only toward the volume\\nobligation for that fuel, but also toward both the advanced\\nbiofuel and total renewable fuel obligations. Thus, “if one\\nmillion gallons of cellulosic biofuel are blended into the fuel\\nsupply, the statute allows those one million gallons to be\\ncredited toward the advanced biofuel and total renewable fuel\\nobligations in addition to the cellulosic biofuel obligation.”\\nACE, 864 F.3d at 698.\\n Obligated parties who have more RINs than they need may\\nsell or trade their excess, 40 C.F.R. § 80.1428(b), or they may\\n“bank” those RINs for use to meet up to 20 percent of their\\nobligations for the following compliance year, Monroe Energy,\\n750 F.3d at 913; see 40 C.F.R. § 80.1427(a)(1), (5); Regulation\\n\\n 12\\nof Fuels and Fuel Additives: Changes to Renewable Fuel\\nStandard Program, 75 Fed. Reg. 14,670, 14,734–35 (Mar. 26,\\n2010). Obligated parties without enough RINs to meet their\\ncompliance obligations may purchase RINs, use banked RINs\\nfrom the prior year, or carry a deficit forward to the following\\nyear to be satisfied together with the following year’s\\nobligations. See ACE, 864 F.3d at 699–700; see also 42 U.S.C.\\n§ 7545(o)(5)(D); 40 C.F.R. § 80.1427(b).\\n\\nB. Procedural Background\\n\\n The procedural history of these cases follows two paths:\\nfirst, the proceedings relevant to the challenge that EPA\\narbitrarily declined to initiate a rulemaking to modify the 2010\\nregulation designating refineries and importers, but not\\nblenders, as obligated parties; and second, the proceedings\\nchallenging the 2017 Rule.\\n\\n 1. 2007, 2010, and 2017 Point of Obligation\\n Proceedings\\n\\n In its 2007 regulations implementing the RFS program,\\nEPA designated refiners and importers, but not blenders, as the\\n“appropriate” parties to meet the renewable fuel obligation. 72\\nFed. Reg. 23,900, 23,923–24 (May 1, 2007). At the time, those\\ndesignations were not challenged in court. EPA reaffirmed its\\ndesignations in a 2010 regulation now commonly known as the\\n“point of obligation rule.” 75 Fed. Reg. at 14,721–22 (codified\\nat 40 C.F.R. § 80.1406(a)(1)). During the 2010 rulemaking,\\nseveral refiners—including petitioner Valero Energy\\nCorporation—argued that failing to obligate blenders, who\\ncombine renewable fuel with fossil fuels, would make the RFS\\nprogram unworkable. EPA concluded that the program was\\nfunctioning adequately and that the burdens and disruption\\nfrom changing the point of obligation would outweigh any\\n\\n 13\\nbenefits. See Summary and Analysis of Comments 3.9.2, Alon\\nJ.A. 287–90. Although other aspects of the 2010 regulations\\nwere challenged in court, see, e.g., Nat’l Chicken Council v.\\nEPA, 687 F.3d 393 (D.C. Cir. 2012); Nat’l Petrochemical &\\nRefiners Ass’n v. EPA, 630 F.3d 145 (D.C. Cir. 2010), the point\\nof obligation rule was not.\\n\\n On December 14, 2015, EPA promulgated the volume\\nrequirements for 2014, 2015, and 2016. Renewable Fuel\\nStandard Program, 80 Fed. Reg. 77,420 (Dec. 14, 2015). In so\\ndoing, EPA exercised its general waiver authority to lower the\\ntotal renewable fuel volumes based on a finding of inadequate\\ndomestic supply due to market factors “affecting the ability to\\ndistribute, blend, dispense, and consume . . . renewable fuels”\\nat the levels required by statute. Id. at 77,435/2. Among those\\nfactors was “the slower than expected development of the\\ncellulosic biofuel industry.” Id. at 77,422. The agency thought\\nan additional “real world constraint[]” was the “E10\\nblendwall”—the difficulty for most American vehicle engines\\nto run on blends containing more than 10% ethanol. Id. at\\n77,423. EPA explained that those factors made the statutory\\nrequirements “impossible to achieve.” Id. at 77,422/2. This\\nCourt later vacated the general waiver on the ground that EPA\\nhad misinterpreted the statutory term “inadequate domestic\\nsupply” to include demand-side constraints such as the E10\\nblendwall. See ACE, 864 F.3d at 704–13.\\n\\n On February 12, 2016, sixty days after EPA promulgated\\nthe volume requirements for 2014–16, the Alon Petitioners\\npetitioned this Court for review of the 2010 point of obligation\\nrule. These petitions contend that the rule was arbitrary and\\ncapricious insofar as it failed to impose the obligation on\\ndownstream blenders—the parties petitioners think are best\\nable to comply with it. The petitions assert jurisdiction under\\nthe after-arising provision in 42 U.S.C. § 7607(b)(1), which\\n\\n 14\\npermits otherwise-untimely challenges to a rule if the\\nchallenges are “based solely on grounds arising after” the sixty-\\nday deadline for seeking judicial review. The petitioners assert\\nthat EPA’s exercise of its general waiver authority in the 2014–\\n16 volume regulations, and its acknowledgment of the RFS\\nprogram’s shortcomings as of that time, provided such an after-\\narising ground.\\n\\n The Alon Petitioners simultaneously petitioned EPA to\\nrevise the point of obligation rule. Some of their requests were\\nstyled as petitions for a rulemaking. Others were styled as\\npetitions for mandatory reconsideration under 42 U.S.C.\\n§ 7607(d)(7)(B), which requires EPA to reconsider a rule if\\ncentrally important objections were impracticable to raise\\nduring the comment period or “arose after” that period “but\\nwithin the time specified for judicial review.” The petitions\\ncited the waiver in the 2014–16 volume regulations and EPA’s\\nacknowledgment of program difficulties as grounds supporting\\nmandatory reconsideration. This Court held in abeyance the\\npetitions for review of the point of obligation rule pending\\nresolution of the petitions to revise it.\\n\\n On November 10, 2016, EPA published a proposed denial\\nof the petitions to revise the point of obligation rule. On\\nNovember 22, 2017, after reviewing more than 18,000\\ncomments on the proposal, EPA denied the petitions. It\\nconcluded that the statutory requirements for mandatory\\nreconsideration were not met, so it treated all the filings as\\npetitions for a rulemaking. Denial of Petitions for Rulemaking\\nto Change the RFS Point of Obligation, EPA-HQ-OAR-2016-\\n0544-0525, at 7 (Nov. 22, 2017) (EPA Denial), Alon J.A. 61.\\nEPA then denied the petitions on the ground that “changing the\\npoint of obligation would . . . likely result in a decrease in the\\nproduction, distribution, and use of [renewable] fuels” and\\nwould “do nothing to incentivize the research, development,\\n\\n 15\\nand commercialization of cellulosic biofuel technologies\\ncritical for the growth of the RFS program in future years.”\\nEPA Denial at 8–9, Alon J.A. 62–63.\\n\\n Within sixty days (in December 2017 and January 2018),\\nthe Alon Petitioners sought judicial review of that denial,\\nwhich it cast as a final agency action under section 7607(b)(1).\\nThe two sets of petitions—the February 2016 petitions for\\nreview of the 2010 point of obligation rule and the 2017–18\\npetitions for review of EPA’s refusal to reconsider the rule—\\nwere consolidated and are now before us.\\n\\n 2. 2017 Annual Volumetric Proceedings\\n\\n EPA issued its 2017 annual volumetric rule on December\\n12, 2016. The 2017 Rule establishes: (1) the applicable volume\\nfor biomass-based diesel for 2018, 81 Fed. Reg. at 89,751/2;\\n(2) the waiver-adjusted applicable volumes for cellulosic\\nbiofuel, advanced biofuel, and total renewable fuel for 2017,\\nid. at 89,747 tbl. I-1; and (3) percentage standards for all four\\nfuel types for 2017, id. at 89,751, tbl. I.B.6-1.\\n EPA exercised its mandatory cellulosic waiver authority to\\ndecrease the 2017 applicable volume for cellulosic biofuel by\\nmore than 94 percent, dropping 5.189 billion gallons from the\\nstatutory target of 5.5 billion gallons, to 311 million gallons.\\nId. at 89,750/2; 42 U.S.C. § 7545(o)(2)(B)(i)(III). EPA then\\nhad discretion under that same waiver authority to cut as much\\nas 5.189 billion gallons off the statutory volumes for advanced\\nbiofuel and total renewable fuel. See 42 U.S.C.\\n§ 7545(o)(7)(D)(i); 81 Fed. Reg. at 89,762 & tbl. IV.A-1. EPA\\npartially exercised that authority, reducing the 9-billion-gallon\\nstatutory target for advanced biofuel by 4.72 billion gallons,\\nresulting in an adjusted applicable volume of 4.28 billion\\ngallons—a greater than 50% decrease. 81 Fed. Reg. at 89,750–\\n51; 42 U.S.C. § 7545(o)(2)(B)(i)(II). EPA reduced the total\\n\\n 16\\nrenewable fuel volume requirements by the same amount,\\nlowering the statutory target of 24 billion gallons to 19.28\\nbillion gallons. 81 Fed. Reg. at 89,751/1; 42 U.S.C.\\n§ 7545(o)(2)(B)(i)(I). EPA considered but decided against also\\nusing its general waiver authority to further lower the\\napplicable volume of total renewable fuel. 81 Fed. Reg. at\\n89,751/1.\\n Using the waiver-adjusted applicable volumes, EPA set\\nthe 2017 percentage standards for each of the four renewable\\nfuel categories. See id. at 89,751, 89,799–801. Finally, EPA\\nset the biomass-based diesel applicable volume for 2018 at 2.1\\nbillion gallons. Id. at 89,751/2. EPA received comments\\nurging it to reassess the point of obligation in the 2017 Rule,\\nbut declined to address them on the grounds that the comments\\nwere “beyond the scope” of the 2017 rulemaking. Response to\\nComments at 542, Coffeyville J.A. 761.\\n After EPA published the 2017 Rule, various parties\\npetitioned for judicial review. The Coffeyville Petitioners\\ncontend that EPA erred by refusing to reconsider which types\\nof parties would bear the direct compliance obligation under\\nthe 2017 Rule. They also argue that EPA arbitrarily calculated\\nthe 2017 production of cellulosic biofuel and arbitrarily\\nexercised its discretionary cellulosic waiver authority, resulting\\nin percentage standards that are too high. NBB argues that\\nEPA set the 2018 applicable volume for biomass-based diesel\\ntoo low by considering factors it should not have and omitting\\nor incorrectly assessing others. Two trade associations\\nrepresenting refineries have intervened in defense of EPA’s\\nbiomass-based diesel decision, and a coalition of trade\\nassociations representing renewable fuel producers and\\nrefineries have intervened to oppose the Coffeyville\\n\\n 17\\nPetitioners’ claims. None of the petitioners’ challenges\\nsucceeds.\\n\\n III. Standard of Review\\n\\n “This court applies the familiar, deferential standard\\nannounced in Chevron, U.S.A., Inc. v. Natural Resources\\nDefense Council, Inc., to sustain any reasonable agency\\ninterpretation of ambiguity in the Clean Air Act.” Nat’l Ass’n\\nfor Surface Finishing v. EPA, 795 F.3d 1, 7 (D.C. Cir. 2015).\\n“We employ the deferential State Farm standard of review\\nwhen reviewing arguments based on allegedly arbitrary or\\nunreasoned agency action.” ACE, 864 F.3d at 726 (citing\\nMotor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut.\\nAuto. Ins. Co., 463 U.S. 29, 43 (1983)). Under that rubric,\\nEPA’s actions are “presumptively valid provided [they] meet[]\\na minimum rationality standard.” Nat. Res. Def. Council, Inc.\\nv. EPA, 194 F.3d 130, 136 (D.C. Cir. 1999). We uphold EPA’s\\nactions so long as they are “reasonable and reasonably\\nexplained.” Jackson v. Mabus, 808 F.3d 933, 936 (D.C. Cir.\\n2015).\\n IV. 2010 Point of Obligation Rule\\n\\n We start with the Alon Petitioners and their challenges to\\nthe 2010 point of obligation rule.\\n\\nA. Jurisdiction\\n\\n We begin, as we must, with our jurisdiction. In general\\nterms, the question presented involves our review of rules\\npromulgated under the Clean Air Act, or EPA’s failure to\\namend them, after the initial window for seeking judicial\\n\\n 18\\nreview has passed. Various statutory provisions frame this\\ninquiry.\\n\\n Section 7607(b)(1) of Title 42 provides for judicial review\\nof regulations promulgated by the Administrator of EPA under\\nthe Clean Air Act. The first sentence of section 7607(b)(1)\\nvests this Court with exclusive jurisdiction to review\\n“nationally applicable regulations promulgated, or final action\\ntaken, by the Administrator” under the Act. The fourth\\nsentence of section 7607(b)(1) specifies the time for seeking\\njudicial review. It imposes a sixty-day time limit, but provides\\nan exception for petitions based on grounds arising after the\\nlimit: “Any petition for review under this subsection shall be\\nfiled within sixty days from the date notice of such\\npromulgation, approval, or action appears in the Federal\\nRegister, except that if such petition is based solely on grounds\\narising after such sixtieth day, then any petition for review\\nunder this subsection shall be filed within sixty days after such\\ngrounds arise.”\\n\\n Section 7607(d) of Title 42 sets forth provisions for\\nvarious rulemakings under the Clean Air Act, including for the\\n“promulgation or revision of any regulation” involving the RFS\\nprogram. Id. § 7607(d)(1)(E). Section 7607(d)(7)(B)\\naddresses various issues regarding exhaustion, agency\\nreconsideration, and judicial review. The first sentence of that\\nprovision imposes a conventional exhaustion requirement,\\nlimiting judicial review to objections “raised with reasonable\\nspecificity during the period for public comment.” The second\\nsentence requires EPA to reconsider regulations in certain\\nnarrow circumstances: “If the person raising an objection can\\ndemonstrate to the Administrator that it was impracticable to\\nraise such objection within such time or if the grounds for such\\nobjection arose after the period for public comment (but within\\nthe time specified for judicial review) and if such objection is\\n\\n 19\\nof central relevance to the outcome of the rule, the\\nAdministrator shall convene a proceeding for reconsideration\\nof the rule and provide the same procedural rights as would\\nhave been afforded had the information been available at the\\ntime the rule was promulgated.” The third sentence of section\\n7607(d)(7)(B) makes any “refusal” to provide such mandatory\\nreconsideration judicially reviewable.\\n\\n At various times in this litigation, the petitioners have\\nasserted three different jurisdictional theories. First, EPA’s\\nrefusal to revise the point of obligation rule in 2017 was a final\\naction reviewable under the first sentence of section\\n7607(b)(1), regardless of the after-arising provision. Second,\\nEPA’s statements and actions in its 2014–16 volume regulation\\nconstitute after-arising grounds permitting a challenge to the\\npoint of obligation rule as promulgated in 2010. Third, these\\nsame EPA statements and actions triggered a right to\\nmandatory reconsideration under section 7607(d)(7)(B), which\\nin turn makes the denial of reconsideration judicially\\nreviewable. As explained below, we conclude that the first\\ncontention is correct, the second has been abandoned, and the\\nthird lacks merit.\\n\\n 1. Final Agency Action Under Section 7607(b)(1)\\n\\n In 2016, various refiners petitioned EPA for a rulemaking\\nto modify the point of obligation rule. The petitions urged that\\nthe need for a modification became evident in 2015, when EPA\\nwaived certain statutory volume requirements and concluded\\nthat changing economic conditions had made the requirements\\n“impossible to achieve.” 2014–16 Rule, 80 Fed. Reg. at\\n77,422/2. In November 2017, EPA denied the rulemaking\\npetitions on the ground that any current problems with the RFS\\nprogram were manageable and that changing the point of\\nobligation at this juncture would be disruptive. EPA Denial at\\n\\n 20\\n8–9, Alon J.A. 62–63. The refiners sought review of the denial\\nin December 2017 and January 2018. As petitioners in this\\nCourt, they contend that the November 2017 denial constituted\\nfinal agency action reviewable under section 7607(b)(1). We\\nagree.\\n\\n As noted above, the first sentence of section 7607(b)(1)\\ngives this Court exclusive jurisdiction to review any nationally\\napplicable “final action” taken by EPA under the Clean Air\\nAct. The parties agree that the denial of the rulemaking\\npetitions was nationally applicable, final, and taken under the\\nClean Air Act. It was also agency “action” within the meaning\\nof the statute. That word “bears the same meaning in [section\\n7607(b)(1)] that it does under the Administrative Procedure\\nAct,” Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 478\\n(2001), which defines “agency” to include EPA, 5 U.S.C.\\n§ 551(1), and “agency action” to include “the whole or a part\\nof an agency rule, order, license, sanction, relief, or the\\nequivalent or denial thereof, or failure to act,” id. § 551(13)\\n(emphases added). So, EPA’s denial of the petitions for\\nrulemaking was a reviewable “action.”\\n\\n The petitions for review were timely. As a general matter,\\nsection 7607(b)(1) requires a petition for review to be filed\\nwithin sixty days of when “notice of such promulgation,\\napproval, or action appears in the Federal Register.” Here, the\\n“action” at issue—denial of the petitions for rulemaking—was\\npublished in the Federal Register on November 22, 2017, and\\nthe petitions for review of that action were filed within sixty\\ndays of that date. Moreover, this conclusion does not depend\\non the after-arising provision. To the contrary, because the\\npetitions for review were filed within sixty days of the “action”\\nunder review, the exception for “grounds arising after such\\nsixtieth day” was not triggered.\\n\\n 21\\n Our caselaw confirms this framing of the jurisdictional\\nissue. In Massachusetts v. EPA, 415 F.3d 50 (D.C. Cir. 2005),\\nthis Court held that EPA’s denial of a petition to regulate\\ngreenhouse gas emissions as air pollutants was itself “final\\naction” reviewable under section 7607(b)(1). See id. at 53–54\\n(opinion of Randolph, J.); id. at 61 (Sentelle, J., concurring in\\nthe judgment). The Supreme Court reversed our judgment on\\nthe merits, but agreed that we had jurisdiction. Massachusetts\\nv. EPA, 549 U.S. 497 (2007). In particular, the Court noted that\\nsection 7607(b)(1) “expressly permits review” of EPA’s\\n“rejection of [a] rulemaking petition.” Id. at 520, 528; see also\\nid. at 517 (section 7607(b)(1) affords “the right to challenge\\n[this] agency action unlawfully withheld”). Likewise, in\\nNatural Resources Defense Council v. EPA, 824 F.2d 1146\\n(D.C. Cir. 1987) (en banc) (NRDC), we held that a 1985\\ndecision to withdraw proposed amendments to certain 1976\\nregulations—which we described as a “decision not to amend”\\nthe regulations—was “reviewable agency action” under\\nsection 7607(b)(1). Id. at 1150 (quotation marks omitted). We\\nconcluded that the petition for review at issue, which on its face\\nchallenged the 1985 withdrawal decision, was not, in\\nsubstance, an untimely “‘back-door’ challenge to the 1976\\nregulations.” Id. On that point, we reasoned that the petitioners\\nclaimed legal errors in the 1985 withdrawal and sought vacatur\\nonly of that order. Likewise, in this case, the petitions for\\nreview filed in 2017 and 2018 raise no back-door challenge to\\nthe 2010 regulation: the petitions contend that EPA in 2017\\narbitrarily refused to take account of changing economic\\nconditions, and they seek vacatur only of the 2017 order\\ndenying a new rulemaking going forward.\\n\\n Background principles of administrative law reinforce our\\nconclusion that the denial of a petition to modify a rule based\\non changed circumstances is itself a reviewable order.\\nOrdinarily, the denial of a petition to amend or rescind a\\n\\n 22\\nregulation is judicially reviewable. See, e.g., NLRB Union v.\\nFLRA, 834 F.2d 191, 195–96 (D.C. Cir. 1987). When the\\npetition to amend attacks defects in the regulation as originally\\npromulgated, and when the time limit for seeking review of the\\nregulation has passed, questions can arise about whether the\\ntime limit is being improperly circumvented. In these\\ncircumstances, we have held that the petitioner cannot raise\\nprocedural challenges to the regulation, but can raise\\nsubstantive arguments that the regulation is unauthorized by or\\nconflicts with a statute. See id. at 196–97. But the\\ncircumvention concern does not even arise when the petitioner\\nraises arguments about changed circumstances or new\\ninformation. Those kinds of arguments—that recent\\ndevelopments compel the amendment of an older regulation—\\nare always cognizable through review of the denial of a petition\\nto amend, though they trigger an “extremely limited” review\\non the merits. Id. at 196. Our decision today harmonizes the\\njudicial-review provisions of the Clean Air Act with this\\ngeneral background principle.\\n\\n EPA recognizes the general rule that, under the NLRB\\nUnion line of cases, the denial of a petition to amend a rule is a\\nreviewable order, which supports both challenges based on\\nrecent developments and substantive challenges to the original\\nregulation. Nonetheless, EPA urges a different rule where the\\napplicable judicial-review provision contains a time limit for\\nseeking review and an exception for grounds arising after the\\ntime limit, as in the Clean Air Act. In those circumstances,\\naccording to EPA, a challenge to the denial of a petition to\\namend is untimely—and the denial is thus entirely\\nunreviewable—unless the after-arising provision is satisfied.\\nEPA rests this conclusion on National Mining Ass’n v.\\nDepartment of Interior, 70 F.3d 1345 (D.C. Cir. 1995), and\\nAmerican Road & Transportation Builders Ass’n v. EPA, 588\\n\\n 23\\nF.3d 1109 (D.C. Cir. 2009) (ARTBA), but neither decision\\nsupports its position.\\n\\n National Mining involved judicial review under the\\nSurface Mining Control and Reclamation Act (SMCRA),\\nwhich requires petitions for review to be filed “within sixty\\ndays” of the agency action at issue “or after such date if the\\npetition is based solely on grounds arising after the sixtieth\\nday.” See 70 F.3d at 1350. In 1986, parties petitioned the\\nDepartment of Interior to rescind a 1979 SMCRA regulation\\non two grounds. First, the petitioners argued that the rule was\\ninconsistent with the statute—an argument attacking the\\nregulation itself and “available” when the regulation was\\noriginally promulgated. See id. After the agency declined to\\nrescind the rule, the challengers sought judicial review. We\\nheld that the after-arising provision made this first challenge\\nuntimely, by explicitly requiring challenges to a regulation\\neither to be filed “within the statutory period” or to “meet the\\nafter-arising test.” Id. at 1350–51. At the same time, however,\\nwe held that the petitioners’ second challenge—to the agency’s\\n1986 decision refusing to “repeal” the regulation based on\\nchanged circumstances—was timely and reviewable. Id. at\\n1352. To be sure, we described the latter challenge as resting\\non “grounds that arose after the sixtieth day.” Id. But we failed\\nto explain what jurisdictional theory that observation\\nsupported: a challenge to the 1979 regulation rendered timely\\nby the after-arising exception; or a challenge, in substance as\\nwell as form, to the 1986 refusal to repeal, akin to the challenge\\nthat we held reviewable in NRDC. Instead, we simply\\nconcluded that, under the “limited scope of our review,” the\\nagency did not “act[] unreasonably in denying the petition for\\nrulemaking.” Id. at 1352–53. Thus, while National Mining\\nblessed jurisdiction to review agency refusals to amend\\nregulations based on changed factual circumstances, it did not\\n\\n 24\\nultimately address what we clarify today—the precise statutory\\nbasis for that jurisdiction.\\n\\n ARTBA applied the reasoning of National Mining to the\\nClean Air Act, which also contains a time limit for judicial\\nreview and an exception for after-arising grounds. ARTBA\\ninvolved a 2002 petition to amend 1997 regulations on the\\nground that they allowed states “to adopt precisely the kinds of\\nregulations that the statute forbids.” 588 F.3d at 1110. As in\\nNational Mining, the challenge was thus a substantive attack\\non a regulation as originally promulgated. We held that, under\\nNational Mining, EPA’s “denial of a revision-seeking petition\\ndoes not allow review of alleged substantive defects in the\\noriginal rule even under the deferential standards applicable to\\nreview of such denials, outside the statutory limitations period\\nrunning from the rule[’s] original promulgation.” Id. at 1113\\n(emphasis added). In other words, National Mining “require[d]\\nus to treat ARTBA’s petition to EPA as a challenge to the\\nregulations it sought revised.” Id. at 1110. As a result,\\ndismissal was necessary unless the petition satisfied the\\nexception for after-arising grounds, which it did not. See id.\\n\\n Neither decision controls here, where the 2017 and 2018\\npetitions for review challenge not the point of obligation\\nregulation as originally promulgated in 2010, but the failure to\\namend the regulation in light of changed circumstances flagged\\nby EPA in 2015. EPA rejected that argument in 2017, and the\\npetitioners sought review of the rejection within sixty days. In\\nsubstance as well as form, the challenge was to the 2017 refusal\\nto amend, not to the underlying 2010 regulation. Under these\\ncircumstances, there was no risk of circumventing the original\\ntime limit. Therefore, there was also no reason to treat the 2010\\npromulgation and the 2017 refusal to amend as one-and-the-\\nsame agency action, despite binding APA definitions treating\\nthem as separate.\\n\\n 25\\n Precedent aside, EPA’s proposed approach—making the\\nafter-arising provision the exclusive vehicle for challenging\\nrefusals to amend regulations based on new information or\\nchanged circumstances—creates various difficulties. For one\\nthing, there is a conceptual mismatch between that provision\\nand these kinds of challenges. Though the after-arising\\nexception and the opportunity to seek rule revision based on\\npost-rulemaking events may seem similar, the first allows an\\nintervening event to secure judicial review on the basis of\\ndefects extant at the time of the rulemaking, whereas the\\nsecond allows review on the question whether intervening\\nevents have fatally undermined the original justification for the\\nrule. The arrow of time runs forward, not backward, so it is at\\nbest awkward—and at worst incoherent—to speak of a later\\ndevelopment rendering unlawful an earlier promulgation.\\nEconomic developments in 2015 may have made it arbitrary\\nfor EPA to adhere to the point of obligation rule in 2017, but\\nthey cannot have retroactively made arbitrary its promulgation\\nin 2010.\\n\\n Even worse, the Clean Air Act’s after-arising provision, if\\nused to judge the timeliness of challenges based on new\\ninformation, would be difficult to apply, capriciously narrow,\\nor both. To satisfy that provision, a petition for review must be\\nboth (i) based “solely” on after-arising grounds and (ii) filed\\n“within sixty days after such grounds arise.” 42 U.S.C.\\n§ 7607(b)(1). Yet the case for changing an environmental\\nregulation will almost never manifest itself at one discrete\\nmoment. Instead, it will accumulate progressively over time,\\nas scientific knowledge advances or economic conditions\\nchange. And so, under EPA’s approach, the relevant filing\\ndeadlines would become practically unknowable. When did\\nsome environmental risk become serious and obvious enough\\nto compel a rulemaking to strengthen an existing regulation?\\nThat will usually be a hard question, and it would be little short\\n\\n 26\\nof miraculous if the answer turned out to be on a date certain\\nwithin sixty days of the filing of a petition for review, as\\nrequired to satisfy the after-arising provision.\\n\\n In contrast, the approach we have sketched out produces\\nsimple questions and discernible deadlines: ask when EPA\\ndenied the rulemaking petition, then add sixty days. If\\npossible, we should avoid trying to fix arbitrarily precise\\naccrual dates for events that develop incrementally over time.\\nSee Nat’l R.R. Passenger Corp. v. Morgan, 536 U.S. 101, 115–\\n21 (2002). And we should avoid jurisdictional tests that are\\ncomplex as opposed to straightforward. See, e.g., Hertz Corp.\\nv. Friend, 559 U.S. 77, 94 (2010). Treating the denial of a\\npetition to amend a rule based on changed circumstances as\\nreviewable agency action honors both principles, while\\nattempting to shoehorn such denials into the after-arising\\nprovision does the opposite.\\n\\n We acknowledge that, in Oljato Chapter of Navajo Tribe\\nv. Train, 515 F.2d 654 (D.C. Cir. 1975), we assumed that the\\nafter-arising provision would govern review of orders denying\\npetitions to modify EPA rules based on changed circumstances.\\nBut our only holdings were that such petitions must first be\\npresented to EPA, id. at 666, and then are reviewable directly\\nin the courts of appeals, id. at 657–65. Moreover, our\\nassumption was understandable in context; the petitioners had\\nmissed the statutory deadline to file a petition for review (then\\nthirty days), see id. at 657, so they pressed alternative\\njurisdictional theories (involving either district-court review or\\nthe after-arising provision) that would avoid that deadline.\\nFurthermore, we expressly reserved the precise bounds of what\\nconstituted a petition based solely on after-arising grounds. See\\nid. at 666–68. Finally, we stressed that review should generally\\nbe available “when new information casts doubt upon the\\nvalidity of a [regulatory] standard.” Id. at 665. At the time, the\\n\\n 27\\nafter-arising provision contained no separate deadline\\nrequiring a petition for review to be filed within sixty days of\\nthe after-arising ground. See id. at 657 n.3. So the difference\\nbetween “direct review” of a regulation through the after-\\narising provision and review of a “refusal to revise” the\\nregulation appeared largely semantic. See id. at 666. But, two\\nyears after Oljato was decided, Congress amended section\\n7607(b)(1) to include the separate filing deadline, see Pub. L.\\nNo. 95-95 § 305(c)(3), 91 Stat. 685, 776 (1977), which, as\\nexplained above, made the after-arising provision a singularly\\npoor vehicle for securing the review that Oljato assumed would\\nbe readily available.\\n\\n After Oljato, we held in Group Against Smog & Pollution\\nv. EPA, 665 F.2d 1284 (D.C. Cir. 1981), that the failure to\\nchallenge a 1974 regulation within the original sixty-day\\ndeadline did not bar “judicial review of the agency’s\\nsubsequent refusal to revise the standard on the basis of new\\ninformation.” Id. at 1291. Quoting liberally from Oljato, we\\nsuggested that such review would proceed through the after-\\narising provision. Id. at 1289. But we permitted review even\\nthough the after-arising provision, as amended, could not have\\napplied. The case involved comments filed with EPA in April\\n1977, which we treated as a petition for a rulemaking. Id. at\\n1290 & n.47. On April 13, 1978, EPA declined to amend the\\nrule as requested. Id. at 1288. The ensuing petition for review\\nwas filed on June 12, 1978—more than a year after the\\ntechnological changes discussed in the comments, but precisely\\nsixty days after the refusal to amend. See id. Because the\\npetition for review was not filed within sixty days of the\\nasserted after-arising grounds, the after-arising provision\\nplainly did not apply. So, review must have rested on the\\ntheory that the refusal to amend was itself a reviewable action\\ntriggering its own sixty-day filing window.\\n\\n 28\\n The approach we follow here—treating denials of\\nrulemakings based on new facts as independently reviewable\\ndecisions—does not reduce the after-arising provision to\\nsurplusage. Our cases have recognized other circumstances\\ntriggering the after-arising provision, including judicial\\ndecisions that significantly “changed the legal landscape”\\nfaced by petitioners, Honeywell Int’l, Inc. v. EPA, 705 F.3d\\n470, 473 (D.C. Cir. 2013), and “the occurrence of an event that\\nripens a claim,” Coalition for Responsible Regulation v. EPA,\\n684 F.3d 102, 129 (D.C. Cir. 2012) (per curiam). In cases like\\nthese—where, for example, an intervening statute, regulation,\\nor judicial decision extends old regulations to new parties—the\\nafter-arising ground is easily dated, the relevant filing\\ndeadlines are clear, and so the provision functions predictably.\\nMoreover, where the after-arising provision does apply, it\\npermits the petitioner to contend not only that changed\\ncircumstances warrant amending an existing regulation but\\nalso that the regulation was unlawful as originally\\npromulgated. See, e.g., Honeywell, 705 F.3d at 473.\\n\\n For these reasons, we conclude that we have jurisdiction\\nto consider the petitioners’ argument that EPA arbitrarily\\nrefused to amend the point of obligation rule based on the\\nchanged circumstances cited by the petitioners.\\n\\n 2. After-Arising Grounds Under Section 7607(b)(1)\\n\\n The February 2016 petitions for review, filed before EPA\\nresolved any of the rulemaking petitions, rested on the\\nalternative jurisdictional theory that EPA’s publication of the\\n2014–16 volume requirements constituted an after-arising\\nground within the meaning of section 7607(b)(1). Our\\nconclusion above does not moot this question, because if this\\nalternative theory were valid, then the petitioners could directly\\nattack the point of obligation rule as originally promulgated.\\n\\n 29\\nNonetheless, the petitioners have abandoned this theory of\\njurisdiction. In their merits briefs, they never actually attack\\nthe 2010 rule as originally promulgated; instead, they challenge\\nonly the 2017 denial of their rulemaking petitions. Moreover,\\nin requesting relief, they do not ask us to set aside the 2010\\nrule, but only to “vacate the [2017] Denial[] and remand to\\nEPA” for a rulemaking to change the point of obligation rule\\ngoing forward. Alon Br. 58. And at oral argument, they\\ndisclaimed any challenge to the 2010 rule itself and confirmed\\nthat their only challenge was to EPA’s 2017 refusal to revise\\nthe point of obligation rule. See Rec. of Oral Argument at\\n2:18:50–2:19:15.\\n\\n 3. Mandatory Reconsideration Under Section\\n 7607(d)(7)(B)\\n\\n Finally, the petitioners assert jurisdiction under section\\n7607(d)(7)(B), on the ground that EPA erroneously denied\\npetitions for mandatory reconsideration of the point of\\nobligation rule. To review, section 7607(d)(7)(B) provides in\\nrelevant part that only objections “raised . . . during the period\\nfor public comment” may be “raised during judicial review.”\\nBut if the objector “can demonstrate to the Administrator that\\nit was impracticable to raise such objection within such time or\\nif the grounds for such objection arose after the period for\\npublic comment (but within the time specified for judicial\\nreview) and if such objection is of central relevance to the\\noutcome of the rule,” then EPA must “convene a proceeding\\nfor reconsideration of the rule,” and any refusal to do so is\\njudicially reviewable. Here, the petitioners assert that the\\nground for their objections—EPA’s statements and actions in\\nits 2014–16 volume regulation—arose “after the period for\\npublic comment” on the 2010 point of obligation rule, and that\\n\\n 30\\nthe objections are “of central relevance to the outcome of the\\nrule.”\\n\\n The petitioners misapprehend the statutory text and\\nstructure. Section 7607(d)(7)(B) does not extend the\\njurisdictional deadline to seek judicial review imposed in\\nsection 7607(b)(1); instead, it specifies a non-jurisdictional\\nexhaustion requirement. See EPA v. EME Homer City\\nGeneration, L.P., 572 U.S. 489, 512 (2014). Its first sentence\\ngenerally requires parties to raise objections “during the period\\nfor public comment” in order to later present them in court. Its\\nsecond sentence allows a narrow exception when a centrally\\nimportant objection cannot feasibly be raised during the\\ncomment period—either because “the grounds for such\\nobjection arose after the period for public comment,” or\\nbecause commenting was otherwise “impracticable.” If an\\nobjection fits within this exception, the consequences are\\nweighty: EPA must grant reconsideration and conduct a new,\\nfull-dress, notice-and-comment rulemaking. And if EPA\\ndenies reconsideration, the objector may seek judicial review.\\n\\n This “limited exception” to the normal exhaustion\\ndeadline, Util. Air Regulatory Grp. v. EPA, 744 F.3d 741, 746\\n(D.C. Cir. 2014), does not come with a free pass from the\\nsubsequent deadline to seek judicial review. To the contrary,\\nthe second sentence of section 7607(d)(7)(B) covers only\\nobjections that arise after the close of the comment period, yet\\nwithin the time specified for judicial review. As noted above,\\nthat time for judicial review—sixty days from the promulgation\\nof the final rule—is specified in section 7607(b)(1). Section\\n7607(d)(7)(B) does not enlarge that filing period, but merely\\nfills a narrow gap within it: allowing orderly exhaustion of\\nimportant objections that “first became known to the petitioner\\nafter the comment period ended, but before the period for\\npetitioning for review expired.” Am. Petroleum Inst. v. Costle,\\n\\n 31\\n665 F.2d 1176, 1192 (D.C. Cir. 1981). We recognize that, as a\\ntextual matter, the statutory phrase “but within the time\\nspecified for review” qualifies the requirement that the grounds\\nmust have arisen “after the period for public comment,” but not\\nthe alternative requirement that “it was impracticable” to raise\\nthe objection “during the period for public comment.”\\nHowever, the petitioners do not invoke the impracticability\\nprong of section 7607(d)(7)(B). Moreover, we have construed\\nthat prong to cover instances when the final rule was not a\\nlogical outgrowth of the proposed rule, Clean Air Council v.\\nPruitt, 862 F.3d 1, 10 (D.C. Cir. 2017) (per curiam), which\\nlikewise involve problems during the period for public\\ncomment on or petitioning for review of the regulation itself—\\nnot problems that arise when circumstances change years or\\ndecades later.\\n\\n The petitioners argued in briefing that the “time specified\\nfor judicial review” referenced in section 7607(d)(7)(B)\\nencompasses not only the initial sixty-day window after a\\nrule’s promulgation, but also the secondary sixty-day limit\\nfrom after-arising grounds in the fourth sentence of section\\n7607(b)(1). But as noted above, the petitioners abandoned at\\noral argument any reliance on the latter after-arising provision.\\nMoreover, their theory would transform what we have\\ndescribed as a “limited” gap-filling provision, Util. Air\\nRegulatory Grp., 744 F.3d at 746, into a perpetually looming\\nthreat of mandatory notice-and-comment reconsideration.\\nTellingly, the petitioners can cite no case employing section\\n7607(d)(7)(B)’s mandatory reconsideration procedure for\\nobjections that arose after the close of the initial window for\\njudicial review. Their interpretation would “make a mockery\\nof Congress’[s] careful effort to force potential litigants to\\nbring challenges to a rule issued under this statute at the\\n\\n 32\\noutset.” Am. Rd. & Transp. Builders Ass’n v. EPA, 705 F.3d\\n453, 458 (D.C. Cir. 2013) (quotation marks omitted).\\n\\n Because the petitioners’ objections to the point of\\nobligation rule did not arise within the initial window for\\njudicial review of the 2010 point of obligation rule, but only\\nsome five years later, EPA properly denied mandatory\\nreconsideration.\\n\\nB. Merits of Challenges to EPA’s Refusal to Revise the\\n 2010 Point of Obligation Rule\\n\\n The Alon Petitioners offer an array of arguments to\\nchallenge the denial. None, however, is persuasive.\\n\\n We are reviewing EPA’s denial of a petition for\\nrulemaking to amend the agency’s point of obligation rule. See\\nsupra Part IV.A.1. Accordingly, our review is “‘extremely\\nlimited’ and ‘highly deferential.’” New York v. EPA, 921 F.3d\\n257, 261 (D.C. Cir. 2019) (quoting Massachusetts, 549 U.S. at\\n527–28). “To set aside the agency’s judgment, [we] must\\nconclude that EPA had not ‘adequately explained the facts and\\npolicy concerns relied on’ or that those facts did not ‘have some\\nbasis in the record.’” Id. (quoting WildEarth Guardians v.\\nEPA, 751 F.3d 649, 653 (D.C. Cir. 2014)). We have no basis\\nfor such a conclusion. In denying the petition for rulemaking,\\nEPA considered the “information currently before” it and\\ndetermined “that the point of obligation is appropriately\\nplaced,” wrestling with the petitioners’ claims to the contrary.\\nEPA Denial at 8, Alon J.A. 62. As is evident from our\\ndiscussion below, EPA did so with enough thoroughness and\\nreasonableness to satisfy our limited, deferential review.\\n\\n We start with EPA’s reasoning, which petitioners say is\\narbitrary and capricious. See 42 U.S.C. § 7607(d)(9)(A).\\n\\n 33\\n At the root of petitioners’ claim is a single premise: that\\nthe current point of obligation misaligns incentives by\\nrequiring those who refine fossil fuel, but not those who blend\\nit, to meet the RFS program’s annual standards. In petitioners’\\nview, this misalignment forces refiners to purchase RINs to\\nsatisfy their RFS obligations, jacking up their costs, while\\ngiving windfall profits to blenders, who produce (but don’t\\nconsume) RINs. From this cycle of “RINsanity,” petitioners\\nsay, flow harms galore. Alon Reply Br. 25. Higher RIN prices\\nnot only threaten the financial viability of refiners (putting our\\neconomy and energy security in jeopardy), see, e.g., Alon Br.\\n46, but they incentivize RIN hoarding, which feeds market\\nvolatility, and gives some market participants an unfair leg up,\\nsee, e.g., id. at 40.\\n\\n The problem with this argument, however, is that EPA\\nreasonably explained why, in its view, there is no misalignment\\nin the RFS program. According to EPA, refiners “recover the\\ncost of the RINs they purchase” by passing that cost along in\\nthe form of “higher prices for the petroleum based fuels they\\nproduce.” EPA Denial at 25, Alon J.A. 79. It grounded that\\nconclusion in studies and data in the record. EPA and the\\nauthors of the pertinent studies took advantage of the fact that\\nthere are pairs of petroleum products in which one variant is\\nsubject to the RIN obligation (such variants being awkwardly\\ncalled “obligated fuels”), whereas its not-quite-identical twin is\\nnot. For example, gasoline and diesel sold for use in the United\\nStates are “obligated,” whereas the same fuels sold for export\\nare not. EPA Denial at 23, Alon J.A. 77. The same goes for\\ndomestic diesel fuel, which is “obligated,” and jet fuel, which\\nis not. Christopher R. Knittel et al., The Pass-Through of RIN\\nPrices to Wholesale and Retail Fuels Under the Renewable\\nFuel Standard 4 (July 2015) (Knittel), Alon J.A. 534.\\nComparing these pairs, the agency found that as RIN prices\\nincreased, a gap “open[ed] up between” the price for obligated\\n\\n 34\\nand unobligated fuels, a gap rather precisely matching the\\ncontemporaneous increase in RIN price—a strong indication\\nthat refiners were “recoup[ing] the costs associated with RIN\\nprices.” EPA Denial at 23, Alon J.A. 77.\\n\\n Further confirming the price relationship, Professor\\nKnittel and his colleagues found that 73% of a change in RIN\\nprice was passed through in the form of higher petroleum prices\\nin the same day, 98% within two business days. Knittel 26,\\nAlon J.A. 536.\\n\\n Reviewing the findings, EPA (accurately) reported that the\\npapers by Knittel and his colleagues, and by Argus Consulting\\nServices, “concluded that the RIN cost was generally included\\nin the sale prices of obligated fuels.” EPA Denial at 25, Alon\\nJ.A. 79; see Knittel 26, Alon J.A. 536; Argus Consulting\\nServices, Do Obligated Parties Include RINs Costs in Product\\nPrices? 15 (Feb. 2017), Alon J.A. 564 (“There are very specific\\ncorrelating price data for diesel that indicate that refiners . . .\\npass along the RINs cost . . . .”).\\n\\n A similar analysis, EPA concluded, reveals that just as\\n(obligated) refiners do not pay excess costs, neither do blenders\\n(who are not obligated under the program) nor integrated\\nrefiners (who perform their own in-house blending) reap\\nwindfall profits. True, both earn RINs, without purchasing\\nthem on the open market, by blending renewable fuel into\\npetroleum blendstock. And true, as well, both can sell those\\nRINs, enjoying whatever revenues market conditions and their\\nown efficiencies permit. But as EPA quite accurately\\nexplained, this is only half the equation. In a competitive\\nmarket there’s no such thing as a free lunch, and blenders and\\nintegrated refiners pay their tab just as others do; they just do\\nso indirectly. To offer finished fuel without attached RINs at a\\ncompetitive price, these entities must discount their blended\\n\\n 35\\nfuel by roughly the value of the RINs that they detached and\\nkept for themselves. EPA Denial at 29, Alon J.A. 83. In other\\nwords, they “sell the finished transportation fuel at a loss,” but\\n“maintain[] profitability through RIN sales.” Id. at 27–28, 29,\\nAlon J.A. 81–82, 83.\\n\\n To be sure, in response to EPA’s proposed denial,\\ncommenters criticized the studies relied on by the agency.\\nThey contended, for example, that Professor Knittel and his\\ncolleagues erred by removing certain spreads from the analysis,\\nby including others, and by pooling the results of various\\ncomparisons. See EPA Denial at 25, Alon J.A. 79. But\\npetitioners have not raised these arguments here, and for that\\nreason we do not consider them. While petitioners do complain\\nthat EPA relied on a “preliminary” analysis, see Alon Br. 54;\\nAlon Reply Br. 23, that objection—whatever its persuasive\\nforce—says nothing about the other studies in the record (for\\nexample, by Professor Knittel et al.).\\n\\n Petitioners try, instead, to trace various refiner problems to\\nEPA’s refusal to obligate blenders. They suggest that the\\nalleged misplacement of the point of obligation causes\\nbankruptcies, see, e.g., Alon Br. 3, 47, and inflicts economic\\nhardship on small refineries, see, e.g., id. at 49, especially in\\nthe form of inflicting wildly disproportionate RIN acquisition\\ncosts on them, see, e.g., Alon Reply Br. 26. But some of these\\nevents occurred after EPA issued its denial, see, e.g., Alon Br.\\n49 (“following the Denial”); Alon Reply Br. 26 (“just after the\\nDenial”), and are therefore not properly before us, see\\nEnvironmental Defense Fund, Inc. v. Costle, 657 F.2d 275, 284\\n(D.C. Cir. 1981). More importantly, the claims presuppose that\\nrefiners cannot recover their RIN costs and that blenders reap\\nwindfall profits—suppositions that, as discussed above, EPA\\nreasonably rejected.\\n\\n 36\\n Petitioners respond by plucking snippets from the denial,\\nstringing them together with contrasting (bolded) conjunctions,\\nand asserting that EPA’s “discussion of RIN prices” is\\n“irreconcilably inconsistent.” Alon Br. 53. But we find no\\ninconsistency on EPA’s part.\\n\\n Take one of petitioners’ examples:\\n\\n . . . RIN prices had no “significant impact on retail\\n gasoline (E10) prices,” JA75;\\n\\n although “RINs . . . provide a price signal to\\n consumers to help achieve . . . greater renewable fuel\\n production and use,” JA75.\\n\\nAlon Br. 53 (second and third alterations in original) (quoting\\nEPA Denial at 21, Alon J.A. 75). At first blush, the two\\ncomments, located on the very same page, seem inconsistent.\\nHow could RIN prices have no “significant impact” on retail\\nprices, while, at the same time, provide “a price signal to\\nconsumers”?\\n\\n They can do so for the simple reason that the remarks refer\\nto different things, a detail omitted from petitioners’ brief. This\\nbecomes apparent when the passage from which petitioners\\nplucked their quotes (bolded and underscored below) is viewed\\nin full:\\n\\n External, non-EPA assessments similarly concluded\\n that increased RIN prices had not had a significant\\n impact on retail gasoline (E10) prices. When RIN\\n prices rise, the market price of the petroleum\\n blendstocks produced by refineries also rise to cover\\n the increased RIN costs, in much the same way as they\\n would rise in response to higher crude oil prices. The\\n effective price of renewable fuels (the price of the\\n\\n 37\\n renewable fuel with attached RIN minus the RIN\\n price), however, decreases as RIN prices increase.\\n When renewable fuels are blended into petroleum\\n fuels these two price impacts generally offset one\\n another for fuel blends such as E10 with a renewable\\n content approximately equal to the required\\n renewable fuel percentage standard. Higher RIN\\n prices also generally result in higher prices for fuels\\n with lower renewable content (such as E0 or\\n petroleum diesel) and lower prices for fuels with\\n higher renewable content (such as E85 or B20). The\\n cost of the RIN therefore serves as a cross-subsidy,\\n reducing the price of renewable fuels and increasing\\n the price of petroleum based fuels in transportation\\n fuel blends, thus incentivizing increased blending of\\n renewable fuels into the transportation fuel pool. In\\n this way the RINs also help provide a price signal to\\n consumers to help achieve the Congressional goals\\n of greater renewable fuel production and use.\\n Fuels with higher renewable content are relatively\\n cheaper to consumers than they would be absent high\\n RIN prices, while fuels with lower renewable content\\n are relatively more expensive when RIN prices are\\n high.\\n\\nEPA Denial at 20–21, Alon J.A. 74–75.\\n\\n As we can see, the first statement (no significant price\\nimpact) is referring to the price of E10—a blend of 90%\\ngasoline, 10% ethanol. As EPA explained, RINs work as a\\n“cross-subsidy,” effectively taxing the use of petroleum-based\\nfuels (e.g., gasoline) and subsidizing the use of renewables\\n(e.g., ethanol) in making a blended transportation fuel like E10.\\nEPA Denial at 21, Alon J.A. 75.\\n\\n 38\\n Before we dig in further, let’s take a step back. We must\\nfirst recognize that EPA assumes that we are talking of a market\\nwhere the RFS program, in effect, mandates minimum levels\\nof renewables in marketed fuels, a mandate that necessarily\\nimpacts fuel prices. EPA is not making a claim that the\\nmandatory inclusion of renewables in transportation fuel\\nrenders a gallon of gasoline lawfully purchased at the pump\\ncheaper than it would have been absent the RFS program.\\n\\n To see what this means, start on the subsidy side: Suppose\\na blender can realize $2.25 on a gallon of ethanol with an\\nattached RIN. If the blender can detach and sell that RIN for\\n$0.05, then the net ethanol value is only $2.20—a $0.05\\nsavings that in a competitive market should pass through to\\nconsumers. Now take the tax side: Because refiners must\\npurchase RINs to satisfy the RFS obligations that arise from\\nselling gasoline to blenders, a blender’s value for a gallon of\\ngasoline may rise, due to the RFS program, from, say, $2.75 to\\n$2.76. See Dallas Burkholder, Office of Transportation & Air\\nQuality, EPA, A Preliminary Assessment of RIN Market\\nDynamics, RIN Prices, and Their Effects 17, EPA-HQ-OAR-\\n2016-0544-0009 (May 14, 2015), Alon J.A. 337. As a result\\nof both price impacts, EPA described, blended fuels with a\\nhigher percentage of renewable content (e.g., 85% ethanol) will\\nbe cheaper than they would have been (absent the program),\\nwhereas fuels with a lower percentage of renewable content\\n(e.g., pure gasoline) will be more costly than they would have\\nbeen (absent the program). EPA Denial at 21, Alon J.A. 75.\\nFor E10, the “two price impacts generally offset one another,”\\nso (back to the first statement) any change in RIN price\\ngenerally has no “significant impact on” the E10 price. Id.\\n\\n But that’s just E10. There is an effect (of differing\\nmagnitude) on, say, E85 or E0. And that is where the second\\nstatement (“provide a price signal”) comes in: the signal arises\\n\\n 39\\nfrom a comparison of relative prices across the spectrum of\\ntransportation fuels. Again, as EPA explains, “[f]uels with\\nhigher renewable content are relatively cheaper to consumers\\nthan they would be absent high RIN prices, while fuels with\\nlower renewable content are relatively more expensive when\\nRIN prices are high.” Id. The two statements are consistent.\\n\\n Continuing the search for inconsistency, petitioners direct\\nour attention to “EPA’s past pronouncements.” Alon Br. 50.\\nIn them, they see an irrational “about-face”—with EPA saying,\\nat first, that “low RIN prices [were] a sign that the [RFS\\nprogram] was working,” but claiming, now, “that high RIN\\nprices are . . . desirable.” Id. at 51–52. Again, that’s not quite\\nright. All EPA originally said was that when it first adopted\\nthe point of obligation, it did so based, in part, on its\\n“expectation at that time that there would be an excess of RINs\\nat low cost.” Regulation of Fuels and Fuel Additives: Changes\\nto Renewable Fuel Standard Program, 74 Fed. Reg. 24,904,\\n24,963/2 (May 26, 2009) (proposed rule); see also Alon Br. 50\\n(citing EPA Denial at 13, Alon J.A. 67 (citing, in turn, 74 Fed.\\nReg. at 24,963)). EPA did not suggest that low RIN prices were\\na sign of market health—nor that high prices were a cause for\\nalarm.\\n\\n In any case, EPA addressed petitioners’ concern over high\\nRIN prices head on; the agency explicitly determined, on the\\ncurrent record, that “higher RIN prices” are not “indicative of\\na dysfunctional RIN market.” EPA Denial at 19, Alon J.A. 73.\\nRather, EPA explained, these prices accurately reflect the\\nincreasing cost associated with “getting ever-greater volumes\\nof renewable fuel into the transportation fuel pool—the explicit\\ngoal [of] the RFS program.” Id. Put more bluntly, the increases\\nin RIN prices are a completely understandable effect of the\\nprogram’s ever-increasing pressure to expand renewable\\nvolumes. Pushing out along the supply curve takes the raw\\n\\n 40\\nmarket price of the RIN-eligible fuel steadily into higher\\nrealms—except to the extent that production innovations or\\neconomies may tend to lower costs. So far as appears, it has\\nnothing to do with EPA’s allocation of the obligation.\\n\\n What about EPA’s concern, petitioners ask, that including\\nblenders in the point of obligation would expand the number of\\nobligated parties and, as a result, ratchet up the program’s\\ncomplexity? Isn’t that hard to square with EPA’s claim, made\\nyears earlier, that “essentially all downstream blenders . . . are\\n[already] regulated parties”? Alon Br. 42 (quoting 75 Fed.\\nReg. at 14,722/2). Again, not at all. Although the participation\\nof all (or nearly all) blenders in the RIN market subjects them\\nto RFS registration, recordkeeping, and reporting\\nrequirements, “the majority of these downstream [regulated]\\nparties are . . . currently not obligated parties.” EPA Denial at\\n69, Alon J.A. 123 (emphasis added). As EPA explained, there\\n“is a significant distinction between being a ‘regulated party’\\nand being an ‘obligated party.’” Id. “Obligated parties must\\nmeet all of [the requirements faced by regulated parties] and\\nalso calculate an annual renewable volume obligation, acquire\\nthe appropriate number of RINs in the market, practic[e] due\\ndiligence to ensure [the RINs’] validity, file annual compliance\\nreports demonstrating compliance, and maintain records to that\\neffect.” Id. at 69 n.205, Alon J.A. 123 (emphasis added); see,\\ne.g., 40 C.F.R. §§ 80.1427(a), 80.1450(a), 80.1451(a),\\n80.1454(a). It was not unreasonable for EPA to conclude that\\nimposing these burdens on additional entities would add to the\\nprogram’s complexity (and therefore be undesirable absent an\\nadequate offsetting benefit).\\n\\n Nor was it unreasonable to find that going down this\\nroute—overhauling a foundational element of the program—\\nwould create “uncertainty in the fuels marketplace.” EPA\\nDenial at 2, Alon J.A. 56. As EPA said, “all parties regulated\\n\\n 41\\nin the RFS program have made significant investments and\\ndecisions about their participation in the program and their\\nposition in the market on the basis of the existing regulations,\\nincluding the definition of obligated parties.” EPA Denial at\\n79, Alon J.A. 133. In these circumstances, it isn’t hard to\\nimagine how changing course could throw players off their\\ngame. Of course, as petitioners note, uncertainty may have\\n“plagued the RFS Program for years.” Alon Br. 37. But true\\nor not, EPA needn’t pile on; the cure for uncertainty isn’t\\nspawning more uncertainty.\\n\\n Taking a step back, petitioners launch a closing broadside\\nagainst the entire process. They assert that EPA “disregarded\\nthis Court’s remand” in ACE, 864 F.3d at 737, and arbitrarily\\ncredited some comments over others. Alon Br. 31–32, 55–56.\\nBut the ACE remand required, at most, that the agency “address\\nthe point of obligation issue.” 864 F.3d at 737. And, as\\ndetailed throughout this opinion, the agency has done so\\nreasonably, analyzing the data and explaining its decision.\\nNothing more was required.\\n\\n\\n * * *\\n\\n We have considered the Alon Petitioners’ other arguments\\nand have found them to be either without merit or, in the case\\nof the argument relying on 42 U.S.C. § 7545(o)(5)(A), see\\nAlon Br. 34, insufficiently developed, see, e.g., Masias v. EPA,\\n906 F.3d 1069, 1077 (D.C. Cir. 2018). For the foregoing\\nreasons, the petitions for review are denied.\\n\\n V. 2017 Annual Volumetric Rule\\n\\n We turn now to the Coffeyville Petitioners’ challenges.\\n\\n 42\\nA. Point of Obligation\\n\\n We first consider the Coffeyville Petitioners’ challenge to\\nEPA’s decision in the 2017 Rule not to reassess which\\ncategories of industry players are “obligated parties” under the\\nrenewable fuel program. As the Coffeyville Petitioners read it,\\nthe statutory provision requiring EPA to set annual renewable\\nfuel percentage standards also imposes on EPA a\\nnondiscretionary duty to reconsider—every year—which types\\nof entities are obligated to demonstrate to EPA compliance\\nwith the percentage standards. See 42 U.S.C. § 7545(o)(3)(B).\\nThey claim EPA shirked that duty when it treated the issue as\\nbeyond the scope of its 2017 annual rulemaking. EPA counters\\nthat it identified the obligated parties in 2007 pursuant to\\nCongress’s mandate to set “compliance provisions” for the new\\nrenewable fuel program, id. § 7545(o)(2), reaffirmed that\\ndecision in 2010, and that nothing in the mandate to calculate\\nthe annual percentage standards requires it to reconsider the\\npoint of obligation each year. EPA also asserts that it\\nappropriately addressed the Coffeyville Petitioners’ complaints\\nthat it obligated the wrong parties in a separate proceeding from\\nits annual volumetric rulemaking.\\n\\n 1. Jurisdiction\\n\\n EPA and a coalition of Respondent-Intervenors\\nrepresenting the renewable fuel and refinery industries assert\\nthat we lack jurisdiction because the Coffeyville Petitioners\\neffectively challenge the compliance rule that has been on the\\nbooks for a decade or so, see 40 C.F.R. § 80.1406(a)(1), and\\ndid not petition within 60 days of its publication or within 60\\ndays of any valid “grounds arising” thereafter. See 42 U.S.C.\\n§ 7607(b)(1); Med. Waste Inst. v. EPA, 645 F.3d 420, 427\\n(D.C. Cir. 2011). But petitioners are not challenging EPA’s\\ndecision to adopt the rule in 2007 or retain it in 2010. Rather,\\n\\n 43\\nthey contend that the provision calling on EPA to set annual\\nvolumes of biofuels “applicable to refineries, blenders, and\\nimporters, as appropriate,” requires EPA to reassess each year\\nwhether the point of obligation set when the agency established\\nthe program is still “appropriate,” or if EPA should re-assign it\\nand restructure the RIN market and other compliance\\ninfrastructure going forward. See 42 U.S.C.\\n§ 7545(o)(3)(B)(ii)(I). That challenge was timely filed within\\n60 days of the promulgation of the annual fuel standards. See\\nsupra Part IV.A.1.\\n\\n 2. Merits\\n\\n This dispute turns on the roles of two provisions of the\\nstatute directing the EPA to establish and run a Renewable Fuel\\nProgram, 42 U.S.C. § 7545(o)—paragraphs (2) and (3).\\n Paragraph (2) directs EPA to “promulgate regulations”\\nsetting up a program to “ensure that transportation fuel sold or\\nintroduced into commerce in the United States . . . contains at\\nleast the applicable volume[s] of renewable fuel,” as specified\\nin subparagraph (2)(B). Id. § 7545(o)(2)(A)(i). Among the\\nparameters Congress required EPA to include were\\n“compliance provisions applicable to refineries, blenders,\\ndistributors, and importers, as appropriate,” to ensure that the\\nrequirements of paragraph (2), including the applicable volume\\nrequirements specified in subparagraph (2)(B), are met. Id.\\n§ 7545(o)(2)(A)(iii)(I), (B). There is no question that EPA has\\nauthority to set those parameters, including the point of\\nobligation, and to adjust them if a change is needed.\\n Paragraph (3), in turn, requires EPA to determine and\\npublish annual renewable fuel obligations designed to\\n“ensure[] that” the applicable volumes specified in paragraph\\n(2) are met. Id. § 7545(o)(3)(B)(i). Those renewable fuel\\nobligations must:\\n\\n 44\\n (I) be applicable to refineries, blenders, and\\n importers, as appropriate;\\n (II) be expressed in terms of a volume percentage\\n of transportation fuel sold or introduced into\\n commerce in the United States; and\\n (III) . . . consist of a single applicable percentage\\n that applies to all categories of persons\\n specified in subclause (I).\\nId. § 7545(o)(3)(B)(ii)(I)–(III).\\n The parties’ dispute centers on the meaning of “as\\nappropriate” in subclause (3)(B)(ii)(I). The Coffeyville\\nPetitioners contend that the phrase unambiguously requires\\nEPA annually to reconsider which parties it is “appropriate” to\\nobligate to meet the renewable fuel obligations. EPA responds\\nthat the statute is, at most, ambiguous as to whether Congress\\nexpected EPA annually to revisit the obligated-parties\\ndesignation, or whether the agency may generally rely on the\\n“appropriate[ness]” finding it made pursuant to its paragraph\\n(2) authority. We begin by asking “whether Congress has\\ndirectly spoken to the precise question at issue,” and conclude\\nthat it has not. Chevron, 467 U.S. at 842.\\n EPA reads “as appropriate” in paragraph (3) to mean that\\nthe agency has “discretion” to decide whether, when, and how\\nto reassess which of three types of industry actors—refineries,\\nblenders, and importers—should continue to bear the point of\\nobligation, as originally designated in the compliance\\nprovisions. See EPA Denial, Coffeyville J.A. 779–80. At oral\\nargument, the agency conceded that its exercise of this\\ndiscretion is reviewable, so that the exclusion of the point of\\nobligation issue from an annual rulemaking could, under other\\ncircumstances, constitute an abuse of discretion. Rec. of Oral\\nArg. 1:22:22–1:24:00; 1:37:45–1:40:11.\\n\\n 45\\n The Coffeyville Petitioners object that the phase\\n“applicable . . . as appropriate” means applicable as\\ncontemporaneously determined to be appropriate in the annual\\nvolumetric rulemakings. Coffeyville Br. 30–33; see also Conc.\\nOp. 4–5, 7. But paragraph (3) does not specify when or in what\\ncontext EPA must make its appropriateness determination, nor\\ndoes the phrase “as appropriate” itself specify a particular\\ntemporal dimension—as between, for example, parties\\nappropriately designated in the past (as EPA interprets it) and\\nparties now appropriately selected (as Coffeyville insists). The\\nterm “appropriate” “naturally and traditionally includes\\nconsideration of all the relevant factors,” Michigan v. EPA,\\n135 S. Ct. 2699, 2707 (2015) (quoting White Stallion Energy\\nCenter, LLC v. EPA, 748 F.3d 1222, 1266 (D.C. Cir. 2014)\\n(Kavanaugh, J., concurring in part and dissenting in part)), but\\nit does not dictate when that consideration must be made. In\\nother words, the requirement that the point of obligation be\\n“appropriate” is at most grounds for assessing whether the\\nagency adequately explained its policy choices regarding the\\nappropriateness determination, not for imposing our own gloss\\non that broad term as a matter of law. See Kisor v. Wilkie, 139\\nS. Ct. 2400, 2448–49 (2019) (Kavanaugh, J., concurring in the\\njudgment) (“[S]ome cases involve regulations that employ\\nbroad and open-ended terms like ‘reasonable,’ ‘appropriate,’\\n‘feasible,’ or ‘practicable.’ Those kinds of terms afford\\nagencies broad policy discretion, and courts allow an agency to\\nreasonably exercise its discretion to choose among the options\\nallowed by the text of the rule. But that is more State Farm\\nthan Auer” or Chevron (emphasis added)). Here, as explained\\nbelow, EPA reasonably exercised its discretion, and explained\\nits decision, to address the point of obligation issue in a\\nseparate proceeding from its annual volumetric rulemaking.\\n The fact that paragraphs (2) and (3) both include the phrase\\n“as appropriate” does not make the Coffeyville Petitioners’ the\\n\\n 46\\nonly permissible interpretation. Even if, as our colleague\\ncontends, Conc. Op. 7, the two phrases bear the exact same\\nmeaning, but see Coffeyville Br. 30–31 (arguing that\\nparagraphs (2) and (3) are “worded differently” and have\\n“different contexts”), paragraph (3) simply does not dictate\\nwhen or in what context EPA must make the appropriateness\\ndetermination. It is the surrounding context, not the phrases\\nthemselves, that suggests when EPA might make that choice.\\n Unable to point to any express textual requirement that\\nEPA annually reconsider the point of obligation, the\\nCoffeyville Petitioners contend that, had Congress intended to\\nallow EPA in annual volumetric rulemakings to rest on its\\nparagraph (2) appropriateness determination, subclause\\n(3)(B)(ii)(I) could have more simply cross-referenced\\nparagraph (2). But replacing subclause (3)(B)(ii)(I) with a\\nsimpler cross-reference would not have achieved quite the\\nsame effect. While refineries, blenders, importers, and\\ndistributors may all be subject to compliance provisions under\\nparagraph (2), paragraph (3)’s applicability provision points to\\na more limited universe of potential obligated parties—to\\nrefineries, blenders, and importers, but not distributors. That\\nsupports EPA’s understanding of subclause (3)(B)(ii)(I) as a\\ncross-reference that also clarifies a limit on EPA’s options in\\nsetting the point of obligation. See 72 Fed. Reg. at 23,923/2\\n(preamble for compliance rule referencing paragraph (3)’s\\napplicability provision).\\n We are unpersuaded by the suggestion that such limitation\\nwas so clear even without subclause (3)(B)(ii)(I) that EPA’s\\nreading of that applicability provision renders it superfluous.\\nThe suggestion is that, because distributors do not “introduce”\\nfuel into commerce, they could not be obligated parties in any\\nevent, with or without the applicability provision. Our\\ncolleague posits that blenders—but not distributors—can in\\nfact “introduce” transportation fuel into commerce by blending\\n\\n 47\\ngasoline and diesel fuel with other fuels that have not already\\nbeen introduced by someone else. Conc. Op. 11. But that\\nargument rests on a complicated series of inferences from spare\\nstatutory text, as well as post-enactment regulations that do not\\nnecessarily show what the statute must have meant. For\\nexample, our colleague reasonably infers that the national\\n“volume[] of transportation fuel,” 42 U.S.C. § 7545(o)(3)(A),\\nmust be counted at the first moment each gallon of fuel enters\\ncommerce, in order to avoid double-counting. But, the statute\\nnowhere states the point directly. Still less clearly does the\\nstatute state our colleague’s corollary—necessary to the\\nsurplusage argument—that the point of obligation must also be\\nplaced, if at all, at the first moment a gallon of fuel enters\\ncommerce. That corollary is less obvious, because the\\nstatutory link between the point of obligation and entry into\\ncommerce is not ironclad: All parties agree that not every\\ngallon of transportation fuel must be subject to the point of\\nobligation upon entry into commerce. The statute plainly\\nallows EPA to obligate an “appropriate” subset of the three\\ncategories of parties. And obligating blenders would involve\\ndouble counting unless the transportation fuel they use to create\\nblends were not already counted upon its importation or sale to\\nthem. Under the circumstances, it is reasonable to read\\nsubclause (3)(B)(ii)(I) as clarifying what is at best a non-\\nobvious inference that distributors cannot be subjected to the\\npoint of obligation.\\n So subclause (3)(B)(ii)(I) as EPA reads it is not a\\nsuperfluity, but makes clear that EPA may have permissibly\\nplaced the point of obligation on refineries, blenders, and\\nimporters, but not distributors. For the same reason, subclause\\n(3)(B)(ii)(III), which requires that the annual standards apply\\n“to all categories of persons specified in subclause (I),” id.\\n§ 7545(o)(3)(B)(ii)(III), does not contain what our colleague\\nviews as an unnecessary double cross-reference to paragraph\\n\\n 48\\n(2), because it, too, is operative in not just cross-referencing,\\nbut also clarifying a limit on the three permissible targets of its\\n“single applicable percentage.”\\n The thrust of the Coffeyville Petitioners’ retort—that if\\nCongress had wanted to confer discretion or provide a limiting\\ncross-reference to paragraph (2), it would have said so more\\nplainly—applies with greater force against their own reading.\\nHad Congress intended EPA to consider on an annual basis\\nwhether to redo the point of obligation designation—a\\ndesignation that no-one disputes is a necessary cornerstone of\\nthe paragraph (2) compliance provisions—it knew how to\\nimpose such a requirement. The Clean Air Act’s provisions on\\nambient air quality, for instance, require EPA to “complete a\\nthorough review” of the air quality standards “at five-year\\nintervals” and “promulgate such new standards as may be\\nappropriate.” Id. § 7409(d)(1). The Act’s provisions\\ncontrolling hazardous air pollutants emitted from major and\\narea sources require EPA to “review, and revise as necessary”\\nthe applicable emission standards “no less often than every 8\\nyears.” Id. § 7412(d)(6). Paragraph (3) of the RFS program,\\nin contrast, does not tell EPA to “complete a thorough review,”\\nor “review, and revise as necessary” its point of obligation\\ndecision—or anything even close.\\n To be sure, EPA’s reading is not ineluctable. We do not\\ndoubt that Congress could have more directly provided that the\\nrenewable fuel obligations do not apply to distributors. See\\nConc. Op. 8. But, for the reasons discussed, we are\\nunconvinced that paragraph (3) plainly requires EPA to\\nconsider adjusting the point of obligation each year. See\\nValero Energy Corp. v. EPA, No. 7:17-cv-00004-O, 2017 WL\\n8780888, at *4 (N.D. Tex. Nov. 28, 2017) (holding that “there\\nis no clear statutory mandate . . . obligating [EPA] to evaluate\\nor adjust . . . what entities are ‘appropriate[ly]’ forced to\\ncomply with” the annual renewable fuel obligations\\n\\n 49\\n(alterations in original) (quoting 42 U.S.C.\\n§ 7545(o)(3)(B)(ii)(I))). Accordingly, we conclude that the\\nmeaning of “as appropriate” in paragraph (3) is ambiguous and\\nturn now to whether EPA’s construction is “based on a\\npermissible construction of the statute.” Chevron, 467 U.S. at\\n843.\\n The difficulty of squaring the Coffeyville Petitioners’\\nreading of “as appropriate” with the structure and purpose of\\nthe statute convinces us of the reasonableness of EPA’s\\ninterpretation. As a structural matter, the RFS program\\ncontains not only “annual” volumetric determinations, Conc.\\nOp. 1, but also a slew of compliance provisions that are not\\nannually re-determined. As a practical matter, the point of\\nobligation is the foundational “compliance provision” of the\\nentire renewable fuels program; EPA could not “ensure” that\\napplicable volumes of renewable fuels are introduced into the\\nnation’s transportation fuel supply without designating the\\nparties responsible for carrying the renewable fuel standards\\ninto operation. Id. § 7545(o)(2)(A)(i). To that end, in writing\\nthe compliance provisions, EPA placed the renewable fuel\\nobligation on the entities at the head of the United States supply\\nchain for nonrenewable fuels—domestic refiners, and\\nimporters of fuel refined elsewhere. See 72 Fed. Reg. at\\n23,923–24. After additional consideration, EPA in 2010\\nadhered to that decision. See 75 Fed. Reg. at 14,721–22\\n(codified at 40 C.F.R. § 80.1406(a)(1)); see also Monroe\\nEnergy, 750 F.3d at 912. No one challenged EPA’s decision\\nin 2007 or 2010, and EPA declined to revisit the issue in\\nresponse to comments in the 2017 annual rulemaking urging it\\nto shift the 2017 point of obligation to blenders. See Response\\nto Comments at 542, Coffeyville J.A. 761.\\n The focus of the annual rulemakings, in contrast, is to\\ntranslate the applicable volumes—as specified in paragraph\\n(2), or set according to the process there described—into\\n\\n 50\\npercentage requirements for each renewable fuel. 42 U.S.C.\\n§ 7545(o)(3)(B)(ii). It would be strange indeed if Congress\\nrequired EPA, as it went about its annual quantitative standard-\\nsetting duties, also to rethink a choice so basic to the RFS\\nprogram’s architecture. This implausibility is illuminated by\\nthe fact that Congress required EPA to facilitate statutory\\ncompliance through a credit trading program, which of\\nnecessity requires some year-to-year stability. See id.\\n§ 7545(o)(5). EPA responded by setting up the RIN system,\\nwith flexibility anchored to a fixed baseline—the point of\\nobligation. The compliance system is flexible in that RINs may\\nbe retired in compliance demonstrations not only in the\\ncompliance year during which they were generated, but also\\nthroughout the ensuing compliance year, 40 C.F.R.\\n§ 80.1427(a)(6), and obligated parties may carry over excess\\nRINs or RIN deficits from year to year, id. § 80.1427(a)(1),\\n(5)–(6); see Monroe Energy, 750 F.3d at 913.\\n Annual changes to the point of obligation could cause\\n“disparities in RIN-holdings,” leaving formerly obligated\\nparties with “significantly more RINs, including carryover\\nRINs, than they desire or can use” and newly obligated parties\\nwith “lower balances than they would desire to protect\\nthemselves against shortfalls in RIN availability or RIN price\\nvolatility.” EPA Denial at 78, Coffeyville J.A. 850. “[A]\\nchange to the point of obligation could also cause volatility in\\nthe [RIN] market,” inhibiting the “ability [of] parties that\\npossess excess carryover RINs to recover the cost of the RINs\\nthey hold by selling them to other parties.” Id. It is not\\nplausible that Congress meant EPA to consider uprooting the\\nbaseline of the RFS program every year. The real stretch is that\\nCongress would have imposed such an onerous and potentially\\ndisruptive duty merely by use of the phrase “as appropriate.”\\n The Coffeyville Petitioners’ reading is not made any more\\nplausible by highlighting the likelihood that, on annual\\n\\n 51\\nconsideration of the point of obligation, EPA would only need\\nto consider recent information, and likely would stay its course.\\nEven if the point of obligation in fact rarely changed, the mere\\n“reconsider[ation]” of the framework would “likely cause\\ndelays to the investments necessary to expand the supply of\\nrenewable fuels in the United States.” See EPA Denial at 2,\\nCoffeyville J.A. 774. EPA reasoned that “fuel[] industry\\nparticipants [would] withhold significant investment decisions\\nuntil the EPA’s final decision and the fallout from the decision\\nare known.” Id. at 81–82, Coffeyville J.A. 853–54. Insisting\\nthat the issue be on the regulatory agenda every year would sow\\n“significant market uncertainty and potential turmoil” into the\\nRFS program without offsetting benefit. Id.\\n Furthermore, any requirement that an agency repeatedly\\ngo through a regulatory process on an issue that promises to\\ndraw a regular parade of criticism from interest groups with\\nample resources is itself burdensome. See AT&T Corp. v. FCC,\\n220 F.3d 607, 630–31 (D.C. Cir. 2000). This issue is no\\nexception. As discussed above, EPA in 2016 and 2017\\nconsidered and decided against reopening its point of\\nobligation rule. In so doing, it received upwards of 18,000\\ncomments and published an exhaustive, 85-page decision. See\\nEPA Denial at 1–85, Coffeyville J.A. 771–857. “Given the\\ntime pressure associated with its annual standards rulemaking,”\\nEPA believes it would not be feasible or worthwhile to\\nundertake such reconsideration annually. Id. at 7 n.10,\\nCoffeyville J.A. 779. Indeed, as EPA acknowledged at oral\\nargument, the agency “has been late on [its annual rules]\\nbefore,” even “when [it hasn’t] taken up the point of\\nobligation.” Rec. of Oral Arg. 1:25:44–52. “[A]dd[ing] on\\nthe” duty to reassess the point of obligation annually, EPA tells\\nus, “would be a significant burden.” Id. at 1:25:55–1:26:05.\\nOur colleague doubts that EPA’s year-to-year burden would be\\nappreciable, but we see no ground to question EPA’s judgment\\n\\n 52\\nto the contrary. It seems unlikely that Congress wrote the\\napplicability provision in order to heap that annual duty onto\\nEPA’s plate. It seems even less likely given the absence of\\nreason to think that yearly second-guessing of program\\nfundamentals makes sense, or that, when and if the need for a\\nprogram restructuring arises, EPA would fail to act. Indeed,\\nthe statute elsewhere explicitly requires EPA to conduct\\n“periodic reviews of . . . the feasibility of achieving compliance\\nwith the [applicable volume] requirements.” 42 U.S.C. §\\n7545(o)(11). That provision has not been briefed, but would\\nappear to require EPA to reconsider the point of obligation if it\\nconcluded that its placement was obstructing compliance.\\n Finally, EPA’s approach coheres with basic principles of\\nadministrative law. In general, the choice between various\\nprocedural channels lies within the “informed discretion of the\\nadministrative agency.” SEC v. Chenery Corp., 332 U.S. 194,\\n203 (1947). That discretion properly includes judgments about\\nthe scope of rulemakings and when to relegate ancillary issues\\nto separate proceedings: “Agencies, like legislatures, do not\\ngenerally resolve massive problems in one fell regulatory\\nswoop.” Massachusetts, 549 U.S. at 524; see, e.g., Grp.\\nAgainst Smog & Pollution, 665 F.2d at 1292 (“. . . EPA cannot\\nsoundly be charged with arbitrariness merely because it chose\\na separate rulemaking proceeding as the process for proposing\\na revised standard in lieu of an undertaking to do so in the\\nnarrower context of the opacity standard proceedings as\\npetitioners requested.”). Once the agency has resolved an issue\\nin a separate proceeding, it may defend against related criticism\\nby “simply refer[ing]” to the other proceeding, so long as the\\n“reasoning remains applicable and adequately refutes the\\nchallenge.” Bechtel v. FCC, 10 F.3d 875, 878 (D.C. Cir. 1993).\\nEPA reasonably reads “as appropriate,” in paragraph (3)(B), to\\nleave undisturbed these background norms of broad but\\nreviewable procedural discretion.\\n\\n 53\\n Our holding today does not give EPA the limitless and\\nunreviewable discretion feared by our colleague. As we have\\nsaid, EPA’s determination as to whether it is “appropriate” to\\nreconsider the point of obligation in the context of an annual\\nvolumetric rulemaking is reviewable for abuse of discretion.\\nEPA did not abuse its discretion in refusing to do so here.\\nIndeed, it considered whether to change the point of obligation\\nrule in a separate, contemporaneous proceeding that yielded a\\nfinal order that we also have reviewed and found to be\\nadequately justified. See supra Part IV.B. We do not address\\nwhether it would be an abuse of discretion for EPA to refuse to\\nreconsider the point of obligation—in an annual volumetric\\nrulemaking or otherwise—in extreme circumstances akin to\\nthose posited by our colleague’s hypothetical about the\\ncontinuing study of an abolished tort. See Conc. Op. 5.\\n In sum, we hold that EPA permissibly rejected the claim\\nthat paragraph (3) requires the agency annually to reassess the\\npoint of obligation in the renewable fuel program. Because\\nEPA has no duty to reconsider the appropriateness of its point\\nof obligation regulation as part of its yearly determination of\\nvolumetric requirements, it was not arbitrary for EPA to treat\\ncomments complaining that it obligated the wrong parties as\\nappropriately assessed in a separate proceeding, and beyond\\nthe scope of proceedings for the 2017 volumetric rulemaking.\\nFor these reasons, we deny the Coffeyville Petitioners’ petition.\\nB. Cellulosic Biofuel Projection\\n\\n The Coffeyville Petitioners lob a variety of challenges at\\nEPA’s cellulosic biofuel projection for 2017. Many of these\\npetitioners, however, raised many of the same arguments\\nbefore. See ACE, 864 F.3d at 727–29 (addressing challenges\\nto EPA’s 2014–16 projection). We rejected those arguments\\nonce—and do so again.\\n\\n 54\\n First, the Coffeyville Petitioners contend that “EPA’s\\n[m]ethodology” for projecting cellulosic biofuel production is\\ninvalid because it “[c]hronically [o]verestimates [a]ctual\\n[p]roduction.” Coffeyville Br. 40. But that argument—that\\nEPA has “repeatedly . . . overshot the mark,” id. at 41—doesn’t\\napply to the methodology EPA actually used here, as we found\\nin ACE, 864 F.3d at 727–28. As we explained when petitioners\\ndeployed this same argument in challenging the 2014–16\\nprojection, “the majority of EPA’s prior overestimations”\\nutilized a different methodology—one that we rejected in API,\\n706 F.3d at 478–81, and that the EPA accordingly abandoned.\\nACE, 864 F.3d at 727. The new methodology—the one EPA\\nused here—has been applied only twice before. At the time\\nEPA made its final evaluation for 2017, that methodology had\\n(as detailed in the table below) undershot for 2015 and overshot\\nfor 2016. See Assessment of the Accuracy of Cellulosic\\nBiofuel Production Projections in 2015 and 2016, EPA-HQ-\\nOAR-2016-0004-3687, at 1–4 (Dec. 12, 2016), Coffeyville\\nJ.A. 515–18. This is hardly a pattern of chronic overestimation.\\n\\n EPA Actual\\n RFS\\n Estimate Production EPA Record\\n Compliance\\n (millions of (millions of Error ** Citation\\n Year\\n RINs) RINs)\\n Q1\\n Q2 [No Data in the Record]\\n 2015\\n Q3\\n Q4 35.00 53.36 - 34.4% J.A. 515\\n Q1\\n Q2 J.A.\\n 2016 230.00 198.39* + 15.9%\\n Q3 516–17\\n\\n Q4\\n\\n 55\\n * At the time of EPA’s assessment, the agency had actual RIN\\nproduction data for only the first nine months of 2016 (123.99 million\\ngallons). To calculate actual production for the year, EPA extrapolated the\\nlikely RIN generation for the last three months of the year based on the\\nhistorical relation (a multiple of 1.8) between the average quarterly\\ngeneration in the first three quarters and that of the last quarter, yielding a\\nfigure of 74.39 million RINs for the last quarter. See Coffeyville J.A. 516–\\n17.\\n\\n ** The EPA error has been calculated as the difference between the\\nEPA estimate and actual production, divided by the actual production.\\n\\n Second, the Coffeyville Petitioners claim that EPA failed\\nto generate a projection “based on” the cellulosic biofuel\\nestimate provided by the Energy Information Administration\\n(EIA), as required by statute. 42 U.S.C. § 7545(o)(7)(D)(i).\\nThat is so, they say, because 96% of EPA’s projected volume\\nwas biogas, a type of cellulosic biofuel that EIA did not include\\nin its estimate. See Coffeyville Br. 43–44. The problem for\\npetitioners, however, is that this closely parallels an argument\\nwe rejected in ACE: “[W]e do not agree that EPA failed to\\ngenerate projections ‘based on’ the [EIA’s] estimates,” even\\nthough those “estimates did not contain figures for [biogas]\\nproduction—production that accounts for the vast majority of\\ncellulosic biofuel” (“around 90 percent”). ACE, 864 F.3d at\\n724, 729. Here, as there, EPA showed sufficient “respect” for\\nEIA’s estimates. Id. at 729. When limited to fuels actually\\nanalyzed by EIA, EPA’s estimates were “very similar” to\\nEIA’s, id.; see 2017 Rule, 81 Fed. Reg. at 89,758/1, a fact that\\nthe Coffeyville Petitioners do not contest.\\n\\n Congress demanded no more. Nothing in the statute\\nrequired EPA to, as the Coffeyville Petitioners insist, “work[]\\nwith the EIA to develop information” about biogas.\\nCoffeyville Br. 44. “[T]he Administrator of the Energy\\nInformation Administration shall provide . . . an estimate,” 42\\nU.S.C. § 7545(o)(3)(A), and EPA shall “respect” it, API, 706\\n\\n 56\\nF.3d at 478. That’s it. In showing such respect, EPA, of\\ncourse, must “understand how EIA derived” its estimate.\\nCoffeyville Reply Br. 24. But, contrary to the Coffeyville\\nPetitioners’ contention, EPA did just that. The agency\\nidentified the types of cellulosic biofuels that EIA considered\\nand then, to test the integrity of its projection, conducted an\\napples-to-apples comparison, “limiting the scope of [its]\\nprojection to the companies assessed by EIA.” See 2017 Rule,\\n81 Fed. Reg. at 89,757–58. Nothing more was required.\\n\\n Third, the Coffeyville Petitioners object that EPA relied on\\n“information from the [biogas] industry”—an industry “with a\\ndirect financial interest in the outcome of the rule.” Coffeyville\\nBr. 45. The Petitioners characterize this as “reliance on\\nundisclosed information.” Id. But EPA did disclose the\\ninformation from the biogas industry—and that the information\\ncame from that industry; it just did so in the aggregate. See\\nOctober 2016 Assessment of Cellulosic Biofuel Production\\nfrom Biogas (2017), EPA-HQ-OAR-2016-0004-3711, at 2–6\\n(Dec. 12, 2016), Coffeyville J.A. 536–40. All the agency\\nwithheld was company-specific information, claiming that it\\nhad to withhold such data as confidential business information,\\nsee id. at 7, Coffeyville J.A. 541; see also 40 C.F.R. § 2.211(b),\\na claim that petitioners never even attempt to rebut, see Masias\\nv. EPA, 906 F.3d 1069, 1077 (D.C. Cir. 2018) (“It is not enough\\nmerely to mention a possible argument in the most skeletal\\nway, leaving the court to do counsel’s work . . . .” (quoting\\nSchneider v. Kissinger, 412 F.3d 190, 200 n.1 (D.C. Cir.\\n2005))).\\n\\n As for the implication of bias, we have previously upheld\\nEPA’s reliance on “biofuel producers’ own forecasts.” ACE,\\n864 F.3d at 728; see also API, 706 F.3d at 478 (recognizing\\nthat producers are an “almost inevitable source of\\ninformation”). Here, as in ACE, EPA did not “blindly adopt[]\\n\\n 57\\nthe facilities’ own forecasts”; it “performed its own\\ninvestigation.” 864 F.3d at 728; see October 2016 Assessment\\nof Cellulosic Biofuel Production from Biogas, supra, at 4,\\nCoffeyville J.A. 538 (“To verify the reasonableness of these\\nprojections, EPA compared the projected volume from each\\nregistered facility to the registered capacity of that facility.”).\\nPetitioners point to no unreasonable step by EPA in its efforts\\nto address “the uncertainty and unreliability identified by the\\n[Coffeyville] Petitioners.” ACE, 864 F.3d at 728.\\n\\n Fourth, the Coffeyville Petitioners protest EPA’s reliance\\non “facilities’ actual production in prior years” as a floor for\\nprojecting future cellulosic biofuel production. Coffeyville Br.\\n46. This was error, they say, because some companies might\\ncease production. Perhaps so. But, as we said in ACE,\\nalthough unforeseen issues “could prevent a producer from\\nmeeting” its prior year’s production, it was “reasonable” for\\nEPA to expect, as a general matter, “that a company’s output\\nwould grow year-over-year as the company gained\\nexperience.” 864 F.3d at 728. This seems especially true in an\\nindustry with the government’s wind surging at its back. And\\neven were EPA’s assumption not true for each company, any\\none facility’s shortfall, EPA explained, could be “off-set” by\\nnew facilities coming online or existing facilities exceeding the\\nhigh end of their projected production range. See Renewable\\nFuel Standard Program—Standards for 2017 and Biomass-\\nBased Diesel Volume for 2018: Response to Comments, EPA-\\nHQ-OAR-2016-0004-3753, at 444 (Dec. 12, 2016),\\nCoffeyville J.A. 707. This explanation fulfills EPA’s “duty to\\narticulate a ‘reasonable and reasonably explained’ approach to\\nsetting the low end of the production ranges.” ACE, 864 F.3d\\nat 729 (quoting Comtys. for a Better Env’t v. EPA, 748 F.3d\\n333, 335 (D.C. Cir. 2014)).\\n\\n 58\\n Fifth, the Coffeyville Petitioners complain that EPA\\nshould have based its cellulosic biofuel projections on “actual”\\nprior production. Coffeyville Br. 47. But this backward\\nlooking approach would have, in EPA’s view, “ignore[d] the\\npotential for facilities . . . to increase their fuel production\\nrates” and would have been “inappropriately conservative” in\\nlight of the “year-over-year increases” that EPA had observed\\n“in recent years.” 2017 Rule, 81 Fed. Reg. at 89,761/1. We\\ncannot say that in rejecting such an approach EPA violated “its\\nduty to take a ‘neutral aim at accuracy.’” ACE, 864 F.3d at 727\\n(quoting API, 706 F.3d at 476).\\n\\n For these reasons, we reject the Coffeyville Petitioners’\\nchallenges to EPA’s cellulosic biofuel projection for 2017. See\\nACE, 864 F.3d at 729.\\n\\nC. Cellulosic Waiver\\n\\n The Coffeyville Petitioners also challenge EPA’s decision\\nto use less than all of its discretionary cellulosic waiver\\nauthority to lower the 2017 requirements for advanced biofuel\\nand total renewable fuel. Having reduced the 2017 cellulosic\\nbiofuel requirement by 5.189 billion gallons, EPA had\\nauthority to reduce the advanced biofuel and total renewable\\nfuel requirements “by the same or a lesser volume.” 42 U.S.C.\\n§ 7545(o)(7)(D)(i). To decide by how much to reduce these\\nstatutory requirements, EPA first determines what reduction in\\nthe advanced biofuel requirement will yield a “reasonably\\nattainable” volume, and it then mechanically applies an\\nequivalent reduction to the total renewable fuel volume. 2017\\nRule, 81 Fed. Reg. at 89,752–53. Petitioners do not directly\\nchallenge this methodology. Instead, they argue that EPA\\napplied it arbitrarily in deciding to waive only 4.719 billion\\ngallons of the advanced biofuel volume for 2017, rather than\\nthe maximum available waiver of 5.189 billion gallons. We\\n\\n 59\\nagain reject some of their arguments as foreclosed by precedent\\nand others on their own terms.\\n\\n First, the Coffeyville Petitioners argue that EPA sought to\\njustify its 2017 advanced biofuel volume in part by making an\\nimpermissible comparison to the statutory volume set by\\nCongress for 2022. In response to a comment expressing\\nconcerns about utilization of non-cellulosic advanced biofuels\\n(which could be food-based) and possible adverse effects on\\nfood availability, EPA noted that its “reasonably attainable”\\nnon-cellulosic advanced biofuel volume for 2017\\n(approximately 4 billion gallons) was “somewhat higher than\\nthe level envisioned in the statute for 2017” (3.5 billion), “but\\nwell below the level of such fuels Congress expected would be\\nused by 2022” (5 billion). Response to Comments at 214,\\nCoffeyville J.A. 689. According to petitioners, “by comparing\\n2017 volumes with 2022 statutory targets, EPA departed from\\nCongress’s intent.” Coffeyville Br. 50.\\n\\n However, nothing in the statute forbids EPA from taking\\naccount of future statutory volumes in this way. Although\\nCongress specified presumptively applicable volumes for\\ncertain years, it also provided waiver authority to depart from\\nthose volumes. Indeed, the discretionary waiver provision\\nnecessarily empowers EPA to depart upward from the statutory\\nlevel of non-cellulosic advanced biofuel for a given year:\\nreducing the advanced biofuel volume by less than the\\nreduction in cellulosic biofuel, as section 7545(o)(7)(D)(i)\\npermits, is mathematically equivalent to increasing the volume\\nof non-cellulosic advanced biofuels, to “partially backfill for\\nmissing cellulosic biofuel.” 2017 Rule, 81 Fed. Reg. at\\n89,763/1. As we have noted, the cellulosic waiver provision\\n“grants EPA ‘broad discretion’ to consider a variety of factors”\\nin exercising this authority to depart from the presumptive\\nstatutory volumes. ACE, 864 F.3d at 733 (quoting Monroe\\n\\n 60\\nEnergy, 750 F.3d at 915). In this case, while deflecting a\\ncomment about food availability, EPA observed that its non-\\ncellulosic advanced biofuel volume for 2017—while higher\\nthan the statutory volume envisioned for that year—was lower\\nthan the presumptive statutory volume for the near future. And\\nit then reasonably concluded that this “somewhat higher\\ninterim volume reflect[ed] [its] assessment that it is appropriate\\nto allow non-cellulosic advanced biofuels to partially backfill\\nfor missing cellulosic volumes in light of the associated\\n[greenhouse gas] and energy security benefits.” Response to\\nComments at 214, Coffeyville J.A. 689.\\n\\n Second, the Coffeyville Petitioners argue that EPA failed\\nto explain its estimate of reasonably attainable 2017 imports of\\nsugarcane ethanol, a type of non-cellulosic advanced biofuel.\\nSugarcane ethanol imports have varied greatly from year to\\nyear, reaching a high of 681 million gallons in 2006 but falling\\nto 64 million gallons in 2014 and 89 million gallons in 2015.\\nSee 2017 Rule, 81 Fed. Reg. at 89,764. At the time of the 2017\\nRule, EPA expected only 76 million gallons to be imported in\\n2016, but it nonetheless adhered to its proposed estimate of 200\\nmillion gallons for 2017—an estimate originally based on\\nEPA’s judgment that circumstances in 2017 were “not . . .\\nsignificantly different” from circumstances in 2016, for which\\nEPA had also projected 200 million gallons. Id. at 89,763/3.\\nEPA acknowledged the “recent low import levels,” but also\\ncited “the difficulty in precisely identifying the reasons” for the\\nhistorical “high variability,” given “uncertainty” as to market\\nfactors including “ongoing growth in gasoline demand in\\nBrazil, and competing world demand for sugar.” Id. at 89,764–\\n65. The agency accordingly reaffirmed that 200 million\\ngallons “reflects a reasonable intermediate point between the\\nlower levels imported recently and the considerably higher\\nlevels that have been achieved in earlier years.” Id. at 89,765/2.\\n\\n 61\\n There is some force to petitioners’ objection to EPA’s\\nadherence to an estimate well over double the actual imports in\\nthe three preceding years. However, our review is “particularly\\ndeferential in matters implicating predictive judgments.”\\nMusic Choice v. Copyright Royalty Bd., 774 F.3d 1000, 1015\\n(D.C. Cir. 2014) (quoting Rural Cellular Ass’n v. FCC, 588\\nF.3d 1095, 1105 (D.C. Cir. 2009)). We accordingly upheld\\nEPA’s identical 2016 sugarcane ethanol estimate as\\n“reasonable and reasonably explained” in ACE, 864 F.3d at 736\\n(quotation marks omitted). In that case, we held that EPA\\nreasonably “concluded that ‘a somewhat lower level of imports\\nwill occur than the historic average’ of 300 million,” based on\\na similar analysis of market factors. Id. (quoting 2014–16\\nRule, 80 Fed. Reg. at 77,478/2). Here, we cannot say that one\\nmore year of low imports made it arbitrary for EPA to adhere\\nto that same projection for 2017.\\n\\n Third, the Coffeyville Petitioners object to EPA’s analysis\\nof supply and demand for regular gasoline (E0) and gasoline\\nwith added ethanol (E15 and E85). However, this analysis\\nplayed no role in EPA’s exercise of its discretionary cellulosic\\nwaiver authority under section 7545(o)(7)(D)(i). As noted\\nabove, EPA’s exercise of that authority rested entirely on its\\ndetermination of reasonably attainable advanced biofuel\\nvolumes. See 2017 Rule, 81 Fed. Reg. at 89,773–74. The\\ndisputed analysis of E0, E15, and E85 supported EPA’s\\nseparate decision not to invoke its “general waiver” authority,\\nunder section 7545(o)(7)(A)(ii), based on “inadequate\\ndomestic supply.” See generally ACE, 864 F.3d at 705–13.\\nBut in their opening brief, petitioners failed to challenge EPA’s\\ndecision not to invoke that separate waiver provision for 2017.\\nAlthough their reply brief gestures at this point, “an argument\\nfirst made in a reply brief is forfeited.” Bartko v. SEC, 845\\nF.3d 1217, 1225 n.7 (D.C. Cir. 2017).\\n\\n 62\\n Finally, the Coffeyville Petitioners take issue with EPA’s\\nresponse to various comments. We have considered these\\narguments and find them to be without merit.\\n\\n For these reasons, we reject the Coffeyville Petitioners’\\nchallenges to EPA’s exercise of its discretionary cellulosic\\nwaiver authority to reduce advanced biofuel and total\\nrenewable fuel volumes for 2017.\\n\\n VI. 2018 Volume for Biomass-Based Diesel\\n\\n Since 2012, EPA, acting in coordination with the\\nSecretaries of Energy and Agriculture, has calculated the\\nannual applicable volume (also known as the “volume\\nrequirement”) for biomass-based diesel based on a holistic,\\nbackward- and forward-looking consideration of relevant\\nfactors. In particular, it has set the volume requirement “based\\non a review of the implementation of the program during\\ncalendar years specified in the tables, and an analysis of” six\\nstatutorily enumerated factors: (1) “the impact of the\\nproduction and use of renewable fuels on the environment”; (2)\\n“the impact of renewable fuels on the energy security of the\\nUnited States”; (3) “the expected annual rate of future\\ncommercial production of renewable fuels, including advanced\\nbiofuels in each category (cellulosic biofuel and biomass-based\\ndiesel)”; (4) “the impact of renewable fuels on the\\ninfrastructure of the United States”; (5) “the impact of the use\\nof renewable fuels on the cost to consumers of transportation\\nfuel and on the cost to transport goods”; and (6) “the impact of\\nthe use of renewable fuels on other factors, including job\\ncreation, the price and supply of agricultural commodities,\\nrural economic development, and food prices.” 42 U.S.C.\\n§ 7545(o)(2)(B)(ii)(I)–(VI).\\n\\n EPA set the 2018 applicable volume for biomass-based\\ndiesel at 2.1 billion gallons, up from 2.0 billion gallons in 2017,\\n\\n 63\\nand 1.1 billion gallons above a statutory minimum that\\nCongress set to plateau at 1 billion gallons as of 2012. 2017\\nRule, 81 Fed. Reg. at 89,798/1; see 42 U.S.C.\\n§ 7545(o)(2)(B)(i)(IV), (v). NBB had asked EPA to set the\\nbiomass-based diesel volume at 2.5 billion gallons, and now\\nchallenges the volume EPA set as arbitrary and capricious and\\ncontrary to the Clean Air Act.\\n\\nA. NBB’s Standing\\n\\n Before considering the merits of NBB’s claims, we must\\nsatisfy ourselves that NBB has standing to assert them.\\nRespondent-Intervenors, the American Fuel & Petrochemical\\nManufacturers and the American Petroleum Institute, contend\\nthat NBB lacks standing because, they say, it has not shown\\nthat the 2017 Rule inflicted a cognizable injury on any of its\\nmembers.\\n\\n NBB has associational standing here for the same reasons\\nwe held it did in National Biodiesel Board v. EPA, 843 F.3d\\n1010, 1015 (D.C. Cir. 2016) (NBB v. EPA), where EPA’s\\nactions “incentivize[d] . . . compet[ition] with [NBB’s\\nmembers’] domestic production.” Here, too, NBB’s members\\n“compete with” the other industry players EPA’s rule is\\ndesigned to affect. Id. at 1016. Recall that biomass-based\\ndiesel is a nested subset of advanced and total renewable fuels,\\nsuch that NBB’s members get (1) a market for compelled\\nbuyers of the specified volume of biomass-based diesel, for\\nwhich they are the exclusive suppliers, plus (2) a market for\\ncompelled buyers of advanced and other renewable fuels\\nalongside a broad array of competing suppliers. See supra at\\n6–8, 11. The 2017 Rule preamble explains that biomass-based\\ndiesel “compet[es] for research and development dollars with\\nother types of advanced biofuels,” and that, “[b]y establishing\\n[the biomass-based diesel] volume requirement[] at [a] level[]\\n\\n 64\\nlower than . . . the expected production of [biomass-based\\ndiesel],” EPA was “creating the potential for some competition\\nbetween [biomass-based diesel] and other advanced biofuels to\\nsatisfy the advanced biofuel” applicable volume and providing\\n“incentives for the continued development of” those\\ncompetitors’ fuels. 81 Fed. Reg. at 89,797; see also EPA\\nCoffeyville Br. 24 (“Above 2.1 billion gallons, biomass-based\\ndiesel will have to compete with other types of advanced\\nbiofuel.”). Such competition will likely “temper to some extent\\n[biomass-based diesel] prices.” Final Statutory Factors\\nAssessment for the 2018 Biomass Based Diesel (BBD)\\nApplicable Volume, EPA-HQ-OAR-2016-0004-3708, at 10\\n(Dec. 12, 2016) (Supplemental Assessment), Coffeyville J.A.\\n533. That is a cognizable injury to NBB’s members. See NBB\\nv. EPA, 843 F.3d at 1015–16; see also Delta Constr. Co. v.\\nEPA, 783 F.3d 1291, 1299 (D.C. Cir. 2015) (per curiam).\\n\\n Though NBB failed to identify any of its members—\\nordinarily a prerequisite for organizations alleging\\nassociational standing, see Summers v. Earth Island Inst., 555\\nU.S. 488, 497–98 (2009)—that omission is not fatal here\\nbecause NBB’s members comprise “the entire biomass-based\\ndiesel category of the Renewable Fuel Standard[s]” and\\nrepresent no other interests. Coffeyville J.A. 134. Consistent\\nwith “the real purpose of the [standing] inquiry—that is, for the\\ncourt to be satisfied that the requisite injury really has occurred\\nor will occur in the future to members of the organization[],”\\nPub. Citizen v. FTC, 869 F.2d 1541, 1552 (D.C. Cir. 1989),\\nthere is no need to identify injured members when “all the\\nmembers of the organization are affected by the challenged\\nactivity,” Summers, 555 U.S. at 499 (citing NAACP v. Ala. ex\\nrel. Patterson, 357 U.S. 449, 459 (1958)). Because EPA’s rule\\nsubjects the biomass-based diesel industry to increased\\ncompetition, with anticipated pricing effects, NBB “meet[s] the\\n\\n 65\\nconstitutional prerequisites of injury, causation, and\\nredressability.” NBB v. EPA, 843 F.3d at 1015.\\n\\nB. Merits of NBB’s Challenges\\n\\n NBB advances two challenges to the applicable volume\\nEPA set for biomass-based diesel: First, that EPA erred in\\nconsidering the interaction of biomass-based diesel with the\\nyet-to-be established 2018 advanced biofuel applicable\\nvolume, and second, that EPA’s consideration of the six\\nstatutory factors was arbitrary and capricious and contrary to\\nlaw. We reject both claims.\\n\\n First, EPA reasonably chose a 2018 biomass-based diesel\\napplicable volume that would “maintain[] support for growth\\nin [biomass-based diesel] volumes” while also encouraging the\\n“development of other advanced biofuels.” 2017 Rule, 81 Fed.\\nReg. at 89,798/1. Congress directed EPA to consider the\\nlessons learned from its retrospective “review” of the program,\\napply them in its prospective “analysis of” the six statutory\\nfactors, and set a biomass-based diesel volume that will apply\\nfourteen months in the future. See 42 U.S.C.\\n§ 7545(o)(2)(B)(ii).\\n\\n EPA’s approach is consistent with the structure and\\npurposes of the statute. Congress set a minimum applicable\\nvolume for biomass-based diesel of one billion gallons for each\\nyear from 2012 forward, id. § 7545(o)(2)(B)(i)(IV), (v), while\\nspecifying statutory minimum volumes for the advanced\\nbiofuel category containing biomass-based diesel that grow\\nyear by year to 21 billion gallons by 2022, id.\\n§ 7545(o)(2)(B)(i)(II), (iii). EPA reasonably concluded that,\\nby nesting biomass-based diesel together with cellulosic (and\\nother unspecified) biofuels within the advanced biofuel\\ncategory, and specifically charting a higher, steeper, and longer\\ninitial growth curve for advanced biofuel, Congress anticipated\\n\\n 66\\nthat production of other types of advanced biofuels could step\\nup to help meet the advanced biofuel volume requirement. See\\n2017 Rule, 81 Fed. Reg. at 89,797/1. EPA also reasonably\\nconcluded that increasing fuel diversity serves one of\\nCongress’s primary goals in establishing the Renewable Fuel\\nStandards program: improving the nation’s “energy\\nindependence and security.” See Pub. L. No. 110-140,\\npreamble; see also 2017 Rule, 81 Fed. Reg. at 89,798/3. EPA\\nalso reasonably anticipated that enhanced competition in the\\nadvanced biofuels market would help “temper to some extent\\n[biomass-based diesel] prices,” Supplemental Assessment 10,\\nCoffeyville J.A. 533, thereby ameliorating Congress’s concern\\nthat, with a too-high target volume, the “price of biomass-based\\ndiesel fuel” would “increase significantly,” 42 U.S.C.\\n§ 7545(o)(7)(E)(ii). And fuel diversity may produce\\nenvironmental benefits insofar as certain advanced biofuels,\\nsuch as ethanol from food waste, will “likely have significantly\\nlower impacts on wetlands, ecosystems, and wildlife habitats”\\nthan would greater reliance on biomass-based diesel.\\nSupplemental Assessment 6, Coffeyville J.A. 529.\\n\\n NBB’s arguments to the contrary turn on reading the\\nstatutory directive that EPA “review . . . the implementation of\\nthe program during calendar years specified in the tables,” id.\\n§ 7545(o)(2)(B)(ii), to confine EPA’s consideration to\\nbiomass-based diesel’s statutory volumes and actual\\nperformance, and to prevent EPA from considering other fuel\\ncategories or future years. In particular, NBB takes issue with\\nEPA’s consideration of the not-yet-finalized 2018 advanced\\nbiofuel applicable volume, which NBB contends led EPA to\\nset the biomass-based diesel volume too low.\\n\\n NBB’s objections are not supported by the text or purpose\\nof the statute. Assuming NBB is right that EPA’s “review of\\nthe implementation of the program” consists of a retrospective\\n\\n 67\\nassessment, the agency must also conduct “an analysis of” six\\nstatutory factors. Id. § 7545(o)(2)(B)(ii). And those factors\\nplainly require a prospective assessment—an assessment that\\nwould likely miss “important aspects of the problem,” State\\nFarm, 463 U.S. at 43, if it ignored the interaction, now and in\\nthe future, of the requirements for all the categories of\\nrenewable fuels. See, e.g., 42 U.S.C. § 7545(o)(2)(B)(ii)(I)\\n(requiring an “analysis of” the “impact of the production and\\nuse of renewable fuels on,” among other things, “the\\nenvironment”). Though EPA set the biomass-based diesel\\nrequirement lower than NBB wished, Congress did not intend\\nto incentivize growth of biomass-based diesel “at all costs.”\\nACE, 864 F.3d at 714 (quoting Am. Express Co. v. Italian\\nColors Rest., 570 U.S. 228, 234 (2013)).\\n\\n NBB objects that setting the 2018 biomass-based diesel\\napplicable volume below expected production might lead to a\\ndepressed advanced biofuel volume for 2018. But the agency\\nspecifically anticipated “that the 2018 advanced biofuel\\nrequirement will be larger than the 2017 advanced biofuel\\nvolume requirement.” 2017 Rule, 81 Fed. Reg. at 89,798/1.\\nEPA has never set the biomass-based diesel applicable volume\\nat the “maximum potential production” level, id. at 89,799/3,\\nyet the “growing supply of” biomass-based diesel has\\nconsistently “allowed EPA to establish higher advanced\\nbiofuel” applicable volumes, id. at 89,797/3. EPA opted for\\n“allowing room within the advanced biofuel volume\\nrequirement for the participation of non-[biomass-based diesel]\\nadvanced fuels” as a reasonable way “to encourage the\\ndevelopment and production of a variety of advanced biofuels\\nover the long term without reducing the incentive for [biomass-\\nbased diesel] beyond the [biomass-based diesel applicable\\nvolume] in 2018.” Id. at 89,797–98.\\n\\n 68\\n Second, in setting the 2018 biomass-based diesel\\napplicable volume, EPA reasonably compared the advantages\\nand disadvantages of biomass-based diesel to those of other\\nfuels. NBB contends that the statute confines EPA to assessing\\nadvantages of biomass-based diesel over petroleum, not\\nconsidering other renewable fuels, and that the agency failed to\\n“meaningfully” consider the six factors. NBB Br. 9–10. Both\\narguments miss the mark.\\n\\n NBB suggests that, because the statute “was intended to\\n‘increase the production of clean renewable fuels’ as a\\nsubstitute for petroleum fuel,” id. at 21 (quoting Pub. L. No.\\n110-140, preamble), the only relevant comparison is to\\npetroleum, not to other categories of renewable fuel. But NBB\\nidentifies nothing in section 7545(o)(2)(B)(ii) or any other\\nsection that requires EPA to assess the performance of a\\nparticular renewable solely by reference to petroleum fuel. Its\\nanalysis would require us to read the term “renewable fuels”\\nused throughout section 7545(o)(2)(B)(ii) to refer to the single\\nrenewable fuel being analyzed, even though the statutory\\ndefinition of “renewable fuel” includes all types of renewables.\\nSee 42 U.S.C. § 7545(o)(1)(J). And if EPA could compare the\\nbenefits of each specific fuel only to petroleum, it might be\\nunable to set rational applicable volumes for each specified\\ncategory of renewable fuel after 2022, when the statute no\\nlonger sets any specific volumes. See id. § 7545(o)(2)(B)(iii)–\\n(v). EPA could easily conclude, for example, that each\\nrenewable fuel had a lower “impact . . . on the environment”\\nthan petroleum fuel, see id. § 7545(o)(2)(B)(ii)(I), but, no\\nmatter their differing merits in serving the statute’s goals, the\\nagency would be barred from making relative judgments\\namong renewable fuel categories.\\n\\n NBB also argues that EPA failed to give meaningful\\nconsideration to the six statutory factors, and instead “pre-\\n\\n 69\\ndetermined the outcome,” NBB Br. 20, but the record shows\\notherwise. EPA considered in detail how setting the biomass-\\nbased diesel applicable volume at a level higher or lower than\\n2.1 billion gallons would affect the six statutory factors. See\\n2017 Rule, 81 Fed. Reg. at 89,798–99. EPA further elaborated\\nits analysis of the factors in an 11-page supplemental\\nmemorandum evaluating effects of its proposed biomass-based\\ndiesel volume on renewable fuel production rates, the\\nenvironment, and the economy. See Supplemental Assessment\\n1–11, Coffeyville J.A. 524–34. EPA concluded that, over the\\nlong term, “[a] variety of different types of advanced biofuels,\\nrather than a single type such as [biomass-based diesel], would\\npositively impact energy security . . . and increase the\\nlikelihood of the development of lower cost advanced biofuels\\nthat meet the same [greenhouse gas] reduction threshold as\\n[biomass-based diesel].” Supplemental Assessment 3,\\nCoffeyville J.A. 526. EPA thus concluded that the statutory\\nfactors supported its biomass-based diesel applicable volume.\\nSee 2017 Rule, 81 Fed. Reg. at 89,798/3.\\n\\n At bottom, NBB’s objections rest on a policy\\ndisagreement: NBB urges that, instead of setting a level that\\nwould support continued investment in the biomass-based\\ndiesel industry while also encouraging producers of other types\\nof advanced biofuel to compete to satisfy the 2018 advanced\\nbiofuel applicable volume at lower cost, EPA should have\\nreserved to biomass-based diesel alone a volume nearer to that\\nindustry’s maximum production potential. But NBB’s\\nproposed “simple solution”—that EPA should have “set[] a\\nmeaningful [biomass-based diesel] volume” while planning to\\n“increas[e] the 2018 advanced-biofuel volume to provide room\\nfor the production of other advanced biofuels when it set that\\nvolume a year later,” NBB Br. 23—describes what EPA\\nactually did. A mere disagreement with the particular\\ncalibration of a line drawn in the exercise of an agency’s\\n\\n 70\\nreasonable judgment is no basis to invalidate a rule. Therefore,\\nwe deny NBB’s petition.\\n\\n VII. Conclusion\\n\\n For these reasons, the petitions for review are denied.\\n\\n So ordered.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4190,"cells":{"id":{"kind":"number","value":4644855,"string":"4,644,855"},"name":{"kind":"string","value":"VIRNETX INC., Appellant v. APPLE INC., Cisco Systems, Inc., Appellees"},"name_abbreviation":{"kind":"string","value":"Virnetx Inc. v. Apple Inc."},"decision_date":{"kind":"string","value":"2019-08-01"},"docket_number":{"kind":"string","value":"2017-1591, 2017-1592, 2017-1593"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Federal Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Prost, Moore, Reyna\", \"parties\": \"\", \"opinions\": [{\"author\": \"Opinion for the court filed by Chief Judge PROST\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n VIRNETX INC.,\\n Appellant\\n\\n v.\\n\\n APPLE INC., CISCO SYSTEMS, INC.,\\n Appellees\\n ______________________\\n\\n 2017-1591, 2017-1592, 2017-1593\\n ______________________\\n\\n Appeals from the United States Patent and Trademark\\nOffice, Patent Trial and Appeal Board in Nos. 95/001,788,\\n95/001,789, 95/001,856.\\n ______________________\\n\\n Decided: August 1, 2019\\n ______________________\\n\\n STEPHEN BLAKE KINNAIRD, Paul Hastings LLP, Wash-\\nington, DC, argued for appellant. Also represented by\\nNAVEEN MODI, JOSEPH PALYS, IGOR VICTOR TIMOFEYEV,\\nMICHAEL WOLFE, DANIEL ZEILBERGER.\\n\\n WILLIAM F. LEE, Wilmer Cutler Pickering Hale and\\nDorr LLP, Boston, MA, argued for appellee Apple Inc. Also\\nrepresented by REBECCA A. BACT, MARK CHRISTOPHER\\nFLEMING, LAUREN B. FLETCHER, DOMINIC E. MASSA;\\nBRITTANY BLUEITT AMADI, Washington, DC; SCOTT\\nBORDER, JEFFREY PAUL KUSHAN, Sidley Austin LLP, Wash-\\nington, DC.\\n\\n2 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n THEODORE M. FOSTER, Haynes & Boone, LLP, Dallas,\\nTX, argued for appellee Cisco Systems, Inc. Also repre-\\nsented by DAVID L. MCCOMBS, ANDREW S. EHMKE, DEBRA\\nJANECE MCCOMAS.\\n ______________________\\n\\n Before PROST, Chief Judge, MOORE and REYNA,\\n Circuit Judges.\\n Opinion for the court filed by Chief Judge PROST.\\nOpinion concurring in part and dissenting in part filed by\\n Circuit Judge REYNA.\\nPROST, Chief Judge.\\n Appellant VirnetX Inc. (“VirnetX”) appeals from deci-\\nsions of the Patent Trial and Appeal Board (“Board”) re-\\nlated to three inter partes reexaminations maintained by\\nApple Inc. (“Apple”) and Cisco Systems, Inc. (“Cisco”). The\\nUnited States Patent and Trademark Office (“PTO”) con-\\ncluded that Apple was not barred from maintaining its\\nreexams by the estoppel provision of the pre-America In-\\nvents Act (“AIA”) version of 35 U.S.C. § 317(b) (2006). The\\nBoard affirmed the Examiner’s determination that the\\nclaims of U.S. Patent Nos. 7,418,504 (“the ’504 patent”) and\\n7,921,211 (“the ’211 patent”) are unpatentable as antici-\\npated or obvious over the prior art of record. For the rea-\\nsons below, we affirm-in-part, vacate-in-part, and remand.\\n BACKGROUND\\n I\\n The ’504 and ’211 patents describe systems and meth-\\nods for “establishing a secure communication link between\\na first computer and a second computer over a computer\\nnetwork, such as the Internet.” ’211 patent col. 6 ll. 36–39.\\nThese systems and methods are “built on top of the existing\\nInternet protocol (IP).” Id. at col. 6 ll. 17–20.\\n\\nVIRNETX INC. v. APPLE INC. 3\\n\\n\\n\\n The Internet uses addressing systems for sending data.\\nIn such systems, physical computers can be identified by a\\nunique IP address (e.g., 123.345.6.7). VirnetX Inc. v. Apple\\nInc., 665 F. App’x 880, 882 (Fed. Cir. 2016).\\n Each IP address corresponds to a domain name (e.g.,\\nwww.Yahoo.com). See ’211 patent col. 38 ll. 58–61, col. 39\\nll. 13–14. A user on one computer can enter a domain name\\nin a web browser to communicate with another computer\\nor server. When the user does so, the computer sends a\\ndomain name service (“DNS”) request to the domain name\\nserver for the IP address corresponding to a given domain\\nname. Id. at col. 38 l. 58–col. 39 l. 3. The domain name\\nserver then looks up the IP address of the requested do-\\nmain name and returns it to the requesting computer. Id.\\nat col. 39 ll. 3–7.\\n Both VirnetX patents claim systems, methods, and me-\\ndia for creating secure communication links via DNS sys-\\ntems. For example, claim 1 of the ’211 patent recites:\\n 1. A system for providing a domain name service\\n for establishing a secure communication link, the\\n system comprising:\\n a domain name service system configured and ar-\\n ranged to\\n [1] be connected to a communication network,\\n [2] store a plurality of domain names and corre-\\n sponding network addresses,\\n [3] receive a query for a network address, and\\n [4] indicate in response to the query whether the\\n domain name service system supports establishing\\n a secure communication link.\\n\\n4 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n Independent claims 36 and 60 of the ’211 patent are\\ndirected to a “machine-readable medium” and a “method,”\\nrespectively. Otherwise, they mirror the requirements of\\nclaim 1. Independent claims 1, 36, and 60 of the ’504 pa-\\ntent are similar to the corresponding independent claims of\\nthe ’211 patent.\\n II\\n In 2010, VirnetX sued Apple in district court. VirnetX\\nalleged infringement of four patents, including the ’504 and\\n’211 patents. 1 VirnetX asserted claims 1, 2, 5, 16, 21, and\\n27 of the ’504 patent and claims 36, 37, 47, and 51 of the\\n’211 patent. VirnetX Inc. v. Apple Inc., 925 F. Supp. 2d 816,\\n824–25 (E.D. Tex. 2013).\\n In October 2011, Apple filed requests for inter partes\\nreexamination of the ’504 and ’211 patents with the PTO.\\nIn Apple’s Reexam Nos. 95/001,788 (“788 case”) and\\n95/001,789 (“789 case”) (collectively, “Apple reexams”), Ap-\\nple challenged all claims as anticipated by the Provino ref-\\nerence or rendered obvious by Provino in view of other prior\\nart. 2\\n The district court action proceeded to trial in late 2012.\\nA jury found the asserted claims infringed and not invalid.\\nThe jury awarded VirnetX $368 million in damages. Vir-\\nnetX, 925 F. Supp. 2d at 825. The district court denied Ap-\\nple’s motion for judgment as a matter of law (“JMOL”) or a\\nnew trial on these issues. Apple appealed.\\n On appeal, we affirmed the jury’s finding of no invalid-\\nity for all four patents. VirnetX, Inc. v. Cisco Sys., Inc., 767\\nF.3d 1308, 1323–24 (Fed. Cir. 2014) (“VirnetX I”). We also\\n\\n\\n\\n 1 VirnetX also alleged that Apple infringed certain\\nclaims of U.S. Patent Nos. 6,502,135 (“the ’135 patent”) and\\n7,490,151 (“the ’151 patent”).\\n 2 U.S. Patent No. 6,557,037 (“Provino”).\\n\\nVIRNETX INC. v. APPLE INC. 5\\n\\n\\n\\naffirmed the jury’s finding of infringement for many of the\\nclaims of the two patents not related to the present appeal\\n(’135 and ’151 patents). 3 Id. at 1320–22. We reversed the\\ndistrict court’s construction of the “secure communication\\nlink” claim term, vacated the related infringement finding\\nfor the two patents in this appeal (’504 and ’211 patents),\\nand vacated the damages award. Id. at 1317–19, 1319,\\n1323–24, 1325–34. We then remanded for further proceed-\\nings. 4 Id. at 1334.\\n Apple did not file a request for rehearing on the inva-\\nlidity or infringement issues affirmed in VirnetX I. Our\\nmandate issued on December 23, 2014. Apple did not seek\\nSupreme Court review. The 90-day period to file a petition\\nfor a writ of certiorari expired.\\n Meanwhile, in the parallel PTO reexamination pro-\\nceedings, the Examiner had found all claims of the ’504 and\\n’211 patents unpatentable. The Examiner issued Right of\\nAppeal Notices (“RANs”) in May 2014. VirnetX appealed\\nthe Examiner’s decisions to the Board.\\n VirnetX also petitioned the PTO to terminate the Apple\\nreexams based on the estoppel provision of § 317(b). The\\nPTO denied VirnetX’s petition in June 2015. J.A. 1659–67,\\n3138–48.\\n In September 2016, the Board affirmed the Examiner’s\\nfindings that all claims of the ’504 and ’211 patents were\\n\\n\\n 3 We reversed the district court’s conclusion related\\nto the doctrine of equivalents for one claim of the ’151 pa-\\ntent. See id. at 1322–23.\\n 4 On remand in 2016, a jury found that Apple in-\\nfringed claims 1, 2, 5, and 27 of the ’504 patent and claims\\n36, 47, and 51 of the ’211 patent, and awarded damages.\\nVirnetX Inc. v. Apple Inc., 324 F. Supp. 3d 836 (E.D. Tex.\\n2017). We recently affirmed. See VirnetX Inc. v. Cisco Sys.,\\nInc., No. 2018-1197, 748 F. App’x 332 (Fed. Cir. 2019).\\n\\n6 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nunpatentable. The Board denied VirnetX’s requests for re-\\nhearing.\\n After appealing the Board’s decisions in the Apple reex-\\nams to this court, VirnetX moved to remand. It argued that\\nthis court’s 2014 opinion was a “final decision” on Apple’s\\nattempt to prove invalidity under § 317(b). ECF No. 27. In\\nJune 2017, a motions panel denied the motion and directed\\nthe parties to address the issue in their merits briefing.\\nECF No. 36.\\n III\\n In December 2011, Cisco also filed a request for inter\\npartes reexamination with the PTO. 5 Cisco’s Reexam No.\\n95/001,856 (“856 case” or “Cisco reexam”) challenged\\nclaims 1–60 of the ’211 patent based on multiple grounds\\nof invalidity. Cisco based several of its arguments on the\\nLendenmann reference. 6\\n After extensive proceedings, the Examiner issued a\\nRAN in January 2015. The RAN rejected claims 36–60 of\\nthe ’211 patent as either anticipated or obvious. VirnetX\\nappealed to the Board. In September 2017, the Board af-\\nfirmed the Examiner, finding claims 36–54 and 57–60 an-\\nticipated by Lendenmann and claims 55 and 56 obvious\\nover Lendenmann and another reference. J.A. 88. VirnetX\\nmoved for rehearing, which was denied. VirnetX then ap-\\npealed.\\n We have jurisdiction under 28 U.S.C. § 1295(a)(4)(A).\\n\\n\\n\\n\\n 5 In 2010, VirnetX also sued Cisco in district court,\\nalleging inter alia infringement of claim 1 of the ’211 pa-\\ntent.\\n 6 Rolf Lendenmann, Understanding OSF DCE 1.1\\nfor AIX and OS/2, IBM International Technical Support\\nOrganization, pp. 1–45 (Oct. 1995) (“Lendenmann”).\\n\\nVIRNETX INC. v. APPLE INC. 7\\n\\n\\n\\n DISCUSSION\\n I\\n VirnetX’s appeal proceeds in two parts. First, VirnetX\\nargues as a threshold matter that Apple was estopped from\\nmaintaining its reexams under the pre-AIA version of\\n35 U.S.C. § 317(b) (2006). Second, VirnetX challenges the\\nmerits of the Board’s conclusions in both the Apple and\\nCisco reexams that the ’504 and ’211 patents are invalid.\\nWe address each in turn.\\n II\\n VirnetX contends that Apple’s reexams were barred by\\n§ 317(b). The PTO refused to terminate Apple’s reexams\\nbased on the conclusion that the provision did not apply.\\nIn VirnetX’s view, the PTO’s decision was inconsistent with\\ncontrolling case law, the statutory text, and Congress’s in-\\ntent. We agree.\\n The applicability of § 317(b) is a question of statutory\\ninterpretation. See Bettcher Indus., Inc. v. Bunzl USA,\\nInc., 661 F.3d 629, 639, 642–48 (Fed. Cir. 2011). Statutory\\ninterpretation is an issue of law we review de novo. Un-\\nwired Planet, LLC v. Google Inc., 841 F.3d 1376, 1379 (Fed.\\nCir. 2016).\\n The Patent Act requires that the PTO terminate a reex-\\namination once there has been a final decision on the pa-\\ntent challenger’s invalidity case in federal court. 35 U.S.C.\\n§ 317(b) (2006). Specifically, pre-AIA § 317(b) provides:\\n Once a final decision has been entered against a\\n party in a civil action arising in whole or in part\\n under section 1338 of title 28, that the party has not\\n sustained its burden of proving the invalidity of any\\n patent claim in suit . . . then neither that party nor\\n its privies may thereafter request an inter partes\\n reexamination of any such patent claim on the ba-\\n sis of issues which that party or its privies raised\\n\\n8 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n or could have raised in such civil action or inter\\n partes reexamination proceeding, and an inter\\n partes reexamination requested by that party or its\\n privies on the basis of such issues may not thereaf-\\n ter be maintained by the Office, notwithstanding\\n any other provision of this chapter.\\n The issue here is very narrow. Apple does not dispute\\nthe underlying procedural facts relevant to § 317(b), which\\nhighlight the advanced stage of these proceedings.\\n In 2010, Apple was sued for patent infringement. There\\nis no dispute that in that “civil action arising . . . under”\\n28 U.S.C. § 1338, Apple raised an affirmative defense of in-\\nvalidity on multiple grounds. 35 U.S.C. § 317(b). Thus,\\nApple had the “burden of proving the invalidity” of the as-\\nserted claims of both the ’504 and ’211 patents. Id.\\n In 2012, Apple presented its invalidity defenses to a\\njury. After a week-long trial, the jury concluded Apple\\nfailed to sustain its burden of proving the invalidity of any\\nasserted claim of the ’504 or ’211 patent.\\n In 2013, the district court determined that in light of\\nthe record at trial, substantial evidence supported the\\njury’s finding. The district court entered final judgment.\\n In 2014, Apple pursued an appeal of that decision here.\\nSpecifically, Apple argued that “the asserted claims are an-\\nticipated by the Kiuchi reference.” VirnetX I, 767 F.3d at\\n1323. After considering Apple’s challenges, we affirmed\\nthe district court’s decision on the issue of invalidity. Id.\\nat 1323–24. Apple did not file a request for rehearing. Ap-\\nple elected not to seek Supreme Court review. The 90-day\\nperiod to file a petition for certiorari expired.\\n In 2015, VirnetX petitioned the PTO to terminate Ap-\\nple’s pending reexaminations based on the estoppel provi-\\nsion of § 317(b). The PTO refused, accepting Apple’s\\nargument that there was not yet a final decision on valid-\\nity. J.A. 1661–63.\\n\\nVIRNETX INC. v. APPLE INC. 9\\n\\n\\n\\n On appeal, the sole issue is whether there has been a\\n“final decision” entered against Apple that it “has not sus-\\ntained its burden of proving” invalidity, triggering estoppel\\nof its parallel PTO reexam proceedings.\\n Apple does not dispute that since it first raised the is-\\nsue in 2010, it has now lost on invalidity before the jury,\\nthe district court, and this court. Nor does it dispute that\\nit allowed the time to petition for certiorari of the validity\\nissues in the first appeal to lapse. Instead, Apple takes the\\nposition that there was no final decision due to a hypothet-\\nical future appeal.\\n Apple’s argument proceeds as follows. In VirnetX I, we\\nvacated and remanded other issues unrelated to invalidity,\\nincluding infringement and damages. In Apple’s view, af-\\nter those issues are decided on remand and reviewed in a\\nsecond appeal, it may choose to file a petition for certiorari\\nof those non-validity issues. In so doing, it may also ask\\nthe Supreme Court to look at any issues from the first ap-\\npeal—including the validity determination.\\n Thus, the Court might elect to examine the validity is-\\nsues at some point in the future. The Court has the ability\\nto consider not only those issues “‘before the court of ap-\\npeals upon the second appeal,’” but also the “questions\\nraised on the first appeal.” Mercer v. Theriot, 377 U.S. 152,\\n153–54 (1964) (per curiam) (quoting Hamilton-Brown Shoe\\nCo. v. Wolf Bros. & Co., 240 U.S. 251, 257 (1916)); see also\\nMajor League Baseball Players Ass’n v. Garvey, 532 U.S.\\n504, 508 n.1 (2001) (per curiam) (noting Supreme Court\\nhas “authority to consider questions determined in earlier\\nstages of the litigation where certiorari is sought from the\\nmost recent of the judgments of the Court of Appeals”). Ac-\\ncording to Apple, the decision is only final when there is no\\npotential that the Supreme Court might some day examine\\nthe invalidity issue during a second appeal.\\n\\n10 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n In response, VirnetX contends there is a “final deci-\\nsion” on a party’s attempt to prove invalidity after the\\nparty fails to petition for certiorari within the 90-day pe-\\nriod.\\n Two reasons compel the conclusion that there has been\\na final decision on validity in this case. First, our decision\\nin Fairchild (Taiwan) Corp. v. Power Integrations, Inc., 854\\nF.3d 1364 (Fed. Cir. 2017), controls and holds that there is\\na “final decision” under § 317(b) once the time to file a cer-\\ntiorari petition has passed. Second, even if Fairchild did\\nnot foreclose Apple’s argument, the plain language of the\\nstatute and its purpose advise against Apple’s definition of\\na “final decision” on the issue of invalidity.\\n A\\n Since the PTO’s decision in 2016, we have ruled on this\\nvery issue. Fairchild directly addressed a situation in\\nwhich validity was affirmed but other issues in the case\\nwere remanded. Fairchild holds that if a finding of no in-\\nvalidity has been affirmed on appeal and remand of other\\nissues will not “have any affect” on validity, the decision is\\n“final” once the 90-day deadline for a certiorari petition on\\nthe validity determination has passed. Id. at 1366.\\nFairchild announced this holding despite the fact that\\nthere is always some “potential” for future Supreme Court\\nreview if the entire case is taken up. Accordingly, Fairchild\\nrequires we reject Apple’s position.\\n In Fairchild, we began our analysis by restating our\\nearlier holding in Bettcher that estoppel “applies when ‘all\\nappeals have terminated.’” Id. at 1365 (quoting Bettcher,\\n661 F.3d at 646). We then concluded that is “precisely the\\nsituation here.” Id. Namely, the district court “entered\\njudgment against Power Integrations, holding that it failed\\nto prove claims 6, 7, 18, and 19 were obvious over [the cited\\nprior art]. This court affirmed the holding, and the time to\\n\\nVIRNETX INC. v. APPLE INC. 11\\n\\n\\n\\npetition for a writ of certiorari has passed.” Id. at 1365–66\\n(emphasis added).\\n As in Fairchild, the district court here entered judg-\\nment against Apple on the issue of invalidity, we affirmed\\nthe no-invalidity finding on appeal in VirnetX I, and the\\ntime to file a petition for certiorari on the issue has long\\nsince passed. Under Fairchild, that constitutes “a final de-\\ncision” under § 317(b) with respect to Apple’s failure to\\nprove invalidity.\\n Resisting this conclusion, Apple argues that if any non-\\ninvalidity issues remain in the district court case related to\\nthe patent, then the validity decision cannot be final under\\n§ 317(b). Since VirnetX I vacated and remanded other is-\\nsues (i.e., infringement and damages) on both patents at\\nissue, Apple contends that it waited until all the issues\\nwith this patent case are decided before bundling it up in a\\npetition for certiorari. Consequently, Apple’s position re-\\nquires resolution of the entire case—or at least all issues on\\neach patent, including infringement, invalidity, and dam-\\nages—before a validity decision is considered “final.”\\n Fairchild rejected such arguments under almost iden-\\ntical circumstances. There, we had also vacated and re-\\nmanded other parts of the case for additional proceedings.\\nId. at 1365 (explaining prior appeal upheld obviousness but\\nreversed jury’s findings on infringement, and “remanded\\nfor further proceedings unrelated to the ’972 patent\\nclaims”). In turn, we rejected the argument that this ren-\\nders the decision not “final” under § 317(b). Id. at 1366.\\nWe reasoned:\\n While it is true that in [the prior appeal] this court\\n vacated and remanded for additional proceedings,\\n we cannot agree with Power Integrations that this\\n renders the decision not “final” for § 317(b) pur-\\n poses. Critically, those proceedings are unrelated\\n to the ’972 patent. By its terms, § 317(b) is\\n\\n12 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n concerned with a final decision ‘that the party has\\n not sustained its burden of proving the invalidity\\n of any patent claim.’ And here, Power Integrations\\n does not suggest, nor is there any reason to believe,\\n that any unresolved issue on remand would have\\n any effect on the now-final ’972 patent validity de-\\n terminations.\\nId. (emphasis added) (quoting 35 U.S.C. § 317(b)).\\n Likewise, here, Apple does not point to any issue in the\\nVirnetX I remand that substantively bears on the finding\\nof no invalidity. Nor could it. Indeed, the only difference\\nhere is that VirnetX I remanded some infringement and\\ndamages issues related to the ’504 and ’211 patents. But\\nApple fails to explain how any of these unresolved issues\\nwould “have any effect on the now-final . . . validity deter-\\nminations” involving those patents. 7 Id. Accordingly,\\nFairchild’s reasoning leaves no doubt that § 317(b) estop-\\npel applies.\\n Finally, Apple contends Fairchild cannot control be-\\ncause no party in that case expressly argued that the Su-\\npreme Court might still take up the invalidity issue on a\\nfuture appeal under Mercer. But Fairchild’s reasoning im-\\nplicitly forecloses such an argument. Fairchild concluded\\nthat “all appeals” had terminated, 854 F.3d at 1365,\\n\\n\\n\\n 7 At oral argument, Apple could not point to a single\\nexample illustrating how an aspect of the infringement re-\\nmand might meaningfully affect or implicate the invalidity\\nruling this court affirmed in 2014 in VirnetX I. Instead,\\nApple offered only speculation. Its argument boiled down\\nto the general conjecture that the district court “could”\\nhave a motion to set aside a validity determination even\\nafter the mandate issues based on arguments made about\\ninfringement. Oral Argument at 29:20–51, No. 2017-1591,\\nhttp://www.cafc.uscourts. gov/oral-argument-recordings.\\n\\nVIRNETX INC. v. APPLE INC. 13\\n\\n\\n\\ndespite the fact that other issues in that case remained\\npending for potential appeal and petition for certiorari. Ap-\\nple can disagree with that conclusion, but that is not a ba-\\nsis for ignoring binding authority. We are bound to follow\\nthe decision of a prior panel, and its reasoning does not\\nleave the question open.\\n In sum, § 317(b) applies here despite the fact that is-\\nsues unrelated to invalidity were remanded.\\n The dissent takes the view that Fairchild is inapplica-\\nble. But it relies on reasons that were not material to the\\noutcome in Fairchild. In addition, the dissent’s attempt to\\ndistinguish Fairchild suffers from several additional infir-\\nmities.\\n According to the dissent, in Fairchild “Power Integra-\\ntions did not appeal the court’s favorable ruling on infringe-\\nment because it prevailed on that issue.” Opinion\\nconcurring in part and dissenting in part (“dissent”) at 4\\n(emphasis added). In turn, the dissent leaps to the conclu-\\nsion that Power Integrations could not have realistically\\npursued certiorari review of the ’972 patent’s validity be-\\ncause it “prevailed” on infringement. Id. In the dissent’s\\nview, “once Power Integrations prevailed on infringement,\\nthe case against it was concluded” and the potential for any\\nfurther Supreme Court review was eliminated. Id. at 4–5.\\nThere is no support for these statements.\\n Fundamentally, infringement and invalidity are sepa-\\nrate “issues.” Pandrol USA, LP v. Airboss Ry. Prod., Inc.,\\n320 F.3d 1354, 1364 (Fed. Cir. 2003) (“Supreme Court prec-\\nedent and our cases make clear that patent infringement\\nand patent validity are treated as separate issues.”).\\nFairchild never held that resolving infringement resolves\\ninvalidity. Indeed, Fairchild was interpreting an estoppel\\nstatute triggered by a decision on invalidity, not infringe-\\nment.\\n\\n14 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n Moreover, the dissent’s position unravels in light of the\\nprocedural posture in Fairchild. Power Integrations did\\nnot simply argue that it did not infringe the ’972 patent.\\nNor did it merely raise invalidity as an affirmative defense\\nto liability. Rather, Power Integrations advanced a sepa-\\nrate counterclaim for declaratory judgment that the ’972\\npatent was invalid. See Plaintiff’s Fourth Amended An-\\nswer and Counterclaims at 23, Power Integrations, Inc. v.\\nFairchild Semiconductor Int’l, Inc. et al, No. 1:08–cv–\\n00309–LPS (D. Del. July 7, 2011) (Dkt. No. 409). There-\\nfore, to say it “prevailed” and its case “concluded” on that\\nissue due to a non-infringement ruling is not correct.\\n It is black letter law that non-infringement does not\\nmoot a counterclaim regarding invalidity. “A party seeking\\na declaratory judgment of invalidity presents a claim inde-\\npendent of the patentee’s charge of infringement.” Cardi-\\nnal Chem. Co. v. Morton Int’l, 508 U.S. 83, 96 (1993). Thus,\\n“appellate affirmance of a judgment of noninfringement,\\neliminating any apprehension of suit, does not moot a de-\\nclaratory judgment counterclaim of patent invalidity.”\\nMedImmune, Inc. v. Genentech, Inc., 549 U.S. 118, 132 n.11\\n(2007); see also Telcordia Techs., Inc. v. Cisco Sys., Inc., 612\\nF.3d 1365, 1375 (Fed. Cir. 2010) (“[T]his court’s affirmance\\nof the district court’s non-infringement findings as to the\\n’306 patent does not moot Cisco’s invalidity counterclaim\\non cross-appeal.”).\\n In addition, the dissent speculates as to why the litigant\\nchose not to seek further review. The dissent appears to\\nassume that if a party is found not liable for certain ac-\\ncused products infringing a patent, as in Fairchild, that\\nparty no longer has the incentive to challenge the underly-\\ning property right’s validity. Dissent at 5–6. This assump-\\ntion fails for several reasons. First, inferring the subjective\\nintent of the litigants—either on remand or in petitioning\\nfor certiorari—is inappropriate. Fairchild never made\\nsuch fact findings. Fairchild never relied on such infer-\\nences either. Put simply, Fairchild did not base its decision\\n\\nVIRNETX INC. v. APPLE INC. 15\\n\\n\\n\\non speculation as to whether remand would impact a\\nparty’s subjective “decision” to pursue certiorari review or\\nwhether its decisions regarding petitioning were “reasona-\\nble,” id. at 5–6. Second, such speculation is irrelevant. The\\ncontrolling inquiry is mandated by statute, which makes\\nno mention of a party’s motives for petitioning for certiorari\\nreview. Third, it is inaccurate. Regardless of a finding of\\nno liability on the presently accused products, an accused\\ninfringer may well have reasons for trying to invalidate the\\npatent (e.g., to guarantee its freedom to operate).\\n But the dissent suffers from a more fundamental flaw.\\nThe dissent appears to suggest that in Fairchild, “potential\\nfuture review” by the Supreme Court was not an option.\\nId. at 4. In light of Mercer, there is no basis for that posi-\\ntion. Indeed, it runs headlong into the dissent’s own insist-\\nence that the Supreme Court’s discretion to take up\\ncertiorari review of earlier issues in a case is broad. On the\\none hand, the dissent acknowledges that a “petition for writ\\nof certiorari can expose the entire case to review,” id. at 10\\n(quoting Christianson v. Colt Indus. Operating Corp., 486\\nU.S. 800, 817 (1988)), and the Supreme Court has “juris-\\ndiction to consider all of the substantial federal questions\\ndetermined in the earlier stages of the litigation,” id. (quot-\\ning Reece v. Georgia, 350 U.S. 85, 87 (1955)). On the other\\nhand, the dissent contends that the litigant’s decision in\\nFairchild “all but foreclosed the possibility that the inva-\\nlidity issue could be subject to potential future review by\\nthe Supreme Court.” Id. at 4. A plain reading of Fairchild,\\nhowever, confirms it never presumed to curtail the Su-\\npreme Court’s discretion under Mercer to review issues\\nraised earlier in a case.\\n Finally, the dissent is concerned that the majority opin-\\nion somehow “intrudes on the Supreme Court’s constitu-\\ntional prerogative whether to take up the invalidity issue.”\\nId. at 11. It does not. Like Fairchild, our decision concerns\\nestoppel of duplicative administrative proceedings before\\nthe PTO. It has no bearing on the Supreme Court’s\\n\\n16 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\ndiscretion to accept or reject a petition for certiorari on any\\nissue—nor could it.\\n In sum, we disagree with the dissent’s efforts to rewrite\\nFairchild. Fairchild speaks for itself. The controlling stat-\\nute is trained on a “final decision” that the party has not\\nsustained its “burden of proving the invalidity” of the pa-\\ntent claim. Fairchild, 843 F.3d at 1366. As the dissent\\nacknowledges, the reasoning at the heart of the Fairchild\\ndecision is that estoppel applied to the duplicative PTO\\nreexam because there was “no suggestion,” dissent at 6,\\nthat “any unresolved issue on remand would have any ef-\\nfect on the now-final patent validity determinations” from\\nthe district court, Fairchild, 854 F.3d at 1366. That is, the\\nremand would have no substantive effect on the validity\\ndetermination. The record is clear that is the same situa-\\ntion here.\\n B\\n Even if Fairchild did not control, the statutory text and\\npurpose of § 317(b) compels us to reject Apple’s theory that\\nthe entire case must be resolved. Apple essentially asks us\\nto rewrite the statute, triggering estoppel only when there\\nis a “final judgment” on all issues in a case.\\n But the statute’s text is clear. The statute only re-\\nquires “a final decision” on a precise issue—whether the\\nparty has “sustained its burden of proving the invalidity of\\nany patent claim in suit.” 35 U.S.C. § 317(b). The statute\\nmakes no mention of resolving the entire case. Nor does it\\nrequire resolution of all other issues related to a patent\\n(e.g., infringement or damages). Instead, the plain lan-\\nguage of the statute demands a party cease duplicative in-\\nvalidity challenges via reexams in the PTO once “a final\\ndecision” has been reached on an invalidity challenge in a\\ncivil action.\\n This reading of the plain text is further reinforced by\\nthe statute’s purpose. The statute is intended to protect\\n\\nVIRNETX INC. v. APPLE INC. 17\\n\\n\\n\\npatent holders against “harassment” from duplicative liti-\\ngation on the issue of invalidity. See 145 Cong. Rec. 29,276\\n(Nov. 9, 1999) (Conference Report on H.R. 1554) (“Subtitle\\nF creates a new section 317 which sets forth certain condi-\\ntions by which inter partes reexamination is prohibited to\\nguard against harassment of a patent holder.”); see also\\nFunction Media, L.L.C. v. Kappos, 508 F. App’x 953, 955–\\n56 (Fed. Cir. 2013) (“This provision existed to prevent du-\\nplicative, harassing actions against a patentee. If a defend-\\nant brought an invalidity challenge in a district court\\nlitigation and was unsuccessful, it is not permitted to bring\\nthe same challenge in an inter partes reexamination.”).\\n Apple concedes that § 317(b) was intended to prevent\\nduplicative litigation. Apple’s Br. 40–41. Yet Apple’s posi-\\ntion increases the likelihood duplicative litigation will oc-\\ncur. If no decision on invalidity is “final” until any\\n“prospect of future Supreme Court review” is eliminated,\\nApple’s Br. 36, estoppel will rarely apply. In light of the\\nlegislative intent, it is doubtful Congress created an estop-\\npel provision that would allow a PTO proceeding to press\\nforward even after all appellate deadlines have elapsed\\nmerely because there is a potential that the Supreme Court\\nmay one day reconsider the invalidity issue—in many\\ncases, years later—while examining the rest of the case. 8\\n Accordingly, the statutory purpose confirms that “a fi-\\nnal decision” triggers estoppel when the invalidity chal-\\nlenge is decided on appeal and the time for petitioning for\\ncertiorari has passed—regardless of the fact review of the\\n\\n\\n\\n 8 Moreover, there is no indication that Congress con-\\nsidered the type of review contemplated in Mercer as part\\nof the definition of a regular appeals process for reaching a\\n“final decision” on invalidity. This is of no surprise given\\nthe statute is trained on the finality of a particular issue,\\nrather than finality of the entire case.\\n\\n18 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nentire case could occur at some point in the future under\\nMercer.\\n C\\n Apple’s remaining arguments against the application\\nof § 317(b) here are unpersuasive. Apple first suggests that\\nFairchild is inconsistent with Bettcher. Apple’s Br. 32–33,\\n37. In Apple’s view, Bettcher held that § 317(b) is not sat-\\nisfied until the Supreme Court can no longer review a case.\\n Bettcher never reached such a holding. Bettcher con-\\ncerned a different estoppel provision, pre-AIA 35 U.S.C.\\n§ 315(c) (2006). 9 The question was whether § 315(c) estop-\\npel attaches at a far earlier stage (before any appeal is pur-\\nsued). In Bettcher, the district court concluded that\\nestoppel barred further civil litigation “as soon as the ex-\\naminer finished the reexamination and the Right of Appeal\\nNotice was sent to the patentee”—i.e., before there is even\\nan appeal to the Board, let alone an appeal to the Federal\\nCircuit. Id. at 636. On appeal, we rejected that conclusion.\\nWe held that “the estoppel provision of § 315 is triggered\\nnot when examination is completed but only after all ap-\\npeal rights have been exhausted.” Id. at 642–43.\\n The panel then rejected the argument that § 317(b)\\nsomehow showed that § 315(c) estoppel must apply as soon\\nas the examiner finishes the reexam. See id. at 643. The\\npanel observed that like § 315(c), estoppel under § 317(b)\\napplies “when reexamination and all appeals have termi-\\nnated.” Id. at 646 (emphasis added).\\n Bettcher addressed appeal rights to the Board and the\\nFederal Circuit. It did not analyze how petitions for\\n\\n\\n 9 Section 315(c) concerns estoppel in the opposite di-\\nrection: when the PTO’s determination in a reexam estops\\na party from asserting invalidity in a civil action. Id. at\\n643.\\n\\nVIRNETX INC. v. APPLE INC. 19\\n\\n\\n\\ncertiorari impact § 317(b). 10 Moreover, Bettcher did not ad-\\ndress the critical inquiry here as to whether a “final deci-\\nsion” regarding the issue of invalidity under § 317(b) is\\naffected by remand of issues unrelated to invalidity.\\n By contrast, Fairchild did reach these issues.\\nFairchild held that estoppel “applies when ‘all appeals’\\nhave terminated,” meaning after the district court’s judg-\\nment of no-invalidity was affirmed on appeal and “the time\\nto petition for a writ of certiorari has passed.” Id. at 1365–\\n66 (quoting Bettcher, 661 F.3d at 646). The panel reached\\nthis conclusion despite the fact that issues unrelated to pa-\\ntent validity were remanded for further proceedings.\\nProperly understood, Fairchild is an extension of Bettcher\\nto Supreme Court certiorari review. Apple’s attempt to\\n\\n\\n 10 Bettcher concluded that the estoppel provision of\\n§ 315 attaches when all appeal rights are exhausted. As\\nthe panel observed, the opening sections of the statute list\\nthose appeal “rights” for both the patent owner and the\\nthird-party requestor, § 315(a)–(b), which consist of ap-\\npeals from (1) the examiner to the Board, and (2) the Board\\nto this court. Id. at 644 (“The estoppel provision of subsec-\\ntion (c) thus falls immediately following the establishment\\nof both parties’ rights of appeal, including appeal to this\\ncourt.”). Bettcher did not address whether petitions for cer-\\ntiorari are considered “appeals,” or whether certiorari re-\\nview by the Supreme Court is a “right.” In VirnetX’s view,\\nan appeal as of right is a different procedural mechanism\\nthan a writ of certiorari. VirnetX Reply Br. 12–13; see also\\nCity & Cty. of San Francisco, Calif. v. Sheehan, 135 S. Ct.\\n1765, 1774 (2015) (“Because certiorari jurisdiction exists to\\nclarify the law, its exercise ‘is not a matter of right, but of\\njudicial discretion.’” (quoting Supreme Court Rule 10)).\\nThough we need not reach such questions, they underscore\\nthe fact that Bettcher simply did not reach the issue of cer-\\ntiorari review now governed by Fairchild.\\n\\n20 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nmanufacture an inconsistency between these two cases has\\nno merit.\\n Apple’s reliance on Fresenius fares no better. Fresenius\\nUSA, Inc. v. Baxter Int’l, Inc., 721 F.3d 1330 (Fed. Cir.\\n2013). Apple concedes that Fresenius “did not construe\\n§ 317(b).” Apple’s Br. 39. As such, Apple admits Fresenius\\nhas no direct bearing on our interpretation of this statutory\\nestoppel provision.\\n In addition, Fresenius concerned an entirely different\\nissue involving res judicata from a “final judgment.” Even\\nwithin the distinct realm of res judicata, Fresenius took\\ngreat care to “distinguish between different concepts of fi-\\nnality.” Id. at 1340. In Fresenius, we specifically addressed\\nwhether a judgment in an infringement case is “sufficiently\\nfinal” so that it is “immune to the effect of the final judg-\\nment in the PTO proceedings, as affirmed by this court in\\nIn re Baxter.” Id. at 1341 (discussing In re Baxter Int’l, Inc.,\\n678 F.3d 1357 (Fed. Cir. 2012)).\\n The issue in Fresenius arose due to the fact that while\\nthe appeal of an infringement action was pending, we af-\\nfirmed the PTO reexamination finding the patent claims\\ninvalid in In re Baxter. As a result, the PTO canceled the\\nclaims. Id. at 1335. Despite the intervening cancellation,\\nthe patent owner argued that the district court’s judgment\\nwas “final” for purposes of res judicata such that the liabil-\\nity finding and past damages award in the pending litiga-\\ntion could not be disturbed. Id. at 1340. We disagreed.\\n While the district court in 2007 had entered “a judg-\\nment final for purposes of appeal,” we concluded that it was\\n“not sufficiently final to preclude application of the inter-\\nvening final judgment in In re Baxter.” Id. at 1341. More-\\nover, the district court’s original judgment had since been\\nset aside and the remand was not sufficiently final for res\\njudicata purposes. Id. “To rise to that level, the litigation\\nmust be entirely concluded so that [the] cause of action\\n[against the infringer] was merged into a final judgment\\n\\nVIRNETX INC. v. APPLE INC. 21\\n\\n\\n\\n. . . one that ‘ends the litigation on the merits and leaves\\nnothing for the court to do but execute the judgment.’” Id.\\n(quoting Mendenhall v. Barber–Greene Co., 26 F.3d 1573,\\n1580 (1994)). We concluded that the remand did not end\\nthe controversy, leaving only the execution of the judg-\\nment, given that several aspects of the litigation (e.g., roy-\\nalty damages, injunctive relief) remained unresolved. Id.\\n Fresenius’s analysis of a “final judgment” for purposes\\nof res judicata does not instruct our understanding of a “fi-\\nnal decision” for purposes of § 317(b) estoppel. Indeed, it\\nonly underscores why we must follow Fairchild. Fairchild\\nfound estoppel absent a final judgment. Indeed, after mul-\\ntiple other issues in the litigation, including infringement\\nand damages for certain patents, were remanded,\\nFairchild concluded nonetheless that there was a “final de-\\ncision” on the narrow issue of invalidity—which is all that\\nthe statute requires to trigger estoppel. The reasoning in\\nFairchild, not Fresenius, governs our application of\\n§ 317(b).\\n Apple also briefly argues that this interpretation of the\\nstatute generates duplicative proceedings for the patent\\nchallenger. Apple’s Br. 41. In Apple’s view, it must seek\\n“interlocutory” review rather than waiting for the case to\\nbe entirely resolved. Id. Apple’s argument misses the\\nmark. Section 317(b) is concerned with preventing dupli-\\ncative litigation against the patent owner. Of course, we\\nrecognize the challenging nature of these proceedings for\\nboth parties. The present actions involve multiple patents\\nand various issues, which create certain complexities for\\npursuing appeals before this court and seeking certiorari\\nbefore the Supreme Court. Some of those complexities\\nmight be ameliorated by requiring final judgment of an en-\\ntire case. But that does not empower us to rewrite a statute\\nthat requires only a final decision on the issue of invalidity.\\nFurthermore, such interlocutory and piecemeal concerns\\nwere also present in Fairchild. Nonetheless, estoppel ap-\\nplied.\\n\\n22 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n Finally, Apple’s arguments based on the PTO’s inter-\\npretations carry no force. They fail for several reasons.\\n First, Apple fails to explain why the PTO’s interpreta-\\ntions of § 317(b) is entitled to Chevron deference. As dis-\\ncussed above, the statutory text and purpose leave no\\nambiguity. Where “the intent of Congress is clear, that is\\nthe end of the matter.” Chevron U.S.A. Inc. v. NRDC, Inc.,\\n467 U.S. 837, 842 (1984). Therefore, Chevron deference\\ndoes not apply. 11 12\\n Second, Skidmore deference is not appropriate either.\\nUnder Skidmore, the weight given to a particular agency\\ninterpretation “will depend upon the thoroughness evident\\nin its consideration, the validity of its reasoning, its con-\\nsistency with earlier and later pronouncements, and all\\nthose factors which give it power to persuade, if lacking\\n\\n\\n 11 Apple also suggests that Bettcher requires defer-\\nence. Not so. Bettcher concluded the statute was unambig-\\nuous. Bettcher, 661 F.3d at 646 (“[W]e believe that § 317(b)\\nis clear . . . .”). Nonetheless, Apple points to Bettcher’s sub-\\nsequent discussion of the deference it “would” give the\\nMPEP had it reached the issue. Id. Apple’s reliance on a\\nsingle line of dicta from Bettcher is unavailing. Even if it\\nwere not dicta, Bettcher’s statement concerned deference to\\nthe Manual of Patent Examining Procedure (“MPEP”)\\n§ 2686.04. Id. As explained later, that section of the MPEP\\ndoes not resolve the issue at hand.\\n 12\\n Apple also suggests the underlying PTO denial of\\nthe petition to terminate here is entitled to Chevron defer-\\nence. In support of this position, Apple cites only to In re\\nAffinity Labs of Texas, LLC, 856 F.3d 883, 894 n.5 (Fed.\\nCir. 2017). The cited footnote from Affinity Labs did not\\nconcern deference. Rather, it concerned whether PTO rul-\\nings on petitions to terminate are reviewable on appeal. Id.\\n(analyzing stages of PTO decision-making process as re-\\nlated to reviewability).\\n\\nVIRNETX INC. v. APPLE INC. 23\\n\\n\\n\\npower to control.” United States v. Mead Corp., 533 U.S.\\n218, 228 (2001) (quoting Skidmore v. Swift & Co., 323 U.S.\\n134, 140 (1944)). None of the considerations favoring def-\\nerence are present here. For example, contrary to Apple’s\\nposition, the PTO has never established a “consistent” con-\\nstruction of § 317(b). Indeed, aside from the decision below,\\nApple cites no other decision in which the PTO has inter-\\npreted § 317(b) to extend passed the deadline for certiorari\\nreview of the invalidity issue to hypothetical petitions aris-\\ning from appeals of non-invalidity issues on remand. Thus,\\ndeference is unwarranted.\\n Third, regardless of whether deference applies, none of\\nApple’s citations resolve the specific issue at hand. For in-\\nstance, no PTO regulation or MPEP section states that a\\ndecision is not “final” for purposes of § 317(b) if non-inva-\\nlidity issues are remanded. Nor do any of these provisions\\nstate that a final judgment is required, or that as long as\\nthe potential for Supreme Court review from a later peti-\\ntion exists, § 317(b) estoppel does not attach. Indeed, none\\nof these provisions even mention Supreme Court certiorari\\nreview. 13\\n For the reasons above, § 317(b) estoppel applies to the\\nApple reexams. Accordingly, we vacate and remand with\\n\\n\\n 13 Other PTO regulations suggest appeals terminate\\nwhen our mandate issues. See, e.g., 37 C.F.R. § 1.197 (“An\\nappeal to the U.S. Court of Appeals for the Federal Circuit,\\nwhether from a decision of the Board or a judgment in a\\ncivil action, is terminated when the mandate is issued by\\nthe Court.”). Notably, 37 C.F.R. § 1.979, which is the reg-\\nulation interpreting pre-AIA § 316(a) permitting the PTO\\nto issue a certificate of cancellation “only after all appeals\\nhave terminated,” Bettcher, 661 F.3d at 644–45, applies the\\nsame interpretation. See 37 C.F.R. § 1.979(b) (providing\\nthat an “appeal [to this Court] is considered terminated\\nwhen the mandate is issued by the Court”).\\n\\n24 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\ninstructions to terminate the Apple reexams with respect\\nto claims 1–35 of the ’504 patent and claims 36–59 of the\\n’211 patent. 14\\n III\\n\\n We now turn to VirnetX’s challenge to the merits of the\\nApple and Cisco reexams that pertain to the claims not\\nsubject to estoppel. The Board affirmed the Examiner’s re-\\njection of all claims of the ’504 and ’211 patents. On appeal,\\nVirnetX takes a kitchen sink approach, arguing that the\\nBoard decisions incorrectly construe certain claims, shift\\nthe burden to the patent owner to prove patentability, and\\nlack substantial evidence. VirnetX’s arguments are unper-\\nsuasive. Therefore, we affirm the Board’s decision on all\\nremaining claims of both patents in the Apple reexam not\\nsubject to § 317(b) estoppel and fully affirm the Board’s de-\\ncision regarding the claims of the ’211 patent in the Cisco\\nreexam. 15\\n\\n\\n\\n 14 Apple appears not to dispute that estoppel extends\\nto the dependent claims here. Section 317(b) bars main-\\ntaining a reexam based on issues which that party “raised”\\nor “could have raised” in the civil action. The PTO found\\nthat Apple “was aware of all prior art raised prior to Octo-\\nber 31, 2012 in the present reexamination proceeding with\\nrespect to claims 1–35 [of the ’504 patent], and could have\\nraised issues with respect to that prior art in the litigation.”\\nJ.A. 1666, 3147 (same regarding claims 36–59 of the ’211\\npatent). No new prior art was added after October 31,\\n2012, when trial began in the litigation. Indeed, Apple\\nnever amended or supplemented its original October 18,\\n2011 reexamination requests to add new prior art.\\n 15 The parties filed supplemental briefing regarding\\nthe relationship between this case and other recently de-\\ncided matters. ECF Nos. 95, 96, 105. Apple primarily ar-\\ngues that collateral estoppel renders various claims of the\\n\\nVIRNETX INC. v. APPLE INC. 25\\n\\n\\n\\n VirnetX contends that the Board improperly construed\\nthe term “indication” by ignoring its disclaimer argument.\\nVirnetX argued before the Board that the patent dis-\\nclaimed “conventional” DNS functions—e.g., looking up\\nand returning an IP address or a public key for the re-\\nquested webpage. In VirnetX’s view, such conventional\\nfunctions therefore do not qualify as “indications” that the\\nsystem “supports establishing a secure communication\\nlink.” J.A. 1348, 2892, 4334–41. But the Board expressly\\nacknowledged VirnetX’s assertion that the patents “dispar-\\naged and disclaimed” systems that use no more than “con-\\nventional” DNS functions. J.A. 26–30, 49–54, 77–80. It\\nthen rejected this argument in a systematic and well-rea-\\nsoned fashion. “Disavowal requires ‘expressions of mani-\\nfest exclusion or restriction, representing a clear disavowal\\nof claim scope.’” Epistar Corp. v. ITC, 566 F.3d 1321, 1335\\n(Fed. Cir. 2009) (quoting Teleflex, Inc. v. Ficosa N. Am.\\nCorp., 299 F.3d 1313, 1325 (Fed. Cir. 2002)). We agree with\\nthe Board that the specification fails to evidence such a\\nclear disavowal here. The Board’s construction is therefore\\nconsistent with the broadest reasonable interpretation of\\n“indication.”\\n Turning to VirnetX’s next challenge, VirnetX contends\\nthat the Board failed to make adequate findings and im-\\npermissibly shifted the burden to VirnetX with respect to\\nthree claim limitations of the ’504 and ’211 patents in the\\nApple reexams. The limitations are “indicating,” “connect-\\nable,” and “authenticate.” Contrary to VirnetX’s\\n\\n\\n\\n’504 and ’211 patents invalid. As an initial matter, Apple’s\\narguments are mooted in part by our holding in the present\\nmatter affirming the Board’s decision that certain claims\\nof the ’504 and ’211 patent are unpatentable. Otherwise,\\nApple’s arguments are unpersuasive. Apple has failed to\\nshow why collateral estoppel is appropriate under these\\ncircumstances.\\n\\n26 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\narguments, the Board’s conclusions are well supported by\\nthe underlying factual findings made by the Examiner.\\nThe Examiner found that Apple had carried its burden of\\nproving unpatentability by “a preponderance of the evi-\\ndence.” See Rambus Inc. v. Rea, 731 F.3d 1248, 1255 (Fed.\\nCir. 2013) (quoting Ethicon, Inc. v. Quigg, 849 F.2d 1422,\\n1427 (Fed. Cir. 1988)). Sitting as an appellate body, the\\nBoard then reviewed those findings and considered\\nwhether VirnetX had raised any issue on appeal that re-\\nquired setting aside the Examiner’s decisions. Although\\n“the examiner retains the burden to show invalidity,” on\\nappeal “the applicant must identify to the Board what the\\nexaminer did wrong.” In re Jung, 637 F.3d 1356, 1365–66\\n(Fed. Cir. 2011). The Board considered VirnetX’s argu-\\nments challenging the Examiner’s rejections and properly\\nexplained why each lacked merit. VirnetX “failed to offer\\nany convincing reason to find otherwise.” In re Cree, Inc.,\\n818 F.3d 694, 702 n.3 (Fed. Cir. 2016). This “is not burden-\\nshifting.” Id.\\n VirnetX further argues that the Board’s findings in the\\nCisco reexam regarding the “indicating” limitations intro-\\nduced improper new arguments. We disagree. According\\nto VirnetX, the Board relied on a new ground for concluding\\nthat Lendenmann’s disclosure of “returning a network ad-\\ndress” and “access control lists” (“ACLs”) meet the claim\\nlimitation. The Board’s findings constitute a new ground\\nof rejection supporting reopening of prosecution if the\\nBoard “finds facts not found by the examiner regarding the\\ndifferences between the prior art and the claimed inven-\\ntion, and these facts are the principal evidence upon which\\nthe Board’s rejection was based.” In re Leithem, 661 F.3d\\n1316, 1320 (Fed. Cir. 2011). “[T]he ultimate criterion of\\nwhether a rejection is considered ‘new’ in a decision by the\\n[B]oard is whether [applicants] have had fair opportunity\\nto react to the thrust of the rejection.” In re Adler, 723 F.3d\\n1322, 1327 (Fed. Cir. 2013) (modifications in original). In\\n\\nVIRNETX INC. v. APPLE INC. 27\\n\\n\\n\\nlight of the proceedings below, we conclude that VirnetX\\nhad a fair opportunity to react to the evidence.\\n VirnetX also contends that the Board’s findings that\\nLendenmann teaches the “indicating” element lack sub-\\nstantial evidence. However, the Board found Lenden-\\nmann’s disclosure of “returning a network address,”\\n“ACLs,” and certain domain names teach the limitation.\\nBased on the record here, there is substantial evidence to\\nsupport the Board’s conclusion that this limitation is\\ntaught.\\n\\n As for the “authenticating” limitation, VirnetX contests\\nthe Board’s determination that Provino discloses a DNS\\nsystem configured “to authenticate the query,” as recited in\\ndependent claims 5, 23, and 47. VirnetX’s arguments do\\nnot unsettle the Board’s conclusions that Provino’s sys-\\ntem—including its “query” and “firewall 30”—teach this\\nlimitation. Furthermore, contrary to VirnetX’s view, no\\nmaterial aspect of the Board’s analysis of the relationship\\nbetween authorization and “authentication” was new or\\nimproper. Based on the record developed before the Exam-\\niner and reviewed by the Board, substantial evidence sup-\\nports the Board’s conclusions regarding the\\n“authenticating” limitation.\\n\\n Next, VirnetX argues that the Board failed to address\\narguments it made in the Cisco reexam as to why depend-\\nent claim 47 of the ’211 patent is patentable over Lenden-\\nmann. There is little indication that VirnetX properly\\nargued claim 47’s patentability separately from independ-\\nent claims 36 and 60. The Board was therefore justified in\\nconcluding “Patent Owner does not provide additional ar-\\nguments in support of” claim 47. J.A. 88. Regardless, the\\nrecord confirms that based on the Examiner’s findings, the\\nBoard properly concluded claim 47 was unpatentable.\\n\\n Finally, VirnetX also argues that the reexams here are\\nunconstitutional. VirnetX took the position that its\\n\\n28 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nchallenge would be resolved by the Supreme Court’s deci-\\nsion regarding the constitutionality of inter partes review\\n(“IPR”) in Oil States Energy Services, LLC v. Greene’s En-\\nergy Group, LLC, 138 S. Ct. 1365 (2018). See VirnetX’s Br.\\n66–67 (citing arguments in Oil States petition for certiorari\\nthat IPRs are unconstitutional under the Seventh Amend-\\nment and Article III). While reexams and IPRs arise under\\ndifferent statutes, VirnetX’s argument treats them as in-\\nterchangeable. The Supreme Court has rejected the\\ngrounds advanced in Oil States. 138 S. Ct. at 1379. Vir-\\nnetX’s challenge is therefore moot.\\n We have considered VirnetX’s remaining arguments\\nand find them unpersuasive.\\n CONCLUSION\\n For the foregoing reasons, we vacate the Board’s deci-\\nsions in the Apple reexams with respect to claims 1–35 of\\nthe ’504 patent and claims 36–59 of the ’211 patent and re-\\nmand with instructions to terminate. However, we affirm\\nthe Board’s decisions in the Apple reexams that claims 36–\\n60 of the ’504 patent and claims 1–35 and 60 of the ’211\\npatent are unpatentable. We further affirm the Board’s de-\\ncision in the Cisco reexam that claims 36–60 of the ’211 pa-\\ntent are unpatentable.\\n AFFIRMED-IN-PART, VACATED-IN-PART, AND\\n REMANDED\\n COSTS\\n The parties shall bear their own costs.\\n\\n United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n VIRNETX INC.,\\n Appellant\\n\\n v.\\n\\n APPLE INC., CISCO SYSTEMS, INC.,\\n Appellees\\n ______________________\\n\\n 2017-1591, 2017-1592, 2017-1593\\n ______________________\\n\\n Appeals from the United States Patent and Trademark\\nOffice, Patent Trial and Appeal Board in Nos. 95/001,788,\\n95/001,789, 95/001,856.\\n ______________________\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4422108/\", \"author_raw\": \"Opinion for the court filed by Chief Judge PROST\"}, {\"author\": \"REYNA, Circuit Judge, concurring\", \"type\": \"concurrence\", \"text\": \"REYNA, Circuit Judge, concurring-in-part and dissenting-\\nin-part.\\n This appeal involves the Supreme Court’s authority to\\ntake appeals involving U.S. patents under Article III of the\\nU.S. Constitution.\\n The majority holds that Apple’s petitions for inter\\npartes reexaminations are barred under 35 U.S.C. § 317(b)\\n(2006). Although I agree with the majority’s disposition of\\nthis case in all other respects, I do not agree that reexami-\\nnation is precluded. Because Apple may still appeal this\\ncourt’s affirmance of no invalidity, there is no “final deci-\\nsion” that triggers a bar under § 317(b). Accordingly, I con-\\ncur-in-part and dissent-in-part.\\n\\n2 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n I.\\n This dispute has reached this court once before. In\\n2010, VirnetX Inc. (“VirnetX”) sued Apple Inc. (“Apple”) in\\ndistrict court, alleging infringement of four patents, includ-\\ning the two at issue in this appeal, U.S. Patent Nos.\\n7,418,504 (“the ’504 patent”) and 7,921,211 (“the ’211 pa-\\ntent”). Apple defended on grounds that the patents were\\ninvalid. After trial, the jury entered verdicts against Apple\\nthat the patents were infringed and not invalid. Apple ap-\\npealed. We affirmed the jury’s finding of no invalidity. Vir-\\nnetX, Inc. v. Cisco Sys., Inc., 767 F.3d 1308, 1323–24 (Fed.\\nCir. 2014) (“VirnetX I”). We reversed the district court’s\\ndecision on claim construction, vacated the jury’s infringe-\\nment finding and damages award, and remanded for a new\\ntrial on those issues. Id. at 1317–19, 1323–34. At that\\ntime, Apple did not petition for a writ of certiorari on the\\nissue of invalidity.\\n During the initial district court litigation, Apple filed\\nrequests for inter partes reexamination of the ’504 and ’211\\npatents. After several years of reexamination proceedings,\\nthe Examiner found all claims of the ’504 and ’211 patents\\ninvalid, and the Patent Trial and Appeal Board (“Board”)\\naffirmed. VirnetX now appeals from the Board’s final writ-\\nten decisions 1 on the basis that reexamination of the ’504\\nand the ’211 patents was barred under § 317(b).\\n Section 317(b) provides in relevant part:\\n Once a final decision has been entered against a\\n party in a civil action arising in whole or in part\\n under section 1338 of title 28, that the party has\\n\\n\\n 1 In a companion case, Apple also appealed from the\\njudgment of the district court in the trial on remand, in\\nwhich a jury once again found that Apple infringed the ’504\\nand ’201 patents. See VirnetX Inc. v. Cisco Sys., Inc., No.\\n2018-1197 (Fed. Cir.) (“VirnetX II”).\\n\\nVIRNETX INC. v. APPLE INC. 3\\n\\n\\n\\n not sustained its burden of proving the invalidity\\n of any patent claim in suit . . . , an inter partes reex-\\n amination requested by that party or its privies on\\n the basis of [issues which that party or its privies\\n raised or could have raised in such civil action] may\\n not thereafter be maintained by the Office . . . .”\\n35 U.S.C. § 317(b) (2006).\\n In this appeal, we are required to determine whether\\nour prior decision that affirmed the jury’s finding of no in-\\nvalidity is a “final decision” under § 317(b), such that reex-\\namination of the two patents is barred.\\n II.\\n The majority relies on this court’s decision in Fairchild\\n(Taiwan) Corp. v. Power Integrations, Inc. to hold that “‘a\\nfinal decision’ triggers estoppel when the invalidity chal-\\nlenge is decided on appeal and the time for petitioning for\\ncertiorari has passed.” Maj. Op. 17; see also id. at 10–11\\n(citing 854 F.3d 1364, 1365–66 (Fed. Cir. 2017)). I agree\\nwith this formulation of the law, but disagree that Apple’s\\ntime for petitioning for certiorari on the issue of invalidity\\nhas passed.\\n The majority interprets Fairchild to mean that the ap-\\npellate process terminates once this court issues a decision\\non any issue and an appeal is not immediately taken. That\\nis not what we held in Fairchild. Rather, we held that\\n§ 317(b)’s “restriction applies when ‘all appeals have termi-\\nnated.’” Fairchild, 854 F.3d at 1365 (quoting Bettcher In-\\ndus., Inc. v. Bunzl USA, Inc., 661 F.3d 629, 646 (Fed. Cir.\\n2011)). We did not hold “terminate” to mean upon the is-\\nsuance of our opinion. Nor did we hold that a party is im-\\nmediately obligated to petition for certiorari once this court\\nresolves an issue of invalidity where other issues involving\\nthe same patent remain undecided. We did not consider in\\nFairchild the finality issue that is before the court in this\\ncase.\\n\\n4 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n Fairchild involved a second appeal in a long-running\\ndispute between Power Integrations and Fairchild. In the\\nfirst appeal of that litigation, we affirmed the jury’s finding\\nof no invalidity and found in favor of Power Integrations on\\ninfringement. Power Integrations, Inc. v. Fairchild Semi-\\nconductor Int’l, Inc., 843 F.3d 1315, 1342–45 (Fed. Cir.\\n2016). We remanded only a cross-appeal by Fairchild that\\nrelated to a wholly different patent. Id. at 1345. Power\\nIntegrations did not appeal the court’s favorable ruling on\\ninfringement because it prevailed on that issue. See Kalka\\nv. Hawk, 215 F.3d 90, 96 n.9 (D.C. Cir. 2000) (cited in Pear-\\nson v. Callahan, 555 U.S. 223, 240 (2009)) (explaining that\\n“[n]ormally, a party may not appeal from a favorable judg-\\nment” and that the Supreme Court “has apparently never\\ngranted the certiorari petition of a party who prevailed in\\nthe appellate court”); see also Bunting v. Mellen, 541 U.S.\\n1019, 1023 (2004) (Scalia, J., dissenting from denial of cer-\\ntiorari) (“[O]ur practice reflects a ‘settled refusal’ to enter-\\ntain an appeal by a party on an issue as to which he\\nprevailed.”).\\n Power Integrations’ decision not to petition for certio-\\nrari all but foreclosed the possibility that the invalidity is-\\nsue could be subject to potential future review by the\\nSupreme Court. 2 Stated differently, once Power\\n\\n\\n\\n\\n 2 The Power Integrations case is still proceeding at\\nthe district court on remand. See Power Integrations, Inc.\\nv. Fairchild Semiconductor Int’l, Inc., No. 1:08-cv-00309-\\nLPS (D. Del.). Yet the theoretical possibility that Power\\nIntegrations would appeal its unfavorable invalidity ruling\\nwas not even argued to this court in Fairchild. See Power\\nIntegrations’ Opposition to Fairchild’s Motion to Remand\\nat 2–3, Fairchild, 854 F.3d 1364 (No. 17-1002); see also\\nFairchild (Taiwan) Corporation’s Reply in Support of its\\nMotion to Remand at 2, Fairchild, 854 F.3d 1364 (No. 17-\\n\\nVIRNETX INC. v. APPLE INC. 5\\n\\n\\n\\nIntegrations prevailed on infringement, the case against it\\nwas concluded because the only issues remaining pertained\\nto a cross-appeal involving a different patent. Hence, all of\\nits appeals terminated at that point. 3 That is not the case\\nhere.\\n In VirnetX I, this court vacated the jury’s finding of in-\\nfringement and remanded for a new trial on damages and\\ninfringement under the correct claim construction. Vir-\\nnetX I, 767 F.3d at 1314. There is no dispute that at that\\npoint, Apple clearly could have petitioned for a writ of cer-\\ntiorari on invalidity. That Apple chose not to appeal is rea-\\nsonable given the potential for the outcome on remand to\\ndrive its decision whether to petition to the Supreme Court\\n\\n\\n\\n1002) (arguing only that a petition for certiorari is not an\\n“appeal” for purposes of § 317(b)).\\n 3 The majority asserts that the case against Power\\nIntegrations did not conclude at this point because the in-\\nvalidity issue was separate from the infringement issue,\\nand arose as a counterclaim by Power Integrations for de-\\nclaratory judgment against Fairchild. Maj. Op. 13–14.\\nThis is incorrect. Power Integrations raised invalidity of\\nthe relevant patent as both an affirmative defense to in-\\nfringement and as a counterclaim. See Plaintiff’s Fourth\\nAmended Answer and Counterclaims at 10, 23, Power In-\\ntegrations, No. 1:08-cv-00309 (D. Del. July 7, 2011), ECF\\nNo. 409. Regardless of the procedural posture of the inva-\\nlidity issue, the majority does not dispute that the remand\\nin Power Integrations concerned only issues not related to\\nthe patent at issue in Fairchild. Power Integrations had\\nno incentive to pursue an appeal on invalidity, because it\\nno longer faced the risk of being found to infringe. The ma-\\njority correctly acknowledges the impropriety of inferring\\nthe subjective intent of litigants in petitioning for certio-\\nrari. Maj. Op. 14. That litigation decision should not be\\ndictated by this court, as we have done in this case.\\n\\n6 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nat all. This is because, had Apple won on infringement on\\nremand, there would then have been no reason for Apple to\\nappeal that decision or the decision on invalidity.\\n Apple, however, lost on infringement on remand. Vir-\\nnetX Inc. v. Apple Inc., 324 F. Supp. 3d 836, 844 (E.D. Tex.\\n2017). We affirmed the jury on appeal. VirnetX II, 748\\nF. App’x 332 (Fed. Cir. 2019). Now, as the majority recog-\\nnizes, Apple has a basis on which to petition for certiorari\\nto the Supreme Court on both infringement and invalidity.\\nThus, unlike in Fairchild, where there was no suggestion\\nthat “any unresolved issue on remand would have any ef-\\nfect on the now-final patent validity determinations,” 854\\nF.3d at 1366, the outcome on remand from VirnetX I di-\\nrectly affects Apple’s decision whether to appeal the inva-\\nlidity issue.\\n Indeed, in this case the issues on remand (infringement\\nand damages) involved the same patent the jury found, and\\nwe affirmed, to be not invalid. But in Fairchild, the re-\\nmand concerned a counterclaim unrelated to the patent\\nclaims challenged in the reexamination proceeding, a fact\\nthis court found to be “critical[]” to its conclusion that the\\ninvalidity decision in Power Integrations was final.\\nFairchild, 854 F.3d at 1365; see also Power Integrations,\\n843 F.3d at 1345 (remanding only with respect to patents\\nthat were not at issue in the reexamination request on ap-\\npeal in Fairchild). By contrast, this court’s remand in Vir-\\nnetX I specifically concerned the same claims we affirmed\\nas not invalid and that were at issue in Apple’s reexamina-\\ntion proceedings, a fact that the majority recognizes. Vir-\\nnetX I, 767 F.3d at 1313–14; Maj. Op. 12. This means that\\nFairchild does not apply here as the majority asserts be-\\ncause the circumstances underlying the Fairchild decision\\non finality are not the same as the circumstances in this\\ncase.\\n Apple’s decision to delay its petition for certiorari to the\\nSupreme Court until resolution of the new trial on\\n\\nVIRNETX INC. v. APPLE INC. 7\\n\\n\\n\\ninfringement was reasonable because Apple can now ap-\\npeal the issues of infringement and invalidity together.\\nWhether Apple’s decision was based on litigation strategy\\nor a desire to conserve resources is not at issue. This court\\nshould not dictate to Apple when or whether it should ap-\\npeal. A party should be free to develop its litigation strat-\\negy within the bounds of the law without this court’s\\ninterference, and nothing in Fairchild suggests otherwise.\\nTo the extent that Apple decided to delay its appeal until\\nthe outcome of its trial on remand, that decision was both\\nreasonable and consistent with the law and principles of\\njudicial economy.\\n III.\\n The majority’s interpretation of Fairchild is at odds\\nwith this court’s decisions in Bettcher Industries, Inc. v.\\nBunzl USA, Inc., 661 F.3d 629 (Fed. Cir. 2011) and Frese-\\nnius USA, Inc. v. Baxter International, Inc., 721 F.3d 1330\\n(Fed. Cir. 2013).\\n In Bettcher, we stated that § 317(b) “prevents further\\nreexamination proceedings once parallel federal court pro-\\nceedings have reached their resolution.” Bettcher, 661 F.3d\\nat 646. In other words, § 317(b) “attaches only after there\\nhas been a final resolution [of the invalidity issue] . . . in\\nfederal court,” and is not triggered “if there remains any\\ntime for an appeal.” Id. (second portion quoting Manual of\\nPatent Examining Procedure § 2686.04). The court further\\nexplained in Bettcher that the legislative history “con-\\nfirm[s] that § 317(b) applies [only] ‘after any appeals.’” Id.\\n(quoting 145 Cong. Rec. 29,276 (Nov. 9, 1999); 145 Cong.\\nRec. 29,973 (Nov. 17, 1999)). The majority’s decision that\\nApple’s reexaminations are barred by § 317(b) is\\n\\n8 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\ninconsistent with Bettcher because time remains for Apple\\nto file a petition for a writ of certiorari. 4\\n The majority’s interpretation of Fairchild is also incon-\\nsistent with Fresenius. This court stated in Fresenius that\\na decision is not final if it “leaves open the question of dam-\\nages or other remedies.” Fresenius, 721 F.3d at 1341–42\\n(citing 18A Charles A. Wright, Arthur R. Miller, & Edward\\nH. Cooper, Federal Practice and Procedure § 4432 (2d ed.\\n2002)); see also id. at 1343 (“[A] decision finding a patent\\nnot invalid but remanding for further damages proceedings\\nis not a final judgment.” (citing Mendenhall v. Barber-\\nGreene Co., 26 F.3d 1573, 1576–77 (Fed. Cir. 1994), as cor-\\nrected on reh’g (Sept. 14, 1994))). This court determined\\nthat there was no finality because the case was remanded\\nto reconsider royalties and injunctive relief. Id. at 1333.\\n In deciding Fresenius, this court relied on the Supreme\\nCourt’s decision in Simmons Co. v. Grier Brothers Co., 258\\nU.S. 82 (1922). We explained that in Simmons, the Su-\\npreme Court concluded that there was no finality with re-\\nspect to the issue of infringement because “even though\\nthere had been an appellate decision entirely resolving the\\npatent infringement claims, . . . there had not yet been a\\n\\n\\n 4 The majority distinguishes Bettcher on the grounds\\nthat it dealt with estoppel under pre-AIA 35 U.S.C. § 315(c)\\nand only addressed appellate review by the Board and this\\ncourt—not the Supreme Court. Maj. Op. 18–19 & n.10.\\nBut Bettcher expressly construed the meaning of “final de-\\ncision” as used in § 317(b) in the context of all federal\\ncourts, and that construction was part of Bettcher’s hold-\\ning. See Fairchild, 854 F.3d at 1365. Although Bettcher\\ndid not specifically mention the Supreme Court, it held\\nthat § 317(b) “applies only when . . . all appeals have ter-\\nminated.” Bettcher, 661 F.3d at 646 (emphasis added).\\nNothing in Bettcher suggests that its holding was limited\\nsolely to appeals to the Board and this court.\\n\\nVIRNETX INC. v. APPLE INC. 9\\n\\n\\n\\nfinal judgment on the unfair competition claims [and an\\naccounting of damages],” and therefore “the Supreme\\nCourt’s intervening decision [in a separate case] was bind-\\ning as to the infringement claims” because “the appellate\\nmandate had not ended the case.” Fresenius, 721 F.3d at\\n1342–43 (citing Simmons, 258 U.S. at 84–91) (emphasis\\nadded); see also id. (“Although the interlocutory decision of\\n[the trial court] on the question of validity and infringe-\\nment was appealable, . . . the decision was not final until\\nthe conclusion of the accounting. Hence the court did not\\nlack power at any time prior to entry of its final judgment\\nat the close of the accounting to reconsider any portion of\\nits decision and reopen any part of the case.” (quoting Mar-\\nconi Wireless Tel. Co. v. United States, 320 U.S. 1, 47–48\\n(1943)). As in Simmons and Marconi, even though this\\ncourt affirmed the jury’s finding of no invalidity in VirnetX\\nI, that decision is not final because the case was remanded\\nfor a new trial on infringement and damages.\\n IV.\\n Other Supreme Court precedent supports the conclu-\\nsion that Apple has not exhausted its appellate rights. Alt-\\nhough Apple did not petition for certiorari on the issue of\\ninvalidity after its first appeal, Supreme Court precedent\\nestablishes that it can do so now. See Major League Base-\\nball Players Ass’n v. Garvey, 532 U.S. 504, 508 n.1 (2001)\\n(per curiam) (“[W]e have authority to consider questions\\ndetermined in earlier stages of the litigation where certio-\\nrari is sought from the most recent of the judgments of the\\nCourt of Appeals.”); Bhd. of Locomotive Firemen & Engine-\\nmen v. Bangor & Aroostook R.R. Co., 389 U.S. 327, 328\\n(1967) (“[B]ecause the Court of Appeals remanded the case,\\nit is not yet ripe for review by this Court.”); Mercer v. The-\\nriot, 377 U.S. 152, 153–54 (1964) (per curiam) (“[I]t is set-\\ntled that we may consider questions raised on the first\\nappeal, as well as ‘those that were before the court of ap-\\npeals upon the second appeal.’” (quoting Hamilton–Brown\\nShoe Co. v. Wolf Bros. & Co., 240 U.S. 251, 257 (1916))). In\\n\\n10 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nReece v. Georgia, the Supreme Court explained that it has\\n“jurisdiction to consider all of the substantial federal ques-\\ntions determined in the earlier stages of the litigation.” 350\\nU.S. 85, 87 (1955) (citing Urie v. Thompson, 337 U.S. 163,\\n172–73 (1949)). This is true even if certiorari was not\\nsought until after a second appeal. Id. The law-of-the-case\\ndoctrine does not bar Supreme Court review because “[a]\\npetition for writ of certiorari can expose the entire case to\\nreview.” Christianson v. Colt Indus. Operating Corp., 486\\nU.S. 800, 817 (1988) (citing Panama R. Co. v. Napier Ship-\\nping Co., 166 U.S. 280, 283–284 (1897)).\\n The majority asserts that § 317(b) concerns only “the\\nfinality of a particular issue, rather than finality of the en-\\ntire case.” Maj. Op. 17 n.8; see also id. at 20–21. Yet, the\\nmajority recognizes that “the Supreme Court may one day\\nreconsider the invalidity issue” decided in VirnetX I. Maj.\\nOp. 17. 5 Thus, even under the majority’s reading of the\\nstatute, finality is lacking with respect to the invalidity is-\\nsue in this case.\\n The majority also suggests that Supreme Court review\\ndoes not qualify as an “appeal” for purposes of § 317(b)’s\\nestoppel provision because there is no right to Supreme\\nCourt review. Maj. Op. 19 n.10. This suggestion is incon-\\nsistent with Article III of the U.S. Constitution, which\\nstates that, aside from cases implicating the Supreme\\nCourt’s original jurisdiction, the Supreme Court “shall\\nhave appellate Jurisdiction, both as to Law and Fact.” U.S.\\nConst. art. III. In any case, Apple has a right to petition for\\na writ of certiorari. See Durham v. United States, 401 U.S.\\n481, 483 n.* (1971), overruled on other grounds by Dove v.\\n\\n\\n 5 In responding to the above line of Supreme Court\\ncases, the majority again confirms that the Court can still\\nreview the invalidity issue in VirnetX I, thus further un-\\ndercutting its insistence that the issue was finally decided\\nin that case. See Maj. Op. 15.\\n\\nVIRNETX INC. v. APPLE INC. 11\\n\\n\\n\\nUnited States, 423 U.S. 325 (1976) (noting the distinction\\nthat “appeals are a matter of right while decisions on cer-\\ntiorari petitions are wholly discretionary,” but explaining\\nthat “Congress, however, has given a right to petition for\\ncertiorari” (emphasis added)). In my view, by determining\\nthat this court’s affirmance of no invalidity in VirnetX I ter-\\nminated Apple’s right to seek certiorari, the majority’s con-\\nclusion intrudes on the Supreme Court’s constitutional\\nprerogative whether to take up the invalidity issue.\\n V.\\n The majority’s decision also intrudes on long estab-\\nlished principles of judicial economy because it fosters\\npiecemeal litigation. Under the majority’s approach, once\\nthis court decides an issue of invalidity on appeal, a party\\nis required to immediately petition for certiorari to the Su-\\npreme Court—regardless of the status of the other issues\\nin the case. This approach stands in stark contrast to “the\\ngeneral judicial bias against piecemeal litigation.” Arizona\\nv. San Carlos Apache Tribe of Arizona, 463 U.S. 545, 570\\n(1983); see also Sears, Roebuck & Co. v. Mackey, 351 U.S.\\n427, 439 (1956) (noting the courts’ longstanding “hostility\\ntoward piecemeal appeals”); Panama, 166 U.S. at 284\\n(“[A]s was recently said in the case of The Conqueror, 17\\nSup. Ct. 510, [the writ of certiorari] is and generally should\\nbe issued only after a final decree.”).\\n In McLish v. Roff, the Supreme Court explained that\\n“[f]rom the very foundation of our judicial system the object\\nand policy . . . have been to save the expense and delays of\\nrepeated appeals in the same suit, and to have the whole\\ncase and every matter in controversy in it decided in a sin-\\ngle appeal.” 141 U.S. 661, 665 (1891).\\n The majority contends that to hold that § 317(b) does\\nnot apply in this case will itself lead to duplicative litiga-\\ntion because waiting “until any prospect of future Supreme\\nCourt review is eliminated” will cause § 317(b) “estoppel\\n[to] rarely apply.” Maj. Op. 17 (internal citation and\\n\\n12 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nquotation marks omitted). But not every appeal that\\nreaches this court results in a remand, and even if that\\nwere the case, estoppel would still apply once all appeals\\nare terminated. It is not rare that the Supreme Court de-\\nnies a party’s last petition for certiorari and litigation con-\\ncludes. But that step is for the Supreme Court to take—\\nnot this court.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4422108/\", \"author_raw\": \"REYNA, Circuit Judge, concurring\"}]}"},"judge1":{"kind":"string","value":"PROST"},"judge2":{"kind":"string","value":"MOORE"},"judge3":{"kind":"string","value":"REYNA"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":1,"string":"1"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4644855/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: VIRNETX INC., Appellant v. APPLE INC., Cisco Systems, Inc., Appellees\nCourt: U.S. Court of Appeals for the Federal Circuit\nDate decided: 2019-08-01\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: Yes\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n for the Federal Circuit\n ______________________\n\n VIRNETX INC.,\n Appellant\n\n v.\n\n APPLE INC., CISCO SYSTEMS, INC.,\n Appellees\n ______________________\n\n 2017-1591, 2017-1592, 2017-1593\n ______________________\n\n Appeals from the United States Patent and Trademark\nOffice, Patent Trial and Appeal Board in Nos. 95/001,788,\n95/001,789, 95/001,856.\n ______________________\n\n Decided: August 1, 2019\n ______________________\n\n STEPHEN BLAKE KINNAIRD, Paul Hastings LLP, Wash-\nington, DC, argued for appellant. Also represented by\nNAVEEN MODI, JOSEPH PALYS, IGOR VICTOR TIMOFEYEV,\nMICHAEL WOLFE, DANIEL ZEILBERGER.\n\n WILLIAM F. LEE, Wilmer Cutler Pickering Hale and\nDorr LLP, Boston, MA, argued for appellee Apple Inc. Also\nrepresented by REBECCA A. BACT, MARK CHRISTOPHER\nFLEMING, LAUREN B. FLETCHER, DOMINIC E. MASSA;\nBRITTANY BLUEITT AMADI, Washington, DC; SCOTT\nBORDER, JEFFREY PAUL KUSHAN, Sidley Austin LLP, Wash-\nington, DC.\n\n2 VIRNETX INC. v. APPLE INC.\n\n\n\n\n THEODORE M. FOSTER, Haynes & Boone, LLP, Dallas,\nTX, argued for appellee Cisco Systems, Inc. Also repre-\nsented by DAVID L. MCCOMBS, ANDREW S. EHMKE, DEBRA\nJANECE MCCOMAS.\n ______________________\n\n Before PROST, Chief Judge, MOORE and REYNA,\n Circuit Judges.\n Opinion for the court filed by Chief Judge PROST.\nOpinion concurring in part and dissenting in part filed by\n Circuit Judge REYNA.\nPROST, Chief Judge.\n Appellant VirnetX Inc. (“VirnetX”) appeals from deci-\nsions of the Patent Trial and Appeal Board (“Board”) re-\nlated to three inter partes reexaminations maintained by\nApple Inc. (“Apple”) and Cisco Systems, Inc. (“Cisco”). The\nUnited States Patent and Trademark Office (“PTO”) con-\ncluded that Apple was not barred from maintaining its\nreexams by the estoppel provision of the pre-America In-\nvents Act (“AIA”) version of 35 U.S.C. § 317(b) (2006). The\nBoard affirmed the Examiner’s determination that the\nclaims of U.S. Patent Nos. 7,418,504 (“the ’504 patent”) and\n7,921,211 (“the ’211 patent”) are unpatentable as antici-\npated or obvious over the prior art of record. For the rea-\nsons below, we affirm-in-part, vacate-in-part, and remand.\n BACKGROUND\n I\n The ’504 and ’211 patents describe systems and meth-\nods for “establishing a secure communication link between\na first computer and a second computer over a computer\nnetwork, such as the Internet.” ’211 patent col. 6 ll. 36–39.\nThese systems and methods are “built on top of the existing\nInternet protocol (IP).” Id. at col. 6 ll. 17–20.\n\nVIRNETX INC. v. APPLE INC. 3\n\n\n\n The Internet uses addressing systems for sending data.\nIn such systems, physical computers can be identified by a\nunique IP address (e.g., 123.345.6.7). VirnetX Inc. v. Apple\nInc., 665 F. App’x 880, 882 (Fed. Cir. 2016).\n Each IP address corresponds to a domain name (e.g.,\nwww.Yahoo.com). See ’211 patent col. 38 ll. 58–61, col. 39\nll. 13–14. A user on one computer can enter a domain name\nin a web browser to communicate with another computer\nor server. When the user does so, the computer sends a\ndomain name service (“DNS”) request to the domain name\nserver for the IP address corresponding to a given domain\nname. Id. at col. 38 l. 58–col. 39 l. 3. The domain name\nserver then looks up the IP address of the requested do-\nmain name and returns it to the requesting computer. Id.\nat col. 39 ll. 3–7.\n Both VirnetX patents claim systems, methods, and me-\ndia for creating secure communication links via DNS sys-\ntems. For example, claim 1 of the ’211 patent recites:\n 1. A system for providing a domain name service\n for establishing a secure communication link, the\n system comprising:\n a domain name service system configured and ar-\n ranged to\n [1] be connected to a communication network,\n [2] store a plurality of domain names and corre-\n sponding network addresses,\n [3] receive a query for a network address, and\n [4] indicate in response to the query whether the\n domain name service system supports establishing\n a secure communication link.\n\n4 VIRNETX INC. v. APPLE INC.\n\n\n\n\n Independent claims 36 and 60 of the ’211 patent are\ndirected to a “machine-readable medium” and a “method,”\nrespectively. Otherwise, they mirror the requirements of\nclaim 1. Independent claims 1, 36, and 60 of the ’504 pa-\ntent are similar to the corresponding independent claims of\nthe ’211 patent.\n II\n In 2010, VirnetX sued Apple in district court. VirnetX\nalleged infringement of four patents, including the ’504 and\n’211 patents. 1 VirnetX asserted claims 1, 2, 5, 16, 21, and\n27 of the ’504 patent and claims 36, 37, 47, and 51 of the\n’211 patent. VirnetX Inc. v. Apple Inc., 925 F. Supp. 2d 816,\n824–25 (E.D. Tex. 2013).\n In October 2011, Apple filed requests for inter partes\nreexamination of the ’504 and ’211 patents with the PTO.\nIn Apple’s Reexam Nos. 95/001,788 (“788 case”) and\n95/001,789 (“789 case”) (collectively, “Apple reexams”), Ap-\nple challenged all claims as anticipated by the Provino ref-\nerence or rendered obvious by Provino in view of other prior\nart. 2\n The district court action proceeded to trial in late 2012.\nA jury found the asserted claims infringed and not invalid.\nThe jury awarded VirnetX $368 million in damages. Vir-\nnetX, 925 F. Supp. 2d at 825. The district court denied Ap-\nple’s motion for judgment as a matter of law (“JMOL”) or a\nnew trial on these issues. Apple appealed.\n On appeal, we affirmed the jury’s finding of no invalid-\nity for all four patents. VirnetX, Inc. v. Cisco Sys., Inc., 767\nF.3d 1308, 1323–24 (Fed. Cir. 2014) (“VirnetX I”). We also\n\n\n\n 1 VirnetX also alleged that Apple infringed certain\nclaims of U.S. Patent Nos. 6,502,135 (“the ’135 patent”) and\n7,490,151 (“the ’151 patent”).\n 2 U.S. Patent No. 6,557,037 (“Provino”).\n\nVIRNETX INC. v. APPLE INC. 5\n\n\n\naffirmed the jury’s finding of infringement for many of the\nclaims of the two patents not related to the present appeal\n(’135 and ’151 patents). 3 Id. at 1320–22. We reversed the\ndistrict court’s construction of the “secure communication\nlink” claim term, vacated the related infringement finding\nfor the two patents in this appeal (’504 and ’211 patents),\nand vacated the damages award. Id. at 1317–19, 1319,\n1323–24, 1325–34. We then remanded for further proceed-\nings. 4 Id. at 1334.\n Apple did not file a request for rehearing on the inva-\nlidity or infringement issues affirmed in VirnetX I. Our\nmandate issued on December 23, 2014. Apple did not seek\nSupreme Court review. The 90-day period to file a petition\nfor a writ of certiorari expired.\n Meanwhile, in the parallel PTO reexamination pro-\nceedings, the Examiner had found all claims of the ’504 and\n’211 patents unpatentable. The Examiner issued Right of\nAppeal Notices (“RANs”) in May 2014. VirnetX appealed\nthe Examiner’s decisions to the Board.\n VirnetX also petitioned the PTO to terminate the Apple\nreexams based on the estoppel provision of § 317(b). The\nPTO denied VirnetX’s petition in June 2015. J.A. 1659–67,\n3138–48.\n In September 2016, the Board affirmed the Examiner’s\nfindings that all claims of the ’504 and ’211 patents were\n\n\n 3 We reversed the district court’s conclusion related\nto the doctrine of equivalents for one claim of the ’151 pa-\ntent. See id. at 1322–23.\n 4 On remand in 2016, a jury found that Apple in-\nfringed claims 1, 2, 5, and 27 of the ’504 patent and claims\n36, 47, and 51 of the ’211 patent, and awarded damages.\nVirnetX Inc. v. Apple Inc., 324 F. Supp. 3d 836 (E.D. Tex.\n2017). We recently affirmed. See VirnetX Inc. v. Cisco Sys.,\nInc., No. 2018-1197, 748 F. App’x 332 (Fed. Cir. 2019).\n\n6 VIRNETX INC. v. APPLE INC.\n\n\n\n\nunpatentable. The Board denied VirnetX’s requests for re-\nhearing.\n After appealing the Board’s decisions in the Apple reex-\nams to this court, VirnetX moved to remand. It argued that\nthis court’s 2014 opinion was a “final decision” on Apple’s\nattempt to prove invalidity under § 317(b). ECF No. 27. In\nJune 2017, a motions panel denied the motion and directed\nthe parties to address the issue in their merits briefing.\nECF No. 36.\n III\n In December 2011, Cisco also filed a request for inter\npartes reexamination with the PTO. 5 Cisco’s Reexam No.\n95/001,856 (“856 case” or “Cisco reexam”) challenged\nclaims 1–60 of the ’211 patent based on multiple grounds\nof invalidity. Cisco based several of its arguments on the\nLendenmann reference. 6\n After extensive proceedings, the Examiner issued a\nRAN in January 2015. The RAN rejected claims 36–60 of\nthe ’211 patent as either anticipated or obvious. VirnetX\nappealed to the Board. In September 2017, the Board af-\nfirmed the Examiner, finding claims 36–54 and 57–60 an-\nticipated by Lendenmann and claims 55 and 56 obvious\nover Lendenmann and another reference. J.A. 88. VirnetX\nmoved for rehearing, which was denied. VirnetX then ap-\npealed.\n We have jurisdiction under 28 U.S.C. § 1295(a)(4)(A).\n\n\n\n\n 5 In 2010, VirnetX also sued Cisco in district court,\nalleging inter alia infringement of claim 1 of the ’211 pa-\ntent.\n 6 Rolf Lendenmann, Understanding OSF DCE 1.1\nfor AIX and OS/2, IBM International Technical Support\nOrganization, pp. 1–45 (Oct. 1995) (“Lendenmann”).\n\nVIRNETX INC. v. APPLE INC. 7\n\n\n\n DISCUSSION\n I\n VirnetX’s appeal proceeds in two parts. First, VirnetX\nargues as a threshold matter that Apple was estopped from\nmaintaining its reexams under the pre-AIA version of\n35 U.S.C. § 317(b) (2006). Second, VirnetX challenges the\nmerits of the Board’s conclusions in both the Apple and\nCisco reexams that the ’504 and ’211 patents are invalid.\nWe address each in turn.\n II\n VirnetX contends that Apple’s reexams were barred by\n§ 317(b). The PTO refused to terminate Apple’s reexams\nbased on the conclusion that the provision did not apply.\nIn VirnetX’s view, the PTO’s decision was inconsistent with\ncontrolling case law, the statutory text, and Congress’s in-\ntent. We agree.\n The applicability of § 317(b) is a question of statutory\ninterpretation. See Bettcher Indus., Inc. v. Bunzl USA,\nInc., 661 F.3d 629, 639, 642–48 (Fed. Cir. 2011). Statutory\ninterpretation is an issue of law we review de novo. Un-\nwired Planet, LLC v. Google Inc., 841 F.3d 1376, 1379 (Fed.\nCir. 2016).\n The Patent Act requires that the PTO terminate a reex-\namination once there has been a final decision on the pa-\ntent challenger’s invalidity case in federal court. 35 U.S.C.\n§ 317(b) (2006). Specifically, pre-AIA § 317(b) provides:\n Once a final decision has been entered against a\n party in a civil action arising in whole or in part\n under section 1338 of title 28, that the party has not\n sustained its burden of proving the invalidity of any\n patent claim in suit . . . then neither that party nor\n its privies may thereafter request an inter partes\n reexamination of any such patent claim on the ba-\n sis of issues which that party or its privies raised\n\n8 VIRNETX INC. v. APPLE INC.\n\n\n\n\n or could have raised in such civil action or inter\n partes reexamination proceeding, and an inter\n partes reexamination requested by that party or its\n privies on the basis of such issues may not thereaf-\n ter be maintained by the Office, notwithstanding\n any other provision of this chapter.\n The issue here is very narrow. Apple does not dispute\nthe underlying procedural facts relevant to § 317(b), which\nhighlight the advanced stage of these proceedings.\n In 2010, Apple was sued for patent infringement. There\nis no dispute that in that “civil action arising . . . under”\n28 U.S.C. § 1338, Apple raised an affirmative defense of in-\nvalidity on multiple grounds. 35 U.S.C. § 317(b). Thus,\nApple had the “burden of proving the invalidity” of the as-\nserted claims of both the ’504 and ’211 patents. Id.\n In 2012, Apple presented its invalidity defenses to a\njury. After a week-long trial, the jury concluded Apple\nfailed to sustain its burden of proving the invalidity of any\nasserted claim of the ’504 or ’211 patent.\n In 2013, the district court determined that in light of\nthe record at trial, substantial evidence supported the\njury’s finding. The district court entered final judgment.\n In 2014, Apple pursued an appeal of that decision here.\nSpecifically, Apple argued that “the asserted claims are an-\nticipated by the Kiuchi reference.” VirnetX I, 767 F.3d at\n1323. After considering Apple’s challenges, we affirmed\nthe district court’s decision on the issue of invalidity. Id.\nat 1323–24. Apple did not file a request for rehearing. Ap-\nple elected not to seek Supreme Court review. The 90-day\nperiod to file a petition for certiorari expired.\n In 2015, VirnetX petitioned the PTO to terminate Ap-\nple’s pending reexaminations based on the estoppel provi-\nsion of § 317(b). The PTO refused, accepting Apple’s\nargument that there was not yet a final decision on valid-\nity. J.A. 1661–63.\n\nVIRNETX INC. v. APPLE INC. 9\n\n\n\n On appeal, the sole issue is whether there has been a\n“final decision” entered against Apple that it “has not sus-\ntained its burden of proving” invalidity, triggering estoppel\nof its parallel PTO reexam proceedings.\n Apple does not dispute that since it first raised the is-\nsue in 2010, it has now lost on invalidity before the jury,\nthe district court, and this court. Nor does it dispute that\nit allowed the time to petition for certiorari of the validity\nissues in the first appeal to lapse. Instead, Apple takes the\nposition that there was no final decision due to a hypothet-\nical future appeal.\n Apple’s argument proceeds as follows. In VirnetX I, we\nvacated and remanded other issues unrelated to invalidity,\nincluding infringement and damages. In Apple’s view, af-\nter those issues are decided on remand and reviewed in a\nsecond appeal, it may choose to file a petition for certiorari\nof those non-validity issues. In so doing, it may also ask\nthe Supreme Court to look at any issues from the first ap-\npeal—including the validity determination.\n Thus, the Court might elect to examine the validity is-\nsues at some point in the future. The Court has the ability\nto consider not only those issues “‘before the court of ap-\npeals upon the second appeal,’” but also the “questions\nraised on the first appeal.” Mercer v. Theriot, 377 U.S. 152,\n153–54 (1964) (per curiam) (quoting Hamilton-Brown Shoe\nCo. v. Wolf Bros. & Co., 240 U.S. 251, 257 (1916)); see also\nMajor League Baseball Players Ass’n v. Garvey, 532 U.S.\n504, 508 n.1 (2001) (per curiam) (noting Supreme Court\nhas “authority to consider questions determined in earlier\nstages of the litigation where certiorari is sought from the\nmost recent of the judgments of the Court of Appeals”). Ac-\ncording to Apple, the decision is only final when there is no\npotential that the Supreme Court might some day examine\nthe invalidity issue during a second appeal.\n\n10 VIRNETX INC. v. APPLE INC.\n\n\n\n\n In response, VirnetX contends there is a “final deci-\nsion” on a party’s attempt to prove invalidity after the\nparty fails to petition for certiorari within the 90-day pe-\nriod.\n Two reasons compel the conclusion that there has been\na final decision on validity in this case. First, our decision\nin Fairchild (Taiwan) Corp. v. Power Integrations, Inc., 854\nF.3d 1364 (Fed. Cir. 2017), controls and holds that there is\na “final decision” under § 317(b) once the time to file a cer-\ntiorari petition has passed. Second, even if Fairchild did\nnot foreclose Apple’s argument, the plain language of the\nstatute and its purpose advise against Apple’s definition of\na “final decision” on the issue of invalidity.\n A\n Since the PTO’s decision in 2016, we have ruled on this\nvery issue. Fairchild directly addressed a situation in\nwhich validity was affirmed but other issues in the case\nwere remanded. Fairchild holds that if a finding of no in-\nvalidity has been affirmed on appeal and remand of other\nissues will not “have any affect” on validity, the decision is\n“final” once the 90-day deadline for a certiorari petition on\nthe validity determination has passed. Id. at 1366.\nFairchild announced this holding despite the fact that\nthere is always some “potential” for future Supreme Court\nreview if the entire case is taken up. Accordingly, Fairchild\nrequires we reject Apple’s position.\n In Fairchild, we began our analysis by restating our\nearlier holding in Bettcher that estoppel “applies when ‘all\nappeals have terminated.’” Id. at 1365 (quoting Bettcher,\n661 F.3d at 646). We then concluded that is “precisely the\nsituation here.” Id. Namely, the district court “entered\njudgment against Power Integrations, holding that it failed\nto prove claims 6, 7, 18, and 19 were obvious over [the cited\nprior art]. This court affirmed the holding, and the time to\n\nVIRNETX INC. v. APPLE INC. 11\n\n\n\npetition for a writ of certiorari has passed.” Id. at 1365–66\n(emphasis added).\n As in Fairchild, the district court here entered judg-\nment against Apple on the issue of invalidity, we affirmed\nthe no-invalidity finding on appeal in VirnetX I, and the\ntime to file a petition for certiorari on the issue has long\nsince passed. Under Fairchild, that constitutes “a final de-\ncision” under § 317(b) with respect to Apple’s failure to\nprove invalidity.\n Resisting this conclusion, Apple argues that if any non-\ninvalidity issues remain in the district court case related to\nthe patent, then the validity decision cannot be final under\n§ 317(b). Since VirnetX I vacated and remanded other is-\nsues (i.e., infringement and damages) on both patents at\nissue, Apple contends that it waited until all the issues\nwith this patent case are decided before bundling it up in a\npetition for certiorari. Consequently, Apple’s position re-\nquires resolution of the entire case—or at least all issues on\neach patent, including infringement, invalidity, and dam-\nages—before a validity decision is considered “final.”\n Fairchild rejected such arguments under almost iden-\ntical circumstances. There, we had also vacated and re-\nmanded other parts of the case for additional proceedings.\nId. at 1365 (explaining prior appeal upheld obviousness but\nreversed jury’s findings on infringement, and “remanded\nfor further proceedings unrelated to the ’972 patent\nclaims”). In turn, we rejected the argument that this ren-\nders the decision not “final” under § 317(b). Id. at 1366.\nWe reasoned:\n While it is true that in [the prior appeal] this court\n vacated and remanded for additional proceedings,\n we cannot agree with Power Integrations that this\n renders the decision not “final” for § 317(b) pur-\n poses. Critically, those proceedings are unrelated\n to the ’972 patent. By its terms, § 317(b) is\n\n12 VIRNETX INC. v. APPLE INC.\n\n\n\n\n concerned with a final decision ‘that the party has\n not sustained its burden of proving the invalidity\n of any patent claim.’ And here, Power Integrations\n does not suggest, nor is there any reason to believe,\n that any unresolved issue on remand would have\n any effect on the now-final ’972 patent validity de-\n terminations.\nId. (emphasis added) (quoting 35 U.S.C. § 317(b)).\n Likewise, here, Apple does not point to any issue in the\nVirnetX I remand that substantively bears on the finding\nof no invalidity. Nor could it. Indeed, the only difference\nhere is that VirnetX I remanded some infringement and\ndamages issues related to the ’504 and ’211 patents. But\nApple fails to explain how any of these unresolved issues\nwould “have any effect on the now-final . . . validity deter-\nminations” involving those patents. 7 Id. Accordingly,\nFairchild’s reasoning leaves no doubt that § 317(b) estop-\npel applies.\n Finally, Apple contends Fairchild cannot control be-\ncause no party in that case expressly argued that the Su-\npreme Court might still take up the invalidity issue on a\nfuture appeal under Mercer. But Fairchild’s reasoning im-\nplicitly forecloses such an argument. Fairchild concluded\nthat “all appeals” had terminated, 854 F.3d at 1365,\n\n\n\n 7 At oral argument, Apple could not point to a single\nexample illustrating how an aspect of the infringement re-\nmand might meaningfully affect or implicate the invalidity\nruling this court affirmed in 2014 in VirnetX I. Instead,\nApple offered only speculation. Its argument boiled down\nto the general conjecture that the district court “could”\nhave a motion to set aside a validity determination even\nafter the mandate issues based on arguments made about\ninfringement. Oral Argument at 29:20–51, No. 2017-1591,\nhttp://www.cafc.uscourts. gov/oral-argument-recordings.\n\nVIRNETX INC. v. APPLE INC. 13\n\n\n\ndespite the fact that other issues in that case remained\npending for potential appeal and petition for certiorari. Ap-\nple can disagree with that conclusion, but that is not a ba-\nsis for ignoring binding authority. We are bound to follow\nthe decision of a prior panel, and its reasoning does not\nleave the question open.\n In sum, § 317(b) applies here despite the fact that is-\nsues unrelated to invalidity were remanded.\n The dissent takes the view that Fairchild is inapplica-\nble. But it relies on reasons that were not material to the\noutcome in Fairchild. In addition, the dissent’s attempt to\ndistinguish Fairchild suffers from several additional infir-\nmities.\n According to the dissent, in Fairchild “Power Integra-\ntions did not appeal the court’s favorable ruling on infringe-\nment because it prevailed on that issue.” Opinion\nconcurring in part and dissenting in part (“dissent”) at 4\n(emphasis added). In turn, the dissent leaps to the conclu-\nsion that Power Integrations could not have realistically\npursued certiorari review of the ’972 patent’s validity be-\ncause it “prevailed” on infringement. Id. In the dissent’s\nview, “once Power Integrations prevailed on infringement,\nthe case against it was concluded” and the potential for any\nfurther Supreme Court review was eliminated. Id. at 4–5.\nThere is no support for these statements.\n Fundamentally, infringement and invalidity are sepa-\nrate “issues.” Pandrol USA, LP v. Airboss Ry. Prod., Inc.,\n320 F.3d 1354, 1364 (Fed. Cir. 2003) (“Supreme Court prec-\nedent and our cases make clear that patent infringement\nand patent validity are treated as separate issues.”).\nFairchild never held that resolving infringement resolves\ninvalidity. Indeed, Fairchild was interpreting an estoppel\nstatute triggered by a decision on invalidity, not infringe-\nment.\n\n14 VIRNETX INC. v. APPLE INC.\n\n\n\n\n Moreover, the dissent’s position unravels in light of the\nprocedural posture in Fairchild. Power Integrations did\nnot simply argue that it did not infringe the ’972 patent.\nNor did it merely raise invalidity as an affirmative defense\nto liability. Rather, Power Integrations advanced a sepa-\nrate counterclaim for declaratory judgment that the ’972\npatent was invalid. See Plaintiff’s Fourth Amended An-\nswer and Counterclaims at 23, Power Integrations, Inc. v.\nFairchild Semiconductor Int’l, Inc. et al, No. 1:08–cv–\n00309–LPS (D. Del. July 7, 2011) (Dkt. No. 409). There-\nfore, to say it “prevailed” and its case “concluded” on that\nissue due to a non-infringement ruling is not correct.\n It is black letter law that non-infringement does not\nmoot a counterclaim regarding invalidity. “A party seeking\na declaratory judgment of invalidity presents a claim inde-\npendent of the patentee’s charge of infringement.” Cardi-\nnal Chem. Co. v. Morton Int’l, 508 U.S. 83, 96 (1993). Thus,\n“appellate affirmance of a judgment of noninfringement,\neliminating any apprehension of suit, does not moot a de-\nclaratory judgment counterclaim of patent invalidity.”\nMedImmune, Inc. v. Genentech, Inc., 549 U.S. 118, 132 n.11\n(2007); see also Telcordia Techs., Inc. v. Cisco Sys., Inc., 612\nF.3d 1365, 1375 (Fed. Cir. 2010) (“[T]his court’s affirmance\nof the district court’s non-infringement findings as to the\n’306 patent does not moot Cisco’s invalidity counterclaim\non cross-appeal.”).\n In addition, the dissent speculates as to why the litigant\nchose not to seek further review. The dissent appears to\nassume that if a party is found not liable for certain ac-\ncused products infringing a patent, as in Fairchild, that\nparty no longer has the incentive to challenge the underly-\ning property right’s validity. Dissent at 5–6. This assump-\ntion fails for several reasons. First, inferring the subjective\nintent of the litigants—either on remand or in petitioning\nfor certiorari—is inappropriate. Fairchild never made\nsuch fact findings. Fairchild never relied on such infer-\nences either. Put simply, Fairchild did not base its decision\n\nVIRNETX INC. v. APPLE INC. 15\n\n\n\non speculation as to whether remand would impact a\nparty’s subjective “decision” to pursue certiorari review or\nwhether its decisions regarding petitioning were “reasona-\nble,” id. at 5–6. Second, such speculation is irrelevant. The\ncontrolling inquiry is mandated by statute, which makes\nno mention of a party’s motives for petitioning for certiorari\nreview. Third, it is inaccurate. Regardless of a finding of\nno liability on the presently accused products, an accused\ninfringer may well have reasons for trying to invalidate the\npatent (e.g., to guarantee its freedom to operate).\n But the dissent suffers from a more fundamental flaw.\nThe dissent appears to suggest that in Fairchild, “potential\nfuture review” by the Supreme Court was not an option.\nId. at 4. In light of Mercer, there is no basis for that posi-\ntion. Indeed, it runs headlong into the dissent’s own insist-\nence that the Supreme Court’s discretion to take up\ncertiorari review of earlier issues in a case is broad. On the\none hand, the dissent acknowledges that a “petition for writ\nof certiorari can expose the entire case to review,” id. at 10\n(quoting Christianson v. Colt Indus. Operating Corp., 486\nU.S. 800, 817 (1988)), and the Supreme Court has “juris-\ndiction to consider all of the substantial federal questions\ndetermined in the earlier stages of the litigation,” id. (quot-\ning Reece v. Georgia, 350 U.S. 85, 87 (1955)). On the other\nhand, the dissent contends that the litigant’s decision in\nFairchild “all but foreclosed the possibility that the inva-\nlidity issue could be subject to potential future review by\nthe Supreme Court.” Id. at 4. A plain reading of Fairchild,\nhowever, confirms it never presumed to curtail the Su-\npreme Court’s discretion under Mercer to review issues\nraised earlier in a case.\n Finally, the dissent is concerned that the majority opin-\nion somehow “intrudes on the Supreme Court’s constitu-\ntional prerogative whether to take up the invalidity issue.”\nId. at 11. It does not. Like Fairchild, our decision concerns\nestoppel of duplicative administrative proceedings before\nthe PTO. It has no bearing on the Supreme Court’s\n\n16 VIRNETX INC. v. APPLE INC.\n\n\n\n\ndiscretion to accept or reject a petition for certiorari on any\nissue—nor could it.\n In sum, we disagree with the dissent’s efforts to rewrite\nFairchild. Fairchild speaks for itself. The controlling stat-\nute is trained on a “final decision” that the party has not\nsustained its “burden of proving the invalidity” of the pa-\ntent claim. Fairchild, 843 F.3d at 1366. As the dissent\nacknowledges, the reasoning at the heart of the Fairchild\ndecision is that estoppel applied to the duplicative PTO\nreexam because there was “no suggestion,” dissent at 6,\nthat “any unresolved issue on remand would have any ef-\nfect on the now-final patent validity determinations” from\nthe district court, Fairchild, 854 F.3d at 1366. That is, the\nremand would have no substantive effect on the validity\ndetermination. The record is clear that is the same situa-\ntion here.\n B\n Even if Fairchild did not control, the statutory text and\npurpose of § 317(b) compels us to reject Apple’s theory that\nthe entire case must be resolved. Apple essentially asks us\nto rewrite the statute, triggering estoppel only when there\nis a “final judgment” on all issues in a case.\n But the statute’s text is clear. The statute only re-\nquires “a final decision” on a precise issue—whether the\nparty has “sustained its burden of proving the invalidity of\nany patent claim in suit.” 35 U.S.C. § 317(b). The statute\nmakes no mention of resolving the entire case. Nor does it\nrequire resolution of all other issues related to a patent\n(e.g., infringement or damages). Instead, the plain lan-\nguage of the statute demands a party cease duplicative in-\nvalidity challenges via reexams in the PTO once “a final\ndecision” has been reached on an invalidity challenge in a\ncivil action.\n This reading of the plain text is further reinforced by\nthe statute’s purpose. The statute is intended to protect\n\nVIRNETX INC. v. APPLE INC. 17\n\n\n\npatent holders against “harassment” from duplicative liti-\ngation on the issue of invalidity. See 145 Cong. Rec. 29,276\n(Nov. 9, 1999) (Conference Report on H.R. 1554) (“Subtitle\nF creates a new section 317 which sets forth certain condi-\ntions by which inter partes reexamination is prohibited to\nguard against harassment of a patent holder.”); see also\nFunction Media, L.L.C. v. Kappos, 508 F. App’x 953, 955–\n56 (Fed. Cir. 2013) (“This provision existed to prevent du-\nplicative, harassing actions against a patentee. If a defend-\nant brought an invalidity challenge in a district court\nlitigation and was unsuccessful, it is not permitted to bring\nthe same challenge in an inter partes reexamination.”).\n Apple concedes that § 317(b) was intended to prevent\nduplicative litigation. Apple’s Br. 40–41. Yet Apple’s posi-\ntion increases the likelihood duplicative litigation will oc-\ncur. If no decision on invalidity is “final” until any\n“prospect of future Supreme Court review” is eliminated,\nApple’s Br. 36, estoppel will rarely apply. In light of the\nlegislative intent, it is doubtful Congress created an estop-\npel provision that would allow a PTO proceeding to press\nforward even after all appellate deadlines have elapsed\nmerely because there is a potential that the Supreme Court\nmay one day reconsider the invalidity issue—in many\ncases, years later—while examining the rest of the case. 8\n Accordingly, the statutory purpose confirms that “a fi-\nnal decision” triggers estoppel when the invalidity chal-\nlenge is decided on appeal and the time for petitioning for\ncertiorari has passed—regardless of the fact review of the\n\n\n\n 8 Moreover, there is no indication that Congress con-\nsidered the type of review contemplated in Mercer as part\nof the definition of a regular appeals process for reaching a\n“final decision” on invalidity. This is of no surprise given\nthe statute is trained on the finality of a particular issue,\nrather than finality of the entire case.\n\n18 VIRNETX INC. v. APPLE INC.\n\n\n\n\nentire case could occur at some point in the future under\nMercer.\n C\n Apple’s remaining arguments against the application\nof § 317(b) here are unpersuasive. Apple first suggests that\nFairchild is inconsistent with Bettcher. Apple’s Br. 32–33,\n37. In Apple’s view, Bettcher held that § 317(b) is not sat-\nisfied until the Supreme Court can no longer review a case.\n Bettcher never reached such a holding. Bettcher con-\ncerned a different estoppel provision, pre-AIA 35 U.S.C.\n§ 315(c) (2006). 9 The question was whether § 315(c) estop-\npel attaches at a far earlier stage (before any appeal is pur-\nsued). In Bettcher, the district court concluded that\nestoppel barred further civil litigation “as soon as the ex-\naminer finished the reexamination and the Right of Appeal\nNotice was sent to the patentee”—i.e., before there is even\nan appeal to the Board, let alone an appeal to the Federal\nCircuit. Id. at 636. On appeal, we rejected that conclusion.\nWe held that “the estoppel provision of § 315 is triggered\nnot when examination is completed but only after all ap-\npeal rights have been exhausted.” Id. at 642–43.\n The panel then rejected the argument that § 317(b)\nsomehow showed that § 315(c) estoppel must apply as soon\nas the examiner finishes the reexam. See id. at 643. The\npanel observed that like § 315(c), estoppel under § 317(b)\napplies “when reexamination and all appeals have termi-\nnated.” Id. at 646 (emphasis added).\n Bettcher addressed appeal rights to the Board and the\nFederal Circuit. It did not analyze how petitions for\n\n\n 9 Section 315(c) concerns estoppel in the opposite di-\nrection: when the PTO’s determination in a reexam estops\na party from asserting invalidity in a civil action. Id. at\n643.\n\nVIRNETX INC. v. APPLE INC. 19\n\n\n\ncertiorari impact § 317(b). 10 Moreover, Bettcher did not ad-\ndress the critical inquiry here as to whether a “final deci-\nsion” regarding the issue of invalidity under § 317(b) is\naffected by remand of issues unrelated to invalidity.\n By contrast, Fairchild did reach these issues.\nFairchild held that estoppel “applies when ‘all appeals’\nhave terminated,” meaning after the district court’s judg-\nment of no-invalidity was affirmed on appeal and “the time\nto petition for a writ of certiorari has passed.” Id. at 1365–\n66 (quoting Bettcher, 661 F.3d at 646). The panel reached\nthis conclusion despite the fact that issues unrelated to pa-\ntent validity were remanded for further proceedings.\nProperly understood, Fairchild is an extension of Bettcher\nto Supreme Court certiorari review. Apple’s attempt to\n\n\n 10 Bettcher concluded that the estoppel provision of\n§ 315 attaches when all appeal rights are exhausted. As\nthe panel observed, the opening sections of the statute list\nthose appeal “rights” for both the patent owner and the\nthird-party requestor, § 315(a)–(b), which consist of ap-\npeals from (1) the examiner to the Board, and (2) the Board\nto this court. Id. at 644 (“The estoppel provision of subsec-\ntion (c) thus falls immediately following the establishment\nof both parties’ rights of appeal, including appeal to this\ncourt.”). Bettcher did not address whether petitions for cer-\ntiorari are considered “appeals,” or whether certiorari re-\nview by the Supreme Court is a “right.” In VirnetX’s view,\nan appeal as of right is a different procedural mechanism\nthan a writ of certiorari. VirnetX Reply Br. 12–13; see also\nCity & Cty. of San Francisco, Calif. v. Sheehan, 135 S. Ct.\n1765, 1774 (2015) (“Because certiorari jurisdiction exists to\nclarify the law, its exercise ‘is not a matter of right, but of\njudicial discretion.’” (quoting Supreme Court Rule 10)).\nThough we need not reach such questions, they underscore\nthe fact that Bettcher simply did not reach the issue of cer-\ntiorari review now governed by Fairchild.\n\n20 VIRNETX INC. v. APPLE INC.\n\n\n\n\nmanufacture an inconsistency between these two cases has\nno merit.\n Apple’s reliance on Fresenius fares no better. Fresenius\nUSA, Inc. v. Baxter Int’l, Inc., 721 F.3d 1330 (Fed. Cir.\n2013). Apple concedes that Fresenius “did not construe\n§ 317(b).” Apple’s Br. 39. As such, Apple admits Fresenius\nhas no direct bearing on our interpretation of this statutory\nestoppel provision.\n In addition, Fresenius concerned an entirely different\nissue involving res judicata from a “final judgment.” Even\nwithin the distinct realm of res judicata, Fresenius took\ngreat care to “distinguish between different concepts of fi-\nnality.” Id. at 1340. In Fresenius, we specifically addressed\nwhether a judgment in an infringement case is “sufficiently\nfinal” so that it is “immune to the effect of the final judg-\nment in the PTO proceedings, as affirmed by this court in\nIn re Baxter.” Id. at 1341 (discussing In re Baxter Int’l, Inc.,\n678 F.3d 1357 (Fed. Cir. 2012)).\n The issue in Fresenius arose due to the fact that while\nthe appeal of an infringement action was pending, we af-\nfirmed the PTO reexamination finding the patent claims\ninvalid in In re Baxter. As a result, the PTO canceled the\nclaims. Id. at 1335. Despite the intervening cancellation,\nthe patent owner argued that the district court’s judgment\nwas “final” for purposes of res judicata such that the liabil-\nity finding and past damages award in the pending litiga-\ntion could not be disturbed. Id. at 1340. We disagreed.\n While the district court in 2007 had entered “a judg-\nment final for purposes of appeal,” we concluded that it was\n“not sufficiently final to preclude application of the inter-\nvening final judgment in In re Baxter.” Id. at 1341. More-\nover, the district court’s original judgment had since been\nset aside and the remand was not sufficiently final for res\njudicata purposes. Id. “To rise to that level, the litigation\nmust be entirely concluded so that [the] cause of action\n[against the infringer] was merged into a final judgment\n\nVIRNETX INC. v. APPLE INC. 21\n\n\n\n. . . one that ‘ends the litigation on the merits and leaves\nnothing for the court to do but execute the judgment.’” Id.\n(quoting Mendenhall v. Barber–Greene Co., 26 F.3d 1573,\n1580 (1994)). We concluded that the remand did not end\nthe controversy, leaving only the execution of the judg-\nment, given that several aspects of the litigation (e.g., roy-\nalty damages, injunctive relief) remained unresolved. Id.\n Fresenius’s analysis of a “final judgment” for purposes\nof res judicata does not instruct our understanding of a “fi-\nnal decision” for purposes of § 317(b) estoppel. Indeed, it\nonly underscores why we must follow Fairchild. Fairchild\nfound estoppel absent a final judgment. Indeed, after mul-\ntiple other issues in the litigation, including infringement\nand damages for certain patents, were remanded,\nFairchild concluded nonetheless that there was a “final de-\ncision” on the narrow issue of invalidity—which is all that\nthe statute requires to trigger estoppel. The reasoning in\nFairchild, not Fresenius, governs our application of\n§ 317(b).\n Apple also briefly argues that this interpretation of the\nstatute generates duplicative proceedings for the patent\nchallenger. Apple’s Br. 41. In Apple’s view, it must seek\n“interlocutory” review rather than waiting for the case to\nbe entirely resolved. Id. Apple’s argument misses the\nmark. Section 317(b) is concerned with preventing dupli-\ncative litigation against the patent owner. Of course, we\nrecognize the challenging nature of these proceedings for\nboth parties. The present actions involve multiple patents\nand various issues, which create certain complexities for\npursuing appeals before this court and seeking certiorari\nbefore the Supreme Court. Some of those complexities\nmight be ameliorated by requiring final judgment of an en-\ntire case. But that does not empower us to rewrite a statute\nthat requires only a final decision on the issue of invalidity.\nFurthermore, such interlocutory and piecemeal concerns\nwere also present in Fairchild. Nonetheless, estoppel ap-\nplied.\n\n22 VIRNETX INC. v. APPLE INC.\n\n\n\n\n Finally, Apple’s arguments based on the PTO’s inter-\npretations carry no force. They fail for several reasons.\n First, Apple fails to explain why the PTO’s interpreta-\ntions of § 317(b) is entitled to Chevron deference. As dis-\ncussed above, the statutory text and purpose leave no\nambiguity. Where “the intent of Congress is clear, that is\nthe end of the matter.” Chevron U.S.A. Inc. v. NRDC, Inc.,\n467 U.S. 837, 842 (1984). Therefore, Chevron deference\ndoes not apply. 11 12\n Second, Skidmore deference is not appropriate either.\nUnder Skidmore, the weight given to a particular agency\ninterpretation “will depend upon the thoroughness evident\nin its consideration, the validity of its reasoning, its con-\nsistency with earlier and later pronouncements, and all\nthose factors which give it power to persuade, if lacking\n\n\n 11 Apple also suggests that Bettcher requires defer-\nence. Not so. Bettcher concluded the statute was unambig-\nuous. Bettcher, 661 F.3d at 646 (“[W]e believe that § 317(b)\nis clear . . . .”). Nonetheless, Apple points to Bettcher’s sub-\nsequent discussion of the deference it “would” give the\nMPEP had it reached the issue. Id. Apple’s reliance on a\nsingle line of dicta from Bettcher is unavailing. Even if it\nwere not dicta, Bettcher’s statement concerned deference to\nthe Manual of Patent Examining Procedure (“MPEP”)\n§ 2686.04. Id. As explained later, that section of the MPEP\ndoes not resolve the issue at hand.\n 12\n Apple also suggests the underlying PTO denial of\nthe petition to terminate here is entitled to Chevron defer-\nence. In support of this position, Apple cites only to In re\nAffinity Labs of Texas, LLC, 856 F.3d 883, 894 n.5 (Fed.\nCir. 2017). The cited footnote from Affinity Labs did not\nconcern deference. Rather, it concerned whether PTO rul-\nings on petitions to terminate are reviewable on appeal. Id.\n(analyzing stages of PTO decision-making process as re-\nlated to reviewability).\n\nVIRNETX INC. v. APPLE INC. 23\n\n\n\npower to control.” United States v. Mead Corp., 533 U.S.\n218, 228 (2001) (quoting Skidmore v. Swift & Co., 323 U.S.\n134, 140 (1944)). None of the considerations favoring def-\nerence are present here. For example, contrary to Apple’s\nposition, the PTO has never established a “consistent” con-\nstruction of § 317(b). Indeed, aside from the decision below,\nApple cites no other decision in which the PTO has inter-\npreted § 317(b) to extend passed the deadline for certiorari\nreview of the invalidity issue to hypothetical petitions aris-\ning from appeals of non-invalidity issues on remand. Thus,\ndeference is unwarranted.\n Third, regardless of whether deference applies, none of\nApple’s citations resolve the specific issue at hand. For in-\nstance, no PTO regulation or MPEP section states that a\ndecision is not “final” for purposes of § 317(b) if non-inva-\nlidity issues are remanded. Nor do any of these provisions\nstate that a final judgment is required, or that as long as\nthe potential for Supreme Court review from a later peti-\ntion exists, § 317(b) estoppel does not attach. Indeed, none\nof these provisions even mention Supreme Court certiorari\nreview. 13\n For the reasons above, § 317(b) estoppel applies to the\nApple reexams. Accordingly, we vacate and remand with\n\n\n 13 Other PTO regulations suggest appeals terminate\nwhen our mandate issues. See, e.g., 37 C.F.R. § 1.197 (“An\nappeal to the U.S. Court of Appeals for the Federal Circuit,\nwhether from a decision of the Board or a judgment in a\ncivil action, is terminated when the mandate is issued by\nthe Court.”). Notably, 37 C.F.R. § 1.979, which is the reg-\nulation interpreting pre-AIA § 316(a) permitting the PTO\nto issue a certificate of cancellation “only after all appeals\nhave terminated,” Bettcher, 661 F.3d at 644–45, applies the\nsame interpretation. See 37 C.F.R. § 1.979(b) (providing\nthat an “appeal [to this Court] is considered terminated\nwhen the mandate is issued by the Court”).\n\n24 VIRNETX INC. v. APPLE INC.\n\n\n\n\ninstructions to terminate the Apple reexams with respect\nto claims 1–35 of the ’504 patent and claims 36–59 of the\n’211 patent. 14\n III\n\n We now turn to VirnetX’s challenge to the merits of the\nApple and Cisco reexams that pertain to the claims not\nsubject to estoppel. The Board affirmed the Examiner’s re-\njection of all claims of the ’504 and ’211 patents. On appeal,\nVirnetX takes a kitchen sink approach, arguing that the\nBoard decisions incorrectly construe certain claims, shift\nthe burden to the patent owner to prove patentability, and\nlack substantial evidence. VirnetX’s arguments are unper-\nsuasive. Therefore, we affirm the Board’s decision on all\nremaining claims of both patents in the Apple reexam not\nsubject to § 317(b) estoppel and fully affirm the Board’s de-\ncision regarding the claims of the ’211 patent in the Cisco\nreexam. 15\n\n\n\n 14 Apple appears not to dispute that estoppel extends\nto the dependent claims here. Section 317(b) bars main-\ntaining a reexam based on issues which that party “raised”\nor “could have raised” in the civil action. The PTO found\nthat Apple “was aware of all prior art raised prior to Octo-\nber 31, 2012 in the present reexamination proceeding with\nrespect to claims 1–35 [of the ’504 patent], and could have\nraised issues with respect to that prior art in the litigation.”\nJ.A. 1666, 3147 (same regarding claims 36–59 of the ’211\npatent). No new prior art was added after October 31,\n2012, when trial began in the litigation. Indeed, Apple\nnever amended or supplemented its original October 18,\n2011 reexamination requests to add new prior art.\n 15 The parties filed supplemental briefing regarding\nthe relationship between this case and other recently de-\ncided matters. ECF Nos. 95, 96, 105. Apple primarily ar-\ngues that collateral estoppel renders various claims of the\n\nVIRNETX INC. v. APPLE INC. 25\n\n\n\n VirnetX contends that the Board improperly construed\nthe term “indication” by ignoring its disclaimer argument.\nVirnetX argued before the Board that the patent dis-\nclaimed “conventional” DNS functions—e.g., looking up\nand returning an IP address or a public key for the re-\nquested webpage. In VirnetX’s view, such conventional\nfunctions therefore do not qualify as “indications” that the\nsystem “supports establishing a secure communication\nlink.” J.A. 1348, 2892, 4334–41. But the Board expressly\nacknowledged VirnetX’s assertion that the patents “dispar-\naged and disclaimed” systems that use no more than “con-\nventional” DNS functions. J.A. 26–30, 49–54, 77–80. It\nthen rejected this argument in a systematic and well-rea-\nsoned fashion. “Disavowal requires ‘expressions of mani-\nfest exclusion or restriction, representing a clear disavowal\nof claim scope.’” Epistar Corp. v. ITC, 566 F.3d 1321, 1335\n(Fed. Cir. 2009) (quoting Teleflex, Inc. v. Ficosa N. Am.\nCorp., 299 F.3d 1313, 1325 (Fed. Cir. 2002)). We agree with\nthe Board that the specification fails to evidence such a\nclear disavowal here. The Board’s construction is therefore\nconsistent with the broadest reasonable interpretation of\n“indication.”\n Turning to VirnetX’s next challenge, VirnetX contends\nthat the Board failed to make adequate findings and im-\npermissibly shifted the burden to VirnetX with respect to\nthree claim limitations of the ’504 and ’211 patents in the\nApple reexams. The limitations are “indicating,” “connect-\nable,” and “authenticate.” Contrary to VirnetX’s\n\n\n\n’504 and ’211 patents invalid. As an initial matter, Apple’s\narguments are mooted in part by our holding in the present\nmatter affirming the Board’s decision that certain claims\nof the ’504 and ’211 patent are unpatentable. Otherwise,\nApple’s arguments are unpersuasive. Apple has failed to\nshow why collateral estoppel is appropriate under these\ncircumstances.\n\n26 VIRNETX INC. v. APPLE INC.\n\n\n\n\narguments, the Board’s conclusions are well supported by\nthe underlying factual findings made by the Examiner.\nThe Examiner found that Apple had carried its burden of\nproving unpatentability by “a preponderance of the evi-\ndence.” See Rambus Inc. v. Rea, 731 F.3d 1248, 1255 (Fed.\nCir. 2013) (quoting Ethicon, Inc. v. Quigg, 849 F.2d 1422,\n1427 (Fed. Cir. 1988)). Sitting as an appellate body, the\nBoard then reviewed those findings and considered\nwhether VirnetX had raised any issue on appeal that re-\nquired setting aside the Examiner’s decisions. Although\n“the examiner retains the burden to show invalidity,” on\nappeal “the applicant must identify to the Board what the\nexaminer did wrong.” In re Jung, 637 F.3d 1356, 1365–66\n(Fed. Cir. 2011). The Board considered VirnetX’s argu-\nments challenging the Examiner’s rejections and properly\nexplained why each lacked merit. VirnetX “failed to offer\nany convincing reason to find otherwise.” In re Cree, Inc.,\n818 F.3d 694, 702 n.3 (Fed. Cir. 2016). This “is not burden-\nshifting.” Id.\n VirnetX further argues that the Board’s findings in the\nCisco reexam regarding the “indicating” limitations intro-\nduced improper new arguments. We disagree. According\nto VirnetX, the Board relied on a new ground for concluding\nthat Lendenmann’s disclosure of “returning a network ad-\ndress” and “access control lists” (“ACLs”) meet the claim\nlimitation. The Board’s findings constitute a new ground\nof rejection supporting reopening of prosecution if the\nBoard “finds facts not found by the examiner regarding the\ndifferences between the prior art and the claimed inven-\ntion, and these facts are the principal evidence upon which\nthe Board’s rejection was based.” In re Leithem, 661 F.3d\n1316, 1320 (Fed. Cir. 2011). “[T]he ultimate criterion of\nwhether a rejection is considered ‘new’ in a decision by the\n[B]oard is whether [applicants] have had fair opportunity\nto react to the thrust of the rejection.” In re Adler, 723 F.3d\n1322, 1327 (Fed. Cir. 2013) (modifications in original). In\n\nVIRNETX INC. v. APPLE INC. 27\n\n\n\nlight of the proceedings below, we conclude that VirnetX\nhad a fair opportunity to react to the evidence.\n VirnetX also contends that the Board’s findings that\nLendenmann teaches the “indicating” element lack sub-\nstantial evidence. However, the Board found Lenden-\nmann’s disclosure of “returning a network address,”\n“ACLs,” and certain domain names teach the limitation.\nBased on the record here, there is substantial evidence to\nsupport the Board’s conclusion that this limitation is\ntaught.\n\n As for the “authenticating” limitation, VirnetX contests\nthe Board’s determination that Provino discloses a DNS\nsystem configured “to authenticate the query,” as recited in\ndependent claims 5, 23, and 47. VirnetX’s arguments do\nnot unsettle the Board’s conclusions that Provino’s sys-\ntem—including its “query” and “firewall 30”—teach this\nlimitation. Furthermore, contrary to VirnetX’s view, no\nmaterial aspect of the Board’s analysis of the relationship\nbetween authorization and “authentication” was new or\nimproper. Based on the record developed before the Exam-\niner and reviewed by the Board, substantial evidence sup-\nports the Board’s conclusions regarding the\n“authenticating” limitation.\n\n Next, VirnetX argues that the Board failed to address\narguments it made in the Cisco reexam as to why depend-\nent claim 47 of the ’211 patent is patentable over Lenden-\nmann. There is little indication that VirnetX properly\nargued claim 47’s patentability separately from independ-\nent claims 36 and 60. The Board was therefore justified in\nconcluding “Patent Owner does not provide additional ar-\nguments in support of” claim 47. J.A. 88. Regardless, the\nrecord confirms that based on the Examiner’s findings, the\nBoard properly concluded claim 47 was unpatentable.\n\n Finally, VirnetX also argues that the reexams here are\nunconstitutional. VirnetX took the position that its\n\n28 VIRNETX INC. v. APPLE INC.\n\n\n\n\nchallenge would be resolved by the Supreme Court’s deci-\nsion regarding the constitutionality of inter partes review\n(“IPR”) in Oil States Energy Services, LLC v. Greene’s En-\nergy Group, LLC, 138 S. Ct. 1365 (2018). See VirnetX’s Br.\n66–67 (citing arguments in Oil States petition for certiorari\nthat IPRs are unconstitutional under the Seventh Amend-\nment and Article III). While reexams and IPRs arise under\ndifferent statutes, VirnetX’s argument treats them as in-\nterchangeable. The Supreme Court has rejected the\ngrounds advanced in Oil States. 138 S. Ct. at 1379. Vir-\nnetX’s challenge is therefore moot.\n We have considered VirnetX’s remaining arguments\nand find them unpersuasive.\n CONCLUSION\n For the foregoing reasons, we vacate the Board’s deci-\nsions in the Apple reexams with respect to claims 1–35 of\nthe ’504 patent and claims 36–59 of the ’211 patent and re-\nmand with instructions to terminate. However, we affirm\nthe Board’s decisions in the Apple reexams that claims 36–\n60 of the ’504 patent and claims 1–35 and 60 of the ’211\npatent are unpatentable. We further affirm the Board’s de-\ncision in the Cisco reexam that claims 36–60 of the ’211 pa-\ntent are unpatentable.\n AFFIRMED-IN-PART, VACATED-IN-PART, AND\n REMANDED\n COSTS\n The parties shall bear their own costs.\n\n United States Court of Appeals\n for the Federal Circuit\n ______________________\n\n VIRNETX INC.,\n Appellant\n\n v.\n\n APPLE INC., CISCO SYSTEMS, INC.,\n Appellees\n ______________________\n\n 2017-1591, 2017-1592, 2017-1593\n ______________________\n\n Appeals from the United States Patent and Trademark\nOffice, Patent Trial and Appeal Board in Nos. 95/001,788,\n95/001,789, 95/001,856.\n ______________________\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: VIRNETX INC., Appellant v. APPLE INC., Cisco Systems, Inc., Appellees\\nCourt: U.S. Court of Appeals for the Federal Circuit\\nDate decided: 2019-08-01\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: Yes\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n VIRNETX INC.,\\n Appellant\\n\\n v.\\n\\n APPLE INC., CISCO SYSTEMS, INC.,\\n Appellees\\n ______________________\\n\\n 2017-1591, 2017-1592, 2017-1593\\n ______________________\\n\\n Appeals from the United States Patent and Trademark\\nOffice, Patent Trial and Appeal Board in Nos. 95/001,788,\\n95/001,789, 95/001,856.\\n ______________________\\n\\n Decided: August 1, 2019\\n ______________________\\n\\n STEPHEN BLAKE KINNAIRD, Paul Hastings LLP, Wash-\\nington, DC, argued for appellant. Also represented by\\nNAVEEN MODI, JOSEPH PALYS, IGOR VICTOR TIMOFEYEV,\\nMICHAEL WOLFE, DANIEL ZEILBERGER.\\n\\n WILLIAM F. LEE, Wilmer Cutler Pickering Hale and\\nDorr LLP, Boston, MA, argued for appellee Apple Inc. Also\\nrepresented by REBECCA A. BACT, MARK CHRISTOPHER\\nFLEMING, LAUREN B. FLETCHER, DOMINIC E. MASSA;\\nBRITTANY BLUEITT AMADI, Washington, DC; SCOTT\\nBORDER, JEFFREY PAUL KUSHAN, Sidley Austin LLP, Wash-\\nington, DC.\\n\\n2 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n THEODORE M. FOSTER, Haynes & Boone, LLP, Dallas,\\nTX, argued for appellee Cisco Systems, Inc. Also repre-\\nsented by DAVID L. MCCOMBS, ANDREW S. EHMKE, DEBRA\\nJANECE MCCOMAS.\\n ______________________\\n\\n Before PROST, Chief Judge, MOORE and REYNA,\\n Circuit Judges.\\n Opinion for the court filed by Chief Judge PROST.\\nOpinion concurring in part and dissenting in part filed by\\n Circuit Judge REYNA.\\nPROST, Chief Judge.\\n Appellant VirnetX Inc. (“VirnetX”) appeals from deci-\\nsions of the Patent Trial and Appeal Board (“Board”) re-\\nlated to three inter partes reexaminations maintained by\\nApple Inc. (“Apple”) and Cisco Systems, Inc. (“Cisco”). The\\nUnited States Patent and Trademark Office (“PTO”) con-\\ncluded that Apple was not barred from maintaining its\\nreexams by the estoppel provision of the pre-America In-\\nvents Act (“AIA”) version of 35 U.S.C. § 317(b) (2006). The\\nBoard affirmed the Examiner’s determination that the\\nclaims of U.S. Patent Nos. 7,418,504 (“the ’504 patent”) and\\n7,921,211 (“the ’211 patent”) are unpatentable as antici-\\npated or obvious over the prior art of record. For the rea-\\nsons below, we affirm-in-part, vacate-in-part, and remand.\\n BACKGROUND\\n I\\n The ’504 and ’211 patents describe systems and meth-\\nods for “establishing a secure communication link between\\na first computer and a second computer over a computer\\nnetwork, such as the Internet.” ’211 patent col. 6 ll. 36–39.\\nThese systems and methods are “built on top of the existing\\nInternet protocol (IP).” Id. at col. 6 ll. 17–20.\\n\\nVIRNETX INC. v. APPLE INC. 3\\n\\n\\n\\n The Internet uses addressing systems for sending data.\\nIn such systems, physical computers can be identified by a\\nunique IP address (e.g., 123.345.6.7). VirnetX Inc. v. Apple\\nInc., 665 F. App’x 880, 882 (Fed. Cir. 2016).\\n Each IP address corresponds to a domain name (e.g.,\\nwww.Yahoo.com). See ’211 patent col. 38 ll. 58–61, col. 39\\nll. 13–14. A user on one computer can enter a domain name\\nin a web browser to communicate with another computer\\nor server. When the user does so, the computer sends a\\ndomain name service (“DNS”) request to the domain name\\nserver for the IP address corresponding to a given domain\\nname. Id. at col. 38 l. 58–col. 39 l. 3. The domain name\\nserver then looks up the IP address of the requested do-\\nmain name and returns it to the requesting computer. Id.\\nat col. 39 ll. 3–7.\\n Both VirnetX patents claim systems, methods, and me-\\ndia for creating secure communication links via DNS sys-\\ntems. For example, claim 1 of the ’211 patent recites:\\n 1. A system for providing a domain name service\\n for establishing a secure communication link, the\\n system comprising:\\n a domain name service system configured and ar-\\n ranged to\\n [1] be connected to a communication network,\\n [2] store a plurality of domain names and corre-\\n sponding network addresses,\\n [3] receive a query for a network address, and\\n [4] indicate in response to the query whether the\\n domain name service system supports establishing\\n a secure communication link.\\n\\n4 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n Independent claims 36 and 60 of the ’211 patent are\\ndirected to a “machine-readable medium” and a “method,”\\nrespectively. Otherwise, they mirror the requirements of\\nclaim 1. Independent claims 1, 36, and 60 of the ’504 pa-\\ntent are similar to the corresponding independent claims of\\nthe ’211 patent.\\n II\\n In 2010, VirnetX sued Apple in district court. VirnetX\\nalleged infringement of four patents, including the ’504 and\\n’211 patents. 1 VirnetX asserted claims 1, 2, 5, 16, 21, and\\n27 of the ’504 patent and claims 36, 37, 47, and 51 of the\\n’211 patent. VirnetX Inc. v. Apple Inc., 925 F. Supp. 2d 816,\\n824–25 (E.D. Tex. 2013).\\n In October 2011, Apple filed requests for inter partes\\nreexamination of the ’504 and ’211 patents with the PTO.\\nIn Apple’s Reexam Nos. 95/001,788 (“788 case”) and\\n95/001,789 (“789 case”) (collectively, “Apple reexams”), Ap-\\nple challenged all claims as anticipated by the Provino ref-\\nerence or rendered obvious by Provino in view of other prior\\nart. 2\\n The district court action proceeded to trial in late 2012.\\nA jury found the asserted claims infringed and not invalid.\\nThe jury awarded VirnetX $368 million in damages. Vir-\\nnetX, 925 F. Supp. 2d at 825. The district court denied Ap-\\nple’s motion for judgment as a matter of law (“JMOL”) or a\\nnew trial on these issues. Apple appealed.\\n On appeal, we affirmed the jury’s finding of no invalid-\\nity for all four patents. VirnetX, Inc. v. Cisco Sys., Inc., 767\\nF.3d 1308, 1323–24 (Fed. Cir. 2014) (“VirnetX I”). We also\\n\\n\\n\\n 1 VirnetX also alleged that Apple infringed certain\\nclaims of U.S. Patent Nos. 6,502,135 (“the ’135 patent”) and\\n7,490,151 (“the ’151 patent”).\\n 2 U.S. Patent No. 6,557,037 (“Provino”).\\n\\nVIRNETX INC. v. APPLE INC. 5\\n\\n\\n\\naffirmed the jury’s finding of infringement for many of the\\nclaims of the two patents not related to the present appeal\\n(’135 and ’151 patents). 3 Id. at 1320–22. We reversed the\\ndistrict court’s construction of the “secure communication\\nlink” claim term, vacated the related infringement finding\\nfor the two patents in this appeal (’504 and ’211 patents),\\nand vacated the damages award. Id. at 1317–19, 1319,\\n1323–24, 1325–34. We then remanded for further proceed-\\nings. 4 Id. at 1334.\\n Apple did not file a request for rehearing on the inva-\\nlidity or infringement issues affirmed in VirnetX I. Our\\nmandate issued on December 23, 2014. Apple did not seek\\nSupreme Court review. The 90-day period to file a petition\\nfor a writ of certiorari expired.\\n Meanwhile, in the parallel PTO reexamination pro-\\nceedings, the Examiner had found all claims of the ’504 and\\n’211 patents unpatentable. The Examiner issued Right of\\nAppeal Notices (“RANs”) in May 2014. VirnetX appealed\\nthe Examiner’s decisions to the Board.\\n VirnetX also petitioned the PTO to terminate the Apple\\nreexams based on the estoppel provision of § 317(b). The\\nPTO denied VirnetX’s petition in June 2015. J.A. 1659–67,\\n3138–48.\\n In September 2016, the Board affirmed the Examiner’s\\nfindings that all claims of the ’504 and ’211 patents were\\n\\n\\n 3 We reversed the district court’s conclusion related\\nto the doctrine of equivalents for one claim of the ’151 pa-\\ntent. See id. at 1322–23.\\n 4 On remand in 2016, a jury found that Apple in-\\nfringed claims 1, 2, 5, and 27 of the ’504 patent and claims\\n36, 47, and 51 of the ’211 patent, and awarded damages.\\nVirnetX Inc. v. Apple Inc., 324 F. Supp. 3d 836 (E.D. Tex.\\n2017). We recently affirmed. See VirnetX Inc. v. Cisco Sys.,\\nInc., No. 2018-1197, 748 F. App’x 332 (Fed. Cir. 2019).\\n\\n6 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nunpatentable. The Board denied VirnetX’s requests for re-\\nhearing.\\n After appealing the Board’s decisions in the Apple reex-\\nams to this court, VirnetX moved to remand. It argued that\\nthis court’s 2014 opinion was a “final decision” on Apple’s\\nattempt to prove invalidity under § 317(b). ECF No. 27. In\\nJune 2017, a motions panel denied the motion and directed\\nthe parties to address the issue in their merits briefing.\\nECF No. 36.\\n III\\n In December 2011, Cisco also filed a request for inter\\npartes reexamination with the PTO. 5 Cisco’s Reexam No.\\n95/001,856 (“856 case” or “Cisco reexam”) challenged\\nclaims 1–60 of the ’211 patent based on multiple grounds\\nof invalidity. Cisco based several of its arguments on the\\nLendenmann reference. 6\\n After extensive proceedings, the Examiner issued a\\nRAN in January 2015. The RAN rejected claims 36–60 of\\nthe ’211 patent as either anticipated or obvious. VirnetX\\nappealed to the Board. In September 2017, the Board af-\\nfirmed the Examiner, finding claims 36–54 and 57–60 an-\\nticipated by Lendenmann and claims 55 and 56 obvious\\nover Lendenmann and another reference. J.A. 88. VirnetX\\nmoved for rehearing, which was denied. VirnetX then ap-\\npealed.\\n We have jurisdiction under 28 U.S.C. § 1295(a)(4)(A).\\n\\n\\n\\n\\n 5 In 2010, VirnetX also sued Cisco in district court,\\nalleging inter alia infringement of claim 1 of the ’211 pa-\\ntent.\\n 6 Rolf Lendenmann, Understanding OSF DCE 1.1\\nfor AIX and OS/2, IBM International Technical Support\\nOrganization, pp. 1–45 (Oct. 1995) (“Lendenmann”).\\n\\nVIRNETX INC. v. APPLE INC. 7\\n\\n\\n\\n DISCUSSION\\n I\\n VirnetX’s appeal proceeds in two parts. First, VirnetX\\nargues as a threshold matter that Apple was estopped from\\nmaintaining its reexams under the pre-AIA version of\\n35 U.S.C. § 317(b) (2006). Second, VirnetX challenges the\\nmerits of the Board’s conclusions in both the Apple and\\nCisco reexams that the ’504 and ’211 patents are invalid.\\nWe address each in turn.\\n II\\n VirnetX contends that Apple’s reexams were barred by\\n§ 317(b). The PTO refused to terminate Apple’s reexams\\nbased on the conclusion that the provision did not apply.\\nIn VirnetX’s view, the PTO’s decision was inconsistent with\\ncontrolling case law, the statutory text, and Congress’s in-\\ntent. We agree.\\n The applicability of § 317(b) is a question of statutory\\ninterpretation. See Bettcher Indus., Inc. v. Bunzl USA,\\nInc., 661 F.3d 629, 639, 642–48 (Fed. Cir. 2011). Statutory\\ninterpretation is an issue of law we review de novo. Un-\\nwired Planet, LLC v. Google Inc., 841 F.3d 1376, 1379 (Fed.\\nCir. 2016).\\n The Patent Act requires that the PTO terminate a reex-\\namination once there has been a final decision on the pa-\\ntent challenger’s invalidity case in federal court. 35 U.S.C.\\n§ 317(b) (2006). Specifically, pre-AIA § 317(b) provides:\\n Once a final decision has been entered against a\\n party in a civil action arising in whole or in part\\n under section 1338 of title 28, that the party has not\\n sustained its burden of proving the invalidity of any\\n patent claim in suit . . . then neither that party nor\\n its privies may thereafter request an inter partes\\n reexamination of any such patent claim on the ba-\\n sis of issues which that party or its privies raised\\n\\n8 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n or could have raised in such civil action or inter\\n partes reexamination proceeding, and an inter\\n partes reexamination requested by that party or its\\n privies on the basis of such issues may not thereaf-\\n ter be maintained by the Office, notwithstanding\\n any other provision of this chapter.\\n The issue here is very narrow. Apple does not dispute\\nthe underlying procedural facts relevant to § 317(b), which\\nhighlight the advanced stage of these proceedings.\\n In 2010, Apple was sued for patent infringement. There\\nis no dispute that in that “civil action arising . . . under”\\n28 U.S.C. § 1338, Apple raised an affirmative defense of in-\\nvalidity on multiple grounds. 35 U.S.C. § 317(b). Thus,\\nApple had the “burden of proving the invalidity” of the as-\\nserted claims of both the ’504 and ’211 patents. Id.\\n In 2012, Apple presented its invalidity defenses to a\\njury. After a week-long trial, the jury concluded Apple\\nfailed to sustain its burden of proving the invalidity of any\\nasserted claim of the ’504 or ’211 patent.\\n In 2013, the district court determined that in light of\\nthe record at trial, substantial evidence supported the\\njury’s finding. The district court entered final judgment.\\n In 2014, Apple pursued an appeal of that decision here.\\nSpecifically, Apple argued that “the asserted claims are an-\\nticipated by the Kiuchi reference.” VirnetX I, 767 F.3d at\\n1323. After considering Apple’s challenges, we affirmed\\nthe district court’s decision on the issue of invalidity. Id.\\nat 1323–24. Apple did not file a request for rehearing. Ap-\\nple elected not to seek Supreme Court review. The 90-day\\nperiod to file a petition for certiorari expired.\\n In 2015, VirnetX petitioned the PTO to terminate Ap-\\nple’s pending reexaminations based on the estoppel provi-\\nsion of § 317(b). The PTO refused, accepting Apple’s\\nargument that there was not yet a final decision on valid-\\nity. J.A. 1661–63.\\n\\nVIRNETX INC. v. APPLE INC. 9\\n\\n\\n\\n On appeal, the sole issue is whether there has been a\\n“final decision” entered against Apple that it “has not sus-\\ntained its burden of proving” invalidity, triggering estoppel\\nof its parallel PTO reexam proceedings.\\n Apple does not dispute that since it first raised the is-\\nsue in 2010, it has now lost on invalidity before the jury,\\nthe district court, and this court. Nor does it dispute that\\nit allowed the time to petition for certiorari of the validity\\nissues in the first appeal to lapse. Instead, Apple takes the\\nposition that there was no final decision due to a hypothet-\\nical future appeal.\\n Apple’s argument proceeds as follows. In VirnetX I, we\\nvacated and remanded other issues unrelated to invalidity,\\nincluding infringement and damages. In Apple’s view, af-\\nter those issues are decided on remand and reviewed in a\\nsecond appeal, it may choose to file a petition for certiorari\\nof those non-validity issues. In so doing, it may also ask\\nthe Supreme Court to look at any issues from the first ap-\\npeal—including the validity determination.\\n Thus, the Court might elect to examine the validity is-\\nsues at some point in the future. The Court has the ability\\nto consider not only those issues “‘before the court of ap-\\npeals upon the second appeal,’” but also the “questions\\nraised on the first appeal.” Mercer v. Theriot, 377 U.S. 152,\\n153–54 (1964) (per curiam) (quoting Hamilton-Brown Shoe\\nCo. v. Wolf Bros. & Co., 240 U.S. 251, 257 (1916)); see also\\nMajor League Baseball Players Ass’n v. Garvey, 532 U.S.\\n504, 508 n.1 (2001) (per curiam) (noting Supreme Court\\nhas “authority to consider questions determined in earlier\\nstages of the litigation where certiorari is sought from the\\nmost recent of the judgments of the Court of Appeals”). Ac-\\ncording to Apple, the decision is only final when there is no\\npotential that the Supreme Court might some day examine\\nthe invalidity issue during a second appeal.\\n\\n10 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n In response, VirnetX contends there is a “final deci-\\nsion” on a party’s attempt to prove invalidity after the\\nparty fails to petition for certiorari within the 90-day pe-\\nriod.\\n Two reasons compel the conclusion that there has been\\na final decision on validity in this case. First, our decision\\nin Fairchild (Taiwan) Corp. v. Power Integrations, Inc., 854\\nF.3d 1364 (Fed. Cir. 2017), controls and holds that there is\\na “final decision” under § 317(b) once the time to file a cer-\\ntiorari petition has passed. Second, even if Fairchild did\\nnot foreclose Apple’s argument, the plain language of the\\nstatute and its purpose advise against Apple’s definition of\\na “final decision” on the issue of invalidity.\\n A\\n Since the PTO’s decision in 2016, we have ruled on this\\nvery issue. Fairchild directly addressed a situation in\\nwhich validity was affirmed but other issues in the case\\nwere remanded. Fairchild holds that if a finding of no in-\\nvalidity has been affirmed on appeal and remand of other\\nissues will not “have any affect” on validity, the decision is\\n“final” once the 90-day deadline for a certiorari petition on\\nthe validity determination has passed. Id. at 1366.\\nFairchild announced this holding despite the fact that\\nthere is always some “potential” for future Supreme Court\\nreview if the entire case is taken up. Accordingly, Fairchild\\nrequires we reject Apple’s position.\\n In Fairchild, we began our analysis by restating our\\nearlier holding in Bettcher that estoppel “applies when ‘all\\nappeals have terminated.’” Id. at 1365 (quoting Bettcher,\\n661 F.3d at 646). We then concluded that is “precisely the\\nsituation here.” Id. Namely, the district court “entered\\njudgment against Power Integrations, holding that it failed\\nto prove claims 6, 7, 18, and 19 were obvious over [the cited\\nprior art]. This court affirmed the holding, and the time to\\n\\nVIRNETX INC. v. APPLE INC. 11\\n\\n\\n\\npetition for a writ of certiorari has passed.” Id. at 1365–66\\n(emphasis added).\\n As in Fairchild, the district court here entered judg-\\nment against Apple on the issue of invalidity, we affirmed\\nthe no-invalidity finding on appeal in VirnetX I, and the\\ntime to file a petition for certiorari on the issue has long\\nsince passed. Under Fairchild, that constitutes “a final de-\\ncision” under § 317(b) with respect to Apple’s failure to\\nprove invalidity.\\n Resisting this conclusion, Apple argues that if any non-\\ninvalidity issues remain in the district court case related to\\nthe patent, then the validity decision cannot be final under\\n§ 317(b). Since VirnetX I vacated and remanded other is-\\nsues (i.e., infringement and damages) on both patents at\\nissue, Apple contends that it waited until all the issues\\nwith this patent case are decided before bundling it up in a\\npetition for certiorari. Consequently, Apple’s position re-\\nquires resolution of the entire case—or at least all issues on\\neach patent, including infringement, invalidity, and dam-\\nages—before a validity decision is considered “final.”\\n Fairchild rejected such arguments under almost iden-\\ntical circumstances. There, we had also vacated and re-\\nmanded other parts of the case for additional proceedings.\\nId. at 1365 (explaining prior appeal upheld obviousness but\\nreversed jury’s findings on infringement, and “remanded\\nfor further proceedings unrelated to the ’972 patent\\nclaims”). In turn, we rejected the argument that this ren-\\nders the decision not “final” under § 317(b). Id. at 1366.\\nWe reasoned:\\n While it is true that in [the prior appeal] this court\\n vacated and remanded for additional proceedings,\\n we cannot agree with Power Integrations that this\\n renders the decision not “final” for § 317(b) pur-\\n poses. Critically, those proceedings are unrelated\\n to the ’972 patent. By its terms, § 317(b) is\\n\\n12 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n concerned with a final decision ‘that the party has\\n not sustained its burden of proving the invalidity\\n of any patent claim.’ And here, Power Integrations\\n does not suggest, nor is there any reason to believe,\\n that any unresolved issue on remand would have\\n any effect on the now-final ’972 patent validity de-\\n terminations.\\nId. (emphasis added) (quoting 35 U.S.C. § 317(b)).\\n Likewise, here, Apple does not point to any issue in the\\nVirnetX I remand that substantively bears on the finding\\nof no invalidity. Nor could it. Indeed, the only difference\\nhere is that VirnetX I remanded some infringement and\\ndamages issues related to the ’504 and ’211 patents. But\\nApple fails to explain how any of these unresolved issues\\nwould “have any effect on the now-final . . . validity deter-\\nminations” involving those patents. 7 Id. Accordingly,\\nFairchild’s reasoning leaves no doubt that § 317(b) estop-\\npel applies.\\n Finally, Apple contends Fairchild cannot control be-\\ncause no party in that case expressly argued that the Su-\\npreme Court might still take up the invalidity issue on a\\nfuture appeal under Mercer. But Fairchild’s reasoning im-\\nplicitly forecloses such an argument. Fairchild concluded\\nthat “all appeals” had terminated, 854 F.3d at 1365,\\n\\n\\n\\n 7 At oral argument, Apple could not point to a single\\nexample illustrating how an aspect of the infringement re-\\nmand might meaningfully affect or implicate the invalidity\\nruling this court affirmed in 2014 in VirnetX I. Instead,\\nApple offered only speculation. Its argument boiled down\\nto the general conjecture that the district court “could”\\nhave a motion to set aside a validity determination even\\nafter the mandate issues based on arguments made about\\ninfringement. Oral Argument at 29:20–51, No. 2017-1591,\\nhttp://www.cafc.uscourts. gov/oral-argument-recordings.\\n\\nVIRNETX INC. v. APPLE INC. 13\\n\\n\\n\\ndespite the fact that other issues in that case remained\\npending for potential appeal and petition for certiorari. Ap-\\nple can disagree with that conclusion, but that is not a ba-\\nsis for ignoring binding authority. We are bound to follow\\nthe decision of a prior panel, and its reasoning does not\\nleave the question open.\\n In sum, § 317(b) applies here despite the fact that is-\\nsues unrelated to invalidity were remanded.\\n The dissent takes the view that Fairchild is inapplica-\\nble. But it relies on reasons that were not material to the\\noutcome in Fairchild. In addition, the dissent’s attempt to\\ndistinguish Fairchild suffers from several additional infir-\\nmities.\\n According to the dissent, in Fairchild “Power Integra-\\ntions did not appeal the court’s favorable ruling on infringe-\\nment because it prevailed on that issue.” Opinion\\nconcurring in part and dissenting in part (“dissent”) at 4\\n(emphasis added). In turn, the dissent leaps to the conclu-\\nsion that Power Integrations could not have realistically\\npursued certiorari review of the ’972 patent’s validity be-\\ncause it “prevailed” on infringement. Id. In the dissent’s\\nview, “once Power Integrations prevailed on infringement,\\nthe case against it was concluded” and the potential for any\\nfurther Supreme Court review was eliminated. Id. at 4–5.\\nThere is no support for these statements.\\n Fundamentally, infringement and invalidity are sepa-\\nrate “issues.” Pandrol USA, LP v. Airboss Ry. Prod., Inc.,\\n320 F.3d 1354, 1364 (Fed. Cir. 2003) (“Supreme Court prec-\\nedent and our cases make clear that patent infringement\\nand patent validity are treated as separate issues.”).\\nFairchild never held that resolving infringement resolves\\ninvalidity. Indeed, Fairchild was interpreting an estoppel\\nstatute triggered by a decision on invalidity, not infringe-\\nment.\\n\\n14 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n Moreover, the dissent’s position unravels in light of the\\nprocedural posture in Fairchild. Power Integrations did\\nnot simply argue that it did not infringe the ’972 patent.\\nNor did it merely raise invalidity as an affirmative defense\\nto liability. Rather, Power Integrations advanced a sepa-\\nrate counterclaim for declaratory judgment that the ’972\\npatent was invalid. See Plaintiff’s Fourth Amended An-\\nswer and Counterclaims at 23, Power Integrations, Inc. v.\\nFairchild Semiconductor Int’l, Inc. et al, No. 1:08–cv–\\n00309–LPS (D. Del. July 7, 2011) (Dkt. No. 409). There-\\nfore, to say it “prevailed” and its case “concluded” on that\\nissue due to a non-infringement ruling is not correct.\\n It is black letter law that non-infringement does not\\nmoot a counterclaim regarding invalidity. “A party seeking\\na declaratory judgment of invalidity presents a claim inde-\\npendent of the patentee’s charge of infringement.” Cardi-\\nnal Chem. Co. v. Morton Int’l, 508 U.S. 83, 96 (1993). Thus,\\n“appellate affirmance of a judgment of noninfringement,\\neliminating any apprehension of suit, does not moot a de-\\nclaratory judgment counterclaim of patent invalidity.”\\nMedImmune, Inc. v. Genentech, Inc., 549 U.S. 118, 132 n.11\\n(2007); see also Telcordia Techs., Inc. v. Cisco Sys., Inc., 612\\nF.3d 1365, 1375 (Fed. Cir. 2010) (“[T]his court’s affirmance\\nof the district court’s non-infringement findings as to the\\n’306 patent does not moot Cisco’s invalidity counterclaim\\non cross-appeal.”).\\n In addition, the dissent speculates as to why the litigant\\nchose not to seek further review. The dissent appears to\\nassume that if a party is found not liable for certain ac-\\ncused products infringing a patent, as in Fairchild, that\\nparty no longer has the incentive to challenge the underly-\\ning property right’s validity. Dissent at 5–6. This assump-\\ntion fails for several reasons. First, inferring the subjective\\nintent of the litigants—either on remand or in petitioning\\nfor certiorari—is inappropriate. Fairchild never made\\nsuch fact findings. Fairchild never relied on such infer-\\nences either. Put simply, Fairchild did not base its decision\\n\\nVIRNETX INC. v. APPLE INC. 15\\n\\n\\n\\non speculation as to whether remand would impact a\\nparty’s subjective “decision” to pursue certiorari review or\\nwhether its decisions regarding petitioning were “reasona-\\nble,” id. at 5–6. Second, such speculation is irrelevant. The\\ncontrolling inquiry is mandated by statute, which makes\\nno mention of a party’s motives for petitioning for certiorari\\nreview. Third, it is inaccurate. Regardless of a finding of\\nno liability on the presently accused products, an accused\\ninfringer may well have reasons for trying to invalidate the\\npatent (e.g., to guarantee its freedom to operate).\\n But the dissent suffers from a more fundamental flaw.\\nThe dissent appears to suggest that in Fairchild, “potential\\nfuture review” by the Supreme Court was not an option.\\nId. at 4. In light of Mercer, there is no basis for that posi-\\ntion. Indeed, it runs headlong into the dissent’s own insist-\\nence that the Supreme Court’s discretion to take up\\ncertiorari review of earlier issues in a case is broad. On the\\none hand, the dissent acknowledges that a “petition for writ\\nof certiorari can expose the entire case to review,” id. at 10\\n(quoting Christianson v. Colt Indus. Operating Corp., 486\\nU.S. 800, 817 (1988)), and the Supreme Court has “juris-\\ndiction to consider all of the substantial federal questions\\ndetermined in the earlier stages of the litigation,” id. (quot-\\ning Reece v. Georgia, 350 U.S. 85, 87 (1955)). On the other\\nhand, the dissent contends that the litigant’s decision in\\nFairchild “all but foreclosed the possibility that the inva-\\nlidity issue could be subject to potential future review by\\nthe Supreme Court.” Id. at 4. A plain reading of Fairchild,\\nhowever, confirms it never presumed to curtail the Su-\\npreme Court’s discretion under Mercer to review issues\\nraised earlier in a case.\\n Finally, the dissent is concerned that the majority opin-\\nion somehow “intrudes on the Supreme Court’s constitu-\\ntional prerogative whether to take up the invalidity issue.”\\nId. at 11. It does not. Like Fairchild, our decision concerns\\nestoppel of duplicative administrative proceedings before\\nthe PTO. It has no bearing on the Supreme Court’s\\n\\n16 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\ndiscretion to accept or reject a petition for certiorari on any\\nissue—nor could it.\\n In sum, we disagree with the dissent’s efforts to rewrite\\nFairchild. Fairchild speaks for itself. The controlling stat-\\nute is trained on a “final decision” that the party has not\\nsustained its “burden of proving the invalidity” of the pa-\\ntent claim. Fairchild, 843 F.3d at 1366. As the dissent\\nacknowledges, the reasoning at the heart of the Fairchild\\ndecision is that estoppel applied to the duplicative PTO\\nreexam because there was “no suggestion,” dissent at 6,\\nthat “any unresolved issue on remand would have any ef-\\nfect on the now-final patent validity determinations” from\\nthe district court, Fairchild, 854 F.3d at 1366. That is, the\\nremand would have no substantive effect on the validity\\ndetermination. The record is clear that is the same situa-\\ntion here.\\n B\\n Even if Fairchild did not control, the statutory text and\\npurpose of § 317(b) compels us to reject Apple’s theory that\\nthe entire case must be resolved. Apple essentially asks us\\nto rewrite the statute, triggering estoppel only when there\\nis a “final judgment” on all issues in a case.\\n But the statute’s text is clear. The statute only re-\\nquires “a final decision” on a precise issue—whether the\\nparty has “sustained its burden of proving the invalidity of\\nany patent claim in suit.” 35 U.S.C. § 317(b). The statute\\nmakes no mention of resolving the entire case. Nor does it\\nrequire resolution of all other issues related to a patent\\n(e.g., infringement or damages). Instead, the plain lan-\\nguage of the statute demands a party cease duplicative in-\\nvalidity challenges via reexams in the PTO once “a final\\ndecision” has been reached on an invalidity challenge in a\\ncivil action.\\n This reading of the plain text is further reinforced by\\nthe statute’s purpose. The statute is intended to protect\\n\\nVIRNETX INC. v. APPLE INC. 17\\n\\n\\n\\npatent holders against “harassment” from duplicative liti-\\ngation on the issue of invalidity. See 145 Cong. Rec. 29,276\\n(Nov. 9, 1999) (Conference Report on H.R. 1554) (“Subtitle\\nF creates a new section 317 which sets forth certain condi-\\ntions by which inter partes reexamination is prohibited to\\nguard against harassment of a patent holder.”); see also\\nFunction Media, L.L.C. v. Kappos, 508 F. App’x 953, 955–\\n56 (Fed. Cir. 2013) (“This provision existed to prevent du-\\nplicative, harassing actions against a patentee. If a defend-\\nant brought an invalidity challenge in a district court\\nlitigation and was unsuccessful, it is not permitted to bring\\nthe same challenge in an inter partes reexamination.”).\\n Apple concedes that § 317(b) was intended to prevent\\nduplicative litigation. Apple’s Br. 40–41. Yet Apple’s posi-\\ntion increases the likelihood duplicative litigation will oc-\\ncur. If no decision on invalidity is “final” until any\\n“prospect of future Supreme Court review” is eliminated,\\nApple’s Br. 36, estoppel will rarely apply. In light of the\\nlegislative intent, it is doubtful Congress created an estop-\\npel provision that would allow a PTO proceeding to press\\nforward even after all appellate deadlines have elapsed\\nmerely because there is a potential that the Supreme Court\\nmay one day reconsider the invalidity issue—in many\\ncases, years later—while examining the rest of the case. 8\\n Accordingly, the statutory purpose confirms that “a fi-\\nnal decision” triggers estoppel when the invalidity chal-\\nlenge is decided on appeal and the time for petitioning for\\ncertiorari has passed—regardless of the fact review of the\\n\\n\\n\\n 8 Moreover, there is no indication that Congress con-\\nsidered the type of review contemplated in Mercer as part\\nof the definition of a regular appeals process for reaching a\\n“final decision” on invalidity. This is of no surprise given\\nthe statute is trained on the finality of a particular issue,\\nrather than finality of the entire case.\\n\\n18 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nentire case could occur at some point in the future under\\nMercer.\\n C\\n Apple’s remaining arguments against the application\\nof § 317(b) here are unpersuasive. Apple first suggests that\\nFairchild is inconsistent with Bettcher. Apple’s Br. 32–33,\\n37. In Apple’s view, Bettcher held that § 317(b) is not sat-\\nisfied until the Supreme Court can no longer review a case.\\n Bettcher never reached such a holding. Bettcher con-\\ncerned a different estoppel provision, pre-AIA 35 U.S.C.\\n§ 315(c) (2006). 9 The question was whether § 315(c) estop-\\npel attaches at a far earlier stage (before any appeal is pur-\\nsued). In Bettcher, the district court concluded that\\nestoppel barred further civil litigation “as soon as the ex-\\naminer finished the reexamination and the Right of Appeal\\nNotice was sent to the patentee”—i.e., before there is even\\nan appeal to the Board, let alone an appeal to the Federal\\nCircuit. Id. at 636. On appeal, we rejected that conclusion.\\nWe held that “the estoppel provision of § 315 is triggered\\nnot when examination is completed but only after all ap-\\npeal rights have been exhausted.” Id. at 642–43.\\n The panel then rejected the argument that § 317(b)\\nsomehow showed that § 315(c) estoppel must apply as soon\\nas the examiner finishes the reexam. See id. at 643. The\\npanel observed that like § 315(c), estoppel under § 317(b)\\napplies “when reexamination and all appeals have termi-\\nnated.” Id. at 646 (emphasis added).\\n Bettcher addressed appeal rights to the Board and the\\nFederal Circuit. It did not analyze how petitions for\\n\\n\\n 9 Section 315(c) concerns estoppel in the opposite di-\\nrection: when the PTO’s determination in a reexam estops\\na party from asserting invalidity in a civil action. Id. at\\n643.\\n\\nVIRNETX INC. v. APPLE INC. 19\\n\\n\\n\\ncertiorari impact § 317(b). 10 Moreover, Bettcher did not ad-\\ndress the critical inquiry here as to whether a “final deci-\\nsion” regarding the issue of invalidity under § 317(b) is\\naffected by remand of issues unrelated to invalidity.\\n By contrast, Fairchild did reach these issues.\\nFairchild held that estoppel “applies when ‘all appeals’\\nhave terminated,” meaning after the district court’s judg-\\nment of no-invalidity was affirmed on appeal and “the time\\nto petition for a writ of certiorari has passed.” Id. at 1365–\\n66 (quoting Bettcher, 661 F.3d at 646). The panel reached\\nthis conclusion despite the fact that issues unrelated to pa-\\ntent validity were remanded for further proceedings.\\nProperly understood, Fairchild is an extension of Bettcher\\nto Supreme Court certiorari review. Apple’s attempt to\\n\\n\\n 10 Bettcher concluded that the estoppel provision of\\n§ 315 attaches when all appeal rights are exhausted. As\\nthe panel observed, the opening sections of the statute list\\nthose appeal “rights” for both the patent owner and the\\nthird-party requestor, § 315(a)–(b), which consist of ap-\\npeals from (1) the examiner to the Board, and (2) the Board\\nto this court. Id. at 644 (“The estoppel provision of subsec-\\ntion (c) thus falls immediately following the establishment\\nof both parties’ rights of appeal, including appeal to this\\ncourt.”). Bettcher did not address whether petitions for cer-\\ntiorari are considered “appeals,” or whether certiorari re-\\nview by the Supreme Court is a “right.” In VirnetX’s view,\\nan appeal as of right is a different procedural mechanism\\nthan a writ of certiorari. VirnetX Reply Br. 12–13; see also\\nCity & Cty. of San Francisco, Calif. v. Sheehan, 135 S. Ct.\\n1765, 1774 (2015) (“Because certiorari jurisdiction exists to\\nclarify the law, its exercise ‘is not a matter of right, but of\\njudicial discretion.’” (quoting Supreme Court Rule 10)).\\nThough we need not reach such questions, they underscore\\nthe fact that Bettcher simply did not reach the issue of cer-\\ntiorari review now governed by Fairchild.\\n\\n20 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nmanufacture an inconsistency between these two cases has\\nno merit.\\n Apple’s reliance on Fresenius fares no better. Fresenius\\nUSA, Inc. v. Baxter Int’l, Inc., 721 F.3d 1330 (Fed. Cir.\\n2013). Apple concedes that Fresenius “did not construe\\n§ 317(b).” Apple’s Br. 39. As such, Apple admits Fresenius\\nhas no direct bearing on our interpretation of this statutory\\nestoppel provision.\\n In addition, Fresenius concerned an entirely different\\nissue involving res judicata from a “final judgment.” Even\\nwithin the distinct realm of res judicata, Fresenius took\\ngreat care to “distinguish between different concepts of fi-\\nnality.” Id. at 1340. In Fresenius, we specifically addressed\\nwhether a judgment in an infringement case is “sufficiently\\nfinal” so that it is “immune to the effect of the final judg-\\nment in the PTO proceedings, as affirmed by this court in\\nIn re Baxter.” Id. at 1341 (discussing In re Baxter Int’l, Inc.,\\n678 F.3d 1357 (Fed. Cir. 2012)).\\n The issue in Fresenius arose due to the fact that while\\nthe appeal of an infringement action was pending, we af-\\nfirmed the PTO reexamination finding the patent claims\\ninvalid in In re Baxter. As a result, the PTO canceled the\\nclaims. Id. at 1335. Despite the intervening cancellation,\\nthe patent owner argued that the district court’s judgment\\nwas “final” for purposes of res judicata such that the liabil-\\nity finding and past damages award in the pending litiga-\\ntion could not be disturbed. Id. at 1340. We disagreed.\\n While the district court in 2007 had entered “a judg-\\nment final for purposes of appeal,” we concluded that it was\\n“not sufficiently final to preclude application of the inter-\\nvening final judgment in In re Baxter.” Id. at 1341. More-\\nover, the district court’s original judgment had since been\\nset aside and the remand was not sufficiently final for res\\njudicata purposes. Id. “To rise to that level, the litigation\\nmust be entirely concluded so that [the] cause of action\\n[against the infringer] was merged into a final judgment\\n\\nVIRNETX INC. v. APPLE INC. 21\\n\\n\\n\\n. . . one that ‘ends the litigation on the merits and leaves\\nnothing for the court to do but execute the judgment.’” Id.\\n(quoting Mendenhall v. Barber–Greene Co., 26 F.3d 1573,\\n1580 (1994)). We concluded that the remand did not end\\nthe controversy, leaving only the execution of the judg-\\nment, given that several aspects of the litigation (e.g., roy-\\nalty damages, injunctive relief) remained unresolved. Id.\\n Fresenius’s analysis of a “final judgment” for purposes\\nof res judicata does not instruct our understanding of a “fi-\\nnal decision” for purposes of § 317(b) estoppel. Indeed, it\\nonly underscores why we must follow Fairchild. Fairchild\\nfound estoppel absent a final judgment. Indeed, after mul-\\ntiple other issues in the litigation, including infringement\\nand damages for certain patents, were remanded,\\nFairchild concluded nonetheless that there was a “final de-\\ncision” on the narrow issue of invalidity—which is all that\\nthe statute requires to trigger estoppel. The reasoning in\\nFairchild, not Fresenius, governs our application of\\n§ 317(b).\\n Apple also briefly argues that this interpretation of the\\nstatute generates duplicative proceedings for the patent\\nchallenger. Apple’s Br. 41. In Apple’s view, it must seek\\n“interlocutory” review rather than waiting for the case to\\nbe entirely resolved. Id. Apple’s argument misses the\\nmark. Section 317(b) is concerned with preventing dupli-\\ncative litigation against the patent owner. Of course, we\\nrecognize the challenging nature of these proceedings for\\nboth parties. The present actions involve multiple patents\\nand various issues, which create certain complexities for\\npursuing appeals before this court and seeking certiorari\\nbefore the Supreme Court. Some of those complexities\\nmight be ameliorated by requiring final judgment of an en-\\ntire case. But that does not empower us to rewrite a statute\\nthat requires only a final decision on the issue of invalidity.\\nFurthermore, such interlocutory and piecemeal concerns\\nwere also present in Fairchild. Nonetheless, estoppel ap-\\nplied.\\n\\n22 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\n Finally, Apple’s arguments based on the PTO’s inter-\\npretations carry no force. They fail for several reasons.\\n First, Apple fails to explain why the PTO’s interpreta-\\ntions of § 317(b) is entitled to Chevron deference. As dis-\\ncussed above, the statutory text and purpose leave no\\nambiguity. Where “the intent of Congress is clear, that is\\nthe end of the matter.” Chevron U.S.A. Inc. v. NRDC, Inc.,\\n467 U.S. 837, 842 (1984). Therefore, Chevron deference\\ndoes not apply. 11 12\\n Second, Skidmore deference is not appropriate either.\\nUnder Skidmore, the weight given to a particular agency\\ninterpretation “will depend upon the thoroughness evident\\nin its consideration, the validity of its reasoning, its con-\\nsistency with earlier and later pronouncements, and all\\nthose factors which give it power to persuade, if lacking\\n\\n\\n 11 Apple also suggests that Bettcher requires defer-\\nence. Not so. Bettcher concluded the statute was unambig-\\nuous. Bettcher, 661 F.3d at 646 (“[W]e believe that § 317(b)\\nis clear . . . .”). Nonetheless, Apple points to Bettcher’s sub-\\nsequent discussion of the deference it “would” give the\\nMPEP had it reached the issue. Id. Apple’s reliance on a\\nsingle line of dicta from Bettcher is unavailing. Even if it\\nwere not dicta, Bettcher’s statement concerned deference to\\nthe Manual of Patent Examining Procedure (“MPEP”)\\n§ 2686.04. Id. As explained later, that section of the MPEP\\ndoes not resolve the issue at hand.\\n 12\\n Apple also suggests the underlying PTO denial of\\nthe petition to terminate here is entitled to Chevron defer-\\nence. In support of this position, Apple cites only to In re\\nAffinity Labs of Texas, LLC, 856 F.3d 883, 894 n.5 (Fed.\\nCir. 2017). The cited footnote from Affinity Labs did not\\nconcern deference. Rather, it concerned whether PTO rul-\\nings on petitions to terminate are reviewable on appeal. Id.\\n(analyzing stages of PTO decision-making process as re-\\nlated to reviewability).\\n\\nVIRNETX INC. v. APPLE INC. 23\\n\\n\\n\\npower to control.” United States v. Mead Corp., 533 U.S.\\n218, 228 (2001) (quoting Skidmore v. Swift & Co., 323 U.S.\\n134, 140 (1944)). None of the considerations favoring def-\\nerence are present here. For example, contrary to Apple’s\\nposition, the PTO has never established a “consistent” con-\\nstruction of § 317(b). Indeed, aside from the decision below,\\nApple cites no other decision in which the PTO has inter-\\npreted § 317(b) to extend passed the deadline for certiorari\\nreview of the invalidity issue to hypothetical petitions aris-\\ning from appeals of non-invalidity issues on remand. Thus,\\ndeference is unwarranted.\\n Third, regardless of whether deference applies, none of\\nApple’s citations resolve the specific issue at hand. For in-\\nstance, no PTO regulation or MPEP section states that a\\ndecision is not “final” for purposes of § 317(b) if non-inva-\\nlidity issues are remanded. Nor do any of these provisions\\nstate that a final judgment is required, or that as long as\\nthe potential for Supreme Court review from a later peti-\\ntion exists, § 317(b) estoppel does not attach. Indeed, none\\nof these provisions even mention Supreme Court certiorari\\nreview. 13\\n For the reasons above, § 317(b) estoppel applies to the\\nApple reexams. Accordingly, we vacate and remand with\\n\\n\\n 13 Other PTO regulations suggest appeals terminate\\nwhen our mandate issues. See, e.g., 37 C.F.R. § 1.197 (“An\\nappeal to the U.S. Court of Appeals for the Federal Circuit,\\nwhether from a decision of the Board or a judgment in a\\ncivil action, is terminated when the mandate is issued by\\nthe Court.”). Notably, 37 C.F.R. § 1.979, which is the reg-\\nulation interpreting pre-AIA § 316(a) permitting the PTO\\nto issue a certificate of cancellation “only after all appeals\\nhave terminated,” Bettcher, 661 F.3d at 644–45, applies the\\nsame interpretation. See 37 C.F.R. § 1.979(b) (providing\\nthat an “appeal [to this Court] is considered terminated\\nwhen the mandate is issued by the Court”).\\n\\n24 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\ninstructions to terminate the Apple reexams with respect\\nto claims 1–35 of the ’504 patent and claims 36–59 of the\\n’211 patent. 14\\n III\\n\\n We now turn to VirnetX’s challenge to the merits of the\\nApple and Cisco reexams that pertain to the claims not\\nsubject to estoppel. The Board affirmed the Examiner’s re-\\njection of all claims of the ’504 and ’211 patents. On appeal,\\nVirnetX takes a kitchen sink approach, arguing that the\\nBoard decisions incorrectly construe certain claims, shift\\nthe burden to the patent owner to prove patentability, and\\nlack substantial evidence. VirnetX’s arguments are unper-\\nsuasive. Therefore, we affirm the Board’s decision on all\\nremaining claims of both patents in the Apple reexam not\\nsubject to § 317(b) estoppel and fully affirm the Board’s de-\\ncision regarding the claims of the ’211 patent in the Cisco\\nreexam. 15\\n\\n\\n\\n 14 Apple appears not to dispute that estoppel extends\\nto the dependent claims here. Section 317(b) bars main-\\ntaining a reexam based on issues which that party “raised”\\nor “could have raised” in the civil action. The PTO found\\nthat Apple “was aware of all prior art raised prior to Octo-\\nber 31, 2012 in the present reexamination proceeding with\\nrespect to claims 1–35 [of the ’504 patent], and could have\\nraised issues with respect to that prior art in the litigation.”\\nJ.A. 1666, 3147 (same regarding claims 36–59 of the ’211\\npatent). No new prior art was added after October 31,\\n2012, when trial began in the litigation. Indeed, Apple\\nnever amended or supplemented its original October 18,\\n2011 reexamination requests to add new prior art.\\n 15 The parties filed supplemental briefing regarding\\nthe relationship between this case and other recently de-\\ncided matters. ECF Nos. 95, 96, 105. Apple primarily ar-\\ngues that collateral estoppel renders various claims of the\\n\\nVIRNETX INC. v. APPLE INC. 25\\n\\n\\n\\n VirnetX contends that the Board improperly construed\\nthe term “indication” by ignoring its disclaimer argument.\\nVirnetX argued before the Board that the patent dis-\\nclaimed “conventional” DNS functions—e.g., looking up\\nand returning an IP address or a public key for the re-\\nquested webpage. In VirnetX’s view, such conventional\\nfunctions therefore do not qualify as “indications” that the\\nsystem “supports establishing a secure communication\\nlink.” J.A. 1348, 2892, 4334–41. But the Board expressly\\nacknowledged VirnetX’s assertion that the patents “dispar-\\naged and disclaimed” systems that use no more than “con-\\nventional” DNS functions. J.A. 26–30, 49–54, 77–80. It\\nthen rejected this argument in a systematic and well-rea-\\nsoned fashion. “Disavowal requires ‘expressions of mani-\\nfest exclusion or restriction, representing a clear disavowal\\nof claim scope.’” Epistar Corp. v. ITC, 566 F.3d 1321, 1335\\n(Fed. Cir. 2009) (quoting Teleflex, Inc. v. Ficosa N. Am.\\nCorp., 299 F.3d 1313, 1325 (Fed. Cir. 2002)). We agree with\\nthe Board that the specification fails to evidence such a\\nclear disavowal here. The Board’s construction is therefore\\nconsistent with the broadest reasonable interpretation of\\n“indication.”\\n Turning to VirnetX’s next challenge, VirnetX contends\\nthat the Board failed to make adequate findings and im-\\npermissibly shifted the burden to VirnetX with respect to\\nthree claim limitations of the ’504 and ’211 patents in the\\nApple reexams. The limitations are “indicating,” “connect-\\nable,” and “authenticate.” Contrary to VirnetX’s\\n\\n\\n\\n’504 and ’211 patents invalid. As an initial matter, Apple’s\\narguments are mooted in part by our holding in the present\\nmatter affirming the Board’s decision that certain claims\\nof the ’504 and ’211 patent are unpatentable. Otherwise,\\nApple’s arguments are unpersuasive. Apple has failed to\\nshow why collateral estoppel is appropriate under these\\ncircumstances.\\n\\n26 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\narguments, the Board’s conclusions are well supported by\\nthe underlying factual findings made by the Examiner.\\nThe Examiner found that Apple had carried its burden of\\nproving unpatentability by “a preponderance of the evi-\\ndence.” See Rambus Inc. v. Rea, 731 F.3d 1248, 1255 (Fed.\\nCir. 2013) (quoting Ethicon, Inc. v. Quigg, 849 F.2d 1422,\\n1427 (Fed. Cir. 1988)). Sitting as an appellate body, the\\nBoard then reviewed those findings and considered\\nwhether VirnetX had raised any issue on appeal that re-\\nquired setting aside the Examiner’s decisions. Although\\n“the examiner retains the burden to show invalidity,” on\\nappeal “the applicant must identify to the Board what the\\nexaminer did wrong.” In re Jung, 637 F.3d 1356, 1365–66\\n(Fed. Cir. 2011). The Board considered VirnetX’s argu-\\nments challenging the Examiner’s rejections and properly\\nexplained why each lacked merit. VirnetX “failed to offer\\nany convincing reason to find otherwise.” In re Cree, Inc.,\\n818 F.3d 694, 702 n.3 (Fed. Cir. 2016). This “is not burden-\\nshifting.” Id.\\n VirnetX further argues that the Board’s findings in the\\nCisco reexam regarding the “indicating” limitations intro-\\nduced improper new arguments. We disagree. According\\nto VirnetX, the Board relied on a new ground for concluding\\nthat Lendenmann’s disclosure of “returning a network ad-\\ndress” and “access control lists” (“ACLs”) meet the claim\\nlimitation. The Board’s findings constitute a new ground\\nof rejection supporting reopening of prosecution if the\\nBoard “finds facts not found by the examiner regarding the\\ndifferences between the prior art and the claimed inven-\\ntion, and these facts are the principal evidence upon which\\nthe Board’s rejection was based.” In re Leithem, 661 F.3d\\n1316, 1320 (Fed. Cir. 2011). “[T]he ultimate criterion of\\nwhether a rejection is considered ‘new’ in a decision by the\\n[B]oard is whether [applicants] have had fair opportunity\\nto react to the thrust of the rejection.” In re Adler, 723 F.3d\\n1322, 1327 (Fed. Cir. 2013) (modifications in original). In\\n\\nVIRNETX INC. v. APPLE INC. 27\\n\\n\\n\\nlight of the proceedings below, we conclude that VirnetX\\nhad a fair opportunity to react to the evidence.\\n VirnetX also contends that the Board’s findings that\\nLendenmann teaches the “indicating” element lack sub-\\nstantial evidence. However, the Board found Lenden-\\nmann’s disclosure of “returning a network address,”\\n“ACLs,” and certain domain names teach the limitation.\\nBased on the record here, there is substantial evidence to\\nsupport the Board’s conclusion that this limitation is\\ntaught.\\n\\n As for the “authenticating” limitation, VirnetX contests\\nthe Board’s determination that Provino discloses a DNS\\nsystem configured “to authenticate the query,” as recited in\\ndependent claims 5, 23, and 47. VirnetX’s arguments do\\nnot unsettle the Board’s conclusions that Provino’s sys-\\ntem—including its “query” and “firewall 30”—teach this\\nlimitation. Furthermore, contrary to VirnetX’s view, no\\nmaterial aspect of the Board’s analysis of the relationship\\nbetween authorization and “authentication” was new or\\nimproper. Based on the record developed before the Exam-\\niner and reviewed by the Board, substantial evidence sup-\\nports the Board’s conclusions regarding the\\n“authenticating” limitation.\\n\\n Next, VirnetX argues that the Board failed to address\\narguments it made in the Cisco reexam as to why depend-\\nent claim 47 of the ’211 patent is patentable over Lenden-\\nmann. There is little indication that VirnetX properly\\nargued claim 47’s patentability separately from independ-\\nent claims 36 and 60. The Board was therefore justified in\\nconcluding “Patent Owner does not provide additional ar-\\nguments in support of” claim 47. J.A. 88. Regardless, the\\nrecord confirms that based on the Examiner’s findings, the\\nBoard properly concluded claim 47 was unpatentable.\\n\\n Finally, VirnetX also argues that the reexams here are\\nunconstitutional. VirnetX took the position that its\\n\\n28 VIRNETX INC. v. APPLE INC.\\n\\n\\n\\n\\nchallenge would be resolved by the Supreme Court’s deci-\\nsion regarding the constitutionality of inter partes review\\n(“IPR”) in Oil States Energy Services, LLC v. Greene’s En-\\nergy Group, LLC, 138 S. Ct. 1365 (2018). See VirnetX’s Br.\\n66–67 (citing arguments in Oil States petition for certiorari\\nthat IPRs are unconstitutional under the Seventh Amend-\\nment and Article III). While reexams and IPRs arise under\\ndifferent statutes, VirnetX’s argument treats them as in-\\nterchangeable. The Supreme Court has rejected the\\ngrounds advanced in Oil States. 138 S. Ct. at 1379. Vir-\\nnetX’s challenge is therefore moot.\\n We have considered VirnetX’s remaining arguments\\nand find them unpersuasive.\\n CONCLUSION\\n For the foregoing reasons, we vacate the Board’s deci-\\nsions in the Apple reexams with respect to claims 1–35 of\\nthe ’504 patent and claims 36–59 of the ’211 patent and re-\\nmand with instructions to terminate. However, we affirm\\nthe Board’s decisions in the Apple reexams that claims 36–\\n60 of the ’504 patent and claims 1–35 and 60 of the ’211\\npatent are unpatentable. We further affirm the Board’s de-\\ncision in the Cisco reexam that claims 36–60 of the ’211 pa-\\ntent are unpatentable.\\n AFFIRMED-IN-PART, VACATED-IN-PART, AND\\n REMANDED\\n COSTS\\n The parties shall bear their own costs.\\n\\n United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n VIRNETX INC.,\\n Appellant\\n\\n v.\\n\\n APPLE INC., CISCO SYSTEMS, INC.,\\n Appellees\\n ______________________\\n\\n 2017-1591, 2017-1592, 2017-1593\\n ______________________\\n\\n Appeals from the United States Patent and Trademark\\nOffice, Patent Trial and Appeal Board in Nos. 95/001,788,\\n95/001,789, 95/001,856.\\n ______________________\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4191,"cells":{"id":{"kind":"number","value":4644857,"string":"4,644,857"},"name":{"kind":"string","value":"HYLETE LLC, Appellant v. HYBRID ATHLETICS, LLC, Appellee"},"name_abbreviation":{"kind":"string","value":"Hylete LLC v. Hybrid Athletics, LLC"},"decision_date":{"kind":"string","value":"2019-08-01"},"docket_number":{"kind":"string","value":"2017-2057"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Federal Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Moore, Reyna, Wallach\", \"parties\": \"\", \"opinions\": [{\"author\": \"REYNA, Circuit Judge.\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n HYLETE LLC,\\n Appellant\\n\\n v.\\n\\n HYBRID ATHLETICS, LLC,\\n Appellee\\n ______________________\\n\\n 2017-2057\\n ______________________\\n\\n Appeal from the United States Patent and Trademark\\nOffice, Trademark Trial and Appeal Board in No.\\n91213057.\\n ______________________\\n\\n Decided: August 1, 2019\\n ______________________\\n\\n PATTRIC RAWLINS, Procopio, Cory, Hargreaves & Sa-\\nvitch LLP, San Diego, CA, argued for appellant. Also rep-\\nresented by DAVE DEONARINE.\\n\\n MICHAEL JOSEPH KOSMA, Whitmyer IP Group LLC,\\nStamford, CT, argued for appellee. Also represented by\\nBENJAMIN N. LUEHRS.\\n ______________________\\n\\n Before MOORE, REYNA, and WALLACH, Circuit Judges.\\n\\f2 HYLETE LLC v. HYBRID ATHLETICS, LLC\\n\\n\\n\\n\\nREYNA, Circuit Judge.\\n Hylete LLC appeals from a decision of the Trademark\\nTrial and Appeal Board sustaining Hybrid Athletics, LLC’s\\nopposition to Hylete’s trademark registration application.\\nWe conclude that Hylete waived the arguments on which\\nits appeal relies because it raises new issues that could\\nhave been raised and were not considered below. We af-\\nfirm.\\n BACKGROUND\\n In January 2013, Hylete applied to register a design\\nmark for a stylized letter “H” in International Class 25 for\\n“[a]thletic apparel, namely, shirts, pants, shorts, jackets,\\nfootwear, hats and caps.” J.A. 76–92. After finding no reg-\\nistrations that would bar registration of the Hylete mark,\\nthe Examining Attorney approved the application, and the\\nTrademark Office published the Hylete mark for opposition\\nin the Trademark Official Gazette on June 18, 2013.\\n On October 16, 2013, Hybrid Athletics, LLC filed a No-\\ntice of Opposition on the grounds of likelihood of confusion\\nwith its mark under Section 2(d) of the Lanham Act, 15\\nU.S.C. § 1052(d). Hybrid’s mark is also a stylized letter\\n“H.” The two marks are shown in the chart below:\\n\\n Hylete’s Mark Hybrid’s Mark\\n\\fHYLETE LLC v. HYBRID ATHLETICS, LLC 3\\n\\n\\n\\n Hybrid’s Notice of Opposition pleaded ownership in Ap-\\nplication No. 86/000,809 (“the ’809 application”) for a de-\\nsign mark of its stylized “H” used “in connection with\\nconducting fitness classes; health club services, namely,\\nproviding instruction and equipment in the field of physical\\nexercise; personal fitness training services and consul-\\ntancy; physical fitness instruction” in International\\nClass 41. J.A. 3, 98–99. Hybrid also pleaded common law\\nrights from its use of the same mark on “athletic apparel,\\nincluding shirts, hats, shorts and socks” since August 1,\\n2008. Id.\\n During opposition proceedings before the Trademark\\nTrial and Appeal Board (“Board”), Hybrid submitted as an\\nexhibit several images depicting the use of its mark on ath-\\nletic apparel, including shirts, shorts, and jackets. The im-\\nages showed Hybrid’s stylized “H” design appearing on the\\napparel above the phrase “Hybrid Athletics” and several\\ndots:\\n\\n\\n\\n\\nJ.A. 12.\\n\\f4 HYLETE LLC v. HYBRID ATHLETICS, LLC\\n\\n\\n\\n\\n In its briefing before the Board, Hylete focused on the\\ndifferences in appearance between the two stylized “H” de-\\nsigns. It argued that its mark was a “highly stylized design\\nlogo” that “is substantially dissimilar from [Hybrid’s] letter\\n‘H’ design logo” and “the lettering style of each mark is sub-\\nstantially dissimilar in appearance and each mark exudes\\nits own distinct commercial expression.” J.A. 405–07\\n(heading capitalization removed). Hylete further argued\\nthat “the stylization of [Hybrid’s] mark[] emphasizes its\\nrepresentation as an ‘H,’ whereas [Hylete’s] mark is a\\nhighly stylized design.” J.A. 407 (graphics removed).\\n On December 15, 2016, the Board issued its final deci-\\nsion sustaining Hybrid’s opposition to Hylete’s registra-\\ntion. J.A. 2–23. Balancing the relevant factors as set forth\\nin In re E.I. DuPont DeNemours & Co., 476 F.2d 1357\\n(CCPA 1973), the Board determined that Hylete’s mark\\nwould likely cause confusion with Hybrid’s “previously-\\nused mark . . . on some of the same goods, namely jackets,\\nshorts, and shirts.” J.A. 22–23. The Board also found that\\nHybrid failed to establish ownership of the ’809 application\\nand based its conclusions only on Hybrid’s prior common\\nlaw rights. J.A. 6.\\n As to the similarity of the marks, the Board recognized\\nthat both marks are stylized versions of the letter “H,” that\\nboth parties’ names begin with the letter “H,” and that even\\nif consumers attribute no specific meaning to the letter “H,”\\nthey may nonetheless view the marks in the same man-\\nner—as an arbitrary use of the stylized letter “H” for ath-\\nletic clothing. J.A. 15–18. According to the Board, the\\naverage consumer would retain a general rather than spe-\\ncific impression of the marks. J.A. 16 (citing Grandpa\\nPidgeon’s of Mo., Inc. v. Borgsmiller, 477 F.2d 586, 587\\n(CCPA 1973) (noting that the marks in question had “a dif-\\nference not likely to be recalled by purchasers seeing the\\nmarks at spaced intervals”)).\\n\\fHYLETE LLC v. HYBRID ATHLETICS, LLC 5\\n\\n\\n\\n The Board considered various design similarities and\\ndifferences between the two marks. It acknowledged that\\nspecific differences can be seen when the marks are com-\\npared side by side, but determined the fact that both are\\nstylized versions of the same letter outweighed those dif-\\nferences. The Board determined that the marks have sim-\\nilar commercial impressions and concluded that Hylete’s\\nmark is likely to cause confusion with Hybrid’s previously\\nused mark.\\n Hylete filed a request for reconsideration of the Board’s\\nfinal decision. Hylete asserted that the Board erred in\\nthree respects, including the Board’s purported “misappre-\\nhension of the commercial impression of [Hylete’s] mark.”\\nJ.A. 25. Hylete argued “[t]here was no record evidence\\ndemonstrating that consumers would view [Hylete’s] mark\\nas a stylized H.” J.A. 32. Hylete focused on its own mark\\nand did not argue that the Board’s analysis should have\\ncompared its mark to anything other than Hybrid’s stylized\\nletter “H” design mark.\\n The Board addressed Hylete’s commercial-impression\\nargument by noting its “stark contrast” with Hylete’s own\\ncharacterization of its mark as a stylized letter “H” in its\\nbriefing:\\n When the Board performs its analysis, it will find\\n two distinct letter “H” marks that already co-exist\\n with one hundred and thirty five (135) other “H”\\n marks registered to International Class 25, thirty-\\n three (33) of which are specifically used in connec-\\n tion with athletic-related clothing.\\nJ.A. 32–33 (quoting Hylete’s trial brief). The Board also\\nnoted that Hylete’s arguments based on how the marks\\nwould be perceived relied on testimony from its CEO, stat-\\ning that he did not “see how anyone looking at these two\\nlogos would think they look alike” but admitting “they both\\nare H’s [sic].” J.A. 33 (emphases added). The Board\\n\\f6 HYLETE LLC v. HYBRID ATHLETICS, LLC\\n\\n\\n\\n\\ntherefore rejected Hylete’s commercial-impression argu-\\nments and denied the request for rehearing. Id.\\n Hylete appeals. We have jurisdiction under 28 U.S.C.\\n§ 1295(a)(4)(B) and 15 U.S.C. § 1071(a)(1).\\n DISCUSSION\\n We review the Board’s legal conclusions de novo and its\\nfactual findings for substantial evidence. In re\\nI.AM.Symbolic, LLC, 866 F.3d 1315, 1322 (Fed. Cir. 2017).\\nLikelihood of confusion is a question of law based on under-\\nlying findings of facts. In re Chatam Int’l Inc., 380 F.3d\\n1340, 1342 (Fed. Cir. 2004). The Board’s underlying find-\\nings of fact as to each DuPont factor are reviewed for sub-\\nstantial evidence. Id. Substantial evidence is “such\\nrelevant evidence [that] a reasonable mind would accept as\\nadequate to support a conclusion.” I.AM.Symbolic, 866\\nF.3d at 1322 (internal quotations omitted).\\n On appeal, Hylete argues that the Board erred in its\\nanalysis by failing to compare Hylete’s stylized “H” mark\\nwith what it now characterizes as Hybrid’s “composite com-\\nmon law mark,” referring to Hybrid’s stylized “H” design\\nappearing above the phrase “Hybrid Athletics” and several\\ndots, as shown below:\\n\\n\\n\\n\\nAppellant Br. 6. According to Hylete, the issues in “this\\nappeal may be summarized into a single question: is\\nHylete’s mark sufficiently similar to [Hybrid’s] composite\\ncommon law mark to be likely to cause confusion on the\\n\\fHYLETE LLC v. HYBRID ATHLETICS, LLC 7\\n\\n\\n\\npart of the ordinary consumer as to the source of the cloth-\\ning items sold under those marks?” Id. As such, Hylete\\nraises on appeal arguments related only to Hybrid’s “com-\\nposite common law mark.”\\n Hybrid responds that Hylete’s “composite common law\\nmark” arguments were never raised before the Board and\\nare therefore waived. Appellee Br. 20–22. We therefore\\nfirst review whether Hylete waived its arguments by not\\nraising them during the proceedings below. We conclude\\nthat those arguments are waived.\\n Generally, federal appellate courts do not consider is-\\nsues “not passed upon below” or entertain arguments not\\npresented to the lower tribunal. Golden Bridge Tech., Inc.\\nv. Nokia, Inc., 527 F.3d 1318, 1322 (Fed. Cir. 2008).\\n“[A]bsent exceptional circumstances, a party cannot raise\\non appeal legal issues not raised and considered in the trial\\nforum.” Finch v. Hughes Aircraft Co., 926 F.2d 1574, 1576\\n(Fed. Cir. 1991); see also Sw. Software, Inc. v. Harlequin\\nInc., 226 F.3d 1280, 1289 n.7 (Fed. Cir. 2000).\\n We have articulated limited circumstances in which\\nconsidering arguments made for the first time on appeal is\\nappropriate: (1) “[w]hen new legislation is passed while an\\nappeal is pending, courts have an obligation to apply the\\nnew law if Congress intended retroactive application even\\nthough the issue was not decided or raised below”; (2)\\n“when there is a change in the jurisprudence of the review-\\ning court or the Supreme Court after consideration of the\\ncase by the lower court”; (3) “appellate courts may apply\\nthe correct law even if the parties did not argue it below\\nand the court below did not decide it, but only if an issue is\\nproperly before the court”; and (4) “where a party appeared\\npro se before the lower court, a court of appeals may appro-\\npriately be less stringent in requiring that the issue have\\nbeen raised explicitly below.” Golden Bridge, 527 F.3d at\\n1322–23 (quoting Forshey v. Principi, 284 F.3d 1335, 1353–\\n57 (Fed. Cir. 2002)).\\n\\f8 HYLETE LLC v. HYBRID ATHLETICS, LLC\\n\\n\\n\\n\\n Hylete does not dispute that it did not present to the\\nBoard the question of whether Hylete’s mark is sufficiently\\nsimilar to Hybrid’s “composite common law mark” to cause\\na likelihood of confusion. Hybrid pleaded common law\\nrights to its mark based on use of the mark on athletic ap-\\nparel and submitted an exhibit showing its use of the so-\\ncalled “composite common law mark” on athletic apparel.\\nYet Hylete’s arguments in both its briefing to the Board\\nand its request for reconsideration remained focused on the\\ndifferences between the two stylized “H” design marks.\\nHylete never contended that Hybrid’s common law rights\\nimplicated a “composite common law mark” that differed\\nfrom the stylized “H” design mark identified in the ’809 ap-\\nplication.\\n Hylete now attempts to avoid waiver by contending\\nthat the Board “sua sponte” raised the issue of Hybrid’s\\ncommon law rights in the final decision and used an incor-\\nrect legal standard in comparing the two marks. Appellant\\nReply Br. 4. Hylete’s failure to raise that argument in the\\nrequest for reconsideration negates its contention. Even\\nafter the Board purportedly raised the issue of Hybrid’s\\ncommon law rights in its final decision, Hylete did not ar-\\ngue in its request for reconsideration that the Board erred\\nby comparing the wrong marks. Hylete’s request for recon-\\nsideration instead focused on purported differences be-\\ntween its mark and Hybrid’s stylized “H” design mark, and\\nnot on Hybrid’s “composite common law mark.” 1\\n\\n\\n\\n 1 Hylete in fact made contradictory arguments fo-\\ncused on the stylized “H” design mark. It first admitted\\nthat its own mark was a letter “H” before the Board’s final\\ndecision. J.A. 32–33. And in the request for reconsidera-\\ntion, Hylete took the position that the Board misappre-\\nhended the commercial impression of Hylete’s mark and\\n“there was no record evidence demonstrating that\\n\\fHYLETE LLC v. HYBRID ATHLETICS, LLC 9\\n\\n\\n\\n Hybrid pleaded and put Hylete on notice of its claim to\\ncommon law trademark rights from the initial filing of the\\nNotice of Opposition and submitted evidence of its use of\\nits mark on athletic apparel. Hylete could have raised the\\nissue of Hybrid’s “composite common law mark” in the op-\\nposition proceedings or in the request for reconsideration\\nbut did not do so. Thus, none of the exceptional circum-\\nstances in which it is appropriate to consider arguments\\nmade for the first time on appeal are present here, and de-\\nclining to consider Hylete’s new arguments does not result\\nin injustice. See Golden Bridge, 527 F.3d at 1323. We hold\\nthat Hylete’s arguments based on Hybrid’s “composite com-\\nmon law mark” are raised for the first time on appeal and\\nare therefore waived.\\n CONCLUSION\\n Hylete waived its argument that Hylete’s mark is suf-\\nficiently different from Hybrid’s “composite common law\\nmark” to avoid a likelihood of confusion as to the source of\\nthe athletic apparel sold bearing those marks. We do not\\naddress the Board’s analysis of the DuPont factors relating\\nto Hybrid’s mark; the only issues Hylete raises on appeal\\nconcern Hybrid’s “composite common law mark.” Accord-\\ningly, we affirm.\\n AFFIRMED\\n COSTS\\n No costs.\\n\\n\\n\\n\\nconsumers would view [Hylete’s] mark as a stylized H.”\\nJ.A. 25, 32.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4422110/\", \"author_raw\": \"REYNA, Circuit Judge.\"}]}"},"judge1":{"kind":"string","value":"MOORE"},"judge2":{"kind":"string","value":"REYNA"},"judge3":{"kind":"string","value":"WALLACH"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4644857/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: HYLETE LLC, Appellant v. HYBRID ATHLETICS, LLC, Appellee\nCourt: U.S. Court of Appeals for the Federal Circuit\nDate decided: 2019-08-01\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n for the Federal Circuit\n ______________________\n\n HYLETE LLC,\n Appellant\n\n v.\n\n HYBRID ATHLETICS, LLC,\n Appellee\n ______________________\n\n 2017-2057\n ______________________\n\n Appeal from the United States Patent and Trademark\nOffice, Trademark Trial and Appeal Board in No.\n91213057.\n ______________________\n\n Decided: August 1, 2019\n ______________________\n\n PATTRIC RAWLINS, Procopio, Cory, Hargreaves & Sa-\nvitch LLP, San Diego, CA, argued for appellant. Also rep-\nresented by DAVE DEONARINE.\n\n MICHAEL JOSEPH KOSMA, Whitmyer IP Group LLC,\nStamford, CT, argued for appellee. Also represented by\nBENJAMIN N. LUEHRS.\n ______________________\n\n Before MOORE, REYNA, and WALLACH, Circuit Judges.\n\f2 HYLETE LLC v. HYBRID ATHLETICS, LLC\n\n\n\n\nREYNA, Circuit Judge.\n Hylete LLC appeals from a decision of the Trademark\nTrial and Appeal Board sustaining Hybrid Athletics, LLC’s\nopposition to Hylete’s trademark registration application.\nWe conclude that Hylete waived the arguments on which\nits appeal relies because it raises new issues that could\nhave been raised and were not considered below. We af-\nfirm.\n BACKGROUND\n In January 2013, Hylete applied to register a design\nmark for a stylized letter “H” in International Class 25 for\n“[a]thletic apparel, namely, shirts, pants, shorts, jackets,\nfootwear, hats and caps.” J.A. 76–92. After finding no reg-\nistrations that would bar registration of the Hylete mark,\nthe Examining Attorney approved the application, and the\nTrademark Office published the Hylete mark for opposition\nin the Trademark Official Gazette on June 18, 2013.\n On October 16, 2013, Hybrid Athletics, LLC filed a No-\ntice of Opposition on the grounds of likelihood of confusion\nwith its mark under Section 2(d) of the Lanham Act, 15\nU.S.C. § 1052(d). Hybrid’s mark is also a stylized letter\n“H.” The two marks are shown in the chart below:\n\n Hylete’s Mark Hybrid’s Mark\n\fHYLETE LLC v. HYBRID ATHLETICS, LLC 3\n\n\n\n Hybrid’s Notice of Opposition pleaded ownership in Ap-\nplication No. 86/000,809 (“the ’809 application”) for a de-\nsign mark of its stylized “H” used “in connection with\nconducting fitness classes; health club services, namely,\nproviding instruction and equipment in the field of physical\nexercise; personal fitness training services and consul-\ntancy; physical fitness instruction” in International\nClass 41. J.A. 3, 98–99. Hybrid also pleaded common law\nrights from its use of the same mark on “athletic apparel,\nincluding shirts, hats, shorts and socks” since August 1,\n2008. Id.\n During opposition proceedings before the Trademark\nTrial and Appeal Board (“Board”), Hybrid submitted as an\nexhibit several images depicting the use of its mark on ath-\nletic apparel, including shirts, shorts, and jackets. The im-\nages showed Hybrid’s stylized “H” design appearing on the\napparel above the phrase “Hybrid Athletics” and several\ndots:\n\n\n\n\nJ.A. 12.\n\f4 HYLETE LLC v. HYBRID ATHLETICS, LLC\n\n\n\n\n In its briefing before the Board, Hylete focused on the\ndifferences in appearance between the two stylized “H” de-\nsigns. It argued that its mark was a “highly stylized design\nlogo” that “is substantially dissimilar from [Hybrid’s] letter\n‘H’ design logo” and “the lettering style of each mark is sub-\nstantially dissimilar in appearance and each mark exudes\nits own distinct commercial expression.” J.A. 405–07\n(heading capitalization removed). Hylete further argued\nthat “the stylization of [Hybrid’s] mark[] emphasizes its\nrepresentation as an ‘H,’ whereas [Hylete’s] mark is a\nhighly stylized design.” J.A. 407 (graphics removed).\n On December 15, 2016, the Board issued its final deci-\nsion sustaining Hybrid’s opposition to Hylete’s registra-\ntion. J.A. 2–23. Balancing the relevant factors as set forth\nin In re E.I. DuPont DeNemours & Co., 476 F.2d 1357\n(CCPA 1973), the Board determined that Hylete’s mark\nwould likely cause confusion with Hybrid’s “previously-\nused mark . . . on some of the same goods, namely jackets,\nshorts, and shirts.” J.A. 22–23. The Board also found that\nHybrid failed to establish ownership of the ’809 application\nand based its conclusions only on Hybrid’s prior common\nlaw rights. J.A. 6.\n As to the similarity of the marks, the Board recognized\nthat both marks are stylized versions of the letter “H,” that\nboth parties’ names begin with the letter “H,” and that even\nif consumers attribute no specific meaning to the letter “H,”\nthey may nonetheless view the marks in the same man-\nner—as an arbitrary use of the stylized letter “H” for ath-\nletic clothing. J.A. 15–18. According to the Board, the\naverage consumer would retain a general rather than spe-\ncific impression of the marks. J.A. 16 (citing Grandpa\nPidgeon’s of Mo., Inc. v. Borgsmiller, 477 F.2d 586, 587\n(CCPA 1973) (noting that the marks in question had “a dif-\nference not likely to be recalled by purchasers seeing the\nmarks at spaced intervals”)).\n\fHYLETE LLC v. HYBRID ATHLETICS, LLC 5\n\n\n\n The Board considered various design similarities and\ndifferences between the two marks. It acknowledged that\nspecific differences can be seen when the marks are com-\npared side by side, but determined the fact that both are\nstylized versions of the same letter outweighed those dif-\nferences. The Board determined that the marks have sim-\nilar commercial impressions and concluded that Hylete’s\nmark is likely to cause confusion with Hybrid’s previously\nused mark.\n Hylete filed a request for reconsideration of the Board’s\nfinal decision. Hylete asserted that the Board erred in\nthree respects, including the Board’s purported “misappre-\nhension of the commercial impression of [Hylete’s] mark.”\nJ.A. 25. Hylete argued “[t]here was no record evidence\ndemonstrating that consumers would view [Hylete’s] mark\nas a stylized H.” J.A. 32. Hylete focused on its own mark\nand did not argue that the Board’s analysis should have\ncompared its mark to anything other than Hybrid’s stylized\nletter “H” design mark.\n The Board addressed Hylete’s commercial-impression\nargument by noting its “stark contrast” with Hylete’s own\ncharacterization of its mark as a stylized letter “H” in its\nbriefing:\n When the Board performs its analysis, it will find\n two distinct letter “H” marks that already co-exist\n with one hundred and thirty five (135) other “H”\n marks registered to International Class 25, thirty-\n three (33) of which are specifically used in connec-\n tion with athletic-related clothing.\nJ.A. 32–33 (quoting Hylete’s trial brief). The Board also\nnoted that Hylete’s arguments based on how the marks\nwould be perceived relied on testimony from its CEO, stat-\ning that he did not “see how anyone looking at these two\nlogos would think they look alike” but admitting “they both\nare H’s [sic].” J.A. 33 (emphases added). The Board\n\f6 HYLETE LLC v. HYBRID ATHLETICS, LLC\n\n\n\n\ntherefore rejected Hylete’s commercial-impression argu-\nments and denied the request for rehearing. Id.\n Hylete appeals. We have jurisdiction under 28 U.S.C.\n§ 1295(a)(4)(B) and 15 U.S.C. § 1071(a)(1).\n DISCUSSION\n We review the Board’s legal conclusions de novo and its\nfactual findings for substantial evidence. In re\nI.AM.Symbolic, LLC, 866 F.3d 1315, 1322 (Fed. Cir. 2017).\nLikelihood of confusion is a question of law based on under-\nlying findings of facts. In re Chatam Int’l Inc., 380 F.3d\n1340, 1342 (Fed. Cir. 2004). The Board’s underlying find-\nings of fact as to each DuPont factor are reviewed for sub-\nstantial evidence. Id. Substantial evidence is “such\nrelevant evidence [that] a reasonable mind would accept as\nadequate to support a conclusion.” I.AM.Symbolic, 866\nF.3d at 1322 (internal quotations omitted).\n On appeal, Hylete argues that the Board erred in its\nanalysis by failing to compare Hylete’s stylized “H” mark\nwith what it now characterizes as Hybrid’s “composite com-\nmon law mark,” referring to Hybrid’s stylized “H” design\nappearing above the phrase “Hybrid Athletics” and several\ndots, as shown below:\n\n\n\n\nAppellant Br. 6. According to Hylete, the issues in “this\nappeal may be summarized into a single question: is\nHylete’s mark sufficiently similar to [Hybrid’s] composite\ncommon law mark to be likely to cause confusion on the\n\fHYLETE LLC v. HYBRID ATHLETICS, LLC 7\n\n\n\npart of the ordinary consumer as to the source of the cloth-\ning items sold under those marks?” Id. As such, Hylete\nraises on appeal arguments related only to Hybrid’s “com-\nposite common law mark.”\n Hybrid responds that Hylete’s “composite common law\nmark” arguments were never raised before the Board and\nare therefore waived. Appellee Br. 20–22. We therefore\nfirst review whether Hylete waived its arguments by not\nraising them during the proceedings below. We conclude\nthat those arguments are waived.\n Generally, federal appellate courts do not consider is-\nsues “not passed upon below” or entertain arguments not\npresented to the lower tribunal. Golden Bridge Tech., Inc.\nv. Nokia, Inc., 527 F.3d 1318, 1322 (Fed. Cir. 2008).\n“[A]bsent exceptional circumstances, a party cannot raise\non appeal legal issues not raised and considered in the trial\nforum.” Finch v. Hughes Aircraft Co., 926 F.2d 1574, 1576\n(Fed. Cir. 1991); see also Sw. Software, Inc. v. Harlequin\nInc., 226 F.3d 1280, 1289 n.7 (Fed. Cir. 2000).\n We have articulated limited circumstances in which\nconsidering arguments made for the first time on appeal is\nappropriate: (1) “[w]hen new legislation is passed while an\nappeal is pending, courts have an obligation to apply the\nnew law if Congress intended retroactive application even\nthough the issue was not decided or raised below”; (2)\n“when there is a change in the jurisprudence of the review-\ning court or the Supreme Court after consideration of the\ncase by the lower court”; (3) “appellate courts may apply\nthe correct law even if the parties did not argue it below\nand the court below did not decide it, but only if an issue is\nproperly before the court”; and (4) “where a party appeared\npro se before the lower court, a court of appeals may appro-\npriately be less stringent in requiring that the issue have\nbeen raised explicitly below.” Golden Bridge, 527 F.3d at\n1322–23 (quoting Forshey v. Principi, 284 F.3d 1335, 1353–\n57 (Fed. Cir. 2002)).\n\f8 HYLETE LLC v. HYBRID ATHLETICS, LLC\n\n\n\n\n Hylete does not dispute that it did not present to the\nBoard the question of whether Hylete’s mark is sufficiently\nsimilar to Hybrid’s “composite common law mark” to cause\na likelihood of confusion. Hybrid pleaded common law\nrights to its mark based on use of the mark on athletic ap-\nparel and submitted an exhibit showing its use of the so-\ncalled “composite common law mark” on athletic apparel.\nYet Hylete’s arguments in both its briefing to the Board\nand its request for reconsideration remained focused on the\ndifferences between the two stylized “H” design marks.\nHylete never contended that Hybrid’s common law rights\nimplicated a “composite common law mark” that differed\nfrom the stylized “H” design mark identified in the ’809 ap-\nplication.\n Hylete now attempts to avoid waiver by contending\nthat the Board “sua sponte” raised the issue of Hybrid’s\ncommon law rights in the final decision and used an incor-\nrect legal standard in comparing the two marks. Appellant\nReply Br. 4. Hylete’s failure to raise that argument in the\nrequest for reconsideration negates its contention. Even\nafter the Board purportedly raised the issue of Hybrid’s\ncommon law rights in its final decision, Hylete did not ar-\ngue in its request for reconsideration that the Board erred\nby comparing the wrong marks. Hylete’s request for recon-\nsideration instead focused on purported differences be-\ntween its mark and Hybrid’s stylized “H” design mark, and\nnot on Hybrid’s “composite common law mark.” 1\n\n\n\n 1 Hylete in fact made contradictory arguments fo-\ncused on the stylized “H” design mark. It first admitted\nthat its own mark was a letter “H” before the Board’s final\ndecision. J.A. 32–33. And in the request for reconsidera-\ntion, Hylete took the position that the Board misappre-\nhended the commercial impression of Hylete’s mark and\n“there was no record evidence demonstrating that\n\fHYLETE LLC v. HYBRID ATHLETICS, LLC 9\n\n\n\n Hybrid pleaded and put Hylete on notice of its claim to\ncommon law trademark rights from the initial filing of the\nNotice of Opposition and submitted evidence of its use of\nits mark on athletic apparel. Hylete could have raised the\nissue of Hybrid’s “composite common law mark” in the op-\nposition proceedings or in the request for reconsideration\nbut did not do so. Thus, none of the exceptional circum-\nstances in which it is appropriate to consider arguments\nmade for the first time on appeal are present here, and de-\nclining to consider Hylete’s new arguments does not result\nin injustice. See Golden Bridge, 527 F.3d at 1323. We hold\nthat Hylete’s arguments based on Hybrid’s “composite com-\nmon law mark” are raised for the first time on appeal and\nare therefore waived.\n CONCLUSION\n Hylete waived its argument that Hylete’s mark is suf-\nficiently different from Hybrid’s “composite common law\nmark” to avoid a likelihood of confusion as to the source of\nthe athletic apparel sold bearing those marks. We do not\naddress the Board’s analysis of the DuPont factors relating\nto Hybrid’s mark; the only issues Hylete raises on appeal\nconcern Hybrid’s “composite common law mark.” Accord-\ningly, we affirm.\n AFFIRMED\n COSTS\n No costs.\n\n\n\n\nconsumers would view [Hylete’s] mark as a stylized H.”\nJ.A. 25, 32.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: HYLETE LLC, Appellant v. HYBRID ATHLETICS, LLC, Appellee\\nCourt: U.S. Court of Appeals for the Federal Circuit\\nDate decided: 2019-08-01\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n HYLETE LLC,\\n Appellant\\n\\n v.\\n\\n HYBRID ATHLETICS, LLC,\\n Appellee\\n ______________________\\n\\n 2017-2057\\n ______________________\\n\\n Appeal from the United States Patent and Trademark\\nOffice, Trademark Trial and Appeal Board in No.\\n91213057.\\n ______________________\\n\\n Decided: August 1, 2019\\n ______________________\\n\\n PATTRIC RAWLINS, Procopio, Cory, Hargreaves & Sa-\\nvitch LLP, San Diego, CA, argued for appellant. Also rep-\\nresented by DAVE DEONARINE.\\n\\n MICHAEL JOSEPH KOSMA, Whitmyer IP Group LLC,\\nStamford, CT, argued for appellee. Also represented by\\nBENJAMIN N. LUEHRS.\\n ______________________\\n\\n Before MOORE, REYNA, and WALLACH, Circuit Judges.\\n\\f2 HYLETE LLC v. HYBRID ATHLETICS, LLC\\n\\n\\n\\n\\nREYNA, Circuit Judge.\\n Hylete LLC appeals from a decision of the Trademark\\nTrial and Appeal Board sustaining Hybrid Athletics, LLC’s\\nopposition to Hylete’s trademark registration application.\\nWe conclude that Hylete waived the arguments on which\\nits appeal relies because it raises new issues that could\\nhave been raised and were not considered below. We af-\\nfirm.\\n BACKGROUND\\n In January 2013, Hylete applied to register a design\\nmark for a stylized letter “H” in International Class 25 for\\n“[a]thletic apparel, namely, shirts, pants, shorts, jackets,\\nfootwear, hats and caps.” J.A. 76–92. After finding no reg-\\nistrations that would bar registration of the Hylete mark,\\nthe Examining Attorney approved the application, and the\\nTrademark Office published the Hylete mark for opposition\\nin the Trademark Official Gazette on June 18, 2013.\\n On October 16, 2013, Hybrid Athletics, LLC filed a No-\\ntice of Opposition on the grounds of likelihood of confusion\\nwith its mark under Section 2(d) of the Lanham Act, 15\\nU.S.C. § 1052(d). Hybrid’s mark is also a stylized letter\\n“H.” The two marks are shown in the chart below:\\n\\n Hylete’s Mark Hybrid’s Mark\\n\\fHYLETE LLC v. HYBRID ATHLETICS, LLC 3\\n\\n\\n\\n Hybrid’s Notice of Opposition pleaded ownership in Ap-\\nplication No. 86/000,809 (“the ’809 application”) for a de-\\nsign mark of its stylized “H” used “in connection with\\nconducting fitness classes; health club services, namely,\\nproviding instruction and equipment in the field of physical\\nexercise; personal fitness training services and consul-\\ntancy; physical fitness instruction” in International\\nClass 41. J.A. 3, 98–99. Hybrid also pleaded common law\\nrights from its use of the same mark on “athletic apparel,\\nincluding shirts, hats, shorts and socks” since August 1,\\n2008. Id.\\n During opposition proceedings before the Trademark\\nTrial and Appeal Board (“Board”), Hybrid submitted as an\\nexhibit several images depicting the use of its mark on ath-\\nletic apparel, including shirts, shorts, and jackets. The im-\\nages showed Hybrid’s stylized “H” design appearing on the\\napparel above the phrase “Hybrid Athletics” and several\\ndots:\\n\\n\\n\\n\\nJ.A. 12.\\n\\f4 HYLETE LLC v. HYBRID ATHLETICS, LLC\\n\\n\\n\\n\\n In its briefing before the Board, Hylete focused on the\\ndifferences in appearance between the two stylized “H” de-\\nsigns. It argued that its mark was a “highly stylized design\\nlogo” that “is substantially dissimilar from [Hybrid’s] letter\\n‘H’ design logo” and “the lettering style of each mark is sub-\\nstantially dissimilar in appearance and each mark exudes\\nits own distinct commercial expression.” J.A. 405–07\\n(heading capitalization removed). Hylete further argued\\nthat “the stylization of [Hybrid’s] mark[] emphasizes its\\nrepresentation as an ‘H,’ whereas [Hylete’s] mark is a\\nhighly stylized design.” J.A. 407 (graphics removed).\\n On December 15, 2016, the Board issued its final deci-\\nsion sustaining Hybrid’s opposition to Hylete’s registra-\\ntion. J.A. 2–23. Balancing the relevant factors as set forth\\nin In re E.I. DuPont DeNemours & Co., 476 F.2d 1357\\n(CCPA 1973), the Board determined that Hylete’s mark\\nwould likely cause confusion with Hybrid’s “previously-\\nused mark . . . on some of the same goods, namely jackets,\\nshorts, and shirts.” J.A. 22–23. The Board also found that\\nHybrid failed to establish ownership of the ’809 application\\nand based its conclusions only on Hybrid’s prior common\\nlaw rights. J.A. 6.\\n As to the similarity of the marks, the Board recognized\\nthat both marks are stylized versions of the letter “H,” that\\nboth parties’ names begin with the letter “H,” and that even\\nif consumers attribute no specific meaning to the letter “H,”\\nthey may nonetheless view the marks in the same man-\\nner—as an arbitrary use of the stylized letter “H” for ath-\\nletic clothing. J.A. 15–18. According to the Board, the\\naverage consumer would retain a general rather than spe-\\ncific impression of the marks. J.A. 16 (citing Grandpa\\nPidgeon’s of Mo., Inc. v. Borgsmiller, 477 F.2d 586, 587\\n(CCPA 1973) (noting that the marks in question had “a dif-\\nference not likely to be recalled by purchasers seeing the\\nmarks at spaced intervals”)).\\n\\fHYLETE LLC v. HYBRID ATHLETICS, LLC 5\\n\\n\\n\\n The Board considered various design similarities and\\ndifferences between the two marks. It acknowledged that\\nspecific differences can be seen when the marks are com-\\npared side by side, but determined the fact that both are\\nstylized versions of the same letter outweighed those dif-\\nferences. The Board determined that the marks have sim-\\nilar commercial impressions and concluded that Hylete’s\\nmark is likely to cause confusion with Hybrid’s previously\\nused mark.\\n Hylete filed a request for reconsideration of the Board’s\\nfinal decision. Hylete asserted that the Board erred in\\nthree respects, including the Board’s purported “misappre-\\nhension of the commercial impression of [Hylete’s] mark.”\\nJ.A. 25. Hylete argued “[t]here was no record evidence\\ndemonstrating that consumers would view [Hylete’s] mark\\nas a stylized H.” J.A. 32. Hylete focused on its own mark\\nand did not argue that the Board’s analysis should have\\ncompared its mark to anything other than Hybrid’s stylized\\nletter “H” design mark.\\n The Board addressed Hylete’s commercial-impression\\nargument by noting its “stark contrast” with Hylete’s own\\ncharacterization of its mark as a stylized letter “H” in its\\nbriefing:\\n When the Board performs its analysis, it will find\\n two distinct letter “H” marks that already co-exist\\n with one hundred and thirty five (135) other “H”\\n marks registered to International Class 25, thirty-\\n three (33) of which are specifically used in connec-\\n tion with athletic-related clothing.\\nJ.A. 32–33 (quoting Hylete’s trial brief). The Board also\\nnoted that Hylete’s arguments based on how the marks\\nwould be perceived relied on testimony from its CEO, stat-\\ning that he did not “see how anyone looking at these two\\nlogos would think they look alike” but admitting “they both\\nare H’s [sic].” J.A. 33 (emphases added). The Board\\n\\f6 HYLETE LLC v. HYBRID ATHLETICS, LLC\\n\\n\\n\\n\\ntherefore rejected Hylete’s commercial-impression argu-\\nments and denied the request for rehearing. Id.\\n Hylete appeals. We have jurisdiction under 28 U.S.C.\\n§ 1295(a)(4)(B) and 15 U.S.C. § 1071(a)(1).\\n DISCUSSION\\n We review the Board’s legal conclusions de novo and its\\nfactual findings for substantial evidence. In re\\nI.AM.Symbolic, LLC, 866 F.3d 1315, 1322 (Fed. Cir. 2017).\\nLikelihood of confusion is a question of law based on under-\\nlying findings of facts. In re Chatam Int’l Inc., 380 F.3d\\n1340, 1342 (Fed. Cir. 2004). The Board’s underlying find-\\nings of fact as to each DuPont factor are reviewed for sub-\\nstantial evidence. Id. Substantial evidence is “such\\nrelevant evidence [that] a reasonable mind would accept as\\nadequate to support a conclusion.” I.AM.Symbolic, 866\\nF.3d at 1322 (internal quotations omitted).\\n On appeal, Hylete argues that the Board erred in its\\nanalysis by failing to compare Hylete’s stylized “H” mark\\nwith what it now characterizes as Hybrid’s “composite com-\\nmon law mark,” referring to Hybrid’s stylized “H” design\\nappearing above the phrase “Hybrid Athletics” and several\\ndots, as shown below:\\n\\n\\n\\n\\nAppellant Br. 6. According to Hylete, the issues in “this\\nappeal may be summarized into a single question: is\\nHylete’s mark sufficiently similar to [Hybrid’s] composite\\ncommon law mark to be likely to cause confusion on the\\n\\fHYLETE LLC v. HYBRID ATHLETICS, LLC 7\\n\\n\\n\\npart of the ordinary consumer as to the source of the cloth-\\ning items sold under those marks?” Id. As such, Hylete\\nraises on appeal arguments related only to Hybrid’s “com-\\nposite common law mark.”\\n Hybrid responds that Hylete’s “composite common law\\nmark” arguments were never raised before the Board and\\nare therefore waived. Appellee Br. 20–22. We therefore\\nfirst review whether Hylete waived its arguments by not\\nraising them during the proceedings below. We conclude\\nthat those arguments are waived.\\n Generally, federal appellate courts do not consider is-\\nsues “not passed upon below” or entertain arguments not\\npresented to the lower tribunal. Golden Bridge Tech., Inc.\\nv. Nokia, Inc., 527 F.3d 1318, 1322 (Fed. Cir. 2008).\\n“[A]bsent exceptional circumstances, a party cannot raise\\non appeal legal issues not raised and considered in the trial\\nforum.” Finch v. Hughes Aircraft Co., 926 F.2d 1574, 1576\\n(Fed. Cir. 1991); see also Sw. Software, Inc. v. Harlequin\\nInc., 226 F.3d 1280, 1289 n.7 (Fed. Cir. 2000).\\n We have articulated limited circumstances in which\\nconsidering arguments made for the first time on appeal is\\nappropriate: (1) “[w]hen new legislation is passed while an\\nappeal is pending, courts have an obligation to apply the\\nnew law if Congress intended retroactive application even\\nthough the issue was not decided or raised below”; (2)\\n“when there is a change in the jurisprudence of the review-\\ning court or the Supreme Court after consideration of the\\ncase by the lower court”; (3) “appellate courts may apply\\nthe correct law even if the parties did not argue it below\\nand the court below did not decide it, but only if an issue is\\nproperly before the court”; and (4) “where a party appeared\\npro se before the lower court, a court of appeals may appro-\\npriately be less stringent in requiring that the issue have\\nbeen raised explicitly below.” Golden Bridge, 527 F.3d at\\n1322–23 (quoting Forshey v. Principi, 284 F.3d 1335, 1353–\\n57 (Fed. Cir. 2002)).\\n\\f8 HYLETE LLC v. HYBRID ATHLETICS, LLC\\n\\n\\n\\n\\n Hylete does not dispute that it did not present to the\\nBoard the question of whether Hylete’s mark is sufficiently\\nsimilar to Hybrid’s “composite common law mark” to cause\\na likelihood of confusion. Hybrid pleaded common law\\nrights to its mark based on use of the mark on athletic ap-\\nparel and submitted an exhibit showing its use of the so-\\ncalled “composite common law mark” on athletic apparel.\\nYet Hylete’s arguments in both its briefing to the Board\\nand its request for reconsideration remained focused on the\\ndifferences between the two stylized “H” design marks.\\nHylete never contended that Hybrid’s common law rights\\nimplicated a “composite common law mark” that differed\\nfrom the stylized “H” design mark identified in the ’809 ap-\\nplication.\\n Hylete now attempts to avoid waiver by contending\\nthat the Board “sua sponte” raised the issue of Hybrid’s\\ncommon law rights in the final decision and used an incor-\\nrect legal standard in comparing the two marks. Appellant\\nReply Br. 4. Hylete’s failure to raise that argument in the\\nrequest for reconsideration negates its contention. Even\\nafter the Board purportedly raised the issue of Hybrid’s\\ncommon law rights in its final decision, Hylete did not ar-\\ngue in its request for reconsideration that the Board erred\\nby comparing the wrong marks. Hylete’s request for recon-\\nsideration instead focused on purported differences be-\\ntween its mark and Hybrid’s stylized “H” design mark, and\\nnot on Hybrid’s “composite common law mark.” 1\\n\\n\\n\\n 1 Hylete in fact made contradictory arguments fo-\\ncused on the stylized “H” design mark. It first admitted\\nthat its own mark was a letter “H” before the Board’s final\\ndecision. J.A. 32–33. And in the request for reconsidera-\\ntion, Hylete took the position that the Board misappre-\\nhended the commercial impression of Hylete’s mark and\\n“there was no record evidence demonstrating that\\n\\fHYLETE LLC v. HYBRID ATHLETICS, LLC 9\\n\\n\\n\\n Hybrid pleaded and put Hylete on notice of its claim to\\ncommon law trademark rights from the initial filing of the\\nNotice of Opposition and submitted evidence of its use of\\nits mark on athletic apparel. Hylete could have raised the\\nissue of Hybrid’s “composite common law mark” in the op-\\nposition proceedings or in the request for reconsideration\\nbut did not do so. Thus, none of the exceptional circum-\\nstances in which it is appropriate to consider arguments\\nmade for the first time on appeal are present here, and de-\\nclining to consider Hylete’s new arguments does not result\\nin injustice. See Golden Bridge, 527 F.3d at 1323. We hold\\nthat Hylete’s arguments based on Hybrid’s “composite com-\\nmon law mark” are raised for the first time on appeal and\\nare therefore waived.\\n CONCLUSION\\n Hylete waived its argument that Hylete’s mark is suf-\\nficiently different from Hybrid’s “composite common law\\nmark” to avoid a likelihood of confusion as to the source of\\nthe athletic apparel sold bearing those marks. We do not\\naddress the Board’s analysis of the DuPont factors relating\\nto Hybrid’s mark; the only issues Hylete raises on appeal\\nconcern Hybrid’s “composite common law mark.” Accord-\\ningly, we affirm.\\n AFFIRMED\\n COSTS\\n No costs.\\n\\n\\n\\n\\nconsumers would view [Hylete’s] mark as a stylized H.”\\nJ.A. 25, 32.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4192,"cells":{"id":{"kind":"number","value":4644858,"string":"4,644,858"},"name":{"kind":"string","value":"Callan CAMPBELL, James H. Chadwick, Judith Strode Chadwick, Kevin C. Chadwick, Individually and Through His Court-Appointed Administrators, James H. Chadwick and Judith Strode Chadwick, Kevin Junso, Niki Junso, Tyler Junso Estate, Through Kevin Junso, Its Personal Representative, Plaintiffs-Appellants v. UNITED STATES, Defendant-Appellee"},"name_abbreviation":{"kind":"string","value":"Campbell v. United States"},"decision_date":{"kind":"string","value":"2019-08-01"},"docket_number":{"kind":"string","value":"2018-2014"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Federal Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Newman, Lourie, Dyk\", \"parties\": \"\", \"opinions\": [{\"author\": \"DYK, Circuit Judge.\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n CALLAN CAMPBELL, JAMES H. CHADWICK,\\n JUDITH STRODE CHADWICK, KEVIN C.\\nCHADWICK, INDIVIDUALLY AND THROUGH HIS\\nCOURT-APPOINTED ADMINISTRATORS, JAMES\\nH. CHADWICK AND JUDITH STRODE CHADWICK,\\n KEVIN JUNSO, NIKI JUNSO, TYLER JUNSO\\n ESTATE, THROUGH KEVIN JUNSO, ITS\\n PERSONAL REPRESENTATIVE,\\n Plaintiffs-Appellants\\n\\n v.\\n\\n UNITED STATES,\\n Defendant-Appellee\\n ______________________\\n\\n 2018-2014\\n ______________________\\n\\n Appeal from the United States Court of Federal Claims\\nin No. 1:15-cv-00717-PEC, Judge Patricia E. Campbell-\\nSmith.\\n ______________________\\n\\n Decided: August 1, 2019\\n ______________________\\n\\n STEVEN R. JAKUBOWSKI, Robbins Salomon & Patt, Ltd.,\\nChicago, IL, argued for plaintiffs-appellants.\\n\\n JOHN JACOB TODOR, Commercial Litigation Branch,\\nCivil Division, United States Department of Justice,\\n\\f2 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nWashington, DC, argued for defendant-appellee. Also rep-\\nresented by JOSEPH H. HUNT, ROBERT EDWARD KIRSCHMAN,\\nJR., FRANKLIN E. WHITE, JR.\\n ______________________\\n\\n Before NEWMAN, LOURIE, and DYK, Circuit Judges.\\nDYK, Circuit Judge.\\n This case arises out of the 2009 bankruptcy of General\\nMotors Corporation (“Old GM”). 1 The plaintiffs compose a\\nputative class of individuals who had asserted personal in-\\njury claims against Old GM, and whose successor liability\\nclaims were extinguished during bankruptcy. Relying on\\nour decision in A & D Auto Sales, Inc. v. United States, 748\\nF.3d 1142 (Fed. Cir. 2014), the plaintiffs sued the United\\nStates on behalf of themselves and others similarly situ-\\nated in the Court of Federal Claims (“Claims Court”), al-\\nleging that the extinguishment of their claims without just\\ncompensation violated the Takings Clause of the Fifth\\nAmendment. The Claims Court dismissed the plaintiffs’\\nclaims, concluding that they were barred by the statute of\\nlimitations and that the plaintiffs had, in any event, failed\\nto state a claim. Because we hold, as to the claims alleging\\ncoercion of Old GM, that the statute of limitations had run\\nwhen the plaintiffs filed their complaint and, with respect\\nto the plaintiffs’ other claims, that the Claims Court also\\nlacks jurisdiction, we affirm.\\n BACKGROUND\\n I\\n In A & D, a group of former automobile dealerships\\nsued the United States, raising Fifth Amendment takings\\nclaims based on the extinguishment of the plaintiffs’\\n\\n\\n\\n 1 “Old GM” refers to the GM entity in existence prior\\nto the sale of assets pursuant to 11 U.S.C. § 363.\\n\\fCAMPBELL v. UNITED STATES 3\\n\\n\\n\\nfranchise agreements with Old GM in a bankruptcy sale\\npursuant to 11 U.S.C. § 363. 748 F.3d at 1147. That sec-\\ntion gives a bankruptcy trustee the power to use, sell, or\\nlease the property of a debtor in bankruptcy. In particular,\\n§ 363(f) (the provision at issue here) provides that “[t]he\\ntrustee may sell property . . . free and clear of any interest\\nin such property of an entity other than the estate” under\\ncertain conditions. 11 U.S.C. § 363(f) (emphasis added).\\n In A & D, the plaintiffs alleged that the government\\nhad conditioned its continued financial assistance to Old\\nGM on the company’s submission for approval of a pro-\\nposed sale order that terminated the plaintiffs’ franchise\\nagreements. 748 F.3d at 1148. The plaintiffs contended\\nthat this purported coercion effected a regulatory taking\\nunder the Fifth Amendment. Id. at 1149. The Claims\\nCourt denied the government’s motion to dismiss for fail-\\nure to state a claim but certified the case for interlocutory\\nappeal pursuant to 28 U.S.C. § 1292(d)(2). Id. at 1150.\\n On appeal, we held that the government may, in some\\ncircumstances, be liable for a regulatory taking of property\\nwhere the government pressures a third party (there, al-\\nlegedly Old GM) to take an “action that affects or elimi-\\nnates the property rights of the plaintiff.” Id. at 1153. We\\ndetermined that such conduct may give rise to a taking\\nwhere the government’s action was “direct and intended”\\nand where the “the third party is acting as the govern-\\nment’s agent or the government’s influence over the third\\nparty was coercive rather than merely persuasive.” Id. at\\n1154. We did not decide whether the government’s actions\\nwith respect to Old GM were coercive or otherwise satisfied\\nthe conditions for takings liability. 2 Id. at 1155–56. We\\n\\n\\n 2 Even if coercion had been established, that would\\nmerely make the third party’s action the equivalent of gov-\\nernment action. The plaintiffs would still be required to\\nengage in the Penn Central analysis and to establish a\\n\\f4 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nexplained that to state a claim, the plaintiffs needed to\\nhave pled that their “property suffered a diminution in\\nvalue or a deprivation of economically beneficial use” as a\\nresult of the government’s action. Id. at 1157. We deter-\\nmined that the plaintiffs’ allegations were insufficient to\\nshow that their franchise agreements had value absent the\\ngovernment action and remanded to the Claims Court to\\npermit the plaintiffs to amend their complaint “to include\\nspecific allegations establishing loss of value” and thereaf-\\nter to determine whether a compensable taking had oc-\\ncurred. Id. at 1158–59.\\n II\\n Relying on A & D, on July 9, 2015, the plaintiffs here\\nsued the government in the Claims Court alleging that the\\ngovernment had coerced Old GM to include in its proposed\\nbankruptcy sale order provisions extinguishing the plain-\\ntiffs’ property interests pursuant to § 363 of the Bank-\\nruptcy Code.\\n The plaintiffs are a group of individuals who alleged\\nthat they are victims of accidents involving GM vehicles (or\\nare the family members or estates of such individuals), and\\nhad personal injury claims against Old GM. The plaintiffs\\nalleged that under Michigan law they possessed successor\\nliability claims at the time the § 363 sale closed. Michigan\\nlaw provides that where there is a sale of assets from one\\nentity to another such that there exists “a continuity of en-\\nterprise between a successor and its predecessor[,] . . . a\\nsuccessor [may be forced] to ‘accept the liability with the\\nbenefits’ of such continuity.” Foster v. Cone-Blanchard\\nMach. Co., 597 N.W.2d 506, 510 (Mich. 1999) (quoting\\nTurner v. Bituminous Cas. Co., 244 N.W.2d 873, 883 (Mich.\\n1976)).\\n\\n\\n\\ndiminution in the value of their property. See Penn. Cent.\\nTransp. Co. v. New York City, 438 U.S. 104 (1978).\\n\\fCAMPBELL v. UNITED STATES 5\\n\\n\\n\\n Under Michigan law,\\n a prima facie case of continuity of enterprise exists\\n where the plaintiff establishes the following facts:\\n (1) there is continuation of the seller corporation, so\\n that there is a continuity of management, personnel,\\n physical location, assets, and general business oper-\\n ations of the predecessor corporation; (2) the prede-\\n cessor corporation ceases its ordinary business\\n operations, liquidates, and dissolves as soon as le-\\n gally and practically possible; and (3) the purchasing\\n corporation assumes those liabilities and obligations\\n of the seller ordinarily necessary for the uninter-\\n rupted continuation of normal business operations\\n of the selling corporation.\\nId. The plaintiffs’ complaint here sets out facts supporting\\neach of these factors. Though the government disputes\\nwhether the plaintiffs’ successor liability claims constitute\\na cognizable property interest for the purposes of a Fifth\\nAmendment taking, we assume, without deciding, that\\nthey do.\\n Here, Old GM filed for bankruptcy on June 1, 2009, and\\nfiled a motion seeking court approval to sell substantially\\nall its assets to a new corporation, referred to as “New GM,”\\npursuant to 11 U.S.C. § 363. In re Gen. Motors Corp., 407\\nB.R. 463, 479–80, 483 (Bankr. S.D.N.Y. 2009). To facilitate\\nthe sale, the government provided financing to Old GM for\\nthe bankruptcy and the company’s ongoing operations. See\\nid. at 473, 479. In return, the government received $8.8\\nbillion in debt and preferred stock of New GM and approx-\\nimately 60 percent of its equity. Id. at 482. According to\\nthe bankruptcy court, without the government’s financing,\\nOld GM would have “face[d] immediate liquidation.” Id. at\\n484. According to the plaintiffs’ complaint, “on the eve of\\nOld GM’s bankruptcy filing, the [g]overnment . . . condi-\\ntion[ed] the closing of the [s]ale on the . . . inclusion of . . .\\n\\f6 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nprovision[s]” concerning successor liability. 3 J.A. 1037; see\\nalso Appellant Br. 8.\\n Because of the government’s insistence, according to\\nthe complaint, the proposed sale order thus included “a\\nnumber of provisions making explicit findings that New\\nGM is not subject to [the plaintiffs’] successor liability\\n[claims].” Gen. Motors Corp., 407 B.R. at 500. Those pro-\\nvisions provide, in relevant part:\\n Except for the Assumed Liabilities, . . . the Pur-\\n chased Assets shall be transferred to the Purchaser\\n . . . free and clear of all liens, claims, encum-\\n brances, and other interests of any kind or nature\\n whatsoever . . . including rights or claims based on\\n any successor or transferee liability . . . .\\n [A]ll persons and entities . . . holding liens, claims,\\n encumbrances, and other interests of any kind or\\n nature whatsoever, including rights or claims\\n based on any successor or transferee liability,\\n . . . are forever barred, estopped, and permanently\\n enjoined . . . .\\nJ.A. 449–50, ¶¶ 7, 8 (emphases added). Allegedly, the gov-\\nernment’s financing was “expressly conditioned upon ap-\\nproval of this motion [seeking approval of the proposed sale\\norder] by July 10.” Gen. Motors Corp., 407 B.R. at 484.\\n The bankruptcy court approved Old GM’s proposed\\nsale order on July 5, 2009, and “permit[ted] GM’s assets to\\npass . . . free and clear of successor liability claims.” Id. at\\n505. The court determined that it would not “gamble on\\nthe notion that the U.S. Government didn’t mean it when\\nit said that it would not keep funding GM.” Id. at 493. In\\n\\n\\n\\n 3 This is somewhat inconsistent with other allega-\\ntions of the complaint stating that extinguishment of the\\nplaintiffs’ claims was not important to the government.\\n\\fCAMPBELL v. UNITED STATES 7\\n\\n\\n\\napproving the sale, the court explained that “GM is hope-\\nlessly insolvent” and that “if GM liquidates, there will . . .\\nbe nothing for stockholders . . . [or] unsecured creditors.”\\nId. at 520.\\n Paragraph 70 of the sale order provided that the order\\n“shall be effective as of 12:00 noon, EDT, on Thursday, July\\n9, 2009.” J.A. 475. The § 363 sale closed thereafter on July\\n10, 2009.\\n III\\n The plaintiffs appealed the bankruptcy court’s order to\\nthe United States District Court for the Southern District\\nof New York. See In re Motors Liquidation Co., 428 B.R. 43\\n(S.D.N.Y. 2010). Somewhat confusingly, the district court\\nboth affirmed the bankruptcy court’s decision and deter-\\nmined that the appeal was moot. Id. at 64. Because the\\nplaintiffs had failed to seek a stay of the § 363 sale pending\\nappeal, and the § 363 sale had closed, the court concluded\\nthat the plaintiffs’ appeal was statutorily moot under\\n§ 363(m) of the Bankruptcy Code. Id. at 56–60. The court\\nentered an order stating that “the appeal is denied as moot,\\nand the judgment of the Bankruptcy Court is AFFIRMED.”\\nId. at 64. Though the district court explained that it was\\n“not unsympathetic to the plight of the [plaintiffs],” it noted\\nthat the plaintiffs’ “position in the bankruptcy appears to\\nbe neither better not worse than that of any other unse-\\ncured contingent creditor.” Id. at 63. The district court\\nconcluded that had the § 363 sale not occurred, Old GM\\nwould have proceeded to liquidation and the plaintiffs “and\\nother unsecured creditors would have received nothing.”\\nId. at 63–64.\\n IV\\n The Claims Court dismissed the plaintiffs’ complaint\\non October 30, 2017, without deciding the question of\\nwhether the alleged class should be certified. The court\\nheld that the plaintiffs’ complaint was not timely because\\n\\f8 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nit was not filed within the Tucker Act’s six-year statute of\\nlimitations. In reaching this conclusion, the court stated\\nthat “[t]he complaint’s specific references to government\\ncoercive action . . . all point to activity that predates the\\nSale Order issued by the bankruptcy court.” J.A. 7. In\\nother words, the court determined that the plaintiffs’ tak-\\nings claim had accrued, at the latest, on July 5, 2009—more\\nthan six years before the date on which the plaintiffs filed\\ntheir complaint in the Claims Court.\\n In the alternative, the Claims Court held that dismis-\\nsal was proper because the plaintiffs had failed to identify\\na cognizable property interest. It characterized the plain-\\ntiffs’ successor liability claims as being “too contingent” to\\ncompose a property interest that would be subject to a com-\\npensable taking under the Fifth Amendment. J.A. 18 The\\ncourt explained that the plaintiffs’ purported property in-\\nterests were “entirely contingent upon two discretionary\\nacts of the federal government: (1) a government financial\\nintervention so that a New GM could be created; and[] (2)\\na government intervention so that Old GM could file for\\nbankruptcy requesting a 363 sale.” J.A. 18. The Claims\\nCourt also concluded that the plaintiffs lacked a cognizable\\nproperty interest because § 363 predated the creation of\\nthe plaintiffs’ purported property interests (the successor\\nliability claims) and therefore that the possibility of extin-\\nguishment in bankruptcy “inhered” in the title to the plain-\\ntiffs’ claims. J.A. 22. The court did not reach the question\\nof whether the plaintiffs had sufficiently alleged a diminu-\\ntion in the value of their property as a result of the govern-\\nment’s action.\\n On November 27, 2017, the plaintiffs filed in the\\nClaims Court a combined motion for reconsideration, mo-\\ntion to amend the judgment, and motion for leave to file a\\nsecond amended complaint. The proposed second amended\\ncomplaint, in the words of the Claims Court, “reshaped[d]\\nthe description of plaintiffs’ takings claims and the facts\\nalready alleged” and contained additional factual\\n\\fCAMPBELL v. UNITED STATES 9\\n\\n\\n\\nallegations. J.A. 24. The plaintiffs added facts purporting\\nto demonstrate that government coercion continued after\\nthe bankruptcy court entered the § 363 sale order. The\\nplaintiffs’ complaint essentially added more detail about\\nthe government’s claimed pressure on the bankruptcy\\ncourt and the district court. The Claims Court considered\\nthe alleged facts presented in the plaintiffs’ second\\namended complaint but determined that the revised alle-\\ngations were insufficient to change the result and denied\\nthe motion, reaffirming both of its initial grounds for dis-\\nmissing the plaintiffs’ complaint.\\n The plaintiffs appealed to this court. 4 We have juris-\\ndiction under 28 U.S.C. § 1295(a)(3). We review the Claims\\nCourt’s decision to dismiss for lack of jurisdiction de novo.\\nDiaz v. United States, 853 F.3d 1355, 1357 (Fed. Cir. 2017).\\n DISCUSSION\\n I\\n We agree with the Claims Court that the plaintiffs’\\ncomplaint was untimely insofar as the complaint alleged\\ncoercion of Old GM. The Tucker Act’s statute of limitations\\nprovides that “[e]very claim of which the United States\\nCourt of Federal Claims has jurisdiction shall be barred\\nunless the petition thereon is filed within six years after\\nsuch claim first accrues.” 28 U.S.C. § 2501. The Supreme\\nCourt has held that the Claims Court’s six-year statute of\\nlimitations is jurisdictional. See John R. Sand & Gravel\\nCo. v. United States, 552 U.S. 130 (2008).\\n The plaintiffs here filed their complaint alleging Fifth\\nAmendment takings claims in the trial court on July 9,\\n\\n\\n\\n 4 We note that the government’s brief before this\\ncourt does not comply with Federal Rule of Appellate Pro-\\ncedure 32(a)(1)(A) because it was filed double-, rather than\\nsingle-sided.\\n\\f10 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\n2015. The question is when the alleged taking occurred.\\nThe plaintiffs alleged in their complaint, and reaffirmed\\nduring oral argument, that the primary conduct which\\ncaused the taking was the government’s coercion of Old GM\\nto secure approval from the bankruptcy court of a proposed\\nsale order extinguishing the plaintiffs’ successor liability\\nclaims. Specifically, the plaintiffs alleged that the govern-\\nment “directed” that the sale order “exclude all Personal\\nInjury Products Liability Claims.” J.A. 398, ¶ 83. In other\\nwords, that the government “condition[ed] the closing of\\nthe Sale on inclusion of a provision within the Sale Order\\nthat expressly extinguished the rights of any Personal In-\\njury Claimant to assert successor liability claims against\\nNew GM.” J.A. 403, ¶ 107. The plaintiffs contended that\\nOld GM was “[l]eft with no option but to comply with the\\n[g]overnment’s mandate or face certain liquidation,” and\\nOld GM consequently filed for bankruptcy and submitted\\nto the bankruptcy court a proposed sale order that “left be-\\nhind the [plaintiffs’] Personal Injury Products Liability\\nClaims.” J.A. 399–400, ¶¶ 88, 98. In light of these allega-\\ntions, we focus our accrual analysis on the government’s\\nalleged coercion of Old GM.\\n The parties present three possible dates on which the\\nplaintiffs’ takings claims may have accrued. The govern-\\nment argues that the claims accrued at least on July 5,\\n2009, when the bankruptcy court approved and entered the\\nsale order (“Entry Date”). The plaintiffs argue to the con-\\ntrary that their claims did not accrue until either July 9,\\n2009, the date on which the § 363 sale became effective\\n(“Effective Date”), or July 10, 2009, the date on which the\\n§ 363 sale closed (“Closing Date”). The plaintiffs contend\\nthat the Entry Date cannot constitute the date on which\\ntheir claims accrued because the value of the plaintiffs’ suc-\\ncessor liability claims had not yet been extinguished as of\\nJuly 5, 2009. We conclude that under the plaintiffs’ theory\\nas to the coercion of Old GM, the alleged taking occurred\\n\\fCAMPBELL v. UNITED STATES 11\\n\\n\\n\\non July 1, 2009—when Old GM filed the proposed sale or-\\nder with the bankruptcy court.\\n The standards for claim accrual in physical takings and\\nregulatory takings cases are distinct, and this distinction\\nis important. As the Supreme Court has held, “it [is] inap-\\npropriate to treat cases involving physical takings as con-\\ntrolling precedents for the evaluation of a claim that there\\nhas been a ‘regulatory taking’, and vice versa.” Tahoe-Si-\\nerra Pres. Council, Inc. v. Tahoe Reg’l Planning Agency, 535\\nU.S. 302, 323 (2002) (footnote omitted).\\n In the case of a physical taking, claim accrual is rela-\\ntively simple to pinpoint. “A physical taking generally oc-\\ncurs when the government directly appropriates private\\nproperty or engages in the functional equivalent of a ‘prac-\\ntical ouster of [the owner’s] possession.’” Katzin v. United\\nStates, 908 F.3d 1350, 1361 (Fed. Cir. 2018) (alteration in\\noriginal) (quoting Washoe Cty. v. United States, 319 F.3d\\n1320, 1326 (Fed. Cir. 2003)); see United States v. Dow, 357\\nU.S. 17, 24–25 (1958) (occupation of property signaled\\nclaim accrual, not the later transfer of a deed). For exam-\\nple, in Casitas Municipal Water District v. United States,\\n708 F.3d 1340 (Fed. Cir. 2013), this court held that the\\nplaintiffs’ physical takings claim had accrued not when the\\nNational Marine Fisheries Service issued an opinion,\\nwhich, if followed, would have resulted in the diversion of\\nsome of the plaintiffs’ water, but rather whenever an actual\\ndiversion of water occurred. Id. at 1358–59; see also Nw.\\nLa. Fish & Game Pres. Comm’n v. United States, 446 F.3d\\n1285, 1290–91 (Fed. Cir. 2006) (where the damage of a pur-\\nported physical taking occurs by a continuing process of\\nphysical events, the damages must be “quantifiable and\\npresent”).\\n In the case of a regulatory taking, however, the taking\\nmay occur before the effect of the regulatory action is felt\\nand actual damage to the property interest is entirely de-\\nterminable. As the Supreme Court recently stated in a\\n\\f12 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nregulatory takings case, “a property owner has a claim for\\na violation of the Takings Clause as soon as [the] govern-\\nment takes his property for public use without paying for\\nit” without regard to post-taking remedies that may be\\navailable. Knick v. Township of Scott, 139 S. Ct. 2162, 2170\\n(2019). In other words, “because a taking without compen-\\nsation violates the self-executing Fifth Amendment at the\\ntime of the taking, the property owner can bring a federal\\nsuit at that time.” Id. at 2172. For example, in Branch v.\\nUnited States, 69 F.3d 1571 (Fed. Cir. 1995), a bankruptcy\\ntrustee challenged the FDIC’s assessment of liability and\\nconsequent seizure of a bank’s assets pursuant to the Fi-\\nnancial Institutions Reform, Recovery, and Enforcement\\nAct. Id. at 1574. This court held that “[t]he seizure and\\nclosure of [a] bank, once the bank became insolvent, did not\\nconstitute a taking.” Id. at 1575. Rather, it was the\\n“FDIC’s assessment of liability,” which directly caused the\\ninsolvency, that was the purported taking and therefore\\nmarked the date of accrual. Id.\\n In Goodrich v. United States, 434 F.3d 1329 (Fed. Cir.\\n2006), a regulatory takings case, the question was whether\\nthe issuance of a Forest Service Record of Decision (“ROD”)\\nand a final Environmental Impact Statement (“EIS”) re-\\nsulted in accrual of the plaintiff’s takings claim. Id. at\\n1331. We held that it did. We determined that the ROD\\nand final EIS were sufficiently final to accrue takings lia-\\nbility because they “are final agency statements of official\\nposition that are published [by the agency] only after years\\nof analysis and consultation with affected parties.” Id. at\\n1335. We explained that the claim accrues when the\\nagency action is complete, “regardless of whether damages\\nare ‘complete and fully calculable.’” Id. at 1336 (quoting\\nFallini v. United States, 56 F.3d 1378, 1382–83 (Fed. Cir.\\n1995)). Here, assuming that the plaintiffs’ claims had\\nvalue in the first place, despite the likelihood of no recovery\\nif Old GM had liquidated, the filing of the proposed\\n\\fCAMPBELL v. UNITED STATES 13\\n\\n\\n\\nbankruptcy sale order clearly inflicted an injury on the\\nplaintiffs by diminishing the value of their claimed prop-\\nerty rights.\\n Notably, the court in Goodrich made clear that where\\na regulatory taking is alleged, it is the final decision of the\\ngovernment actor alleged to have caused the taking that\\ntriggers accrual of a takings claim, not the ultimate impact\\nof that decision. In Goodrich, though “it took the Forest\\nService over three years to implement [the ROD] transfer-\\nring Kennedy’s cattle to the Whitetail Allotment”—i.e., for\\nthe government’s decision to be implemented—the court\\nnevertheless determined that the issuance of the ROD, ra-\\nther than the physical appropriation by cattle of water was\\n“a better place to deem any taking occurred.” Id.\\n The plaintiffs argue that the government’s pressure of\\nOld GM was not complete (that is, not “final”) at the Entry\\nDate. According to the plaintiffs, the government could\\nhave changed its mind regarding the plaintiffs’ successor\\nliability claims at any point before the § 363 sale closed on\\nJuly 10, 2009. But the plaintiffs do not cite any authority\\nin support of their theory that a government actor’s ability\\nto change its mind prevents claim accrual. To the contrary,\\nin Ladd v. United States, 630 F.3d 1015, 1023–24 (Fed. Cir.\\n2010); Barclay v. United States, 443 F.3d 1368, 1378 (Fed.\\nCir. 2006), cert. denied, 549 U.S. 1209 (2007); and Caldwell\\nv. United States, 391 F.3d 1226, 1234–35 (Fed. Cir. 2004),\\ncert. denied, 546 U.S. 826 (2005), in the context of an al-\\nleged physical taking effected by an action taken pursuant\\nto the National Trail Systems Act, we held that the govern-\\nment’s issuance of Notices of Interim Trail Use or Aban-\\ndonment (“NITUs”) effected a taking even though the\\nNITUs could have been vacated by the agency or set aside\\nby a court.\\n Similarly, in Cuban Truck & Equipment Co. v. United\\nStates, 333 F.2d 873 (Ct. Cl. 1964), our predecessor court\\nrejected a somewhat similar theory. There the plaintiffs\\n\\f14 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nasserted a takings claim seeking compensation for the\\nvalue of certain vehicles. Id. at 875. The alleged taking\\nwas effected when the German government (at the insist-\\nence of the United States) directed a quasi-public company\\nin possession of the vehicles to “exclude from sale or deliv-\\nery [those vehicles] at its depots” (“the freeze”). Id. at 878\\n(footnote omitted). The takings claim was held to accrue\\non the date of the freeze. The plaintiffs argued for a later\\naccrual date because the German government could have\\nunilaterally decided to lift the freeze. Id. at 879. Our pre-\\ndecessor court held that this possibility did not impact the\\ndate of accrual. Id. Though Ladd, Barclay, Caldwell, and\\nCuban Truck are physical takings cases, we conclude that\\ntheir reasoning is equally applicable in the regulatory tak-\\nings context. The possibility that the government or a third\\nparty would change its mind does not affect the date on\\nwhich the plaintiffs’ takings claims accrued.\\n Such a takings theory, moreover, would be unworkable.\\nAgencies generally have broad power to reconsider their\\ndecisions. Medtronic, Inc. v. Robert Bosch Healthcare Sys.,\\nInc., 839 F.3d 1382, 1385 (Fed. Cir. 2016). Thus, determin-\\ning accrual based on the possibility the government actor\\nwould change its mind would make the date of accrual en-\\ntirely indeterminate in many situations.\\n The plaintiffs also contend that the Entry Date cannot\\nserve as the date of accrual because the order “could have\\nbeen overturned in advance of the ‘Effective Date’ by a\\n‘higher body’ (i.e., the District Court at the July 9, 2009\\nhearing on whether to stay the effectiveness of the Sale Or-\\nder[)].” Appellant Br. 23. We disagree. This argument ig-\\nnores the fact that the government action purported to\\nhave effected a taking is the government’s coercion of Old\\nGM, not the bankruptcy court’s order. Any action by the\\nbankruptcy court or any higher judicial body merely goes\\nto the ultimate effect of the alleged government taking.\\nSuch collateral action does not alter the finality of the gov-\\nernment’s action for the purpose of accrual of a takings\\n\\fCAMPBELL v. UNITED STATES 15\\n\\n\\n\\nclaim. See Palazzolo v. Rhode Island, 533 U.S. 606, 618\\n(2001) (“The central question in resolving the ripeness is-\\nsue . . . is whether petitioner obtained a final decision from\\nthe [government] determining the permitted use for the\\nland.”); Williamson Cty. Reg’l Planning Comm’n v. Hamil-\\nton Bank of Johnson City, 473 U.S. 172, 186 (1985) (“[A]\\nclaim that the application of government regulations ef-\\nfects a taking of a property interest is not ripe until the\\ngovernment entity charged with implementing the regula-\\ntions has reached a final decision regarding the application\\nof the regulations to the property at issue.”). As the Su-\\npreme Court in Williamson made clear, “the finality re-\\nquirement is concerned with whether the initial\\ndecisionmaker has arrived at a definitive position on the\\nissue that inflicts an actual, concrete injury.” 5 473 U.S. at\\n193 (emphasis added).\\n II\\n The plaintiffs also argue that the government effected\\na taking by pressuring the bankruptcy court and the dis-\\ntrict court to approve the proposed sale order, citing color-\\nful language from a book by Steve Rattner (the leader of\\nthe government team responsible for assessing the viabil-\\nity of Old GM’s restructuring plans) that the government’s\\nthreats were “‘the financial equivalent of holding a gun to\\nthe head’ of the courts.” Appellant Br. 5, 12. These actions\\nallegedly occurred during the bankruptcy court’s and dis-\\ntrict court’s consideration of the proposed sale order, i.e.,\\nwithin the limitations period. But the coercion that could\\ngive rise to a regulatory takings claim does not include “co-\\nercion” of the court system by making an argument for a\\nparticular result. It is well established that the Claims\\nCourt “cannot entertain a taking[s] claim that requires the\\ncourt to scrutinize the actions of another tribunal.” Petro-\\n\\n\\n 5 This aspect of Williamson remains good law under\\nKnick. See 138 S. Ct. 2162 (2019).\\n\\f16 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nHunt, L.L.C. v. United States, 862 F.3d 1370, 1386 (Fed.\\nCir. 2017) (alteration in original) (quoting Shinnecock In-\\ndian Nation v. United States, 782 F.3d 1345, 1353 (Fed.\\nCir. 2015)); accord Allustiarte v. United States, 256 F.3d\\n1349, 1351–52 (Fed. Cir. 2001).\\n In Allustiarte, we considered a similar claim that sev-\\neral plaintiffs had suffered a “taking of their property at\\nthe hands of the bankruptcy trustees and courts.” 256 F.3d\\nat 1351. We held that the Claims Court lacked jurisdiction\\nover such an action because it “would require the court to\\nscrutinize the actions of the bankruptcy trustees and\\ncourts” and we noted that “permit[ting such] collateral at-\\ntacks on bankruptcy court judgments would ‘seriously un-\\ndercut[ ] the orderly process of the law.’” Id. at 1351–52\\n(second alteration in original) (quoting Celotex Corp. v. Ed-\\nwards, 514 U.S. 300, 313 (1995)). We explained that “[t]he\\nproper forum for [the plaintiffs’] challenges . . . lies in the\\nNinth Circuit [the appropriate appellate forum in Allusti-\\narte], not the [Claims Court].” Id. at 1352.\\n The same reasoning applies here. The plaintiffs’ alle-\\ngations that the government coerced the bankruptcy court\\nand the district court amount to no more than a mine-run\\nchallenge to the bases for those court’s decisions regarding\\nthe sale order. The supposed coercion here was not extra-\\njudicial, as was the case in A & D. The plaintiffs cannot\\nmaintain a collateral attack on the decisions of the bank-\\nruptcy court and district court on a takings theory. The\\nproper forum for such a challenge is the judicial appellate\\nprocess.\\n To the extent that the plaintiffs argue that there was a\\ntaking because § 363 did not permit the extinguishment of\\ntheir successor liability claims, we disagree that this is a\\ncognizable takings action. As we held in Lion Raisins, Inc.\\nv. United States, 416 F.3d 1356 (Fed. Cir. 2005), a plain-\\ntiff’s claim that “it is entitled to prevail because the agency\\nacted in violation of a statute . . . [does] not give the\\n\\fCAMPBELL v. UNITED STATES 17\\n\\n\\n\\nplaintiff the right to litigate that issue in a takings action\\nrather than in the congressionally mandated . . . review\\nproceeding.” Id. at 1369 (emphasis omitted) (quoting Rith\\nEnergy, Inc. v. United States, 247 F.3d 1355, 1366 (Fed.\\nCir. 2001)); see also Petro-Hunt, 862 F.3d at 1386. 6 We see\\nno difference, in this regard, between a takings claim based\\non an agency action allegedly in violation of a statute and\\nthat of a federal court. Again, the plaintiffs’ remedy based\\non their challenge to the lawfulness of the § 363 sale was\\nto pursue their successor liability claims through the usual\\n\\n\\n 6 We note that the circuits appear to be split on the\\nissue of whether a bankruptcy court has authority to extin-\\nguish successor liability claims pursuant to a § 363 sale.\\nCompare Precision Indus., Inc. v. Qualitech Steel SBQ,\\nLLC, 327 F.3d 537, 545–46 (7th Cir. 2003) (concluding that\\nthe “right to possess . . . property as a lessee qualifies as an\\ninterest for the purposes of section 363(f)” but limiting the\\ndefinition of “interest” to include only a “right to the prop-\\nerty itself” rather than “a right that is connected to or aris-\\ning from the property”); and In re Wolverine Radio Co., 930\\nF.2d 1132, 1147 n.23 (6th Cir. 1991) (questioning whether\\n“general unsecured interests fall within the scope of those\\ninterests that can be discharged pursuant to section\\n363(f)”), with In re Trans World Airlines, Inc., 322 F.3d 283,\\n288–93 (3d Cir. 2003) (citing 3 Collier on Bankruptcy\\n¶ 363.06[1]) (holding that a sale free and clear of successor\\nliability claims based on employment discrimination was\\nauthorized under § 363); and In re Leckie Smokeless Coal\\nCo., 99 F.3d 575, 581–82 (4th Cir. 1996) (affirming sale free\\nand clear of successor liability for claims under Coal Act).\\nSee generally Charles Jordan Tabb, Law of Bankruptcy\\n450–51 (4th ed. 2016) (noting that the “prevailing trend” is\\nthat courts find that a § 363 sale can be made free and clear\\nof successor liability claims but that there is no consensus).\\nWe need not decide this issue because its outcome does not\\naffect our analysis.\\n\\f18 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nappeal process, as they attempted to do. See Motors Liqui-\\ndation Co., 428 B.R. 43.\\n III\\n The plaintiffs briefly contend that, at worst, they\\nshould now be permitted to amend their complaint to cure\\nany deficiency identified by this court as the plaintiffs in A\\n& D were permitted to do. But the plaintiffs here already\\namended their complaint once as of right and then sought\\nleave to amend their complaint again in conjunction with\\ntheir motion requesting that the Claims Court reconsider\\nits dismissal of this case. The Claims Court denied the\\nplaintiffs’ request to amend their complaint, deeming it fu-\\ntile in attempting to overcome the deficiencies on which the\\nClaims Court’s dismissal was based. We agree with the\\nClaims Court and therefore likewise decline the plaintiffs’\\nrequest to amend. The proposed second amended com-\\nplaint simply supplies additional detail about the extent\\nand timing of the alleged pressure on the bankruptcy court\\nand district court. There are no new allegations in the pro-\\nposed second amended complaint that would alter the date\\nof accrual concerning the claim based on coercion of Old\\nGM or that bolster the plaintiffs’ theory concerning coer-\\ncion of the courts.\\n CONCLUSION\\n We conclude that the plaintiffs’ takings claims based\\non the alleged coercion of Old GM accrued when Old GM\\nsubmitted the proposed sale order to the bankruptcy court\\non July 1, 2009, and that the plaintiffs’ complaint in this\\nrespect was untimely because it was filed more than six\\nyears after their claims accrued. With respect to the plain-\\ntiffs’ claims that the government had coerced the bank-\\nruptcy court and the district court, we conclude that the\\nplaintiffs’ claims are not within the Claims Court’s juris-\\ndiction. Under the circumstances, we need not decide the\\nquestion of whether the plaintiffs have sufficiently alleged\\n\\fCAMPBELL v. UNITED STATES 19\\n\\n\\n\\na loss of value of their alleged property interests. Accord-\\ningly, we affirm.\\n AFFIRMED\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4422111/\", \"author_raw\": \"DYK, Circuit Judge.\"}]}"},"judge1":{"kind":"string","value":"NEWMAN"},"judge2":{"kind":"string","value":"LOURIE"},"judge3":{"kind":"string","value":"DYK"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4644858/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: Callan CAMPBELL, James H. Chadwick, Judith Strode Chadwick, Kevin C. Chadwick, Individually and Through His Court-Appointed Administrators, James H. Chadwick and Judith Strode Chadwick, Kevin Junso, Niki Junso, Tyler Junso Estate, Through Kevin Junso, Its Personal Representative, Plaintiffs-Appellants v. UNITED STATES, Defendant-Appellee\nCourt: U.S. Court of Appeals for the Federal Circuit\nDate decided: 2019-08-01\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n for the Federal Circuit\n ______________________\n\n CALLAN CAMPBELL, JAMES H. CHADWICK,\n JUDITH STRODE CHADWICK, KEVIN C.\nCHADWICK, INDIVIDUALLY AND THROUGH HIS\nCOURT-APPOINTED ADMINISTRATORS, JAMES\nH. CHADWICK AND JUDITH STRODE CHADWICK,\n KEVIN JUNSO, NIKI JUNSO, TYLER JUNSO\n ESTATE, THROUGH KEVIN JUNSO, ITS\n PERSONAL REPRESENTATIVE,\n Plaintiffs-Appellants\n\n v.\n\n UNITED STATES,\n Defendant-Appellee\n ______________________\n\n 2018-2014\n ______________________\n\n Appeal from the United States Court of Federal Claims\nin No. 1:15-cv-00717-PEC, Judge Patricia E. Campbell-\nSmith.\n ______________________\n\n Decided: August 1, 2019\n ______________________\n\n STEVEN R. JAKUBOWSKI, Robbins Salomon & Patt, Ltd.,\nChicago, IL, argued for plaintiffs-appellants.\n\n JOHN JACOB TODOR, Commercial Litigation Branch,\nCivil Division, United States Department of Justice,\n\f2 CAMPBELL v. UNITED STATES\n\n\n\n\nWashington, DC, argued for defendant-appellee. Also rep-\nresented by JOSEPH H. HUNT, ROBERT EDWARD KIRSCHMAN,\nJR., FRANKLIN E. WHITE, JR.\n ______________________\n\n Before NEWMAN, LOURIE, and DYK, Circuit Judges.\nDYK, Circuit Judge.\n This case arises out of the 2009 bankruptcy of General\nMotors Corporation (“Old GM”). 1 The plaintiffs compose a\nputative class of individuals who had asserted personal in-\njury claims against Old GM, and whose successor liability\nclaims were extinguished during bankruptcy. Relying on\nour decision in A & D Auto Sales, Inc. v. United States, 748\nF.3d 1142 (Fed. Cir. 2014), the plaintiffs sued the United\nStates on behalf of themselves and others similarly situ-\nated in the Court of Federal Claims (“Claims Court”), al-\nleging that the extinguishment of their claims without just\ncompensation violated the Takings Clause of the Fifth\nAmendment. The Claims Court dismissed the plaintiffs’\nclaims, concluding that they were barred by the statute of\nlimitations and that the plaintiffs had, in any event, failed\nto state a claim. Because we hold, as to the claims alleging\ncoercion of Old GM, that the statute of limitations had run\nwhen the plaintiffs filed their complaint and, with respect\nto the plaintiffs’ other claims, that the Claims Court also\nlacks jurisdiction, we affirm.\n BACKGROUND\n I\n In A & D, a group of former automobile dealerships\nsued the United States, raising Fifth Amendment takings\nclaims based on the extinguishment of the plaintiffs’\n\n\n\n 1 “Old GM” refers to the GM entity in existence prior\nto the sale of assets pursuant to 11 U.S.C. § 363.\n\fCAMPBELL v. UNITED STATES 3\n\n\n\nfranchise agreements with Old GM in a bankruptcy sale\npursuant to 11 U.S.C. § 363. 748 F.3d at 1147. That sec-\ntion gives a bankruptcy trustee the power to use, sell, or\nlease the property of a debtor in bankruptcy. In particular,\n§ 363(f) (the provision at issue here) provides that “[t]he\ntrustee may sell property . . . free and clear of any interest\nin such property of an entity other than the estate” under\ncertain conditions. 11 U.S.C. § 363(f) (emphasis added).\n In A & D, the plaintiffs alleged that the government\nhad conditioned its continued financial assistance to Old\nGM on the company’s submission for approval of a pro-\nposed sale order that terminated the plaintiffs’ franchise\nagreements. 748 F.3d at 1148. The plaintiffs contended\nthat this purported coercion effected a regulatory taking\nunder the Fifth Amendment. Id. at 1149. The Claims\nCourt denied the government’s motion to dismiss for fail-\nure to state a claim but certified the case for interlocutory\nappeal pursuant to 28 U.S.C. § 1292(d)(2). Id. at 1150.\n On appeal, we held that the government may, in some\ncircumstances, be liable for a regulatory taking of property\nwhere the government pressures a third party (there, al-\nlegedly Old GM) to take an “action that affects or elimi-\nnates the property rights of the plaintiff.” Id. at 1153. We\ndetermined that such conduct may give rise to a taking\nwhere the government’s action was “direct and intended”\nand where the “the third party is acting as the govern-\nment’s agent or the government’s influence over the third\nparty was coercive rather than merely persuasive.” Id. at\n1154. We did not decide whether the government’s actions\nwith respect to Old GM were coercive or otherwise satisfied\nthe conditions for takings liability. 2 Id. at 1155–56. We\n\n\n 2 Even if coercion had been established, that would\nmerely make the third party’s action the equivalent of gov-\nernment action. The plaintiffs would still be required to\nengage in the Penn Central analysis and to establish a\n\f4 CAMPBELL v. UNITED STATES\n\n\n\n\nexplained that to state a claim, the plaintiffs needed to\nhave pled that their “property suffered a diminution in\nvalue or a deprivation of economically beneficial use” as a\nresult of the government’s action. Id. at 1157. We deter-\nmined that the plaintiffs’ allegations were insufficient to\nshow that their franchise agreements had value absent the\ngovernment action and remanded to the Claims Court to\npermit the plaintiffs to amend their complaint “to include\nspecific allegations establishing loss of value” and thereaf-\nter to determine whether a compensable taking had oc-\ncurred. Id. at 1158–59.\n II\n Relying on A & D, on July 9, 2015, the plaintiffs here\nsued the government in the Claims Court alleging that the\ngovernment had coerced Old GM to include in its proposed\nbankruptcy sale order provisions extinguishing the plain-\ntiffs’ property interests pursuant to § 363 of the Bank-\nruptcy Code.\n The plaintiffs are a group of individuals who alleged\nthat they are victims of accidents involving GM vehicles (or\nare the family members or estates of such individuals), and\nhad personal injury claims against Old GM. The plaintiffs\nalleged that under Michigan law they possessed successor\nliability claims at the time the § 363 sale closed. Michigan\nlaw provides that where there is a sale of assets from one\nentity to another such that there exists “a continuity of en-\nterprise between a successor and its predecessor[,] . . . a\nsuccessor [may be forced] to ‘accept the liability with the\nbenefits’ of such continuity.” Foster v. Cone-Blanchard\nMach. Co., 597 N.W.2d 506, 510 (Mich. 1999) (quoting\nTurner v. Bituminous Cas. Co., 244 N.W.2d 873, 883 (Mich.\n1976)).\n\n\n\ndiminution in the value of their property. See Penn. Cent.\nTransp. Co. v. New York City, 438 U.S. 104 (1978).\n\fCAMPBELL v. UNITED STATES 5\n\n\n\n Under Michigan law,\n a prima facie case of continuity of enterprise exists\n where the plaintiff establishes the following facts:\n (1) there is continuation of the seller corporation, so\n that there is a continuity of management, personnel,\n physical location, assets, and general business oper-\n ations of the predecessor corporation; (2) the prede-\n cessor corporation ceases its ordinary business\n operations, liquidates, and dissolves as soon as le-\n gally and practically possible; and (3) the purchasing\n corporation assumes those liabilities and obligations\n of the seller ordinarily necessary for the uninter-\n rupted continuation of normal business operations\n of the selling corporation.\nId. The plaintiffs’ complaint here sets out facts supporting\neach of these factors. Though the government disputes\nwhether the plaintiffs’ successor liability claims constitute\na cognizable property interest for the purposes of a Fifth\nAmendment taking, we assume, without deciding, that\nthey do.\n Here, Old GM filed for bankruptcy on June 1, 2009, and\nfiled a motion seeking court approval to sell substantially\nall its assets to a new corporation, referred to as “New GM,”\npursuant to 11 U.S.C. § 363. In re Gen. Motors Corp., 407\nB.R. 463, 479–80, 483 (Bankr. S.D.N.Y. 2009). To facilitate\nthe sale, the government provided financing to Old GM for\nthe bankruptcy and the company’s ongoing operations. See\nid. at 473, 479. In return, the government received $8.8\nbillion in debt and preferred stock of New GM and approx-\nimately 60 percent of its equity. Id. at 482. According to\nthe bankruptcy court, without the government’s financing,\nOld GM would have “face[d] immediate liquidation.” Id. at\n484. According to the plaintiffs’ complaint, “on the eve of\nOld GM’s bankruptcy filing, the [g]overnment . . . condi-\ntion[ed] the closing of the [s]ale on the . . . inclusion of . . .\n\f6 CAMPBELL v. UNITED STATES\n\n\n\n\nprovision[s]” concerning successor liability. 3 J.A. 1037; see\nalso Appellant Br. 8.\n Because of the government’s insistence, according to\nthe complaint, the proposed sale order thus included “a\nnumber of provisions making explicit findings that New\nGM is not subject to [the plaintiffs’] successor liability\n[claims].” Gen. Motors Corp., 407 B.R. at 500. Those pro-\nvisions provide, in relevant part:\n Except for the Assumed Liabilities, . . . the Pur-\n chased Assets shall be transferred to the Purchaser\n . . . free and clear of all liens, claims, encum-\n brances, and other interests of any kind or nature\n whatsoever . . . including rights or claims based on\n any successor or transferee liability . . . .\n [A]ll persons and entities . . . holding liens, claims,\n encumbrances, and other interests of any kind or\n nature whatsoever, including rights or claims\n based on any successor or transferee liability,\n . . . are forever barred, estopped, and permanently\n enjoined . . . .\nJ.A. 449–50, ¶¶ 7, 8 (emphases added). Allegedly, the gov-\nernment’s financing was “expressly conditioned upon ap-\nproval of this motion [seeking approval of the proposed sale\norder] by July 10.” Gen. Motors Corp., 407 B.R. at 484.\n The bankruptcy court approved Old GM’s proposed\nsale order on July 5, 2009, and “permit[ted] GM’s assets to\npass . . . free and clear of successor liability claims.” Id. at\n505. The court determined that it would not “gamble on\nthe notion that the U.S. Government didn’t mean it when\nit said that it would not keep funding GM.” Id. at 493. In\n\n\n\n 3 This is somewhat inconsistent with other allega-\ntions of the complaint stating that extinguishment of the\nplaintiffs’ claims was not important to the government.\n\fCAMPBELL v. UNITED STATES 7\n\n\n\napproving the sale, the court explained that “GM is hope-\nlessly insolvent” and that “if GM liquidates, there will . . .\nbe nothing for stockholders . . . [or] unsecured creditors.”\nId. at 520.\n Paragraph 70 of the sale order provided that the order\n“shall be effective as of 12:00 noon, EDT, on Thursday, July\n9, 2009.” J.A. 475. The § 363 sale closed thereafter on July\n10, 2009.\n III\n The plaintiffs appealed the bankruptcy court’s order to\nthe United States District Court for the Southern District\nof New York. See In re Motors Liquidation Co., 428 B.R. 43\n(S.D.N.Y. 2010). Somewhat confusingly, the district court\nboth affirmed the bankruptcy court’s decision and deter-\nmined that the appeal was moot. Id. at 64. Because the\nplaintiffs had failed to seek a stay of the § 363 sale pending\nappeal, and the § 363 sale had closed, the court concluded\nthat the plaintiffs’ appeal was statutorily moot under\n§ 363(m) of the Bankruptcy Code. Id. at 56–60. The court\nentered an order stating that “the appeal is denied as moot,\nand the judgment of the Bankruptcy Court is AFFIRMED.”\nId. at 64. Though the district court explained that it was\n“not unsympathetic to the plight of the [plaintiffs],” it noted\nthat the plaintiffs’ “position in the bankruptcy appears to\nbe neither better not worse than that of any other unse-\ncured contingent creditor.” Id. at 63. The district court\nconcluded that had the § 363 sale not occurred, Old GM\nwould have proceeded to liquidation and the plaintiffs “and\nother unsecured creditors would have received nothing.”\nId. at 63–64.\n IV\n The Claims Court dismissed the plaintiffs’ complaint\non October 30, 2017, without deciding the question of\nwhether the alleged class should be certified. The court\nheld that the plaintiffs’ complaint was not timely because\n\f8 CAMPBELL v. UNITED STATES\n\n\n\n\nit was not filed within the Tucker Act’s six-year statute of\nlimitations. In reaching this conclusion, the court stated\nthat “[t]he complaint’s specific references to government\ncoercive action . . . all point to activity that predates the\nSale Order issued by the bankruptcy court.” J.A. 7. In\nother words, the court determined that the plaintiffs’ tak-\nings claim had accrued, at the latest, on July 5, 2009—more\nthan six years before the date on which the plaintiffs filed\ntheir complaint in the Claims Court.\n In the alternative, the Claims Court held that dismis-\nsal was proper because the plaintiffs had failed to identify\na cognizable property interest. It characterized the plain-\ntiffs’ successor liability claims as being “too contingent” to\ncompose a property interest that would be subject to a com-\npensable taking under the Fifth Amendment. J.A. 18 The\ncourt explained that the plaintiffs’ purported property in-\nterests were “entirely contingent upon two discretionary\nacts of the federal government: (1) a government financial\nintervention so that a New GM could be created; and[] (2)\na government intervention so that Old GM could file for\nbankruptcy requesting a 363 sale.” J.A. 18. The Claims\nCourt also concluded that the plaintiffs lacked a cognizable\nproperty interest because § 363 predated the creation of\nthe plaintiffs’ purported property interests (the successor\nliability claims) and therefore that the possibility of extin-\nguishment in bankruptcy “inhered” in the title to the plain-\ntiffs’ claims. J.A. 22. The court did not reach the question\nof whether the plaintiffs had sufficiently alleged a diminu-\ntion in the value of their property as a result of the govern-\nment’s action.\n On November 27, 2017, the plaintiffs filed in the\nClaims Court a combined motion for reconsideration, mo-\ntion to amend the judgment, and motion for leave to file a\nsecond amended complaint. The proposed second amended\ncomplaint, in the words of the Claims Court, “reshaped[d]\nthe description of plaintiffs’ takings claims and the facts\nalready alleged” and contained additional factual\n\fCAMPBELL v. UNITED STATES 9\n\n\n\nallegations. J.A. 24. The plaintiffs added facts purporting\nto demonstrate that government coercion continued after\nthe bankruptcy court entered the § 363 sale order. The\nplaintiffs’ complaint essentially added more detail about\nthe government’s claimed pressure on the bankruptcy\ncourt and the district court. The Claims Court considered\nthe alleged facts presented in the plaintiffs’ second\namended complaint but determined that the revised alle-\ngations were insufficient to change the result and denied\nthe motion, reaffirming both of its initial grounds for dis-\nmissing the plaintiffs’ complaint.\n The plaintiffs appealed to this court. 4 We have juris-\ndiction under 28 U.S.C. § 1295(a)(3). We review the Claims\nCourt’s decision to dismiss for lack of jurisdiction de novo.\nDiaz v. United States, 853 F.3d 1355, 1357 (Fed. Cir. 2017).\n DISCUSSION\n I\n We agree with the Claims Court that the plaintiffs’\ncomplaint was untimely insofar as the complaint alleged\ncoercion of Old GM. The Tucker Act’s statute of limitations\nprovides that “[e]very claim of which the United States\nCourt of Federal Claims has jurisdiction shall be barred\nunless the petition thereon is filed within six years after\nsuch claim first accrues.” 28 U.S.C. § 2501. The Supreme\nCourt has held that the Claims Court’s six-year statute of\nlimitations is jurisdictional. See John R. Sand & Gravel\nCo. v. United States, 552 U.S. 130 (2008).\n The plaintiffs here filed their complaint alleging Fifth\nAmendment takings claims in the trial court on July 9,\n\n\n\n 4 We note that the government’s brief before this\ncourt does not comply with Federal Rule of Appellate Pro-\ncedure 32(a)(1)(A) because it was filed double-, rather than\nsingle-sided.\n\f10 CAMPBELL v. UNITED STATES\n\n\n\n\n2015. The question is when the alleged taking occurred.\nThe plaintiffs alleged in their complaint, and reaffirmed\nduring oral argument, that the primary conduct which\ncaused the taking was the government’s coercion of Old GM\nto secure approval from the bankruptcy court of a proposed\nsale order extinguishing the plaintiffs’ successor liability\nclaims. Specifically, the plaintiffs alleged that the govern-\nment “directed” that the sale order “exclude all Personal\nInjury Products Liability Claims.” J.A. 398, ¶ 83. In other\nwords, that the government “condition[ed] the closing of\nthe Sale on inclusion of a provision within the Sale Order\nthat expressly extinguished the rights of any Personal In-\njury Claimant to assert successor liability claims against\nNew GM.” J.A. 403, ¶ 107. The plaintiffs contended that\nOld GM was “[l]eft with no option but to comply with the\n[g]overnment’s mandate or face certain liquidation,” and\nOld GM consequently filed for bankruptcy and submitted\nto the bankruptcy court a proposed sale order that “left be-\nhind the [plaintiffs’] Personal Injury Products Liability\nClaims.” J.A. 399–400, ¶¶ 88, 98. In light of these allega-\ntions, we focus our accrual analysis on the government’s\nalleged coercion of Old GM.\n The parties present three possible dates on which the\nplaintiffs’ takings claims may have accrued. The govern-\nment argues that the claims accrued at least on July 5,\n2009, when the bankruptcy court approved and entered the\nsale order (“Entry Date”). The plaintiffs argue to the con-\ntrary that their claims did not accrue until either July 9,\n2009, the date on which the § 363 sale became effective\n(“Effective Date”), or July 10, 2009, the date on which the\n§ 363 sale closed (“Closing Date”). The plaintiffs contend\nthat the Entry Date cannot constitute the date on which\ntheir claims accrued because the value of the plaintiffs’ suc-\ncessor liability claims had not yet been extinguished as of\nJuly 5, 2009. We conclude that under the plaintiffs’ theory\nas to the coercion of Old GM, the alleged taking occurred\n\fCAMPBELL v. UNITED STATES 11\n\n\n\non July 1, 2009—when Old GM filed the proposed sale or-\nder with the bankruptcy court.\n The standards for claim accrual in physical takings and\nregulatory takings cases are distinct, and this distinction\nis important. As the Supreme Court has held, “it [is] inap-\npropriate to treat cases involving physical takings as con-\ntrolling precedents for the evaluation of a claim that there\nhas been a ‘regulatory taking’, and vice versa.” Tahoe-Si-\nerra Pres. Council, Inc. v. Tahoe Reg’l Planning Agency, 535\nU.S. 302, 323 (2002) (footnote omitted).\n In the case of a physical taking, claim accrual is rela-\ntively simple to pinpoint. “A physical taking generally oc-\ncurs when the government directly appropriates private\nproperty or engages in the functional equivalent of a ‘prac-\ntical ouster of [the owner’s] possession.’” Katzin v. United\nStates, 908 F.3d 1350, 1361 (Fed. Cir. 2018) (alteration in\noriginal) (quoting Washoe Cty. v. United States, 319 F.3d\n1320, 1326 (Fed. Cir. 2003)); see United States v. Dow, 357\nU.S. 17, 24–25 (1958) (occupation of property signaled\nclaim accrual, not the later transfer of a deed). For exam-\nple, in Casitas Municipal Water District v. United States,\n708 F.3d 1340 (Fed. Cir. 2013), this court held that the\nplaintiffs’ physical takings claim had accrued not when the\nNational Marine Fisheries Service issued an opinion,\nwhich, if followed, would have resulted in the diversion of\nsome of the plaintiffs’ water, but rather whenever an actual\ndiversion of water occurred. Id. at 1358–59; see also Nw.\nLa. Fish & Game Pres. Comm’n v. United States, 446 F.3d\n1285, 1290–91 (Fed. Cir. 2006) (where the damage of a pur-\nported physical taking occurs by a continuing process of\nphysical events, the damages must be “quantifiable and\npresent”).\n In the case of a regulatory taking, however, the taking\nmay occur before the effect of the regulatory action is felt\nand actual damage to the property interest is entirely de-\nterminable. As the Supreme Court recently stated in a\n\f12 CAMPBELL v. UNITED STATES\n\n\n\n\nregulatory takings case, “a property owner has a claim for\na violation of the Takings Clause as soon as [the] govern-\nment takes his property for public use without paying for\nit” without regard to post-taking remedies that may be\navailable. Knick v. Township of Scott, 139 S. Ct. 2162, 2170\n(2019). In other words, “because a taking without compen-\nsation violates the self-executing Fifth Amendment at the\ntime of the taking, the property owner can bring a federal\nsuit at that time.” Id. at 2172. For example, in Branch v.\nUnited States, 69 F.3d 1571 (Fed. Cir. 1995), a bankruptcy\ntrustee challenged the FDIC’s assessment of liability and\nconsequent seizure of a bank’s assets pursuant to the Fi-\nnancial Institutions Reform, Recovery, and Enforcement\nAct. Id. at 1574. This court held that “[t]he seizure and\nclosure of [a] bank, once the bank became insolvent, did not\nconstitute a taking.” Id. at 1575. Rather, it was the\n“FDIC’s assessment of liability,” which directly caused the\ninsolvency, that was the purported taking and therefore\nmarked the date of accrual. Id.\n In Goodrich v. United States, 434 F.3d 1329 (Fed. Cir.\n2006), a regulatory takings case, the question was whether\nthe issuance of a Forest Service Record of Decision (“ROD”)\nand a final Environmental Impact Statement (“EIS”) re-\nsulted in accrual of the plaintiff’s takings claim. Id. at\n1331. We held that it did. We determined that the ROD\nand final EIS were sufficiently final to accrue takings lia-\nbility because they “are final agency statements of official\nposition that are published [by the agency] only after years\nof analysis and consultation with affected parties.” Id. at\n1335. We explained that the claim accrues when the\nagency action is complete, “regardless of whether damages\nare ‘complete and fully calculable.’” Id. at 1336 (quoting\nFallini v. United States, 56 F.3d 1378, 1382–83 (Fed. Cir.\n1995)). Here, assuming that the plaintiffs’ claims had\nvalue in the first place, despite the likelihood of no recovery\nif Old GM had liquidated, the filing of the proposed\n\fCAMPBELL v. UNITED STATES 13\n\n\n\nbankruptcy sale order clearly inflicted an injury on the\nplaintiffs by diminishing the value of their claimed prop-\nerty rights.\n Notably, the court in Goodrich made clear that where\na regulatory taking is alleged, it is the final decision of the\ngovernment actor alleged to have caused the taking that\ntriggers accrual of a takings claim, not the ultimate impact\nof that decision. In Goodrich, though “it took the Forest\nService over three years to implement [the ROD] transfer-\nring Kennedy’s cattle to the Whitetail Allotment”—i.e., for\nthe government’s decision to be implemented—the court\nnevertheless determined that the issuance of the ROD, ra-\nther than the physical appropriation by cattle of water was\n“a better place to deem any taking occurred.” Id.\n The plaintiffs argue that the government’s pressure of\nOld GM was not complete (that is, not “final”) at the Entry\nDate. According to the plaintiffs, the government could\nhave changed its mind regarding the plaintiffs’ successor\nliability claims at any point before the § 363 sale closed on\nJuly 10, 2009. But the plaintiffs do not cite any authority\nin support of their theory that a government actor’s ability\nto change its mind prevents claim accrual. To the contrary,\nin Ladd v. United States, 630 F.3d 1015, 1023–24 (Fed. Cir.\n2010); Barclay v. United States, 443 F.3d 1368, 1378 (Fed.\nCir. 2006), cert. denied, 549 U.S. 1209 (2007); and Caldwell\nv. United States, 391 F.3d 1226, 1234–35 (Fed. Cir. 2004),\ncert. denied, 546 U.S. 826 (2005), in the context of an al-\nleged physical taking effected by an action taken pursuant\nto the National Trail Systems Act, we held that the govern-\nment’s issuance of Notices of Interim Trail Use or Aban-\ndonment (“NITUs”) effected a taking even though the\nNITUs could have been vacated by the agency or set aside\nby a court.\n Similarly, in Cuban Truck & Equipment Co. v. United\nStates, 333 F.2d 873 (Ct. Cl. 1964), our predecessor court\nrejected a somewhat similar theory. There the plaintiffs\n\f14 CAMPBELL v. UNITED STATES\n\n\n\n\nasserted a takings claim seeking compensation for the\nvalue of certain vehicles. Id. at 875. The alleged taking\nwas effected when the German government (at the insist-\nence of the United States) directed a quasi-public company\nin possession of the vehicles to “exclude from sale or deliv-\nery [those vehicles] at its depots” (“the freeze”). Id. at 878\n(footnote omitted). The takings claim was held to accrue\non the date of the freeze. The plaintiffs argued for a later\naccrual date because the German government could have\nunilaterally decided to lift the freeze. Id. at 879. Our pre-\ndecessor court held that this possibility did not impact the\ndate of accrual. Id. Though Ladd, Barclay, Caldwell, and\nCuban Truck are physical takings cases, we conclude that\ntheir reasoning is equally applicable in the regulatory tak-\nings context. The possibility that the government or a third\nparty would change its mind does not affect the date on\nwhich the plaintiffs’ takings claims accrued.\n Such a takings theory, moreover, would be unworkable.\nAgencies generally have broad power to reconsider their\ndecisions. Medtronic, Inc. v. Robert Bosch Healthcare Sys.,\nInc., 839 F.3d 1382, 1385 (Fed. Cir. 2016). Thus, determin-\ning accrual based on the possibility the government actor\nwould change its mind would make the date of accrual en-\ntirely indeterminate in many situations.\n The plaintiffs also contend that the Entry Date cannot\nserve as the date of accrual because the order “could have\nbeen overturned in advance of the ‘Effective Date’ by a\n‘higher body’ (i.e., the District Court at the July 9, 2009\nhearing on whether to stay the effectiveness of the Sale Or-\nder[)].” Appellant Br. 23. We disagree. This argument ig-\nnores the fact that the government action purported to\nhave effected a taking is the government’s coercion of Old\nGM, not the bankruptcy court’s order. Any action by the\nbankruptcy court or any higher judicial body merely goes\nto the ultimate effect of the alleged government taking.\nSuch collateral action does not alter the finality of the gov-\nernment’s action for the purpose of accrual of a takings\n\fCAMPBELL v. UNITED STATES 15\n\n\n\nclaim. See Palazzolo v. Rhode Island, 533 U.S. 606, 618\n(2001) (“The central question in resolving the ripeness is-\nsue . . . is whether petitioner obtained a final decision from\nthe [government] determining the permitted use for the\nland.”); Williamson Cty. Reg’l Planning Comm’n v. Hamil-\nton Bank of Johnson City, 473 U.S. 172, 186 (1985) (“[A]\nclaim that the application of government regulations ef-\nfects a taking of a property interest is not ripe until the\ngovernment entity charged with implementing the regula-\ntions has reached a final decision regarding the application\nof the regulations to the property at issue.”). As the Su-\npreme Court in Williamson made clear, “the finality re-\nquirement is concerned with whether the initial\ndecisionmaker has arrived at a definitive position on the\nissue that inflicts an actual, concrete injury.” 5 473 U.S. at\n193 (emphasis added).\n II\n The plaintiffs also argue that the government effected\na taking by pressuring the bankruptcy court and the dis-\ntrict court to approve the proposed sale order, citing color-\nful language from a book by Steve Rattner (the leader of\nthe government team responsible for assessing the viabil-\nity of Old GM’s restructuring plans) that the government’s\nthreats were “‘the financial equivalent of holding a gun to\nthe head’ of the courts.” Appellant Br. 5, 12. These actions\nallegedly occurred during the bankruptcy court’s and dis-\ntrict court’s consideration of the proposed sale order, i.e.,\nwithin the limitations period. But the coercion that could\ngive rise to a regulatory takings claim does not include “co-\nercion” of the court system by making an argument for a\nparticular result. It is well established that the Claims\nCourt “cannot entertain a taking[s] claim that requires the\ncourt to scrutinize the actions of another tribunal.” Petro-\n\n\n 5 This aspect of Williamson remains good law under\nKnick. See 138 S. Ct. 2162 (2019).\n\f16 CAMPBELL v. UNITED STATES\n\n\n\n\nHunt, L.L.C. v. United States, 862 F.3d 1370, 1386 (Fed.\nCir. 2017) (alteration in original) (quoting Shinnecock In-\ndian Nation v. United States, 782 F.3d 1345, 1353 (Fed.\nCir. 2015)); accord Allustiarte v. United States, 256 F.3d\n1349, 1351–52 (Fed. Cir. 2001).\n In Allustiarte, we considered a similar claim that sev-\neral plaintiffs had suffered a “taking of their property at\nthe hands of the bankruptcy trustees and courts.” 256 F.3d\nat 1351. We held that the Claims Court lacked jurisdiction\nover such an action because it “would require the court to\nscrutinize the actions of the bankruptcy trustees and\ncourts” and we noted that “permit[ting such] collateral at-\ntacks on bankruptcy court judgments would ‘seriously un-\ndercut[ ] the orderly process of the law.’” Id. at 1351–52\n(second alteration in original) (quoting Celotex Corp. v. Ed-\nwards, 514 U.S. 300, 313 (1995)). We explained that “[t]he\nproper forum for [the plaintiffs’] challenges . . . lies in the\nNinth Circuit [the appropriate appellate forum in Allusti-\narte], not the [Claims Court].” Id. at 1352.\n The same reasoning applies here. The plaintiffs’ alle-\ngations that the government coerced the bankruptcy court\nand the district court amount to no more than a mine-run\nchallenge to the bases for those court’s decisions regarding\nthe sale order. The supposed coercion here was not extra-\njudicial, as was the case in A & D. The plaintiffs cannot\nmaintain a collateral attack on the decisions of the bank-\nruptcy court and district court on a takings theory. The\nproper forum for such a challenge is the judicial appellate\nprocess.\n To the extent that the plaintiffs argue that there was a\ntaking because § 363 did not permit the extinguishment of\ntheir successor liability claims, we disagree that this is a\ncognizable takings action. As we held in Lion Raisins, Inc.\nv. United States, 416 F.3d 1356 (Fed. Cir. 2005), a plain-\ntiff’s claim that “it is entitled to prevail because the agency\nacted in violation of a statute . . . [does] not give the\n\fCAMPBELL v. UNITED STATES 17\n\n\n\nplaintiff the right to litigate that issue in a takings action\nrather than in the congressionally mandated . . . review\nproceeding.” Id. at 1369 (emphasis omitted) (quoting Rith\nEnergy, Inc. v. United States, 247 F.3d 1355, 1366 (Fed.\nCir. 2001)); see also Petro-Hunt, 862 F.3d at 1386. 6 We see\nno difference, in this regard, between a takings claim based\non an agency action allegedly in violation of a statute and\nthat of a federal court. Again, the plaintiffs’ remedy based\non their challenge to the lawfulness of the § 363 sale was\nto pursue their successor liability claims through the usual\n\n\n 6 We note that the circuits appear to be split on the\nissue of whether a bankruptcy court has authority to extin-\nguish successor liability claims pursuant to a § 363 sale.\nCompare Precision Indus., Inc. v. Qualitech Steel SBQ,\nLLC, 327 F.3d 537, 545–46 (7th Cir. 2003) (concluding that\nthe “right to possess . . . property as a lessee qualifies as an\ninterest for the purposes of section 363(f)” but limiting the\ndefinition of “interest” to include only a “right to the prop-\nerty itself” rather than “a right that is connected to or aris-\ning from the property”); and In re Wolverine Radio Co., 930\nF.2d 1132, 1147 n.23 (6th Cir. 1991) (questioning whether\n“general unsecured interests fall within the scope of those\ninterests that can be discharged pursuant to section\n363(f)”), with In re Trans World Airlines, Inc., 322 F.3d 283,\n288–93 (3d Cir. 2003) (citing 3 Collier on Bankruptcy\n¶ 363.06[1]) (holding that a sale free and clear of successor\nliability claims based on employment discrimination was\nauthorized under § 363); and In re Leckie Smokeless Coal\nCo., 99 F.3d 575, 581–82 (4th Cir. 1996) (affirming sale free\nand clear of successor liability for claims under Coal Act).\nSee generally Charles Jordan Tabb, Law of Bankruptcy\n450–51 (4th ed. 2016) (noting that the “prevailing trend” is\nthat courts find that a § 363 sale can be made free and clear\nof successor liability claims but that there is no consensus).\nWe need not decide this issue because its outcome does not\naffect our analysis.\n\f18 CAMPBELL v. UNITED STATES\n\n\n\n\nappeal process, as they attempted to do. See Motors Liqui-\ndation Co., 428 B.R. 43.\n III\n The plaintiffs briefly contend that, at worst, they\nshould now be permitted to amend their complaint to cure\nany deficiency identified by this court as the plaintiffs in A\n& D were permitted to do. But the plaintiffs here already\namended their complaint once as of right and then sought\nleave to amend their complaint again in conjunction with\ntheir motion requesting that the Claims Court reconsider\nits dismissal of this case. The Claims Court denied the\nplaintiffs’ request to amend their complaint, deeming it fu-\ntile in attempting to overcome the deficiencies on which the\nClaims Court’s dismissal was based. We agree with the\nClaims Court and therefore likewise decline the plaintiffs’\nrequest to amend. The proposed second amended com-\nplaint simply supplies additional detail about the extent\nand timing of the alleged pressure on the bankruptcy court\nand district court. There are no new allegations in the pro-\nposed second amended complaint that would alter the date\nof accrual concerning the claim based on coercion of Old\nGM or that bolster the plaintiffs’ theory concerning coer-\ncion of the courts.\n CONCLUSION\n We conclude that the plaintiffs’ takings claims based\non the alleged coercion of Old GM accrued when Old GM\nsubmitted the proposed sale order to the bankruptcy court\non July 1, 2009, and that the plaintiffs’ complaint in this\nrespect was untimely because it was filed more than six\nyears after their claims accrued. With respect to the plain-\ntiffs’ claims that the government had coerced the bank-\nruptcy court and the district court, we conclude that the\nplaintiffs’ claims are not within the Claims Court’s juris-\ndiction. Under the circumstances, we need not decide the\nquestion of whether the plaintiffs have sufficiently alleged\n\fCAMPBELL v. UNITED STATES 19\n\n\n\na loss of value of their alleged property interests. Accord-\ningly, we affirm.\n AFFIRMED\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: Callan CAMPBELL, James H. Chadwick, Judith Strode Chadwick, Kevin C. Chadwick, Individually and Through His Court-Appointed Administrators, James H. Chadwick and Judith Strode Chadwick, Kevin Junso, Niki Junso, Tyler Junso Estate, Through Kevin Junso, Its Personal Representative, Plaintiffs-Appellants v. UNITED STATES, Defendant-Appellee\\nCourt: U.S. Court of Appeals for the Federal Circuit\\nDate decided: 2019-08-01\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n CALLAN CAMPBELL, JAMES H. CHADWICK,\\n JUDITH STRODE CHADWICK, KEVIN C.\\nCHADWICK, INDIVIDUALLY AND THROUGH HIS\\nCOURT-APPOINTED ADMINISTRATORS, JAMES\\nH. CHADWICK AND JUDITH STRODE CHADWICK,\\n KEVIN JUNSO, NIKI JUNSO, TYLER JUNSO\\n ESTATE, THROUGH KEVIN JUNSO, ITS\\n PERSONAL REPRESENTATIVE,\\n Plaintiffs-Appellants\\n\\n v.\\n\\n UNITED STATES,\\n Defendant-Appellee\\n ______________________\\n\\n 2018-2014\\n ______________________\\n\\n Appeal from the United States Court of Federal Claims\\nin No. 1:15-cv-00717-PEC, Judge Patricia E. Campbell-\\nSmith.\\n ______________________\\n\\n Decided: August 1, 2019\\n ______________________\\n\\n STEVEN R. JAKUBOWSKI, Robbins Salomon & Patt, Ltd.,\\nChicago, IL, argued for plaintiffs-appellants.\\n\\n JOHN JACOB TODOR, Commercial Litigation Branch,\\nCivil Division, United States Department of Justice,\\n\\f2 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nWashington, DC, argued for defendant-appellee. Also rep-\\nresented by JOSEPH H. HUNT, ROBERT EDWARD KIRSCHMAN,\\nJR., FRANKLIN E. WHITE, JR.\\n ______________________\\n\\n Before NEWMAN, LOURIE, and DYK, Circuit Judges.\\nDYK, Circuit Judge.\\n This case arises out of the 2009 bankruptcy of General\\nMotors Corporation (“Old GM”). 1 The plaintiffs compose a\\nputative class of individuals who had asserted personal in-\\njury claims against Old GM, and whose successor liability\\nclaims were extinguished during bankruptcy. Relying on\\nour decision in A & D Auto Sales, Inc. v. United States, 748\\nF.3d 1142 (Fed. Cir. 2014), the plaintiffs sued the United\\nStates on behalf of themselves and others similarly situ-\\nated in the Court of Federal Claims (“Claims Court”), al-\\nleging that the extinguishment of their claims without just\\ncompensation violated the Takings Clause of the Fifth\\nAmendment. The Claims Court dismissed the plaintiffs’\\nclaims, concluding that they were barred by the statute of\\nlimitations and that the plaintiffs had, in any event, failed\\nto state a claim. Because we hold, as to the claims alleging\\ncoercion of Old GM, that the statute of limitations had run\\nwhen the plaintiffs filed their complaint and, with respect\\nto the plaintiffs’ other claims, that the Claims Court also\\nlacks jurisdiction, we affirm.\\n BACKGROUND\\n I\\n In A & D, a group of former automobile dealerships\\nsued the United States, raising Fifth Amendment takings\\nclaims based on the extinguishment of the plaintiffs’\\n\\n\\n\\n 1 “Old GM” refers to the GM entity in existence prior\\nto the sale of assets pursuant to 11 U.S.C. § 363.\\n\\fCAMPBELL v. UNITED STATES 3\\n\\n\\n\\nfranchise agreements with Old GM in a bankruptcy sale\\npursuant to 11 U.S.C. § 363. 748 F.3d at 1147. That sec-\\ntion gives a bankruptcy trustee the power to use, sell, or\\nlease the property of a debtor in bankruptcy. In particular,\\n§ 363(f) (the provision at issue here) provides that “[t]he\\ntrustee may sell property . . . free and clear of any interest\\nin such property of an entity other than the estate” under\\ncertain conditions. 11 U.S.C. § 363(f) (emphasis added).\\n In A & D, the plaintiffs alleged that the government\\nhad conditioned its continued financial assistance to Old\\nGM on the company’s submission for approval of a pro-\\nposed sale order that terminated the plaintiffs’ franchise\\nagreements. 748 F.3d at 1148. The plaintiffs contended\\nthat this purported coercion effected a regulatory taking\\nunder the Fifth Amendment. Id. at 1149. The Claims\\nCourt denied the government’s motion to dismiss for fail-\\nure to state a claim but certified the case for interlocutory\\nappeal pursuant to 28 U.S.C. § 1292(d)(2). Id. at 1150.\\n On appeal, we held that the government may, in some\\ncircumstances, be liable for a regulatory taking of property\\nwhere the government pressures a third party (there, al-\\nlegedly Old GM) to take an “action that affects or elimi-\\nnates the property rights of the plaintiff.” Id. at 1153. We\\ndetermined that such conduct may give rise to a taking\\nwhere the government’s action was “direct and intended”\\nand where the “the third party is acting as the govern-\\nment’s agent or the government’s influence over the third\\nparty was coercive rather than merely persuasive.” Id. at\\n1154. We did not decide whether the government’s actions\\nwith respect to Old GM were coercive or otherwise satisfied\\nthe conditions for takings liability. 2 Id. at 1155–56. We\\n\\n\\n 2 Even if coercion had been established, that would\\nmerely make the third party’s action the equivalent of gov-\\nernment action. The plaintiffs would still be required to\\nengage in the Penn Central analysis and to establish a\\n\\f4 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nexplained that to state a claim, the plaintiffs needed to\\nhave pled that their “property suffered a diminution in\\nvalue or a deprivation of economically beneficial use” as a\\nresult of the government’s action. Id. at 1157. We deter-\\nmined that the plaintiffs’ allegations were insufficient to\\nshow that their franchise agreements had value absent the\\ngovernment action and remanded to the Claims Court to\\npermit the plaintiffs to amend their complaint “to include\\nspecific allegations establishing loss of value” and thereaf-\\nter to determine whether a compensable taking had oc-\\ncurred. Id. at 1158–59.\\n II\\n Relying on A & D, on July 9, 2015, the plaintiffs here\\nsued the government in the Claims Court alleging that the\\ngovernment had coerced Old GM to include in its proposed\\nbankruptcy sale order provisions extinguishing the plain-\\ntiffs’ property interests pursuant to § 363 of the Bank-\\nruptcy Code.\\n The plaintiffs are a group of individuals who alleged\\nthat they are victims of accidents involving GM vehicles (or\\nare the family members or estates of such individuals), and\\nhad personal injury claims against Old GM. The plaintiffs\\nalleged that under Michigan law they possessed successor\\nliability claims at the time the § 363 sale closed. Michigan\\nlaw provides that where there is a sale of assets from one\\nentity to another such that there exists “a continuity of en-\\nterprise between a successor and its predecessor[,] . . . a\\nsuccessor [may be forced] to ‘accept the liability with the\\nbenefits’ of such continuity.” Foster v. Cone-Blanchard\\nMach. Co., 597 N.W.2d 506, 510 (Mich. 1999) (quoting\\nTurner v. Bituminous Cas. Co., 244 N.W.2d 873, 883 (Mich.\\n1976)).\\n\\n\\n\\ndiminution in the value of their property. See Penn. Cent.\\nTransp. Co. v. New York City, 438 U.S. 104 (1978).\\n\\fCAMPBELL v. UNITED STATES 5\\n\\n\\n\\n Under Michigan law,\\n a prima facie case of continuity of enterprise exists\\n where the plaintiff establishes the following facts:\\n (1) there is continuation of the seller corporation, so\\n that there is a continuity of management, personnel,\\n physical location, assets, and general business oper-\\n ations of the predecessor corporation; (2) the prede-\\n cessor corporation ceases its ordinary business\\n operations, liquidates, and dissolves as soon as le-\\n gally and practically possible; and (3) the purchasing\\n corporation assumes those liabilities and obligations\\n of the seller ordinarily necessary for the uninter-\\n rupted continuation of normal business operations\\n of the selling corporation.\\nId. The plaintiffs’ complaint here sets out facts supporting\\neach of these factors. Though the government disputes\\nwhether the plaintiffs’ successor liability claims constitute\\na cognizable property interest for the purposes of a Fifth\\nAmendment taking, we assume, without deciding, that\\nthey do.\\n Here, Old GM filed for bankruptcy on June 1, 2009, and\\nfiled a motion seeking court approval to sell substantially\\nall its assets to a new corporation, referred to as “New GM,”\\npursuant to 11 U.S.C. § 363. In re Gen. Motors Corp., 407\\nB.R. 463, 479–80, 483 (Bankr. S.D.N.Y. 2009). To facilitate\\nthe sale, the government provided financing to Old GM for\\nthe bankruptcy and the company’s ongoing operations. See\\nid. at 473, 479. In return, the government received $8.8\\nbillion in debt and preferred stock of New GM and approx-\\nimately 60 percent of its equity. Id. at 482. According to\\nthe bankruptcy court, without the government’s financing,\\nOld GM would have “face[d] immediate liquidation.” Id. at\\n484. According to the plaintiffs’ complaint, “on the eve of\\nOld GM’s bankruptcy filing, the [g]overnment . . . condi-\\ntion[ed] the closing of the [s]ale on the . . . inclusion of . . .\\n\\f6 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nprovision[s]” concerning successor liability. 3 J.A. 1037; see\\nalso Appellant Br. 8.\\n Because of the government’s insistence, according to\\nthe complaint, the proposed sale order thus included “a\\nnumber of provisions making explicit findings that New\\nGM is not subject to [the plaintiffs’] successor liability\\n[claims].” Gen. Motors Corp., 407 B.R. at 500. Those pro-\\nvisions provide, in relevant part:\\n Except for the Assumed Liabilities, . . . the Pur-\\n chased Assets shall be transferred to the Purchaser\\n . . . free and clear of all liens, claims, encum-\\n brances, and other interests of any kind or nature\\n whatsoever . . . including rights or claims based on\\n any successor or transferee liability . . . .\\n [A]ll persons and entities . . . holding liens, claims,\\n encumbrances, and other interests of any kind or\\n nature whatsoever, including rights or claims\\n based on any successor or transferee liability,\\n . . . are forever barred, estopped, and permanently\\n enjoined . . . .\\nJ.A. 449–50, ¶¶ 7, 8 (emphases added). Allegedly, the gov-\\nernment’s financing was “expressly conditioned upon ap-\\nproval of this motion [seeking approval of the proposed sale\\norder] by July 10.” Gen. Motors Corp., 407 B.R. at 484.\\n The bankruptcy court approved Old GM’s proposed\\nsale order on July 5, 2009, and “permit[ted] GM’s assets to\\npass . . . free and clear of successor liability claims.” Id. at\\n505. The court determined that it would not “gamble on\\nthe notion that the U.S. Government didn’t mean it when\\nit said that it would not keep funding GM.” Id. at 493. In\\n\\n\\n\\n 3 This is somewhat inconsistent with other allega-\\ntions of the complaint stating that extinguishment of the\\nplaintiffs’ claims was not important to the government.\\n\\fCAMPBELL v. UNITED STATES 7\\n\\n\\n\\napproving the sale, the court explained that “GM is hope-\\nlessly insolvent” and that “if GM liquidates, there will . . .\\nbe nothing for stockholders . . . [or] unsecured creditors.”\\nId. at 520.\\n Paragraph 70 of the sale order provided that the order\\n“shall be effective as of 12:00 noon, EDT, on Thursday, July\\n9, 2009.” J.A. 475. The § 363 sale closed thereafter on July\\n10, 2009.\\n III\\n The plaintiffs appealed the bankruptcy court’s order to\\nthe United States District Court for the Southern District\\nof New York. See In re Motors Liquidation Co., 428 B.R. 43\\n(S.D.N.Y. 2010). Somewhat confusingly, the district court\\nboth affirmed the bankruptcy court’s decision and deter-\\nmined that the appeal was moot. Id. at 64. Because the\\nplaintiffs had failed to seek a stay of the § 363 sale pending\\nappeal, and the § 363 sale had closed, the court concluded\\nthat the plaintiffs’ appeal was statutorily moot under\\n§ 363(m) of the Bankruptcy Code. Id. at 56–60. The court\\nentered an order stating that “the appeal is denied as moot,\\nand the judgment of the Bankruptcy Court is AFFIRMED.”\\nId. at 64. Though the district court explained that it was\\n“not unsympathetic to the plight of the [plaintiffs],” it noted\\nthat the plaintiffs’ “position in the bankruptcy appears to\\nbe neither better not worse than that of any other unse-\\ncured contingent creditor.” Id. at 63. The district court\\nconcluded that had the § 363 sale not occurred, Old GM\\nwould have proceeded to liquidation and the plaintiffs “and\\nother unsecured creditors would have received nothing.”\\nId. at 63–64.\\n IV\\n The Claims Court dismissed the plaintiffs’ complaint\\non October 30, 2017, without deciding the question of\\nwhether the alleged class should be certified. The court\\nheld that the plaintiffs’ complaint was not timely because\\n\\f8 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nit was not filed within the Tucker Act’s six-year statute of\\nlimitations. In reaching this conclusion, the court stated\\nthat “[t]he complaint’s specific references to government\\ncoercive action . . . all point to activity that predates the\\nSale Order issued by the bankruptcy court.” J.A. 7. In\\nother words, the court determined that the plaintiffs’ tak-\\nings claim had accrued, at the latest, on July 5, 2009—more\\nthan six years before the date on which the plaintiffs filed\\ntheir complaint in the Claims Court.\\n In the alternative, the Claims Court held that dismis-\\nsal was proper because the plaintiffs had failed to identify\\na cognizable property interest. It characterized the plain-\\ntiffs’ successor liability claims as being “too contingent” to\\ncompose a property interest that would be subject to a com-\\npensable taking under the Fifth Amendment. J.A. 18 The\\ncourt explained that the plaintiffs’ purported property in-\\nterests were “entirely contingent upon two discretionary\\nacts of the federal government: (1) a government financial\\nintervention so that a New GM could be created; and[] (2)\\na government intervention so that Old GM could file for\\nbankruptcy requesting a 363 sale.” J.A. 18. The Claims\\nCourt also concluded that the plaintiffs lacked a cognizable\\nproperty interest because § 363 predated the creation of\\nthe plaintiffs’ purported property interests (the successor\\nliability claims) and therefore that the possibility of extin-\\nguishment in bankruptcy “inhered” in the title to the plain-\\ntiffs’ claims. J.A. 22. The court did not reach the question\\nof whether the plaintiffs had sufficiently alleged a diminu-\\ntion in the value of their property as a result of the govern-\\nment’s action.\\n On November 27, 2017, the plaintiffs filed in the\\nClaims Court a combined motion for reconsideration, mo-\\ntion to amend the judgment, and motion for leave to file a\\nsecond amended complaint. The proposed second amended\\ncomplaint, in the words of the Claims Court, “reshaped[d]\\nthe description of plaintiffs’ takings claims and the facts\\nalready alleged” and contained additional factual\\n\\fCAMPBELL v. UNITED STATES 9\\n\\n\\n\\nallegations. J.A. 24. The plaintiffs added facts purporting\\nto demonstrate that government coercion continued after\\nthe bankruptcy court entered the § 363 sale order. The\\nplaintiffs’ complaint essentially added more detail about\\nthe government’s claimed pressure on the bankruptcy\\ncourt and the district court. The Claims Court considered\\nthe alleged facts presented in the plaintiffs’ second\\namended complaint but determined that the revised alle-\\ngations were insufficient to change the result and denied\\nthe motion, reaffirming both of its initial grounds for dis-\\nmissing the plaintiffs’ complaint.\\n The plaintiffs appealed to this court. 4 We have juris-\\ndiction under 28 U.S.C. § 1295(a)(3). We review the Claims\\nCourt’s decision to dismiss for lack of jurisdiction de novo.\\nDiaz v. United States, 853 F.3d 1355, 1357 (Fed. Cir. 2017).\\n DISCUSSION\\n I\\n We agree with the Claims Court that the plaintiffs’\\ncomplaint was untimely insofar as the complaint alleged\\ncoercion of Old GM. The Tucker Act’s statute of limitations\\nprovides that “[e]very claim of which the United States\\nCourt of Federal Claims has jurisdiction shall be barred\\nunless the petition thereon is filed within six years after\\nsuch claim first accrues.” 28 U.S.C. § 2501. The Supreme\\nCourt has held that the Claims Court’s six-year statute of\\nlimitations is jurisdictional. See John R. Sand & Gravel\\nCo. v. United States, 552 U.S. 130 (2008).\\n The plaintiffs here filed their complaint alleging Fifth\\nAmendment takings claims in the trial court on July 9,\\n\\n\\n\\n 4 We note that the government’s brief before this\\ncourt does not comply with Federal Rule of Appellate Pro-\\ncedure 32(a)(1)(A) because it was filed double-, rather than\\nsingle-sided.\\n\\f10 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\n2015. The question is when the alleged taking occurred.\\nThe plaintiffs alleged in their complaint, and reaffirmed\\nduring oral argument, that the primary conduct which\\ncaused the taking was the government’s coercion of Old GM\\nto secure approval from the bankruptcy court of a proposed\\nsale order extinguishing the plaintiffs’ successor liability\\nclaims. Specifically, the plaintiffs alleged that the govern-\\nment “directed” that the sale order “exclude all Personal\\nInjury Products Liability Claims.” J.A. 398, ¶ 83. In other\\nwords, that the government “condition[ed] the closing of\\nthe Sale on inclusion of a provision within the Sale Order\\nthat expressly extinguished the rights of any Personal In-\\njury Claimant to assert successor liability claims against\\nNew GM.” J.A. 403, ¶ 107. The plaintiffs contended that\\nOld GM was “[l]eft with no option but to comply with the\\n[g]overnment’s mandate or face certain liquidation,” and\\nOld GM consequently filed for bankruptcy and submitted\\nto the bankruptcy court a proposed sale order that “left be-\\nhind the [plaintiffs’] Personal Injury Products Liability\\nClaims.” J.A. 399–400, ¶¶ 88, 98. In light of these allega-\\ntions, we focus our accrual analysis on the government’s\\nalleged coercion of Old GM.\\n The parties present three possible dates on which the\\nplaintiffs’ takings claims may have accrued. The govern-\\nment argues that the claims accrued at least on July 5,\\n2009, when the bankruptcy court approved and entered the\\nsale order (“Entry Date”). The plaintiffs argue to the con-\\ntrary that their claims did not accrue until either July 9,\\n2009, the date on which the § 363 sale became effective\\n(“Effective Date”), or July 10, 2009, the date on which the\\n§ 363 sale closed (“Closing Date”). The plaintiffs contend\\nthat the Entry Date cannot constitute the date on which\\ntheir claims accrued because the value of the plaintiffs’ suc-\\ncessor liability claims had not yet been extinguished as of\\nJuly 5, 2009. We conclude that under the plaintiffs’ theory\\nas to the coercion of Old GM, the alleged taking occurred\\n\\fCAMPBELL v. UNITED STATES 11\\n\\n\\n\\non July 1, 2009—when Old GM filed the proposed sale or-\\nder with the bankruptcy court.\\n The standards for claim accrual in physical takings and\\nregulatory takings cases are distinct, and this distinction\\nis important. As the Supreme Court has held, “it [is] inap-\\npropriate to treat cases involving physical takings as con-\\ntrolling precedents for the evaluation of a claim that there\\nhas been a ‘regulatory taking’, and vice versa.” Tahoe-Si-\\nerra Pres. Council, Inc. v. Tahoe Reg’l Planning Agency, 535\\nU.S. 302, 323 (2002) (footnote omitted).\\n In the case of a physical taking, claim accrual is rela-\\ntively simple to pinpoint. “A physical taking generally oc-\\ncurs when the government directly appropriates private\\nproperty or engages in the functional equivalent of a ‘prac-\\ntical ouster of [the owner’s] possession.’” Katzin v. United\\nStates, 908 F.3d 1350, 1361 (Fed. Cir. 2018) (alteration in\\noriginal) (quoting Washoe Cty. v. United States, 319 F.3d\\n1320, 1326 (Fed. Cir. 2003)); see United States v. Dow, 357\\nU.S. 17, 24–25 (1958) (occupation of property signaled\\nclaim accrual, not the later transfer of a deed). For exam-\\nple, in Casitas Municipal Water District v. United States,\\n708 F.3d 1340 (Fed. Cir. 2013), this court held that the\\nplaintiffs’ physical takings claim had accrued not when the\\nNational Marine Fisheries Service issued an opinion,\\nwhich, if followed, would have resulted in the diversion of\\nsome of the plaintiffs’ water, but rather whenever an actual\\ndiversion of water occurred. Id. at 1358–59; see also Nw.\\nLa. Fish & Game Pres. Comm’n v. United States, 446 F.3d\\n1285, 1290–91 (Fed. Cir. 2006) (where the damage of a pur-\\nported physical taking occurs by a continuing process of\\nphysical events, the damages must be “quantifiable and\\npresent”).\\n In the case of a regulatory taking, however, the taking\\nmay occur before the effect of the regulatory action is felt\\nand actual damage to the property interest is entirely de-\\nterminable. As the Supreme Court recently stated in a\\n\\f12 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nregulatory takings case, “a property owner has a claim for\\na violation of the Takings Clause as soon as [the] govern-\\nment takes his property for public use without paying for\\nit” without regard to post-taking remedies that may be\\navailable. Knick v. Township of Scott, 139 S. Ct. 2162, 2170\\n(2019). In other words, “because a taking without compen-\\nsation violates the self-executing Fifth Amendment at the\\ntime of the taking, the property owner can bring a federal\\nsuit at that time.” Id. at 2172. For example, in Branch v.\\nUnited States, 69 F.3d 1571 (Fed. Cir. 1995), a bankruptcy\\ntrustee challenged the FDIC’s assessment of liability and\\nconsequent seizure of a bank’s assets pursuant to the Fi-\\nnancial Institutions Reform, Recovery, and Enforcement\\nAct. Id. at 1574. This court held that “[t]he seizure and\\nclosure of [a] bank, once the bank became insolvent, did not\\nconstitute a taking.” Id. at 1575. Rather, it was the\\n“FDIC’s assessment of liability,” which directly caused the\\ninsolvency, that was the purported taking and therefore\\nmarked the date of accrual. Id.\\n In Goodrich v. United States, 434 F.3d 1329 (Fed. Cir.\\n2006), a regulatory takings case, the question was whether\\nthe issuance of a Forest Service Record of Decision (“ROD”)\\nand a final Environmental Impact Statement (“EIS”) re-\\nsulted in accrual of the plaintiff’s takings claim. Id. at\\n1331. We held that it did. We determined that the ROD\\nand final EIS were sufficiently final to accrue takings lia-\\nbility because they “are final agency statements of official\\nposition that are published [by the agency] only after years\\nof analysis and consultation with affected parties.” Id. at\\n1335. We explained that the claim accrues when the\\nagency action is complete, “regardless of whether damages\\nare ‘complete and fully calculable.’” Id. at 1336 (quoting\\nFallini v. United States, 56 F.3d 1378, 1382–83 (Fed. Cir.\\n1995)). Here, assuming that the plaintiffs’ claims had\\nvalue in the first place, despite the likelihood of no recovery\\nif Old GM had liquidated, the filing of the proposed\\n\\fCAMPBELL v. UNITED STATES 13\\n\\n\\n\\nbankruptcy sale order clearly inflicted an injury on the\\nplaintiffs by diminishing the value of their claimed prop-\\nerty rights.\\n Notably, the court in Goodrich made clear that where\\na regulatory taking is alleged, it is the final decision of the\\ngovernment actor alleged to have caused the taking that\\ntriggers accrual of a takings claim, not the ultimate impact\\nof that decision. In Goodrich, though “it took the Forest\\nService over three years to implement [the ROD] transfer-\\nring Kennedy’s cattle to the Whitetail Allotment”—i.e., for\\nthe government’s decision to be implemented—the court\\nnevertheless determined that the issuance of the ROD, ra-\\nther than the physical appropriation by cattle of water was\\n“a better place to deem any taking occurred.” Id.\\n The plaintiffs argue that the government’s pressure of\\nOld GM was not complete (that is, not “final”) at the Entry\\nDate. According to the plaintiffs, the government could\\nhave changed its mind regarding the plaintiffs’ successor\\nliability claims at any point before the § 363 sale closed on\\nJuly 10, 2009. But the plaintiffs do not cite any authority\\nin support of their theory that a government actor’s ability\\nto change its mind prevents claim accrual. To the contrary,\\nin Ladd v. United States, 630 F.3d 1015, 1023–24 (Fed. Cir.\\n2010); Barclay v. United States, 443 F.3d 1368, 1378 (Fed.\\nCir. 2006), cert. denied, 549 U.S. 1209 (2007); and Caldwell\\nv. United States, 391 F.3d 1226, 1234–35 (Fed. Cir. 2004),\\ncert. denied, 546 U.S. 826 (2005), in the context of an al-\\nleged physical taking effected by an action taken pursuant\\nto the National Trail Systems Act, we held that the govern-\\nment’s issuance of Notices of Interim Trail Use or Aban-\\ndonment (“NITUs”) effected a taking even though the\\nNITUs could have been vacated by the agency or set aside\\nby a court.\\n Similarly, in Cuban Truck & Equipment Co. v. United\\nStates, 333 F.2d 873 (Ct. Cl. 1964), our predecessor court\\nrejected a somewhat similar theory. There the plaintiffs\\n\\f14 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nasserted a takings claim seeking compensation for the\\nvalue of certain vehicles. Id. at 875. The alleged taking\\nwas effected when the German government (at the insist-\\nence of the United States) directed a quasi-public company\\nin possession of the vehicles to “exclude from sale or deliv-\\nery [those vehicles] at its depots” (“the freeze”). Id. at 878\\n(footnote omitted). The takings claim was held to accrue\\non the date of the freeze. The plaintiffs argued for a later\\naccrual date because the German government could have\\nunilaterally decided to lift the freeze. Id. at 879. Our pre-\\ndecessor court held that this possibility did not impact the\\ndate of accrual. Id. Though Ladd, Barclay, Caldwell, and\\nCuban Truck are physical takings cases, we conclude that\\ntheir reasoning is equally applicable in the regulatory tak-\\nings context. The possibility that the government or a third\\nparty would change its mind does not affect the date on\\nwhich the plaintiffs’ takings claims accrued.\\n Such a takings theory, moreover, would be unworkable.\\nAgencies generally have broad power to reconsider their\\ndecisions. Medtronic, Inc. v. Robert Bosch Healthcare Sys.,\\nInc., 839 F.3d 1382, 1385 (Fed. Cir. 2016). Thus, determin-\\ning accrual based on the possibility the government actor\\nwould change its mind would make the date of accrual en-\\ntirely indeterminate in many situations.\\n The plaintiffs also contend that the Entry Date cannot\\nserve as the date of accrual because the order “could have\\nbeen overturned in advance of the ‘Effective Date’ by a\\n‘higher body’ (i.e., the District Court at the July 9, 2009\\nhearing on whether to stay the effectiveness of the Sale Or-\\nder[)].” Appellant Br. 23. We disagree. This argument ig-\\nnores the fact that the government action purported to\\nhave effected a taking is the government’s coercion of Old\\nGM, not the bankruptcy court’s order. Any action by the\\nbankruptcy court or any higher judicial body merely goes\\nto the ultimate effect of the alleged government taking.\\nSuch collateral action does not alter the finality of the gov-\\nernment’s action for the purpose of accrual of a takings\\n\\fCAMPBELL v. UNITED STATES 15\\n\\n\\n\\nclaim. See Palazzolo v. Rhode Island, 533 U.S. 606, 618\\n(2001) (“The central question in resolving the ripeness is-\\nsue . . . is whether petitioner obtained a final decision from\\nthe [government] determining the permitted use for the\\nland.”); Williamson Cty. Reg’l Planning Comm’n v. Hamil-\\nton Bank of Johnson City, 473 U.S. 172, 186 (1985) (“[A]\\nclaim that the application of government regulations ef-\\nfects a taking of a property interest is not ripe until the\\ngovernment entity charged with implementing the regula-\\ntions has reached a final decision regarding the application\\nof the regulations to the property at issue.”). As the Su-\\npreme Court in Williamson made clear, “the finality re-\\nquirement is concerned with whether the initial\\ndecisionmaker has arrived at a definitive position on the\\nissue that inflicts an actual, concrete injury.” 5 473 U.S. at\\n193 (emphasis added).\\n II\\n The plaintiffs also argue that the government effected\\na taking by pressuring the bankruptcy court and the dis-\\ntrict court to approve the proposed sale order, citing color-\\nful language from a book by Steve Rattner (the leader of\\nthe government team responsible for assessing the viabil-\\nity of Old GM’s restructuring plans) that the government’s\\nthreats were “‘the financial equivalent of holding a gun to\\nthe head’ of the courts.” Appellant Br. 5, 12. These actions\\nallegedly occurred during the bankruptcy court’s and dis-\\ntrict court’s consideration of the proposed sale order, i.e.,\\nwithin the limitations period. But the coercion that could\\ngive rise to a regulatory takings claim does not include “co-\\nercion” of the court system by making an argument for a\\nparticular result. It is well established that the Claims\\nCourt “cannot entertain a taking[s] claim that requires the\\ncourt to scrutinize the actions of another tribunal.” Petro-\\n\\n\\n 5 This aspect of Williamson remains good law under\\nKnick. See 138 S. Ct. 2162 (2019).\\n\\f16 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nHunt, L.L.C. v. United States, 862 F.3d 1370, 1386 (Fed.\\nCir. 2017) (alteration in original) (quoting Shinnecock In-\\ndian Nation v. United States, 782 F.3d 1345, 1353 (Fed.\\nCir. 2015)); accord Allustiarte v. United States, 256 F.3d\\n1349, 1351–52 (Fed. Cir. 2001).\\n In Allustiarte, we considered a similar claim that sev-\\neral plaintiffs had suffered a “taking of their property at\\nthe hands of the bankruptcy trustees and courts.” 256 F.3d\\nat 1351. We held that the Claims Court lacked jurisdiction\\nover such an action because it “would require the court to\\nscrutinize the actions of the bankruptcy trustees and\\ncourts” and we noted that “permit[ting such] collateral at-\\ntacks on bankruptcy court judgments would ‘seriously un-\\ndercut[ ] the orderly process of the law.’” Id. at 1351–52\\n(second alteration in original) (quoting Celotex Corp. v. Ed-\\nwards, 514 U.S. 300, 313 (1995)). We explained that “[t]he\\nproper forum for [the plaintiffs’] challenges . . . lies in the\\nNinth Circuit [the appropriate appellate forum in Allusti-\\narte], not the [Claims Court].” Id. at 1352.\\n The same reasoning applies here. The plaintiffs’ alle-\\ngations that the government coerced the bankruptcy court\\nand the district court amount to no more than a mine-run\\nchallenge to the bases for those court’s decisions regarding\\nthe sale order. The supposed coercion here was not extra-\\njudicial, as was the case in A & D. The plaintiffs cannot\\nmaintain a collateral attack on the decisions of the bank-\\nruptcy court and district court on a takings theory. The\\nproper forum for such a challenge is the judicial appellate\\nprocess.\\n To the extent that the plaintiffs argue that there was a\\ntaking because § 363 did not permit the extinguishment of\\ntheir successor liability claims, we disagree that this is a\\ncognizable takings action. As we held in Lion Raisins, Inc.\\nv. United States, 416 F.3d 1356 (Fed. Cir. 2005), a plain-\\ntiff’s claim that “it is entitled to prevail because the agency\\nacted in violation of a statute . . . [does] not give the\\n\\fCAMPBELL v. UNITED STATES 17\\n\\n\\n\\nplaintiff the right to litigate that issue in a takings action\\nrather than in the congressionally mandated . . . review\\nproceeding.” Id. at 1369 (emphasis omitted) (quoting Rith\\nEnergy, Inc. v. United States, 247 F.3d 1355, 1366 (Fed.\\nCir. 2001)); see also Petro-Hunt, 862 F.3d at 1386. 6 We see\\nno difference, in this regard, between a takings claim based\\non an agency action allegedly in violation of a statute and\\nthat of a federal court. Again, the plaintiffs’ remedy based\\non their challenge to the lawfulness of the § 363 sale was\\nto pursue their successor liability claims through the usual\\n\\n\\n 6 We note that the circuits appear to be split on the\\nissue of whether a bankruptcy court has authority to extin-\\nguish successor liability claims pursuant to a § 363 sale.\\nCompare Precision Indus., Inc. v. Qualitech Steel SBQ,\\nLLC, 327 F.3d 537, 545–46 (7th Cir. 2003) (concluding that\\nthe “right to possess . . . property as a lessee qualifies as an\\ninterest for the purposes of section 363(f)” but limiting the\\ndefinition of “interest” to include only a “right to the prop-\\nerty itself” rather than “a right that is connected to or aris-\\ning from the property”); and In re Wolverine Radio Co., 930\\nF.2d 1132, 1147 n.23 (6th Cir. 1991) (questioning whether\\n“general unsecured interests fall within the scope of those\\ninterests that can be discharged pursuant to section\\n363(f)”), with In re Trans World Airlines, Inc., 322 F.3d 283,\\n288–93 (3d Cir. 2003) (citing 3 Collier on Bankruptcy\\n¶ 363.06[1]) (holding that a sale free and clear of successor\\nliability claims based on employment discrimination was\\nauthorized under § 363); and In re Leckie Smokeless Coal\\nCo., 99 F.3d 575, 581–82 (4th Cir. 1996) (affirming sale free\\nand clear of successor liability for claims under Coal Act).\\nSee generally Charles Jordan Tabb, Law of Bankruptcy\\n450–51 (4th ed. 2016) (noting that the “prevailing trend” is\\nthat courts find that a § 363 sale can be made free and clear\\nof successor liability claims but that there is no consensus).\\nWe need not decide this issue because its outcome does not\\naffect our analysis.\\n\\f18 CAMPBELL v. UNITED STATES\\n\\n\\n\\n\\nappeal process, as they attempted to do. See Motors Liqui-\\ndation Co., 428 B.R. 43.\\n III\\n The plaintiffs briefly contend that, at worst, they\\nshould now be permitted to amend their complaint to cure\\nany deficiency identified by this court as the plaintiffs in A\\n& D were permitted to do. But the plaintiffs here already\\namended their complaint once as of right and then sought\\nleave to amend their complaint again in conjunction with\\ntheir motion requesting that the Claims Court reconsider\\nits dismissal of this case. The Claims Court denied the\\nplaintiffs’ request to amend their complaint, deeming it fu-\\ntile in attempting to overcome the deficiencies on which the\\nClaims Court’s dismissal was based. We agree with the\\nClaims Court and therefore likewise decline the plaintiffs’\\nrequest to amend. The proposed second amended com-\\nplaint simply supplies additional detail about the extent\\nand timing of the alleged pressure on the bankruptcy court\\nand district court. There are no new allegations in the pro-\\nposed second amended complaint that would alter the date\\nof accrual concerning the claim based on coercion of Old\\nGM or that bolster the plaintiffs’ theory concerning coer-\\ncion of the courts.\\n CONCLUSION\\n We conclude that the plaintiffs’ takings claims based\\non the alleged coercion of Old GM accrued when Old GM\\nsubmitted the proposed sale order to the bankruptcy court\\non July 1, 2009, and that the plaintiffs’ complaint in this\\nrespect was untimely because it was filed more than six\\nyears after their claims accrued. With respect to the plain-\\ntiffs’ claims that the government had coerced the bank-\\nruptcy court and the district court, we conclude that the\\nplaintiffs’ claims are not within the Claims Court’s juris-\\ndiction. Under the circumstances, we need not decide the\\nquestion of whether the plaintiffs have sufficiently alleged\\n\\fCAMPBELL v. UNITED STATES 19\\n\\n\\n\\na loss of value of their alleged property interests. Accord-\\ningly, we affirm.\\n AFFIRMED\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4193,"cells":{"id":{"kind":"number","value":4645829,"string":"4,645,829"},"name":{"kind":"string","value":"ATEN INTERNATIONAL CO., LTD., Plaintiff-Appellee v. UNICLASS TECHNOLOGY CO., LTD., Electronic Technology Co., Ltd., Airlink 101, Phoebe Micro, Inc., Broadtech International Co., Ltd., Dba Linkskey, Black Box Corporation, Black Box Corporation of Pennsylvania, Defendants-Appellants"},"name_abbreviation":{"kind":"string","value":"Aten International Co., Ltd. v. Uniclass Technology Co., Ltd."},"decision_date":{"kind":"string","value":"2019-08-06"},"docket_number":{"kind":"string","value":"2018-1922"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Federal Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Moore, Wallach, Taranto\", \"parties\": \"\", \"opinions\": [{\"author\": \"MOORE, Circuit Judge.\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n ATEN INTERNATIONAL CO., LTD.,\\n Plaintiff-Appellee\\n\\n v.\\n\\n UNICLASS TECHNOLOGY CO., LTD.,\\nELECTRONIC TECHNOLOGY CO., LTD., AIRLINK\\n 101, PHOEBE MICRO, INC., BROADTECH\\n INTERNATIONAL CO., LTD., DBA LINKSKEY,\\n BLACK BOX CORPORATION, BLACK BOX\\n CORPORATION OF PENNSYLVANIA,\\n Defendants-Appellants\\n ______________________\\n\\n 2018-1922\\n ______________________\\n\\n Appeal from the United States District Court for the\\nCentral District of California in No. 2:15-cv-04424-AG-\\nAJW, Judge Andrew J. Guilford.\\n ______________________\\n\\n Decided: August 6, 2019\\n ______________________\\n\\n EDWARD NAIDICH, Mei & Mark LLP, Washington, DC,\\nargued for plaintiff-appellee. Also represented by LEI MEI,\\nRICHARD HADORN, PHILIP ANDREW RILEY, LAURENCE M.\\nSANDELL.\\n\\n JOSEPH PIA, Pia Anderson Moss Hoyt, Salt Lake City,\\n\\f2 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\nUT, argued for defendants-appellants. Also represented by\\nROBERT AYCOCK.\\n ______________________\\n\\n Before MOORE, WALLACH, and TARANTO, Circuit Judges.\\nMOORE, Circuit Judge.\\n Uniclass Technology Co., Ltd. (“Uniclass”), as well as\\nElectronic Technology Co., Ltd.; Airlink 101; Phoebe Micro,\\nInc.; Broadtech International Co., Ltd., d/b/a Linkskey;\\nBlack Box Corporation; and Black Box Corporation of\\nPennsylvania (collectively the “customer defendants”) (all\\ncollectively “Appellants”) appeal the United States District\\nCourt for the Central District of California’s order denying\\nAppellants’ motion for attorney fees. Because we hold the\\ndistrict court did not abuse its discretion in declining to\\nfind this an exceptional case under 35 U.S.C. § 285, we af-\\nfirm.\\n BACKGROUND\\n Uniclass and ATEN International Co., Ltd. (“ATEN”)\\nare involved in making and selling keyboard-video-mouse\\nswitch systems that allow a user to control multiple com-\\nputers from a single keyboard, video device, and mouse. In\\n2011, Uniclass stopped making payments on a license\\nagreement it entered with ATEN in 2009. In 2014, ATEN\\nsued Uniclass and the customer defendants alleging in-\\nfringement, seeking damages and injunctive relief. Uni-\\nclass moved for summary judgment on ATEN’s lost profits\\ntheory of damages, which the district court granted in April\\n2017. ATEN proceeded to trial based on a reasonable roy-\\nalty theory of damages, under which its expert testified\\nthat the maximum recovery (not including its requested\\ntreble damages) was $678,337. At trial, a jury found that\\nUniclass did not infringe the asserted claims of U.S. Patent\\nNo. 8,589,141 or U.S. Patent No. 7,640,289. It also found\\nthe asserted claims of the ’141 patent invalid. ATEN\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 3\\nCO., LTD.\\n\\n\\nappealed various aspects of the September 2017 jury ver-\\ndict separately in No. 18-1606, which we also decide today.\\n After trial, Uniclass moved to declare this case excep-\\ntional under 35 U.S.C. § 285, arguing that ATEN did not\\nconduct an adequate pre-filing investigation, unnecessarily\\nincreased the costs of claim construction, drastically in-\\ncreased discovery costs by frequently changing counsel and\\ninfringement positions, and engaged in unreasonable liti-\\ngation behavior requiring additional motion practice and\\nleading to an expensive and disproportionate trial. The\\ndistrict court denied the motion.\\n Uniclass timely appealed. We have jurisdiction under\\n28 U.S.C. § 1295(a)(1).\\n DISCUSSION\\n Under § 285, “[t]he court in exceptional cases may\\naward reasonable attorney fees to the prevailing party.”\\n“[A]n ‘exceptional’ case is simply one that stands out from\\nothers with respect to the substantive strength of a party’s\\nlitigating position (considering both the governing law and\\nthe facts of the case) or the unreasonable manner in which\\nthe case was litigated.” Octane Fitness, LLC v. ICON\\nHealth & Fitness, Inc., 572 U.S. 545, 554 (2014). District\\ncourts should determine whether a case is exceptional on a\\ncase-by-case basis, considering the totality of the circum-\\nstances. Id. We review the district court’s § 285 determi-\\nnation for abuse of discretion. Highmark Inc. v. Allcare\\nHealth Mgmt. Sys., Inc., 572 U.S. 559, 561 (2014). “The\\nabuse-of-discretion standard does not preclude an appel-\\nlate court’s correction of a district court’s legal or factual\\nerror: A district court would necessarily abuse its discre-\\ntion if it based its ruling on an erroneous view of the law or\\non a clearly erroneous assessment of the evidence.” Id. at\\n563 n.2 (internal quotation marks omitted).\\n\\f4 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\n I\\n Appellants focus their arguments on appeal on a theory\\nthat the district court erred in not finding this case excep-\\ntional based on ATEN’s disregard for the “foundational pol-\\nicy” of proportionate litigation. Appellants’ Br. 15–16.\\nAppellants summarize ATEN’s expenses as including over\\n$700,000 in expert witness fees alone, without considering\\nother expenses including attorney fees. Appellants argue\\nthat ATEN could recover, at most, $678,337 in reasonable\\nroyalty damages. See J.A. 902 at 7:20–8:4; J.A. 870 at\\n23:8–11; J.A. 923 at 90:14–15. Accordingly, Appellants ar-\\ngue this case is exceptional because the cost of litigating\\nthe case exceeded ATEN’s potential recovery at trial. See,\\ne.g., Appellants’ Br. 15–20, 25–26. We reject this argu-\\nment. We see no error in the court’s legal analysis and no\\nclear error in its fact findings.\\n There is no per se rule that a case is exceptional if liti-\\ngation costs exceed the potential damages. The District of\\nDelaware’s post-Octane decision in EON is persuasive on\\nthis point: “[The court] is unaware of any de minimis ex-\\nception for infringement. It cannot be the case that a plain-\\ntiff may be subjected to monetary sanctions for failing to\\ndrop a case against a defendant if the cost of litigation ex-\\nceeds the potential recovery.” EON Corp. IP Holdings,\\nLLC v. FLO TV Inc., No. 10-812-RGA, 2014 WL 2196418,\\nat *2 (D. Del. May 27, 2014). Litigation strategies motivat-\\ning a patent suit are many, with monetary damages being\\none. We routinely hear cases in which damages are not at\\nissue—for example, in appeals from the International\\nTrade Commission, in some Hatch-Waxman cases, and\\nother cases where only an injunction is sought. A patentee\\nmay also bring suit to deter other competitors from in-\\nfringement, encourage licensing, or test a patent’s ability\\nto withstand validity challenges. Thermolife Int’l LLC v.\\nGNC Corp., 922 F.3d 1347, 1363–64 (Fed. Cir. 2019).\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 5\\nCO., LTD.\\n\\n\\n Moreover, ATEN sought injunctive relief against Uni-\\nclass, its direct competitor. This alone undermines Appel-\\nlants’ argument. The case could have proceeded to trial\\nrequesting only an injunction, and there would be no po-\\ntential damages to compare to ATEN’s expenses. We can-\\nnot conclude that the district court’s decision denying\\nAppellants’ motion for fees under these circumstances\\namounts to an abuse of discretion.\\n Appellants argue the district court should be bound by\\nits statements during trial that the court was troubled by\\ndisproportionate aspects of the case. But Appellants admit\\nthat such comments were largely directed to the amount of\\npublic resources expended on the limited recovery at issue.\\nAppellants’ Br. 17–20 (citing J.A. 955 at 12:16–20 (regard-\\ning public resources); J.A. 1052 at 73:19–25 (regarding\\ntime spent with one witness); J.A. 1053 at 80:17–22 (re-\\ngarding the length of trial compared to damages at issue);\\nJ.A. 1153 at 102:15–25 (regarding ATEN’s attorney fees\\nand the amount of time spent with the witness)). Appel-\\nlants cite an exchange in which the court recognized a sce-\\nnario in which expenditures might understandably exceed\\nrecovery: “I just don’t understand [expert expenditures of]\\n$400,000 in a $650,000 case. When I was a lawyer, if I did\\nthat, I would be fired, or frankly, I’d be working for a client\\nwho said: They are competitors. Spend every penny.” Ap-\\npellants’ Br. 18 (citing J.A. 956 at 13:2–6 (emphasis\\nadded)). We cannot conclude that the court’s comments\\nduring trial regarding public resources render its subse-\\nquent decision denying fees an abuse of discretion. We see\\nno abuse of discretion in the court’s conclusion that under\\nthe totality of the circumstances this litigation was not ex-\\nceptional.\\n II\\n Appellants also argue the district court erred by failing\\nto weigh ATEN’s frivolous damages methodology in its\\nanalysis. Specifically, Appellants argue the district court\\n\\f6 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\nshould have weighed the strength of ATEN’s lost profits\\ntheory (rejected at the summary judgment stage), which\\nAppellants allege was knowingly baseless and brought in\\nbad faith. Appellants’ Br. 22–25.\\n The district court did refer to Uniclass’ argument that\\nATEN presented a weak case for lost profits that relied on\\nguesswork, which it listed among the “whole slew” of Uni-\\nclass’ arguments it considered. J.A. 3. The district court\\nfound there were “perhaps . . . shortcomings” in ATEN’s po-\\nsitions that did not “amount to a showing that [ATEN’s]\\nbehavior was exceptional, amounting to the type of unrea-\\nsonableness discussed in Octane.” J.A. 4.\\n Though ATEN’s lost profits theory was not strong\\nenough to withstand summary judgment, “[a] party’s posi-\\ntion . . . ultimately need not be correct for them to not\\n‘stand[ ] out.’” SFA Sys., LLC v. Newegg Inc., 793 F.3d\\n1344, 1348 (Fed. Cir. 2015) (quoting Octane, 572 U.S. at\\n544). The district court “had no obligation to write an opin-\\nion that reveals [its] assessment of every consideration.”\\nUniv. of Utah v. Max-Planck-Gesellschaft, 851 F.3d 1317,\\n1323 (Fed. Cir. 2017). We can find no abuse of discretion\\nin the district court’s failure to discuss the strength of\\nATEN’s lost profits theory.\\n III\\n To the extent Appellants argue weakness in ATEN’s\\nother substantive positions justifies reversing the district\\ncourt’s determination that this case was not exceptional,\\nwe do not agree. Today in No. 18-1606, we reverse the dis-\\ntrict court’s denial of JMOL because the jury’s finding of\\nanticipation was not supported by substantial evidence.\\nThough we affirm as to noninfringement of the ’141 and\\n’289 patents, we note that ATEN’s primary argument on\\nappeal—that the court improperly gave claim construction\\ndisputes to the jury—was rejected because ATEN did not\\ntimely raise this issue with the court below. We find no\\nabuse of discretion in the district court’s determination\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 7\\nCO., LTD.\\n\\n\\nthat ATEN’s positions were not so objectively unreasonable\\nor exceptionally meritless as to stand out from other cases.\\n We have considered Appellants’ other arguments and\\nfind them without merit.\\n CONCLUSION\\n We hold the district court did not abuse its discretion\\nin declining to find this an exceptional case under 35 U.S.C.\\n§ 285. For the foregoing reasons, we affirm.\\n AFFIRMED\\n COSTS\\n Costs to ATEN.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4423082/\", \"author_raw\": \"MOORE, Circuit Judge.\"}]}"},"judge1":{"kind":"string","value":"MOORE"},"judge2":{"kind":"string","value":"WALLACH"},"judge3":{"kind":"string","value":"TARANTO"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4645829/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: ATEN INTERNATIONAL CO., LTD., Plaintiff-Appellee v. UNICLASS TECHNOLOGY CO., LTD., Electronic Technology Co., Ltd., Airlink 101, Phoebe Micro, Inc., Broadtech International Co., Ltd., Dba Linkskey, Black Box Corporation, Black Box Corporation of Pennsylvania, Defendants-Appellants\nCourt: U.S. Court of Appeals for the Federal Circuit\nDate decided: 2019-08-06\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n for the Federal Circuit\n ______________________\n\n ATEN INTERNATIONAL CO., LTD.,\n Plaintiff-Appellee\n\n v.\n\n UNICLASS TECHNOLOGY CO., LTD.,\nELECTRONIC TECHNOLOGY CO., LTD., AIRLINK\n 101, PHOEBE MICRO, INC., BROADTECH\n INTERNATIONAL CO., LTD., DBA LINKSKEY,\n BLACK BOX CORPORATION, BLACK BOX\n CORPORATION OF PENNSYLVANIA,\n Defendants-Appellants\n ______________________\n\n 2018-1922\n ______________________\n\n Appeal from the United States District Court for the\nCentral District of California in No. 2:15-cv-04424-AG-\nAJW, Judge Andrew J. Guilford.\n ______________________\n\n Decided: August 6, 2019\n ______________________\n\n EDWARD NAIDICH, Mei & Mark LLP, Washington, DC,\nargued for plaintiff-appellee. Also represented by LEI MEI,\nRICHARD HADORN, PHILIP ANDREW RILEY, LAURENCE M.\nSANDELL.\n\n JOSEPH PIA, Pia Anderson Moss Hoyt, Salt Lake City,\n\f2 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\n CO., LTD.\n\n\nUT, argued for defendants-appellants. Also represented by\nROBERT AYCOCK.\n ______________________\n\n Before MOORE, WALLACH, and TARANTO, Circuit Judges.\nMOORE, Circuit Judge.\n Uniclass Technology Co., Ltd. (“Uniclass”), as well as\nElectronic Technology Co., Ltd.; Airlink 101; Phoebe Micro,\nInc.; Broadtech International Co., Ltd., d/b/a Linkskey;\nBlack Box Corporation; and Black Box Corporation of\nPennsylvania (collectively the “customer defendants”) (all\ncollectively “Appellants”) appeal the United States District\nCourt for the Central District of California’s order denying\nAppellants’ motion for attorney fees. Because we hold the\ndistrict court did not abuse its discretion in declining to\nfind this an exceptional case under 35 U.S.C. § 285, we af-\nfirm.\n BACKGROUND\n Uniclass and ATEN International Co., Ltd. (“ATEN”)\nare involved in making and selling keyboard-video-mouse\nswitch systems that allow a user to control multiple com-\nputers from a single keyboard, video device, and mouse. In\n2011, Uniclass stopped making payments on a license\nagreement it entered with ATEN in 2009. In 2014, ATEN\nsued Uniclass and the customer defendants alleging in-\nfringement, seeking damages and injunctive relief. Uni-\nclass moved for summary judgment on ATEN’s lost profits\ntheory of damages, which the district court granted in April\n2017. ATEN proceeded to trial based on a reasonable roy-\nalty theory of damages, under which its expert testified\nthat the maximum recovery (not including its requested\ntreble damages) was $678,337. At trial, a jury found that\nUniclass did not infringe the asserted claims of U.S. Patent\nNo. 8,589,141 or U.S. Patent No. 7,640,289. It also found\nthe asserted claims of the ’141 patent invalid. ATEN\n\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 3\nCO., LTD.\n\n\nappealed various aspects of the September 2017 jury ver-\ndict separately in No. 18-1606, which we also decide today.\n After trial, Uniclass moved to declare this case excep-\ntional under 35 U.S.C. § 285, arguing that ATEN did not\nconduct an adequate pre-filing investigation, unnecessarily\nincreased the costs of claim construction, drastically in-\ncreased discovery costs by frequently changing counsel and\ninfringement positions, and engaged in unreasonable liti-\ngation behavior requiring additional motion practice and\nleading to an expensive and disproportionate trial. The\ndistrict court denied the motion.\n Uniclass timely appealed. We have jurisdiction under\n28 U.S.C. § 1295(a)(1).\n DISCUSSION\n Under § 285, “[t]he court in exceptional cases may\naward reasonable attorney fees to the prevailing party.”\n“[A]n ‘exceptional’ case is simply one that stands out from\nothers with respect to the substantive strength of a party’s\nlitigating position (considering both the governing law and\nthe facts of the case) or the unreasonable manner in which\nthe case was litigated.” Octane Fitness, LLC v. ICON\nHealth & Fitness, Inc., 572 U.S. 545, 554 (2014). District\ncourts should determine whether a case is exceptional on a\ncase-by-case basis, considering the totality of the circum-\nstances. Id. We review the district court’s § 285 determi-\nnation for abuse of discretion. Highmark Inc. v. Allcare\nHealth Mgmt. Sys., Inc., 572 U.S. 559, 561 (2014). “The\nabuse-of-discretion standard does not preclude an appel-\nlate court’s correction of a district court’s legal or factual\nerror: A district court would necessarily abuse its discre-\ntion if it based its ruling on an erroneous view of the law or\non a clearly erroneous assessment of the evidence.” Id. at\n563 n.2 (internal quotation marks omitted).\n\f4 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\n CO., LTD.\n\n\n I\n Appellants focus their arguments on appeal on a theory\nthat the district court erred in not finding this case excep-\ntional based on ATEN’s disregard for the “foundational pol-\nicy” of proportionate litigation. Appellants’ Br. 15–16.\nAppellants summarize ATEN’s expenses as including over\n$700,000 in expert witness fees alone, without considering\nother expenses including attorney fees. Appellants argue\nthat ATEN could recover, at most, $678,337 in reasonable\nroyalty damages. See J.A. 902 at 7:20–8:4; J.A. 870 at\n23:8–11; J.A. 923 at 90:14–15. Accordingly, Appellants ar-\ngue this case is exceptional because the cost of litigating\nthe case exceeded ATEN’s potential recovery at trial. See,\ne.g., Appellants’ Br. 15–20, 25–26. We reject this argu-\nment. We see no error in the court’s legal analysis and no\nclear error in its fact findings.\n There is no per se rule that a case is exceptional if liti-\ngation costs exceed the potential damages. The District of\nDelaware’s post-Octane decision in EON is persuasive on\nthis point: “[The court] is unaware of any de minimis ex-\nception for infringement. It cannot be the case that a plain-\ntiff may be subjected to monetary sanctions for failing to\ndrop a case against a defendant if the cost of litigation ex-\nceeds the potential recovery.” EON Corp. IP Holdings,\nLLC v. FLO TV Inc., No. 10-812-RGA, 2014 WL 2196418,\nat *2 (D. Del. May 27, 2014). Litigation strategies motivat-\ning a patent suit are many, with monetary damages being\none. We routinely hear cases in which damages are not at\nissue—for example, in appeals from the International\nTrade Commission, in some Hatch-Waxman cases, and\nother cases where only an injunction is sought. A patentee\nmay also bring suit to deter other competitors from in-\nfringement, encourage licensing, or test a patent’s ability\nto withstand validity challenges. Thermolife Int’l LLC v.\nGNC Corp., 922 F.3d 1347, 1363–64 (Fed. Cir. 2019).\n\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 5\nCO., LTD.\n\n\n Moreover, ATEN sought injunctive relief against Uni-\nclass, its direct competitor. This alone undermines Appel-\nlants’ argument. The case could have proceeded to trial\nrequesting only an injunction, and there would be no po-\ntential damages to compare to ATEN’s expenses. We can-\nnot conclude that the district court’s decision denying\nAppellants’ motion for fees under these circumstances\namounts to an abuse of discretion.\n Appellants argue the district court should be bound by\nits statements during trial that the court was troubled by\ndisproportionate aspects of the case. But Appellants admit\nthat such comments were largely directed to the amount of\npublic resources expended on the limited recovery at issue.\nAppellants’ Br. 17–20 (citing J.A. 955 at 12:16–20 (regard-\ning public resources); J.A. 1052 at 73:19–25 (regarding\ntime spent with one witness); J.A. 1053 at 80:17–22 (re-\ngarding the length of trial compared to damages at issue);\nJ.A. 1153 at 102:15–25 (regarding ATEN’s attorney fees\nand the amount of time spent with the witness)). Appel-\nlants cite an exchange in which the court recognized a sce-\nnario in which expenditures might understandably exceed\nrecovery: “I just don’t understand [expert expenditures of]\n$400,000 in a $650,000 case. When I was a lawyer, if I did\nthat, I would be fired, or frankly, I’d be working for a client\nwho said: They are competitors. Spend every penny.” Ap-\npellants’ Br. 18 (citing J.A. 956 at 13:2–6 (emphasis\nadded)). We cannot conclude that the court’s comments\nduring trial regarding public resources render its subse-\nquent decision denying fees an abuse of discretion. We see\nno abuse of discretion in the court’s conclusion that under\nthe totality of the circumstances this litigation was not ex-\nceptional.\n II\n Appellants also argue the district court erred by failing\nto weigh ATEN’s frivolous damages methodology in its\nanalysis. Specifically, Appellants argue the district court\n\f6 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\n CO., LTD.\n\n\nshould have weighed the strength of ATEN’s lost profits\ntheory (rejected at the summary judgment stage), which\nAppellants allege was knowingly baseless and brought in\nbad faith. Appellants’ Br. 22–25.\n The district court did refer to Uniclass’ argument that\nATEN presented a weak case for lost profits that relied on\nguesswork, which it listed among the “whole slew” of Uni-\nclass’ arguments it considered. J.A. 3. The district court\nfound there were “perhaps . . . shortcomings” in ATEN’s po-\nsitions that did not “amount to a showing that [ATEN’s]\nbehavior was exceptional, amounting to the type of unrea-\nsonableness discussed in Octane.” J.A. 4.\n Though ATEN’s lost profits theory was not strong\nenough to withstand summary judgment, “[a] party’s posi-\ntion . . . ultimately need not be correct for them to not\n‘stand[ ] out.’” SFA Sys., LLC v. Newegg Inc., 793 F.3d\n1344, 1348 (Fed. Cir. 2015) (quoting Octane, 572 U.S. at\n544). The district court “had no obligation to write an opin-\nion that reveals [its] assessment of every consideration.”\nUniv. of Utah v. Max-Planck-Gesellschaft, 851 F.3d 1317,\n1323 (Fed. Cir. 2017). We can find no abuse of discretion\nin the district court’s failure to discuss the strength of\nATEN’s lost profits theory.\n III\n To the extent Appellants argue weakness in ATEN’s\nother substantive positions justifies reversing the district\ncourt’s determination that this case was not exceptional,\nwe do not agree. Today in No. 18-1606, we reverse the dis-\ntrict court’s denial of JMOL because the jury’s finding of\nanticipation was not supported by substantial evidence.\nThough we affirm as to noninfringement of the ’141 and\n’289 patents, we note that ATEN’s primary argument on\nappeal—that the court improperly gave claim construction\ndisputes to the jury—was rejected because ATEN did not\ntimely raise this issue with the court below. We find no\nabuse of discretion in the district court’s determination\n\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 7\nCO., LTD.\n\n\nthat ATEN’s positions were not so objectively unreasonable\nor exceptionally meritless as to stand out from other cases.\n We have considered Appellants’ other arguments and\nfind them without merit.\n CONCLUSION\n We hold the district court did not abuse its discretion\nin declining to find this an exceptional case under 35 U.S.C.\n§ 285. For the foregoing reasons, we affirm.\n AFFIRMED\n COSTS\n Costs to ATEN.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: ATEN INTERNATIONAL CO., LTD., Plaintiff-Appellee v. UNICLASS TECHNOLOGY CO., LTD., Electronic Technology Co., Ltd., Airlink 101, Phoebe Micro, Inc., Broadtech International Co., Ltd., Dba Linkskey, Black Box Corporation, Black Box Corporation of Pennsylvania, Defendants-Appellants\\nCourt: U.S. Court of Appeals for the Federal Circuit\\nDate decided: 2019-08-06\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n ATEN INTERNATIONAL CO., LTD.,\\n Plaintiff-Appellee\\n\\n v.\\n\\n UNICLASS TECHNOLOGY CO., LTD.,\\nELECTRONIC TECHNOLOGY CO., LTD., AIRLINK\\n 101, PHOEBE MICRO, INC., BROADTECH\\n INTERNATIONAL CO., LTD., DBA LINKSKEY,\\n BLACK BOX CORPORATION, BLACK BOX\\n CORPORATION OF PENNSYLVANIA,\\n Defendants-Appellants\\n ______________________\\n\\n 2018-1922\\n ______________________\\n\\n Appeal from the United States District Court for the\\nCentral District of California in No. 2:15-cv-04424-AG-\\nAJW, Judge Andrew J. Guilford.\\n ______________________\\n\\n Decided: August 6, 2019\\n ______________________\\n\\n EDWARD NAIDICH, Mei & Mark LLP, Washington, DC,\\nargued for plaintiff-appellee. Also represented by LEI MEI,\\nRICHARD HADORN, PHILIP ANDREW RILEY, LAURENCE M.\\nSANDELL.\\n\\n JOSEPH PIA, Pia Anderson Moss Hoyt, Salt Lake City,\\n\\f2 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\nUT, argued for defendants-appellants. Also represented by\\nROBERT AYCOCK.\\n ______________________\\n\\n Before MOORE, WALLACH, and TARANTO, Circuit Judges.\\nMOORE, Circuit Judge.\\n Uniclass Technology Co., Ltd. (“Uniclass”), as well as\\nElectronic Technology Co., Ltd.; Airlink 101; Phoebe Micro,\\nInc.; Broadtech International Co., Ltd., d/b/a Linkskey;\\nBlack Box Corporation; and Black Box Corporation of\\nPennsylvania (collectively the “customer defendants”) (all\\ncollectively “Appellants”) appeal the United States District\\nCourt for the Central District of California’s order denying\\nAppellants’ motion for attorney fees. Because we hold the\\ndistrict court did not abuse its discretion in declining to\\nfind this an exceptional case under 35 U.S.C. § 285, we af-\\nfirm.\\n BACKGROUND\\n Uniclass and ATEN International Co., Ltd. (“ATEN”)\\nare involved in making and selling keyboard-video-mouse\\nswitch systems that allow a user to control multiple com-\\nputers from a single keyboard, video device, and mouse. In\\n2011, Uniclass stopped making payments on a license\\nagreement it entered with ATEN in 2009. In 2014, ATEN\\nsued Uniclass and the customer defendants alleging in-\\nfringement, seeking damages and injunctive relief. Uni-\\nclass moved for summary judgment on ATEN’s lost profits\\ntheory of damages, which the district court granted in April\\n2017. ATEN proceeded to trial based on a reasonable roy-\\nalty theory of damages, under which its expert testified\\nthat the maximum recovery (not including its requested\\ntreble damages) was $678,337. At trial, a jury found that\\nUniclass did not infringe the asserted claims of U.S. Patent\\nNo. 8,589,141 or U.S. Patent No. 7,640,289. It also found\\nthe asserted claims of the ’141 patent invalid. ATEN\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 3\\nCO., LTD.\\n\\n\\nappealed various aspects of the September 2017 jury ver-\\ndict separately in No. 18-1606, which we also decide today.\\n After trial, Uniclass moved to declare this case excep-\\ntional under 35 U.S.C. § 285, arguing that ATEN did not\\nconduct an adequate pre-filing investigation, unnecessarily\\nincreased the costs of claim construction, drastically in-\\ncreased discovery costs by frequently changing counsel and\\ninfringement positions, and engaged in unreasonable liti-\\ngation behavior requiring additional motion practice and\\nleading to an expensive and disproportionate trial. The\\ndistrict court denied the motion.\\n Uniclass timely appealed. We have jurisdiction under\\n28 U.S.C. § 1295(a)(1).\\n DISCUSSION\\n Under § 285, “[t]he court in exceptional cases may\\naward reasonable attorney fees to the prevailing party.”\\n“[A]n ‘exceptional’ case is simply one that stands out from\\nothers with respect to the substantive strength of a party’s\\nlitigating position (considering both the governing law and\\nthe facts of the case) or the unreasonable manner in which\\nthe case was litigated.” Octane Fitness, LLC v. ICON\\nHealth & Fitness, Inc., 572 U.S. 545, 554 (2014). District\\ncourts should determine whether a case is exceptional on a\\ncase-by-case basis, considering the totality of the circum-\\nstances. Id. We review the district court’s § 285 determi-\\nnation for abuse of discretion. Highmark Inc. v. Allcare\\nHealth Mgmt. Sys., Inc., 572 U.S. 559, 561 (2014). “The\\nabuse-of-discretion standard does not preclude an appel-\\nlate court’s correction of a district court’s legal or factual\\nerror: A district court would necessarily abuse its discre-\\ntion if it based its ruling on an erroneous view of the law or\\non a clearly erroneous assessment of the evidence.” Id. at\\n563 n.2 (internal quotation marks omitted).\\n\\f4 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\n I\\n Appellants focus their arguments on appeal on a theory\\nthat the district court erred in not finding this case excep-\\ntional based on ATEN’s disregard for the “foundational pol-\\nicy” of proportionate litigation. Appellants’ Br. 15–16.\\nAppellants summarize ATEN’s expenses as including over\\n$700,000 in expert witness fees alone, without considering\\nother expenses including attorney fees. Appellants argue\\nthat ATEN could recover, at most, $678,337 in reasonable\\nroyalty damages. See J.A. 902 at 7:20–8:4; J.A. 870 at\\n23:8–11; J.A. 923 at 90:14–15. Accordingly, Appellants ar-\\ngue this case is exceptional because the cost of litigating\\nthe case exceeded ATEN’s potential recovery at trial. See,\\ne.g., Appellants’ Br. 15–20, 25–26. We reject this argu-\\nment. We see no error in the court’s legal analysis and no\\nclear error in its fact findings.\\n There is no per se rule that a case is exceptional if liti-\\ngation costs exceed the potential damages. The District of\\nDelaware’s post-Octane decision in EON is persuasive on\\nthis point: “[The court] is unaware of any de minimis ex-\\nception for infringement. It cannot be the case that a plain-\\ntiff may be subjected to monetary sanctions for failing to\\ndrop a case against a defendant if the cost of litigation ex-\\nceeds the potential recovery.” EON Corp. IP Holdings,\\nLLC v. FLO TV Inc., No. 10-812-RGA, 2014 WL 2196418,\\nat *2 (D. Del. May 27, 2014). Litigation strategies motivat-\\ning a patent suit are many, with monetary damages being\\none. We routinely hear cases in which damages are not at\\nissue—for example, in appeals from the International\\nTrade Commission, in some Hatch-Waxman cases, and\\nother cases where only an injunction is sought. A patentee\\nmay also bring suit to deter other competitors from in-\\nfringement, encourage licensing, or test a patent’s ability\\nto withstand validity challenges. Thermolife Int’l LLC v.\\nGNC Corp., 922 F.3d 1347, 1363–64 (Fed. Cir. 2019).\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 5\\nCO., LTD.\\n\\n\\n Moreover, ATEN sought injunctive relief against Uni-\\nclass, its direct competitor. This alone undermines Appel-\\nlants’ argument. The case could have proceeded to trial\\nrequesting only an injunction, and there would be no po-\\ntential damages to compare to ATEN’s expenses. We can-\\nnot conclude that the district court’s decision denying\\nAppellants’ motion for fees under these circumstances\\namounts to an abuse of discretion.\\n Appellants argue the district court should be bound by\\nits statements during trial that the court was troubled by\\ndisproportionate aspects of the case. But Appellants admit\\nthat such comments were largely directed to the amount of\\npublic resources expended on the limited recovery at issue.\\nAppellants’ Br. 17–20 (citing J.A. 955 at 12:16–20 (regard-\\ning public resources); J.A. 1052 at 73:19–25 (regarding\\ntime spent with one witness); J.A. 1053 at 80:17–22 (re-\\ngarding the length of trial compared to damages at issue);\\nJ.A. 1153 at 102:15–25 (regarding ATEN’s attorney fees\\nand the amount of time spent with the witness)). Appel-\\nlants cite an exchange in which the court recognized a sce-\\nnario in which expenditures might understandably exceed\\nrecovery: “I just don’t understand [expert expenditures of]\\n$400,000 in a $650,000 case. When I was a lawyer, if I did\\nthat, I would be fired, or frankly, I’d be working for a client\\nwho said: They are competitors. Spend every penny.” Ap-\\npellants’ Br. 18 (citing J.A. 956 at 13:2–6 (emphasis\\nadded)). We cannot conclude that the court’s comments\\nduring trial regarding public resources render its subse-\\nquent decision denying fees an abuse of discretion. We see\\nno abuse of discretion in the court’s conclusion that under\\nthe totality of the circumstances this litigation was not ex-\\nceptional.\\n II\\n Appellants also argue the district court erred by failing\\nto weigh ATEN’s frivolous damages methodology in its\\nanalysis. Specifically, Appellants argue the district court\\n\\f6 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\nshould have weighed the strength of ATEN’s lost profits\\ntheory (rejected at the summary judgment stage), which\\nAppellants allege was knowingly baseless and brought in\\nbad faith. Appellants’ Br. 22–25.\\n The district court did refer to Uniclass’ argument that\\nATEN presented a weak case for lost profits that relied on\\nguesswork, which it listed among the “whole slew” of Uni-\\nclass’ arguments it considered. J.A. 3. The district court\\nfound there were “perhaps . . . shortcomings” in ATEN’s po-\\nsitions that did not “amount to a showing that [ATEN’s]\\nbehavior was exceptional, amounting to the type of unrea-\\nsonableness discussed in Octane.” J.A. 4.\\n Though ATEN’s lost profits theory was not strong\\nenough to withstand summary judgment, “[a] party’s posi-\\ntion . . . ultimately need not be correct for them to not\\n‘stand[ ] out.’” SFA Sys., LLC v. Newegg Inc., 793 F.3d\\n1344, 1348 (Fed. Cir. 2015) (quoting Octane, 572 U.S. at\\n544). The district court “had no obligation to write an opin-\\nion that reveals [its] assessment of every consideration.”\\nUniv. of Utah v. Max-Planck-Gesellschaft, 851 F.3d 1317,\\n1323 (Fed. Cir. 2017). We can find no abuse of discretion\\nin the district court’s failure to discuss the strength of\\nATEN’s lost profits theory.\\n III\\n To the extent Appellants argue weakness in ATEN’s\\nother substantive positions justifies reversing the district\\ncourt’s determination that this case was not exceptional,\\nwe do not agree. Today in No. 18-1606, we reverse the dis-\\ntrict court’s denial of JMOL because the jury’s finding of\\nanticipation was not supported by substantial evidence.\\nThough we affirm as to noninfringement of the ’141 and\\n’289 patents, we note that ATEN’s primary argument on\\nappeal—that the court improperly gave claim construction\\ndisputes to the jury—was rejected because ATEN did not\\ntimely raise this issue with the court below. We find no\\nabuse of discretion in the district court’s determination\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 7\\nCO., LTD.\\n\\n\\nthat ATEN’s positions were not so objectively unreasonable\\nor exceptionally meritless as to stand out from other cases.\\n We have considered Appellants’ other arguments and\\nfind them without merit.\\n CONCLUSION\\n We hold the district court did not abuse its discretion\\nin declining to find this an exceptional case under 35 U.S.C.\\n§ 285. For the foregoing reasons, we affirm.\\n AFFIRMED\\n COSTS\\n Costs to ATEN.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4194,"cells":{"id":{"kind":"number","value":4645830,"string":"4,645,830"},"name":{"kind":"string","value":"ATEN INTERNATIONAL CO., LTD., Plaintiff-Appellant v. UNICLASS TECHNOLOGY CO., LTD., Electronic Technology Co., Ltd., Airlink 101, Phoebe Micro, Inc., Broadtech International Co., Ltd., Dba Linkskey, Black Box Corporation, Black Box Corporation of Pennsylvania, Defendants-Appellees"},"name_abbreviation":{"kind":"string","value":"Aten International Co., Ltd. v. Uniclass Technology Co., Ltd."},"decision_date":{"kind":"string","value":"2019-08-06"},"docket_number":{"kind":"string","value":"2018-1606"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Federal Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Moore, Wallach, Taranto\", \"parties\": \"\", \"opinions\": [{\"author\": \"MOORE, Circuit Judge.\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n ATEN INTERNATIONAL CO., LTD.,\\n Plaintiff-Appellant\\n\\n v.\\n\\n UNICLASS TECHNOLOGY CO., LTD.,\\nELECTRONIC TECHNOLOGY CO., LTD., AIRLINK\\n 101, PHOEBE MICRO, INC., BROADTECH\\n INTERNATIONAL CO., LTD., DBA LINKSKEY,\\n BLACK BOX CORPORATION, BLACK BOX\\n CORPORATION OF PENNSYLVANIA,\\n Defendants-Appellees\\n ______________________\\n\\n 2018-1606\\n ______________________\\n\\n Appeal from the United States District Court for the\\nCentral District of California in No. 2:15-cv-04424-AG-\\nAJW, Judge Andrew J. Guilford.\\n ______________________\\n\\n Decided: August 6, 2019\\n ______________________\\n\\n LAURENCE M. SANDELL, Mei & Mark LLP, Washington,\\nDC, argued for plaintiff-appellant. Also represented by LEI\\nMEI.\\n\\n JOSEPH PIA, Pia Anderson Moss Hoyt, Salt Lake City,\\nUT, argued for defendants-appellees. Also represented by\\nROBERT AYCOCK.\\n\\f2 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\n ______________________\\n\\n Before MOORE, WALLACH, and TARANTO, Circuit Judges.\\nMOORE, Circuit Judge.\\n ATEN International Co., Ltd. (“ATEN”) appeals the\\nUnited States District Court for the Central District of Cal-\\nifornia’s denial of judgment as a matter of law (“JMOL”),\\nwhich declined to overturn the jury’s findings that the as-\\nserted claims of U.S. Patent No. 8,589,141 are invalid as\\nanticipated under 35 U.S.C. § 102 and not infringed; and\\nthat the asserted claims of U.S. Patent No. 7,640,289 are\\nnot infringed. For the reasons discussed below, we reverse\\nas to invalidity and affirm as to noninfringement.\\n BACKGROUND\\n Uniclass Technology Co., Ltd. (“Uniclass”) and ATEN\\nare involved in making and selling keyboard-video-mouse\\n(“KVM”) switch systems that allow a user to control multi-\\nple computers from a single keyboard, video device, and\\nmouse. ATEN sued Uniclass as well as Electronic Technol-\\nogy Co., Ltd.; Airlink 101; Phoebe Micro, Inc.; Broadtech\\nInternational Co., Ltd. d/b/a Linkskey; Black Box Corpora-\\ntion; and Black Box Corporation of Pennsylvania (collec-\\ntively, the “customer defendants”) alleging, as relevant\\nhere, infringement of claims 3, 8, and 10 of the ’141 patent\\nand claims 1–20 of the ’289 patent. The ’141 patent is di-\\nrected to technology for switching between computers that\\nshare a keyboard, monitor, and mouse through a KVM\\nswitch, such as a keyboard shortcut. ’141 patent at 2:51–\\n61. Claim 3 depends from claims 1 and 2, and claim 8 de-\\npends from claim 1. Independent claim 1 recites:\\n 1. A method for controlling a resource sharing ap-\\n paratus coupling at least one input device to a plu-\\n rality of hosts including a first host, the method\\n comprising:\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 3\\nCO., LTD.\\n\\n\\n connecting the input device to the first\\n host;\\n while the input device is connected to the\\n first host, acquiring a first input signal\\n from the input device and determining\\n whether the first input signal comprises a\\n standby indication of a switch command\\n and wherein the standby indication is for\\n indicating that connection between the in-\\n put device and the first host can be\\n changed; and\\n in response to a determination that the\\n first input signal comprises the standby in-\\n dication, disconnecting the input device\\n from the first host without connecting the\\n input device to any other host and starting\\n emulating the input device to the first host;\\n and\\n acquiring a second input signal from the in-\\n put device, wherein the second input signal\\n is not transferred to the hosts when it is in-\\n putted to the resource sharing apparatus.\\n The ’289 patent is directed to technology for stringing\\ntogether several KVM switches. ’289 patent at 2:1–6. It\\nprovides that each KVM switch can detect whether it is a\\nmaster or slave by, for example, detecting whether the port\\nused to connect other KVM switches is occupied. Id. at 4:1–\\n4. Independent claim 1 recites:\\n 1. A computer switch comprising:\\n a set of peripheral device ports for connect-\\n ing to a set of peripheral devices;\\n a first port;\\n a second port;\\n\\f4 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\n a control device coupled to the first port and\\n the second port, the control device repeat-\\n edly detecting whether the first port is oc-\\n cupied to repeatedly determine a master or\\n slave status of the computer switch,\\n wherein\\n if the first port is occupied, the control de-\\n vice determines the computer switch to be\\n a slave and provides data to a first external\\n computer switch connected to the first port,\\n and\\n if the first port is unoccupied, the control\\n device determines the computer switch to\\n be a master and controls a second external\\n computer switch connected to the second\\n port and obtains data from the second ex-\\n ternal computer switch; and\\n a plurality of computer connection ports\\n coupled to the control device for connecting\\n to and controlling a plurality of computers,\\n wherein the plurality of computers are con-\\n trolled by the set of peripheral devices if the\\n computer switch is determined to be a mas-\\n ter, and are controlled by the first external\\n computer switch connected to the first port\\n if the computer switch is determined to be\\n a slave.\\n At trial, the jury found that Uniclass did not infringe\\nthe asserted claims of the ’141 or ’289 patents. It also found\\nthe asserted claims of the ’141 patent were invalid as an-\\nticipated without specifying which reference was the basis\\nfor its finding. ATEN moved for JMOL, which the district\\ncourt denied in the aspects relevant to this appeal.\\n ATEN timely appealed. We have jurisdiction under 28\\nU.S.C. § 1295(a)(1).\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 5\\nCO., LTD.\\n\\n\\n DISCUSSION\\n We review denials of JMOL under the law of the re-\\ngional circuit. TVIIM, LLC v. McAfee, Inc., 851 F.3d 1356,\\n1362 (Fed. Cir. 2017). The Ninth Circuit reviews a denial\\nof JMOL de novo. Harper v. City of Los Angeles, 533 F.3d\\n1010, 1021 (9th Cir. 2008). JMOL is proper when the evi-\\ndence permits only one reasonable conclusion which is con-\\ntrary to the jury’s verdict, but the jury’s verdict must be\\nupheld if it is supported by substantial evidence. Id. An-\\nticipation and infringement are questions of fact that we\\nreview for substantial evidence. Cordis Corp. v. Bos. Sci.\\nCorp., 561 F.3d 1319, 1330, 1335 (Fed. Cir. 2009). Whether\\na reference is prior art is a question of law based on under-\\nlying factual questions. TypeRight Keyboard Corp. v. Mi-\\ncrosoft Corp., 374 F.3d 1151, 1157 (Fed. Cir. 2004).\\nWhether a reference is publicly accessible is a question of\\nfact. In re NTP, Inc., 654 F.3d 1279, 1296 (Fed. Cir. 2011).\\n I\\n At trial, Uniclass asserted anticipation of the claims of\\nthe ’141 patent based on two references (1) CS-1762, an\\nearlier product by ATEN, and/or its user manual; and (2)\\nGreat Britain Patent No. 2,352,540 (“GB ’540”). The jury\\nfound that the asserted claims of the ’141 patent were in-\\nvalid as anticipated, but there was no special verdict indi-\\ncating whether one or both references formed the basis for\\nthe jury’s decision. J.A. 9769. The district court denied\\nATEN’s JMOL motion, upholding the jury’s finding of an-\\nticipation based on the first theory, that the CS-1762 refer-\\nence anticipated the asserted claims. It did not address the\\nsecond theory. Because the jury’s verdict of anticipation\\nwas not supported by substantial record evidence as to ei-\\nther reference, we reverse the denial of JMOL on anticipa-\\ntion.\\n\\f6 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\n A\\n To establish that an asserted reference is prior art un-\\nder § 102(b), the patent challenger must prove, by clear and\\nconvincing evidence, that it predates the critical date. Ma-\\nhurkar v. C.R. Bard, Inc., 79 F.3d 1572, 1578 (Fed. Cir.\\n1996).\\n The undisputed critical date for the ’141 patent is July\\n24, 2006—one year prior to its filing date. J.A. 4; J.A. 24;\\nJ.A. 9302–9303 at 176:24–177:3. Uniclass’ expert\\nMr. Dezmelyk relied on his testing of a 2009 CS-1762 spec-\\nimen to meet certain limitations of claim 1 of the ’141 pa-\\ntent. Uniclass argues the 2009 specimen qualifies as prior\\nart because it uses firmware that was released in 2006 be-\\nfore the critical date of the ’141 patent. See Appellee’s Br.\\n20. But Mr. Dezmelyk testified merely that the firmware\\ndated to 2006, without specifying a day or month. J.A.\\n9313 at 14–17 (“I then looked on the wayback machine to\\nfind when that firmware update came out and the last time\\nit was changed, and that was in 2006. So this firmware is\\nprior art firmware operating on this, the test hardware.”).\\n To establish that the CS-1762 qualifies as prior art,\\nUniclass was required to prove by clear and convincing ev-\\nidence that the firmware existed prior to July 24, 2006.\\nMerely establishing that the firmware existed in the same\\nyear as the critical date is insufficient. This deficiency was\\nhighlighted on cross-examination:\\n Q. Yes or no, you did not testify to a particular\\n month in 2006?\\n A. No, I don’t believe I mentioned that in my testi-\\n mony.\\n Q. And you did not testify to a specific date in 2006?\\n A. I don’t believe I testified to a particular day ei-\\n ther.\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 7\\nCO., LTD.\\n\\n\\nJ.A. 9338 at 18–22. Mr. Dezmelyk did not provide any ad-\\nditional date-related information, and Uniclass did not con-\\nduct re-direct examination on this issue.\\n With a critical date of July 24, 2006, testimony that the\\nfirmware running on the CS-1762 device existed in 2006\\nalone is not enough to support the jury’s finding that the\\nfirmware pre-dated the critical date and thus qualifies as\\nprior art. We hold that the district court erred in denying\\nJMOL because substantial evidence does not support the\\njury’s finding of anticipation based on the CS-1762 device.\\n B\\n A finding of anticipation requires clear and convincing\\nevidence that “each and every element is found within a\\nsingle prior art reference, arranged as claimed.” Summit 6,\\nLLC v. Samsung Elecs. Co., 802 F.3d 1283, 1294 (Fed. Cir.\\n2015). Testimony concerning anticipation must typically\\n“explain in detail how each claim element is disclosed in\\nthe prior art reference. The testimony is insufficient if it is\\nmerely conclusory.” Schumer v. Lab. Comput. Sys., Inc.,\\n308 F.3d 1304, 1315–16 (Fed. Cir. 2002).\\n The jury verdict of anticipation cannot be upheld with\\nregard to the GB ’540 reference. The asserted claims of the\\n’141 patent require “emulating the input device to the first\\nhost.” Uniclass’ expert Mr. Dezmelyk testified about the\\nemulation described in GB ’540 in the context of other as-\\nserted patents and claim limitations. J.A. 9296 at 170:10–\\n20 (regarding “emulating the console devices according to the\\nindustry standard”); J.A. 9301–9302 at 175:20–176:6 (de-\\nscribing that the reference “has USB device emulators in it”);\\nJ.A. 9285–9286 at 159:16–160:18 (describing the “USB host\\nemulator”). But Uniclass offered no testimony explaining\\nhow this emulation described in GB ’540 meets claim 1’s\\nrequirement of “emulating . . . to the first host.” We also\\ncannot identify any place in the three pages of GB ’540 pro-\\nvided in the joint appendix that would support Uniclass’\\ntheory that GB ’540 discloses this limitation. J.A. 9789;\\n\\f8 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\nJ.A. 9790; J.A. 9793. We see no record evidence that GB\\n’540 discloses “emulating the input device to the first host.”\\nTo the extent that the jury finding of anticipation was\\nbased on the GB ’540 reference, that finding lacks substan-\\ntial evidence in this record.\\n The district court erred in its denial of JMOL because\\nthe jury’s finding of anticipation was not supported by sub-\\nstantial evidence as to either reference.\\n II\\n On appeal, ATEN challenges the district court’s denial\\nof JMOL as to infringement of the ’141 and ’289 patents.\\nATEN argued in its motion for JMOL that the evidence\\ncompels a finding of infringement, but the jury was con-\\nfused due to Uniclass’ expert Mr. Dezmelyk’s testimony on\\nthe scope of the claim terms. E.g., J.A. 8–12; J.A. 7937–\\n7938. The district court denied the motion, holding ATEN\\nwaived any right to contest Mr. Dezmelyk’s claim construc-\\ntion testimony because it failed “to raise concerns regard-\\ning the proper construction of claim terms at trial itself.”\\nJ.A. 10.\\n The record below indicates that Mr. Dezmelyk offered\\ntrial testimony regarding the construction of the terms\\n“disconnecting” and “connecting” in the ’141 patent. At the\\nclaim construction stage, the district court gave the term\\n“disconnecting the input device from the first host without\\nconnecting the input device to any other host and starting\\nemulating the input device to the first host” its plain and\\nordinary meaning. ATEN did not object, nor did it proffer\\na different construction. In fact, it was ATEN who urged\\nthe court to adopt the plain and ordinary meaning. At trial,\\nMr. Dezmelyk opined that the accused devices “never con-\\nnect the input device to the host computer” based on his\\ninterpretation of the plain and ordinary meaning of the\\nterm “connected” as requiring a direct communication\\npathway. J.A. 9237 at 15–16; see also J.A. 9240 at 16–18;\\nJ.A. 9241 at 15–21. To inform his interpretation, Mr.\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 9\\nCO., LTD.\\n\\n\\nDezmelyk testified about the import of the prosecution his-\\ntory. J.A. 9242 at 13–23 (“I look at the prosecution history\\nto understand what ATEN . . . was telling the patent office\\n. . . . In this case one of the statements they made to the\\nexaminer was the applicant’s invention . . . has a funda-\\nmental difference in comparison to the teaching of [a refer-\\nence] with regard to the concepts of connected and\\ndisconnected between the input device and the host de-\\nvices.”). According to ATEN, this expert was testifying as\\nto the proper claim construction to be given to these terms.\\n The record also indicates that Mr. Dezmelyk offered\\ntrial testimony regarding the meaning of “detecting\\nwhether the first port is occupied” in the ’289 patent.\\nMr. Dezmelyk opined that he believed “the ordinary mean-\\ning of the phrase, that is what an engineer would think\\nwhen they read this patent, what was meant by that\\nphrase when he says detecting the first port is occupied, is\\ndetecting if a cable was plugged into the port.” J.A. 9223\\nat 1–5; see also J.A. 9223 at 17–18; J.A. 9221 at 20–24. To\\ninform his interpretation, Mr. Dezmelyk again referred to\\nthe prosecution history. J.A. 9228 at 7–12 (“[T]he con-\\nnector is occupied in the [reference], but notice there is no\\ndiscussion of messages . . . . In fact, . . . the [reference] does\\nmeet the messages test. So it has to be a determination\\nbased on the connector itself actually having a plug in it.”);\\nsee also J.A. 9226 at 13–14.\\n Mr. Dezmelyk’s testimony on both patents amounts to\\nclaim construction testimony before the jury. ATEN did\\nnot object to any of this testimony during the trial, nor did\\nATEN move for the district court to resolve these claim con-\\nstruction disputes. The only objection to the expert’s testi-\\nmony came after the close of evidence when ATEN orally\\nmoved for a “post-testifying Daubert on Mr. Dezmelyk’s\\ntestimony on invalidity with respect to anything that\\n\\f10 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\nrequired an ATEN claim construction.” 1 J.A. 9357–9358 at\\n231:23–232:1 (italics added). ATEN now argues its objec-\\ntion regarding invalidity should be broadened to apply to\\nclaim construction for infringement purposes. We do not\\nagree. Even if ATEN’s oral motion was timely, we agree\\nwith the district court that ATEN did not object to\\nMr. Dezmelyk’s claim construction testimony with respect\\nto infringement. J.A. 9–10. By failing to object, ATEN has\\nwaived any challenge to the jury’s finding of infringement\\nbased on this testimony.\\n It is beyond dispute that claim construction issues are\\nto be decided by the court. It is thus improper for an expert\\nwitness to testify before the jury regarding claim construc-\\ntion. CytoLogix Corp. v. Ventana Med. Sys., Inc., 424 F.3d\\n1168, 1172 (Fed. Cir. 2005). In CytoLogix, the parties\\nagreed to present expert witness testimony regarding\\nclaim construction to the jury and counsel argued conflict-\\ning claim constructions to the jury. Id. We noted the dis-\\ntrict court should have refused to allow such testimony,\\ndespite the agreement of the parties, because the risk of\\nconfusing the jury was high. Id. But here, as in CytoLogix,\\n“there is no ground for reversal since there was no objection\\nto the expert testimony as to claim construction.” Id. at\\n1173. We do not fault the district court here for allowing\\nclaim construction to go to the jury. It is the parties’ obli-\\ngation to raise a dispute regarding the proper scope of\\nclaims to the court. O2 Micro Int’l Ltd. v. Beyond Innova-\\ntion Tech. Co., 521 F.3d 1351, 1362 (Fed. Cir. 2008) (“When\\nthe parties present a fundamental dispute regarding the\\n\\n\\n\\n 1 ATEN also filed a motion for JMOL on the final day\\nof trial that briefly discussed the term “occupied” of the ’289\\npatent. J.A. 7937. ATEN did not raise this issue at the\\nhearing. J.A. 10–11. We agree with the district court that\\nATEN failed to preserve its claim construction challenge\\nvia this motion. J.A. 11.\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 11\\nCO., LTD.\\n\\n\\nscope of a claim term, it is the court’s duty to resolve it.”).\\nWe therefore affirm the district court’s denial of JMOL as\\nto noninfringement.\\n CONCLUSION\\n For the reasons discussed above, we reverse as to inva-\\nlidity and affirm as to noninfringement. Because we affirm\\nas to noninfringement, we need not reach lost profits.\\n AFFIRMED-IN-PART, REVERSED-IN-PART\\n COSTS\\n No costs.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4423083/\", \"author_raw\": \"MOORE, Circuit Judge.\"}]}"},"judge1":{"kind":"string","value":"MOORE"},"judge2":{"kind":"string","value":"WALLACH"},"judge3":{"kind":"string","value":"TARANTO"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4645830/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: ATEN INTERNATIONAL CO., LTD., Plaintiff-Appellant v. UNICLASS TECHNOLOGY CO., LTD., Electronic Technology Co., Ltd., Airlink 101, Phoebe Micro, Inc., Broadtech International Co., Ltd., Dba Linkskey, Black Box Corporation, Black Box Corporation of Pennsylvania, Defendants-Appellees\nCourt: U.S. Court of Appeals for the Federal Circuit\nDate decided: 2019-08-06\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n for the Federal Circuit\n ______________________\n\n ATEN INTERNATIONAL CO., LTD.,\n Plaintiff-Appellant\n\n v.\n\n UNICLASS TECHNOLOGY CO., LTD.,\nELECTRONIC TECHNOLOGY CO., LTD., AIRLINK\n 101, PHOEBE MICRO, INC., BROADTECH\n INTERNATIONAL CO., LTD., DBA LINKSKEY,\n BLACK BOX CORPORATION, BLACK BOX\n CORPORATION OF PENNSYLVANIA,\n Defendants-Appellees\n ______________________\n\n 2018-1606\n ______________________\n\n Appeal from the United States District Court for the\nCentral District of California in No. 2:15-cv-04424-AG-\nAJW, Judge Andrew J. Guilford.\n ______________________\n\n Decided: August 6, 2019\n ______________________\n\n LAURENCE M. SANDELL, Mei & Mark LLP, Washington,\nDC, argued for plaintiff-appellant. Also represented by LEI\nMEI.\n\n JOSEPH PIA, Pia Anderson Moss Hoyt, Salt Lake City,\nUT, argued for defendants-appellees. Also represented by\nROBERT AYCOCK.\n\f2 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\n CO., LTD.\n\n\n ______________________\n\n Before MOORE, WALLACH, and TARANTO, Circuit Judges.\nMOORE, Circuit Judge.\n ATEN International Co., Ltd. (“ATEN”) appeals the\nUnited States District Court for the Central District of Cal-\nifornia’s denial of judgment as a matter of law (“JMOL”),\nwhich declined to overturn the jury’s findings that the as-\nserted claims of U.S. Patent No. 8,589,141 are invalid as\nanticipated under 35 U.S.C. § 102 and not infringed; and\nthat the asserted claims of U.S. Patent No. 7,640,289 are\nnot infringed. For the reasons discussed below, we reverse\nas to invalidity and affirm as to noninfringement.\n BACKGROUND\n Uniclass Technology Co., Ltd. (“Uniclass”) and ATEN\nare involved in making and selling keyboard-video-mouse\n(“KVM”) switch systems that allow a user to control multi-\nple computers from a single keyboard, video device, and\nmouse. ATEN sued Uniclass as well as Electronic Technol-\nogy Co., Ltd.; Airlink 101; Phoebe Micro, Inc.; Broadtech\nInternational Co., Ltd. d/b/a Linkskey; Black Box Corpora-\ntion; and Black Box Corporation of Pennsylvania (collec-\ntively, the “customer defendants”) alleging, as relevant\nhere, infringement of claims 3, 8, and 10 of the ’141 patent\nand claims 1–20 of the ’289 patent. The ’141 patent is di-\nrected to technology for switching between computers that\nshare a keyboard, monitor, and mouse through a KVM\nswitch, such as a keyboard shortcut. ’141 patent at 2:51–\n61. Claim 3 depends from claims 1 and 2, and claim 8 de-\npends from claim 1. Independent claim 1 recites:\n 1. A method for controlling a resource sharing ap-\n paratus coupling at least one input device to a plu-\n rality of hosts including a first host, the method\n comprising:\n\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 3\nCO., LTD.\n\n\n connecting the input device to the first\n host;\n while the input device is connected to the\n first host, acquiring a first input signal\n from the input device and determining\n whether the first input signal comprises a\n standby indication of a switch command\n and wherein the standby indication is for\n indicating that connection between the in-\n put device and the first host can be\n changed; and\n in response to a determination that the\n first input signal comprises the standby in-\n dication, disconnecting the input device\n from the first host without connecting the\n input device to any other host and starting\n emulating the input device to the first host;\n and\n acquiring a second input signal from the in-\n put device, wherein the second input signal\n is not transferred to the hosts when it is in-\n putted to the resource sharing apparatus.\n The ’289 patent is directed to technology for stringing\ntogether several KVM switches. ’289 patent at 2:1–6. It\nprovides that each KVM switch can detect whether it is a\nmaster or slave by, for example, detecting whether the port\nused to connect other KVM switches is occupied. Id. at 4:1–\n4. Independent claim 1 recites:\n 1. A computer switch comprising:\n a set of peripheral device ports for connect-\n ing to a set of peripheral devices;\n a first port;\n a second port;\n\f4 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\n CO., LTD.\n\n\n a control device coupled to the first port and\n the second port, the control device repeat-\n edly detecting whether the first port is oc-\n cupied to repeatedly determine a master or\n slave status of the computer switch,\n wherein\n if the first port is occupied, the control de-\n vice determines the computer switch to be\n a slave and provides data to a first external\n computer switch connected to the first port,\n and\n if the first port is unoccupied, the control\n device determines the computer switch to\n be a master and controls a second external\n computer switch connected to the second\n port and obtains data from the second ex-\n ternal computer switch; and\n a plurality of computer connection ports\n coupled to the control device for connecting\n to and controlling a plurality of computers,\n wherein the plurality of computers are con-\n trolled by the set of peripheral devices if the\n computer switch is determined to be a mas-\n ter, and are controlled by the first external\n computer switch connected to the first port\n if the computer switch is determined to be\n a slave.\n At trial, the jury found that Uniclass did not infringe\nthe asserted claims of the ’141 or ’289 patents. It also found\nthe asserted claims of the ’141 patent were invalid as an-\nticipated without specifying which reference was the basis\nfor its finding. ATEN moved for JMOL, which the district\ncourt denied in the aspects relevant to this appeal.\n ATEN timely appealed. We have jurisdiction under 28\nU.S.C. § 1295(a)(1).\n\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 5\nCO., LTD.\n\n\n DISCUSSION\n We review denials of JMOL under the law of the re-\ngional circuit. TVIIM, LLC v. McAfee, Inc., 851 F.3d 1356,\n1362 (Fed. Cir. 2017). The Ninth Circuit reviews a denial\nof JMOL de novo. Harper v. City of Los Angeles, 533 F.3d\n1010, 1021 (9th Cir. 2008). JMOL is proper when the evi-\ndence permits only one reasonable conclusion which is con-\ntrary to the jury’s verdict, but the jury’s verdict must be\nupheld if it is supported by substantial evidence. Id. An-\nticipation and infringement are questions of fact that we\nreview for substantial evidence. Cordis Corp. v. Bos. Sci.\nCorp., 561 F.3d 1319, 1330, 1335 (Fed. Cir. 2009). Whether\na reference is prior art is a question of law based on under-\nlying factual questions. TypeRight Keyboard Corp. v. Mi-\ncrosoft Corp., 374 F.3d 1151, 1157 (Fed. Cir. 2004).\nWhether a reference is publicly accessible is a question of\nfact. In re NTP, Inc., 654 F.3d 1279, 1296 (Fed. Cir. 2011).\n I\n At trial, Uniclass asserted anticipation of the claims of\nthe ’141 patent based on two references (1) CS-1762, an\nearlier product by ATEN, and/or its user manual; and (2)\nGreat Britain Patent No. 2,352,540 (“GB ’540”). The jury\nfound that the asserted claims of the ’141 patent were in-\nvalid as anticipated, but there was no special verdict indi-\ncating whether one or both references formed the basis for\nthe jury’s decision. J.A. 9769. The district court denied\nATEN’s JMOL motion, upholding the jury’s finding of an-\nticipation based on the first theory, that the CS-1762 refer-\nence anticipated the asserted claims. It did not address the\nsecond theory. Because the jury’s verdict of anticipation\nwas not supported by substantial record evidence as to ei-\nther reference, we reverse the denial of JMOL on anticipa-\ntion.\n\f6 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\n CO., LTD.\n\n\n A\n To establish that an asserted reference is prior art un-\nder § 102(b), the patent challenger must prove, by clear and\nconvincing evidence, that it predates the critical date. Ma-\nhurkar v. C.R. Bard, Inc., 79 F.3d 1572, 1578 (Fed. Cir.\n1996).\n The undisputed critical date for the ’141 patent is July\n24, 2006—one year prior to its filing date. J.A. 4; J.A. 24;\nJ.A. 9302–9303 at 176:24–177:3. Uniclass’ expert\nMr. Dezmelyk relied on his testing of a 2009 CS-1762 spec-\nimen to meet certain limitations of claim 1 of the ’141 pa-\ntent. Uniclass argues the 2009 specimen qualifies as prior\nart because it uses firmware that was released in 2006 be-\nfore the critical date of the ’141 patent. See Appellee’s Br.\n20. But Mr. Dezmelyk testified merely that the firmware\ndated to 2006, without specifying a day or month. J.A.\n9313 at 14–17 (“I then looked on the wayback machine to\nfind when that firmware update came out and the last time\nit was changed, and that was in 2006. So this firmware is\nprior art firmware operating on this, the test hardware.”).\n To establish that the CS-1762 qualifies as prior art,\nUniclass was required to prove by clear and convincing ev-\nidence that the firmware existed prior to July 24, 2006.\nMerely establishing that the firmware existed in the same\nyear as the critical date is insufficient. This deficiency was\nhighlighted on cross-examination:\n Q. Yes or no, you did not testify to a particular\n month in 2006?\n A. No, I don’t believe I mentioned that in my testi-\n mony.\n Q. And you did not testify to a specific date in 2006?\n A. I don’t believe I testified to a particular day ei-\n ther.\n\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 7\nCO., LTD.\n\n\nJ.A. 9338 at 18–22. Mr. Dezmelyk did not provide any ad-\nditional date-related information, and Uniclass did not con-\nduct re-direct examination on this issue.\n With a critical date of July 24, 2006, testimony that the\nfirmware running on the CS-1762 device existed in 2006\nalone is not enough to support the jury’s finding that the\nfirmware pre-dated the critical date and thus qualifies as\nprior art. We hold that the district court erred in denying\nJMOL because substantial evidence does not support the\njury’s finding of anticipation based on the CS-1762 device.\n B\n A finding of anticipation requires clear and convincing\nevidence that “each and every element is found within a\nsingle prior art reference, arranged as claimed.” Summit 6,\nLLC v. Samsung Elecs. Co., 802 F.3d 1283, 1294 (Fed. Cir.\n2015). Testimony concerning anticipation must typically\n“explain in detail how each claim element is disclosed in\nthe prior art reference. The testimony is insufficient if it is\nmerely conclusory.” Schumer v. Lab. Comput. Sys., Inc.,\n308 F.3d 1304, 1315–16 (Fed. Cir. 2002).\n The jury verdict of anticipation cannot be upheld with\nregard to the GB ’540 reference. The asserted claims of the\n’141 patent require “emulating the input device to the first\nhost.” Uniclass’ expert Mr. Dezmelyk testified about the\nemulation described in GB ’540 in the context of other as-\nserted patents and claim limitations. J.A. 9296 at 170:10–\n20 (regarding “emulating the console devices according to the\nindustry standard”); J.A. 9301–9302 at 175:20–176:6 (de-\nscribing that the reference “has USB device emulators in it”);\nJ.A. 9285–9286 at 159:16–160:18 (describing the “USB host\nemulator”). But Uniclass offered no testimony explaining\nhow this emulation described in GB ’540 meets claim 1’s\nrequirement of “emulating . . . to the first host.” We also\ncannot identify any place in the three pages of GB ’540 pro-\nvided in the joint appendix that would support Uniclass’\ntheory that GB ’540 discloses this limitation. J.A. 9789;\n\f8 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\n CO., LTD.\n\n\nJ.A. 9790; J.A. 9793. We see no record evidence that GB\n’540 discloses “emulating the input device to the first host.”\nTo the extent that the jury finding of anticipation was\nbased on the GB ’540 reference, that finding lacks substan-\ntial evidence in this record.\n The district court erred in its denial of JMOL because\nthe jury’s finding of anticipation was not supported by sub-\nstantial evidence as to either reference.\n II\n On appeal, ATEN challenges the district court’s denial\nof JMOL as to infringement of the ’141 and ’289 patents.\nATEN argued in its motion for JMOL that the evidence\ncompels a finding of infringement, but the jury was con-\nfused due to Uniclass’ expert Mr. Dezmelyk’s testimony on\nthe scope of the claim terms. E.g., J.A. 8–12; J.A. 7937–\n7938. The district court denied the motion, holding ATEN\nwaived any right to contest Mr. Dezmelyk’s claim construc-\ntion testimony because it failed “to raise concerns regard-\ning the proper construction of claim terms at trial itself.”\nJ.A. 10.\n The record below indicates that Mr. Dezmelyk offered\ntrial testimony regarding the construction of the terms\n“disconnecting” and “connecting” in the ’141 patent. At the\nclaim construction stage, the district court gave the term\n“disconnecting the input device from the first host without\nconnecting the input device to any other host and starting\nemulating the input device to the first host” its plain and\nordinary meaning. ATEN did not object, nor did it proffer\na different construction. In fact, it was ATEN who urged\nthe court to adopt the plain and ordinary meaning. At trial,\nMr. Dezmelyk opined that the accused devices “never con-\nnect the input device to the host computer” based on his\ninterpretation of the plain and ordinary meaning of the\nterm “connected” as requiring a direct communication\npathway. J.A. 9237 at 15–16; see also J.A. 9240 at 16–18;\nJ.A. 9241 at 15–21. To inform his interpretation, Mr.\n\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 9\nCO., LTD.\n\n\nDezmelyk testified about the import of the prosecution his-\ntory. J.A. 9242 at 13–23 (“I look at the prosecution history\nto understand what ATEN . . . was telling the patent office\n. . . . In this case one of the statements they made to the\nexaminer was the applicant’s invention . . . has a funda-\nmental difference in comparison to the teaching of [a refer-\nence] with regard to the concepts of connected and\ndisconnected between the input device and the host de-\nvices.”). According to ATEN, this expert was testifying as\nto the proper claim construction to be given to these terms.\n The record also indicates that Mr. Dezmelyk offered\ntrial testimony regarding the meaning of “detecting\nwhether the first port is occupied” in the ’289 patent.\nMr. Dezmelyk opined that he believed “the ordinary mean-\ning of the phrase, that is what an engineer would think\nwhen they read this patent, what was meant by that\nphrase when he says detecting the first port is occupied, is\ndetecting if a cable was plugged into the port.” J.A. 9223\nat 1–5; see also J.A. 9223 at 17–18; J.A. 9221 at 20–24. To\ninform his interpretation, Mr. Dezmelyk again referred to\nthe prosecution history. J.A. 9228 at 7–12 (“[T]he con-\nnector is occupied in the [reference], but notice there is no\ndiscussion of messages . . . . In fact, . . . the [reference] does\nmeet the messages test. So it has to be a determination\nbased on the connector itself actually having a plug in it.”);\nsee also J.A. 9226 at 13–14.\n Mr. Dezmelyk’s testimony on both patents amounts to\nclaim construction testimony before the jury. ATEN did\nnot object to any of this testimony during the trial, nor did\nATEN move for the district court to resolve these claim con-\nstruction disputes. The only objection to the expert’s testi-\nmony came after the close of evidence when ATEN orally\nmoved for a “post-testifying Daubert on Mr. Dezmelyk’s\ntestimony on invalidity with respect to anything that\n\f10 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\n CO., LTD.\n\n\nrequired an ATEN claim construction.” 1 J.A. 9357–9358 at\n231:23–232:1 (italics added). ATEN now argues its objec-\ntion regarding invalidity should be broadened to apply to\nclaim construction for infringement purposes. We do not\nagree. Even if ATEN’s oral motion was timely, we agree\nwith the district court that ATEN did not object to\nMr. Dezmelyk’s claim construction testimony with respect\nto infringement. J.A. 9–10. By failing to object, ATEN has\nwaived any challenge to the jury’s finding of infringement\nbased on this testimony.\n It is beyond dispute that claim construction issues are\nto be decided by the court. It is thus improper for an expert\nwitness to testify before the jury regarding claim construc-\ntion. CytoLogix Corp. v. Ventana Med. Sys., Inc., 424 F.3d\n1168, 1172 (Fed. Cir. 2005). In CytoLogix, the parties\nagreed to present expert witness testimony regarding\nclaim construction to the jury and counsel argued conflict-\ning claim constructions to the jury. Id. We noted the dis-\ntrict court should have refused to allow such testimony,\ndespite the agreement of the parties, because the risk of\nconfusing the jury was high. Id. But here, as in CytoLogix,\n“there is no ground for reversal since there was no objection\nto the expert testimony as to claim construction.” Id. at\n1173. We do not fault the district court here for allowing\nclaim construction to go to the jury. It is the parties’ obli-\ngation to raise a dispute regarding the proper scope of\nclaims to the court. O2 Micro Int’l Ltd. v. Beyond Innova-\ntion Tech. Co., 521 F.3d 1351, 1362 (Fed. Cir. 2008) (“When\nthe parties present a fundamental dispute regarding the\n\n\n\n 1 ATEN also filed a motion for JMOL on the final day\nof trial that briefly discussed the term “occupied” of the ’289\npatent. J.A. 7937. ATEN did not raise this issue at the\nhearing. J.A. 10–11. We agree with the district court that\nATEN failed to preserve its claim construction challenge\nvia this motion. J.A. 11.\n\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 11\nCO., LTD.\n\n\nscope of a claim term, it is the court’s duty to resolve it.”).\nWe therefore affirm the district court’s denial of JMOL as\nto noninfringement.\n CONCLUSION\n For the reasons discussed above, we reverse as to inva-\nlidity and affirm as to noninfringement. Because we affirm\nas to noninfringement, we need not reach lost profits.\n AFFIRMED-IN-PART, REVERSED-IN-PART\n COSTS\n No costs.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: ATEN INTERNATIONAL CO., LTD., Plaintiff-Appellant v. UNICLASS TECHNOLOGY CO., LTD., Electronic Technology Co., Ltd., Airlink 101, Phoebe Micro, Inc., Broadtech International Co., Ltd., Dba Linkskey, Black Box Corporation, Black Box Corporation of Pennsylvania, Defendants-Appellees\\nCourt: U.S. Court of Appeals for the Federal Circuit\\nDate decided: 2019-08-06\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n ATEN INTERNATIONAL CO., LTD.,\\n Plaintiff-Appellant\\n\\n v.\\n\\n UNICLASS TECHNOLOGY CO., LTD.,\\nELECTRONIC TECHNOLOGY CO., LTD., AIRLINK\\n 101, PHOEBE MICRO, INC., BROADTECH\\n INTERNATIONAL CO., LTD., DBA LINKSKEY,\\n BLACK BOX CORPORATION, BLACK BOX\\n CORPORATION OF PENNSYLVANIA,\\n Defendants-Appellees\\n ______________________\\n\\n 2018-1606\\n ______________________\\n\\n Appeal from the United States District Court for the\\nCentral District of California in No. 2:15-cv-04424-AG-\\nAJW, Judge Andrew J. Guilford.\\n ______________________\\n\\n Decided: August 6, 2019\\n ______________________\\n\\n LAURENCE M. SANDELL, Mei & Mark LLP, Washington,\\nDC, argued for plaintiff-appellant. Also represented by LEI\\nMEI.\\n\\n JOSEPH PIA, Pia Anderson Moss Hoyt, Salt Lake City,\\nUT, argued for defendants-appellees. Also represented by\\nROBERT AYCOCK.\\n\\f2 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\n ______________________\\n\\n Before MOORE, WALLACH, and TARANTO, Circuit Judges.\\nMOORE, Circuit Judge.\\n ATEN International Co., Ltd. (“ATEN”) appeals the\\nUnited States District Court for the Central District of Cal-\\nifornia’s denial of judgment as a matter of law (“JMOL”),\\nwhich declined to overturn the jury’s findings that the as-\\nserted claims of U.S. Patent No. 8,589,141 are invalid as\\nanticipated under 35 U.S.C. § 102 and not infringed; and\\nthat the asserted claims of U.S. Patent No. 7,640,289 are\\nnot infringed. For the reasons discussed below, we reverse\\nas to invalidity and affirm as to noninfringement.\\n BACKGROUND\\n Uniclass Technology Co., Ltd. (“Uniclass”) and ATEN\\nare involved in making and selling keyboard-video-mouse\\n(“KVM”) switch systems that allow a user to control multi-\\nple computers from a single keyboard, video device, and\\nmouse. ATEN sued Uniclass as well as Electronic Technol-\\nogy Co., Ltd.; Airlink 101; Phoebe Micro, Inc.; Broadtech\\nInternational Co., Ltd. d/b/a Linkskey; Black Box Corpora-\\ntion; and Black Box Corporation of Pennsylvania (collec-\\ntively, the “customer defendants”) alleging, as relevant\\nhere, infringement of claims 3, 8, and 10 of the ’141 patent\\nand claims 1–20 of the ’289 patent. The ’141 patent is di-\\nrected to technology for switching between computers that\\nshare a keyboard, monitor, and mouse through a KVM\\nswitch, such as a keyboard shortcut. ’141 patent at 2:51–\\n61. Claim 3 depends from claims 1 and 2, and claim 8 de-\\npends from claim 1. Independent claim 1 recites:\\n 1. A method for controlling a resource sharing ap-\\n paratus coupling at least one input device to a plu-\\n rality of hosts including a first host, the method\\n comprising:\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 3\\nCO., LTD.\\n\\n\\n connecting the input device to the first\\n host;\\n while the input device is connected to the\\n first host, acquiring a first input signal\\n from the input device and determining\\n whether the first input signal comprises a\\n standby indication of a switch command\\n and wherein the standby indication is for\\n indicating that connection between the in-\\n put device and the first host can be\\n changed; and\\n in response to a determination that the\\n first input signal comprises the standby in-\\n dication, disconnecting the input device\\n from the first host without connecting the\\n input device to any other host and starting\\n emulating the input device to the first host;\\n and\\n acquiring a second input signal from the in-\\n put device, wherein the second input signal\\n is not transferred to the hosts when it is in-\\n putted to the resource sharing apparatus.\\n The ’289 patent is directed to technology for stringing\\ntogether several KVM switches. ’289 patent at 2:1–6. It\\nprovides that each KVM switch can detect whether it is a\\nmaster or slave by, for example, detecting whether the port\\nused to connect other KVM switches is occupied. Id. at 4:1–\\n4. Independent claim 1 recites:\\n 1. A computer switch comprising:\\n a set of peripheral device ports for connect-\\n ing to a set of peripheral devices;\\n a first port;\\n a second port;\\n\\f4 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\n a control device coupled to the first port and\\n the second port, the control device repeat-\\n edly detecting whether the first port is oc-\\n cupied to repeatedly determine a master or\\n slave status of the computer switch,\\n wherein\\n if the first port is occupied, the control de-\\n vice determines the computer switch to be\\n a slave and provides data to a first external\\n computer switch connected to the first port,\\n and\\n if the first port is unoccupied, the control\\n device determines the computer switch to\\n be a master and controls a second external\\n computer switch connected to the second\\n port and obtains data from the second ex-\\n ternal computer switch; and\\n a plurality of computer connection ports\\n coupled to the control device for connecting\\n to and controlling a plurality of computers,\\n wherein the plurality of computers are con-\\n trolled by the set of peripheral devices if the\\n computer switch is determined to be a mas-\\n ter, and are controlled by the first external\\n computer switch connected to the first port\\n if the computer switch is determined to be\\n a slave.\\n At trial, the jury found that Uniclass did not infringe\\nthe asserted claims of the ’141 or ’289 patents. It also found\\nthe asserted claims of the ’141 patent were invalid as an-\\nticipated without specifying which reference was the basis\\nfor its finding. ATEN moved for JMOL, which the district\\ncourt denied in the aspects relevant to this appeal.\\n ATEN timely appealed. We have jurisdiction under 28\\nU.S.C. § 1295(a)(1).\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 5\\nCO., LTD.\\n\\n\\n DISCUSSION\\n We review denials of JMOL under the law of the re-\\ngional circuit. TVIIM, LLC v. McAfee, Inc., 851 F.3d 1356,\\n1362 (Fed. Cir. 2017). The Ninth Circuit reviews a denial\\nof JMOL de novo. Harper v. City of Los Angeles, 533 F.3d\\n1010, 1021 (9th Cir. 2008). JMOL is proper when the evi-\\ndence permits only one reasonable conclusion which is con-\\ntrary to the jury’s verdict, but the jury’s verdict must be\\nupheld if it is supported by substantial evidence. Id. An-\\nticipation and infringement are questions of fact that we\\nreview for substantial evidence. Cordis Corp. v. Bos. Sci.\\nCorp., 561 F.3d 1319, 1330, 1335 (Fed. Cir. 2009). Whether\\na reference is prior art is a question of law based on under-\\nlying factual questions. TypeRight Keyboard Corp. v. Mi-\\ncrosoft Corp., 374 F.3d 1151, 1157 (Fed. Cir. 2004).\\nWhether a reference is publicly accessible is a question of\\nfact. In re NTP, Inc., 654 F.3d 1279, 1296 (Fed. Cir. 2011).\\n I\\n At trial, Uniclass asserted anticipation of the claims of\\nthe ’141 patent based on two references (1) CS-1762, an\\nearlier product by ATEN, and/or its user manual; and (2)\\nGreat Britain Patent No. 2,352,540 (“GB ’540”). The jury\\nfound that the asserted claims of the ’141 patent were in-\\nvalid as anticipated, but there was no special verdict indi-\\ncating whether one or both references formed the basis for\\nthe jury’s decision. J.A. 9769. The district court denied\\nATEN’s JMOL motion, upholding the jury’s finding of an-\\nticipation based on the first theory, that the CS-1762 refer-\\nence anticipated the asserted claims. It did not address the\\nsecond theory. Because the jury’s verdict of anticipation\\nwas not supported by substantial record evidence as to ei-\\nther reference, we reverse the denial of JMOL on anticipa-\\ntion.\\n\\f6 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\n A\\n To establish that an asserted reference is prior art un-\\nder § 102(b), the patent challenger must prove, by clear and\\nconvincing evidence, that it predates the critical date. Ma-\\nhurkar v. C.R. Bard, Inc., 79 F.3d 1572, 1578 (Fed. Cir.\\n1996).\\n The undisputed critical date for the ’141 patent is July\\n24, 2006—one year prior to its filing date. J.A. 4; J.A. 24;\\nJ.A. 9302–9303 at 176:24–177:3. Uniclass’ expert\\nMr. Dezmelyk relied on his testing of a 2009 CS-1762 spec-\\nimen to meet certain limitations of claim 1 of the ’141 pa-\\ntent. Uniclass argues the 2009 specimen qualifies as prior\\nart because it uses firmware that was released in 2006 be-\\nfore the critical date of the ’141 patent. See Appellee’s Br.\\n20. But Mr. Dezmelyk testified merely that the firmware\\ndated to 2006, without specifying a day or month. J.A.\\n9313 at 14–17 (“I then looked on the wayback machine to\\nfind when that firmware update came out and the last time\\nit was changed, and that was in 2006. So this firmware is\\nprior art firmware operating on this, the test hardware.”).\\n To establish that the CS-1762 qualifies as prior art,\\nUniclass was required to prove by clear and convincing ev-\\nidence that the firmware existed prior to July 24, 2006.\\nMerely establishing that the firmware existed in the same\\nyear as the critical date is insufficient. This deficiency was\\nhighlighted on cross-examination:\\n Q. Yes or no, you did not testify to a particular\\n month in 2006?\\n A. No, I don’t believe I mentioned that in my testi-\\n mony.\\n Q. And you did not testify to a specific date in 2006?\\n A. I don’t believe I testified to a particular day ei-\\n ther.\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 7\\nCO., LTD.\\n\\n\\nJ.A. 9338 at 18–22. Mr. Dezmelyk did not provide any ad-\\nditional date-related information, and Uniclass did not con-\\nduct re-direct examination on this issue.\\n With a critical date of July 24, 2006, testimony that the\\nfirmware running on the CS-1762 device existed in 2006\\nalone is not enough to support the jury’s finding that the\\nfirmware pre-dated the critical date and thus qualifies as\\nprior art. We hold that the district court erred in denying\\nJMOL because substantial evidence does not support the\\njury’s finding of anticipation based on the CS-1762 device.\\n B\\n A finding of anticipation requires clear and convincing\\nevidence that “each and every element is found within a\\nsingle prior art reference, arranged as claimed.” Summit 6,\\nLLC v. Samsung Elecs. Co., 802 F.3d 1283, 1294 (Fed. Cir.\\n2015). Testimony concerning anticipation must typically\\n“explain in detail how each claim element is disclosed in\\nthe prior art reference. The testimony is insufficient if it is\\nmerely conclusory.” Schumer v. Lab. Comput. Sys., Inc.,\\n308 F.3d 1304, 1315–16 (Fed. Cir. 2002).\\n The jury verdict of anticipation cannot be upheld with\\nregard to the GB ’540 reference. The asserted claims of the\\n’141 patent require “emulating the input device to the first\\nhost.” Uniclass’ expert Mr. Dezmelyk testified about the\\nemulation described in GB ’540 in the context of other as-\\nserted patents and claim limitations. J.A. 9296 at 170:10–\\n20 (regarding “emulating the console devices according to the\\nindustry standard”); J.A. 9301–9302 at 175:20–176:6 (de-\\nscribing that the reference “has USB device emulators in it”);\\nJ.A. 9285–9286 at 159:16–160:18 (describing the “USB host\\nemulator”). But Uniclass offered no testimony explaining\\nhow this emulation described in GB ’540 meets claim 1’s\\nrequirement of “emulating . . . to the first host.” We also\\ncannot identify any place in the three pages of GB ’540 pro-\\nvided in the joint appendix that would support Uniclass’\\ntheory that GB ’540 discloses this limitation. J.A. 9789;\\n\\f8 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\nJ.A. 9790; J.A. 9793. We see no record evidence that GB\\n’540 discloses “emulating the input device to the first host.”\\nTo the extent that the jury finding of anticipation was\\nbased on the GB ’540 reference, that finding lacks substan-\\ntial evidence in this record.\\n The district court erred in its denial of JMOL because\\nthe jury’s finding of anticipation was not supported by sub-\\nstantial evidence as to either reference.\\n II\\n On appeal, ATEN challenges the district court’s denial\\nof JMOL as to infringement of the ’141 and ’289 patents.\\nATEN argued in its motion for JMOL that the evidence\\ncompels a finding of infringement, but the jury was con-\\nfused due to Uniclass’ expert Mr. Dezmelyk’s testimony on\\nthe scope of the claim terms. E.g., J.A. 8–12; J.A. 7937–\\n7938. The district court denied the motion, holding ATEN\\nwaived any right to contest Mr. Dezmelyk’s claim construc-\\ntion testimony because it failed “to raise concerns regard-\\ning the proper construction of claim terms at trial itself.”\\nJ.A. 10.\\n The record below indicates that Mr. Dezmelyk offered\\ntrial testimony regarding the construction of the terms\\n“disconnecting” and “connecting” in the ’141 patent. At the\\nclaim construction stage, the district court gave the term\\n“disconnecting the input device from the first host without\\nconnecting the input device to any other host and starting\\nemulating the input device to the first host” its plain and\\nordinary meaning. ATEN did not object, nor did it proffer\\na different construction. In fact, it was ATEN who urged\\nthe court to adopt the plain and ordinary meaning. At trial,\\nMr. Dezmelyk opined that the accused devices “never con-\\nnect the input device to the host computer” based on his\\ninterpretation of the plain and ordinary meaning of the\\nterm “connected” as requiring a direct communication\\npathway. J.A. 9237 at 15–16; see also J.A. 9240 at 16–18;\\nJ.A. 9241 at 15–21. To inform his interpretation, Mr.\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 9\\nCO., LTD.\\n\\n\\nDezmelyk testified about the import of the prosecution his-\\ntory. J.A. 9242 at 13–23 (“I look at the prosecution history\\nto understand what ATEN . . . was telling the patent office\\n. . . . In this case one of the statements they made to the\\nexaminer was the applicant’s invention . . . has a funda-\\nmental difference in comparison to the teaching of [a refer-\\nence] with regard to the concepts of connected and\\ndisconnected between the input device and the host de-\\nvices.”). According to ATEN, this expert was testifying as\\nto the proper claim construction to be given to these terms.\\n The record also indicates that Mr. Dezmelyk offered\\ntrial testimony regarding the meaning of “detecting\\nwhether the first port is occupied” in the ’289 patent.\\nMr. Dezmelyk opined that he believed “the ordinary mean-\\ning of the phrase, that is what an engineer would think\\nwhen they read this patent, what was meant by that\\nphrase when he says detecting the first port is occupied, is\\ndetecting if a cable was plugged into the port.” J.A. 9223\\nat 1–5; see also J.A. 9223 at 17–18; J.A. 9221 at 20–24. To\\ninform his interpretation, Mr. Dezmelyk again referred to\\nthe prosecution history. J.A. 9228 at 7–12 (“[T]he con-\\nnector is occupied in the [reference], but notice there is no\\ndiscussion of messages . . . . In fact, . . . the [reference] does\\nmeet the messages test. So it has to be a determination\\nbased on the connector itself actually having a plug in it.”);\\nsee also J.A. 9226 at 13–14.\\n Mr. Dezmelyk’s testimony on both patents amounts to\\nclaim construction testimony before the jury. ATEN did\\nnot object to any of this testimony during the trial, nor did\\nATEN move for the district court to resolve these claim con-\\nstruction disputes. The only objection to the expert’s testi-\\nmony came after the close of evidence when ATEN orally\\nmoved for a “post-testifying Daubert on Mr. Dezmelyk’s\\ntestimony on invalidity with respect to anything that\\n\\f10 ATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY\\n CO., LTD.\\n\\n\\nrequired an ATEN claim construction.” 1 J.A. 9357–9358 at\\n231:23–232:1 (italics added). ATEN now argues its objec-\\ntion regarding invalidity should be broadened to apply to\\nclaim construction for infringement purposes. We do not\\nagree. Even if ATEN’s oral motion was timely, we agree\\nwith the district court that ATEN did not object to\\nMr. Dezmelyk’s claim construction testimony with respect\\nto infringement. J.A. 9–10. By failing to object, ATEN has\\nwaived any challenge to the jury’s finding of infringement\\nbased on this testimony.\\n It is beyond dispute that claim construction issues are\\nto be decided by the court. It is thus improper for an expert\\nwitness to testify before the jury regarding claim construc-\\ntion. CytoLogix Corp. v. Ventana Med. Sys., Inc., 424 F.3d\\n1168, 1172 (Fed. Cir. 2005). In CytoLogix, the parties\\nagreed to present expert witness testimony regarding\\nclaim construction to the jury and counsel argued conflict-\\ning claim constructions to the jury. Id. We noted the dis-\\ntrict court should have refused to allow such testimony,\\ndespite the agreement of the parties, because the risk of\\nconfusing the jury was high. Id. But here, as in CytoLogix,\\n“there is no ground for reversal since there was no objection\\nto the expert testimony as to claim construction.” Id. at\\n1173. We do not fault the district court here for allowing\\nclaim construction to go to the jury. It is the parties’ obli-\\ngation to raise a dispute regarding the proper scope of\\nclaims to the court. O2 Micro Int’l Ltd. v. Beyond Innova-\\ntion Tech. Co., 521 F.3d 1351, 1362 (Fed. Cir. 2008) (“When\\nthe parties present a fundamental dispute regarding the\\n\\n\\n\\n 1 ATEN also filed a motion for JMOL on the final day\\nof trial that briefly discussed the term “occupied” of the ’289\\npatent. J.A. 7937. ATEN did not raise this issue at the\\nhearing. J.A. 10–11. We agree with the district court that\\nATEN failed to preserve its claim construction challenge\\nvia this motion. J.A. 11.\\n\\fATEN INTERNATIONAL CO., LTD. v. UNICLASS TECHNOLOGY 11\\nCO., LTD.\\n\\n\\nscope of a claim term, it is the court’s duty to resolve it.”).\\nWe therefore affirm the district court’s denial of JMOL as\\nto noninfringement.\\n CONCLUSION\\n For the reasons discussed above, we reverse as to inva-\\nlidity and affirm as to noninfringement. Because we affirm\\nas to noninfringement, we need not reach lost profits.\\n AFFIRMED-IN-PART, REVERSED-IN-PART\\n COSTS\\n No costs.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4195,"cells":{"id":{"kind":"number","value":4645831,"string":"4,645,831"},"name":{"kind":"string","value":"AJINOMOTO CO., INC., Ajinomoto Heartland Inc., Appellants v. INTERNATIONAL TRADE COMMISSION, Appellee CJ CheilJedang Corp., CJ America, Inc., PT CheilJedang Indonesia, Intervenors CJ CheilJedang Corp., CJ America, Inc., PT CheilJedang Indonesia, Appellants v. International Trade Commission, Appellee Ajinomoto Co., Inc., Ajinomoto Heartland Inc., Intervenors"},"name_abbreviation":{"kind":"string","value":"Ajinomoto Co. v. Int'l Trade Comm'n"},"decision_date":{"kind":"string","value":"2019-08-06"},"docket_number":{"kind":"string","value":"2018-1590, 2018-1629"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Federal Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Dyk, Moore, Taranto\", \"parties\": \"\", \"opinions\": [{\"author\": \"TARANTO, Circuit Judge.\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\\n INC.,\\n Appellants\\n\\n v.\\n\\n INTERNATIONAL TRADE COMMISSION,\\n Appellee\\n\\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\\n CHEILJEDANG INDONESIA,\\n Intervenors\\n\\n--------------------------------------------\\n\\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\\n CHEILJEDANG INDONESIA,\\n Appellants\\n\\n v.\\n\\n INTERNATIONAL TRADE COMMISSION,\\n Appellee\\n\\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\\n INC.,\\n Intervenors\\n ______________________\\n\\n 2018-1590, 2018-1629\\n ______________________\\n\\n2 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\n Appeals from the United States International Trade\\nCommission in Investigation No. 337-TA-1005.\\n ______________________\\n\\n Decided: August 6, 2019\\n ______________________\\n\\n JOHN D. LIVINGSTONE, Finnegan, Henderson, Farabow,\\nGarrett & Dunner, LLP, Atlanta, GA, argued for Ajinomoto\\nCo., Inc., Ajinomoto Heartland Inc. Also represented by\\nMARTIN DAVID WEINGARTEN; CHARLES E. LIPSEY, Reston,\\nVA; MAREESA ARNITA FREDERICK, CORA RENAE HOLT,\\nBARBARA RUDOLPH, Washington, DC.\\n\\n HOUDA MORAD, Office of General Counsel, United\\nStates International Trade Commission, Washington, DC,\\nargued for appellee. Also represented by SIDNEY A.\\nROSENZWEIG, DOMINIC L. BIANCHI, WAYNE W. HERRINGTON.\\n\\n JAMES F. HALEY, JR., Haley Guiliano LLP, New York,\\nNY, argued for CJ CheilJedang Corp., CJ America, Inc., PT\\nCheiJedang Indonesia. Also represented by STEVEN PEPE,\\nRopes & Gray LLP, New York, NY; MATTHEW RIZZOLO,\\nWashington, DC.\\n ______________________\\n\\n Before DYK, MOORE, and TARANTO, Circuit Judges.\\n Opinion for the court filed by Circuit Judge TARANTO.\\nOpinion concurring in part and dissenting in part filed by\\n Circuit Judge DYK.\\nTARANTO, Circuit Judge.\\n Ajinomoto Co., Inc. and Ajinomoto Heartland Inc. (col-\\nlectively, Ajinomoto) filed a complaint against CJ\\nCheilJedang Corp., CJ America, Inc., and PT CheilJedang\\nIndonesia (collectively, CJ) with the International Trade\\nCommission, alleging that CJ was importing certain\\n\\nAJINOMOTO CO., INC. v. ITC 3\\n\\n\\n\\nproducts that infringed Ajinomoto’s U.S. Patent No.\\n7,666,655. CJ used several strains of Escherichia coli bac-\\nteria to produce L-tryptophan products, which it then im-\\nported into the United States. The Commission\\ndetermined that CJ’s earlier strains did not infringe but\\nthat CJ’s two later strains did. The Commission also found\\nthat the relevant claim of the ’655 patent is not invalid for\\nlack of an adequate written description.\\n Ajinomoto appeals the Commission’s claim construc-\\ntion underlying the determination of no infringement by\\nthe earlier strains. CJ cross-appeals aspects of the deter-\\nmination of infringement by the later strains and the rejec-\\ntion of the invalidity challenge. We affirm.\\n I\\n A\\n The ’655 patent claims E. coli bacteria that have been\\ngenetically engineered to increase their production of aro-\\nmatic L-amino acids, such as L-tryptophan, during fermen-\\ntation, as well as methods of producing aromatic L-amino\\nacids using such bacteria. See ’655 patent, col. 2, lines 40–\\n45. In particular, the ’655 patent identifies a specific gene\\nin the E. coli genome, the yddG gene, that encodes a mem-\\nbrane protein, the YddG protein. Id., col. 2, lines 46–48.\\nThat protein transports aromatic L-amino acids out of the\\nbacterial cell and into the surrounding culture medium,\\nwhere they can be collected. See id., col. 7, lines 11–16.\\nWhen yddG gene activity in bacteria is enhanced so that\\nmore YddG protein is produced, the bacteria show in-\\ncreased production of, and increased resistance to, aro-\\nmatic L-amino acids. Id., col. 2, lines 49–57. 1\\n\\n\\n\\n 1 The specification defines a bacterium’s “resistance”\\nto an amino acid as its ability “to grow on a minimal me-\\ndium containing” the amino acid on “which unmodified or\\n\\n4 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\n The ’655 patent describes three ways to enhance the\\nactivity of the yddG gene. First, plasmids containing addi-\\ntional copies of the yddG gene can be introduced into the\\nbacterium. Id., col. 2, lines 50–52; id., col. 5, line 62,\\nthrough col. 6, line 2. Second, additional copies of the yddG\\ngene can be inserted into the bacterial chromosome. Id.,\\ncol. 2, lines 52–54; id., col. 6, lines 3–6. Third, a stronger\\n“promoter” than the one native to the E. coli yddG gene can\\nbe used. Id., col. 2, lines 54–57; id., col. 6, lines 12–15. 2\\n Claim 20, the only claim of the ’655 patent still asserted\\nwhen the Commission issued its decision, claims “[a]\\nmethod for producing an aromatic L-amino acid, which\\ncomprises cultivating the bacterium according to any\\none of claims 9–12, 13, 14, 15–18, or 19.” Id., col. 24, lines\\n\\n\\n\\nthe wild type, or the parental strain of the bacterium can-\\nnot grow,” or its ability “to grow faster” on such a medium\\n“than unmodified or the wild type, or the parental strain of\\nthe bacterium.” ’655 patent, col. 4, lines 49–56.\\n 2 A promoter is a nucleotide sequence within a DNA\\nmolecule, located adjacent to the nucleotide sequence that\\nconstitutes the gene to be expressed. The Lewin textbook\\ncited by Ajinomoto shows a “typical promoter” around 41\\nnucleotides long. J.A. 6043; see also J.A. 6177 (article by\\nDeuschle et al., cited at ’655 patent, col. 6, lines 18–21,\\nshowing longer promoters). The promoter is the binding\\nsite for RNA polymerase, which initiates transcription (the\\nfirst step in gene expression) by separating the two strands\\nof DNA. The ’655 patent’s specification defines “[s]trength\\nof promoter” with reference to the “frequency of acts of the\\nRNA synthesis initiation.” ’655 patent, col. 6, lines 15–16.\\n The promoter is only one part of a gene’s “expression\\nregulation sequence,” which controls expression of the\\ngene. See id., col. 3, line 14; id., col. 5, line 2. Besides pro-\\nmoters, the “expression regulation sequence” can include,\\ne.g., operators, enhancers, terminators, and silencers.\\n\\nAJINOMOTO CO., INC. v. ITC 5\\n\\n\\n\\n4–6 (emphasis added). Of the claims in that list, claims 9\\nand 15 are the independent claims, and they are the two\\nalternatives, under claim 20, of importance in this case.\\n Claim 9 recites:\\n 9. A recombinant Escherichia coli bacterium,\\n which has the ability to accumulate aromatic L-\\n amino acid in a medium, wherein the aromatic L-\\n amino acid production by said bacterium is en-\\n hanced by enhancing activity of a protein in a cell\\n of said bacterium beyond the levels observed in a\\n wild-type of said bacterium,\\n [1] and in which said protein consists of the\\n amino acid sequence of SEQ ID NO: 2\\n [2] and said protein has the activity to make\\n the bacterium resistant to L-phenylalanine, fluoro-\\n phenylalanine or 5[-]fluoro-DL-tryptophan,\\n [3] wherein the activity of the protein is en-\\n hanced by [3a] transformation of the bacterium\\n with a DNA encoding the protein to express the\\n protein in the bacterium, [3b] by replacing the na-\\n tive promoter which precedes the DNA on the chro-\\n mosome of the bacterium with a more potent\\n promoter, [3c] or by introduction of multiple copies\\n of the DNA encoding said protein into the chromo-\\n some of said bacterium to express the protein in\\n said bacterium.\\nId., col. 22, lines 51–67 (paragraph breaks and bold num-\\nbering added). The Commission referred to limitation [1]\\nas the “protein limitation,” limitation [2] as the “resistance\\nlimitation,” and limitation [3] as the “enhancement limita-\\ntion.” Claim 15 is materially identical to claim 9, except for\\nthe protein limitation. Whereas claim 9 identifies the\\nclaimed protein by a specific amino-acid sequence, claim 15\\nidentifies it by reference to a corresponding DNA se-\\nquence—a protein “encoded by the nucleotide sequence\\n\\n6 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nwhich hybridizes with the complement of the nucleotide se-\\nquence of SEQ ID NO: 1 under” certain conditions. See id.,\\ncol. 23, lines 14–32.\\n B\\n In May 2016, Ajinomoto filed a complaint against CJ\\nwith the Commission under 19 U.S.C. § 1337. Ajinomoto\\nalleged that CJ violated § 1337(a)(1)(B)(ii) by importing an-\\nimal-feed-grade L-tryptophan products produced by a pro-\\ncess covered by the ’655 patent. 3 The Commission\\ninstituted an investigation based on Ajinomoto’s com-\\nplaint.\\n The parties before us, including the Commission, agree\\nthat whether the accused products were produced by a pro-\\ncess covered by the patent is a question of infringement.\\nThe proceeding focused on three groups of E. coli strains\\nthat CJ has used to produce tryptophan. First, CJ’s “ear-\\nlier strains” contained both the native E. coli yddG gene\\nand the native E. coli yddG promoter, except that the first\\nnucleotide of the promoter was changed through chemical\\nmutagenesis, resulting in a stronger promoter. Second, in\\nNovember 2016, several months after Ajinomoto filed its\\ncomplaint, CJ began using its first “later strain,” which\\ncontained two copies of a yddG gene: (1) the native E. coli\\nyddG gene with the native E. coli yddG promoter; and (2) a\\nnon-E. coli yddG gene with two promoters—(2a) a native\\nnon-E. coli yddG promoter and (2b) an rmf promoter. 4\\nThird, in December 2016, CJ started using its second “later\\n\\n\\n\\n 3 Ajinomoto also alleged that CJ infringed U.S. Pa-\\ntent No. 6,180,373, which similarly claims methods of pro-\\nducing tryptophan using genetically engineered bacteria.\\nThe ’373 patent expired on January 30, 2018, and is not at\\nissue in this court.\\n 4 The rmf and rhtB promoters are promoters associ-\\nated with other genes in the E. coli genome.\\n\\nAJINOMOTO CO., INC. v. ITC 7\\n\\n\\n\\nstrain,” which also contained two copies of a yddG gene:\\n(1) the native E. coli yddG gene with the native E. coli\\nyddG promoter; and (2) a codon-randomized non-E. coli\\nyddG gene with two promoters—(2a) an rmf promoter and\\n(2b) an rhtB promoter. 5\\n In August 2017, the administrative law judge (ALJ) is-\\nsued a final initial determination. The ALJ construed “re-\\nplacing the native promoter . . . with a more potent\\npromoter” in the enhancement limitation to mean “remov-\\ning the native upstream region of the yddG gene and in-\\nserting one of a class of promoters that controls expression\\nof a different gene.” J.A. 90–91. Using that construction,\\nthe ALJ found that CJ’s earlier strains did not infringe; he\\nfound that they failed to meet the enhancement limitation\\nbecause CJ created the more potent promoter in those\\nstrains by mutagenesis of a single nucleotide rather than\\nremoval of the entire native promoter and insertion of a\\nnew promoter. As to CJ’s later strains, the ALJ found that\\n(a) the first later strain did not infringe because Ajinomoto\\nhad failed to prove that it met the resistance limitation,\\nand (b) the second later strain also did not infringe because\\n\\n\\n\\n 5 Each particular codon (three nucleotides in a row\\non a DNA molecule) that encodes for an amino acid always\\nencodes for the same amino acid, but many of the 20 amino\\nacids are encoded by more than one of the 64 codons. See\\nAmgen, Inc. v. Chugai Pharm. Co., 927 F.2d 1200, 1208 n.4\\n(Fed. Cir. 1991) (discussing “redundancy” of genetic code).\\nFor instance, the DNA sequences TTA and TTG both code\\nfor the amino acid leucine. “Codon randomization” refers\\nto creation of DNA molecules that use different codons (e.g.,\\nTTA or TTG) to code for the same amino acid (e.g., leucine)\\nin building the same protein. See Mycogen Plant Sci. v.\\nMonsanto Co., 243 F.3d 1316, 1323 (Fed. Cir. 2001) (“[O]ne\\ncodon can be substituted for another in the gene without\\nchanging the amino acid and resulting protein.”).\\n\\n8 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nits non-E. coli YddG protein was not equivalent to the\\nclaimed E. coli YddG protein under the doctrine of equiva-\\nlents. Finally, the ALJ found that claim 20 of the ’655 pa-\\ntent is invalid for lack of an adequate written description\\nof the “more potent promoter” limitation incorporated into\\nthat claim.\\n In October 2017, the full Commission decided to review\\nthe ALJ’s final initial determination in its entirety, and in\\nDecember 2017, the Commission issued its decision. It af-\\nfirmed the ALJ’s construction of “replacing the native pro-\\nmoter . . . with a more potent promoter” and accordingly\\naffirmed the ALJ’s finding that CJ’s earlier strains did not\\ninfringe. But the Commission reversed several of the ALJ’s\\nother findings. Specifically, it determined that both of CJ’s\\nlater strains met all disputed claim limitations and thus\\ninfringed claim 20 and that claim 20 was not proved to lack\\nan adequate written description. The Commission accord-\\ningly entered a limited exclusion order against CJ’s infring-\\ning products, i.e., those made by both of CJ’s later strains\\nbut not its earlier strains. The Commission also issued a\\ncease-and-desist order against CJ America, which held in-\\nventory of the infringing products.\\n Ajinomoto and CJ both timely appealed. We have ju-\\nrisdiction under 28 U.S.C. § 1295(a)(6).\\n II\\n We begin with Ajinomoto’s appeal of the Commission’s\\nfinding of no infringement by the earlier strains.\\nAjinomoto challenges that finding solely by arguing that\\nthe Commission erred in its claim construction of “replac-\\ning the native promoter . . . with a more potent promoter.”\\nAjinomoto argues that, properly construed, the phrase is\\nnot limited to removing the entire native promoter and in-\\nserting a new promoter, as the Commission concluded, but\\nencompasses mutagenesis of individual nucleotides within\\nthe native promoter. We review the Commission’s claim\\nconstruction de novo, as the Commission relied on only\\n\\nAJINOMOTO CO., INC. v. ITC 9\\n\\n\\n\\nintrinsic evidence and made no factual findings based on\\nextrinsic evidence. Teva Pharm. USA, Inc. v. Sandoz, Inc.,\\n135 S. Ct. 831, 841 (2015); see Cont’l Circuits LLC v. Intel\\nCorp., 915 F.3d 788, 795 (Fed. Cir. 2019). We agree with\\nthe Commission’s claim construction and therefore affirm\\nthe non-infringement finding.\\n The ordinary and customary meaning of the claim lan-\\nguage provides support for the Commission’s claim con-\\nstruction. The language of “replacing the native promoter\\n. . . with a more potent promoter” suggests, in ordinary par-\\nlance, an operation at the level of the entire promoter as a\\nunit, not at the level of a single nucleotide that is just one\\nsmall component of the promoter. To say that one is “re-\\nplacing” an object (e.g., a laptop computer, a bicycle, a sail-\\nboat, a blender) suggests that one is doing more than\\naltering one small part of it. That suggestion is bolstered\\nwhen one also uses language (here, a “more potent pro-\\nmoter”) referring to the replacement at the level of the\\noverall object. The suggestion is further reinforced by the\\nmost apt of the dictionary definitions of “replace” intro-\\nduced before the Commission—“to provide a substitute\\nfor.” J.A. 10361; see also J.A. 5622 (patent applicants ex-\\nplaining “replacing” as “substitut[ing]”). In many contexts,\\none would not refer to swapping out one small component\\nof a larger unit as “replacing” the unit or as providing a\\n“substitute” for the unit, even though the net result is a\\ndifferently constituted larger unit. Context matters, of\\ncourse, but here, Ajinomoto has not shown a contrary com-\\nmon understanding (or even one of several common under-\\nstandings) among relevant artisans in the specific context\\nof replacing a promoter with a more potent promoter. Ac-\\ncordingly, the claim language, though hardly establishing\\na plain meaning, supports the Commission’s construction.\\n The specification offers additional support, though it\\ntoo is hardly plain insofar as it bears on the particular con-\\nstruction issue. The specification states that “the enhance-\\nment of gene expression can be achieved by locating the\\n\\n10 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nDNA of the present invention under control of more potent\\npromoter instead of the native promoter.” ’655 patent,\\ncol. 6, lines 12–15. That statement speaks of a promoter as\\na unit, but it does not use the language of “replacing.” In-\\ndeed, the specification nowhere uses that language. But it\\ndoes discuss “substituting” promoters, using a term that,\\nas indicated above, is an apt definition of “replacing” here.\\nThe specification describes “[t]he present inventions” as in-\\ncluding “[t]he bacterium according to the above bacterium,\\nwherein native promoter of said DNA is substituted with\\nmore potent promoter.” Id., col. 3, lines 19–21. The term\\nis then used in Example 4, which is titled “Substitution of\\nthe Native Upstream Region of yddG Gene by the Hybrid\\nRegulatory Element Carrying the PL Promoter and SDlacZ\\nin E. coli Chromosome,” and which involves removing the\\nentire native promoter and inserting a new promoter. See\\nid., col. 11, line 5, through col. 12, line 46. The sole specifi-\\ncation example of “substitution” thus fits the Commission’s\\nclaim construction. And while the specification discusses\\nmutagenesis, it does so only in the context of the protein-\\ncoding region of the yddG gene, not the promoter. See id.,\\ncol. 5, lines 18–30.\\n We turn finally to the prosecution history—on which\\nthe parties to this case have focused most of their compet-\\ning analyses. We conclude that the best understanding of\\nwhat transpired before the examiner further supports the\\nCommission’s construction. Because the prosecution his-\\ntory reinforces what is already suggested by the claim lan-\\nguage and specification, this case provides no occasion,\\ncontrary to Ajinomoto’s contention (Ajinomoto Br. 34), for\\nrequiring clear and unmistakable disavowal or disclaimer\\nto justify a claim construction contrary to a meaning evi-\\ndent from the claim language and specification.\\n What was claim 2 of the original application recited\\n“[t]he bacterium according to claim 1, wherein said activi-\\nties of proteins . . . is enhanced by transformation of said\\nbacterium with DNA coding for the protein . . . or by\\n\\nAJINOMOTO CO., INC. v. ITC 11\\n\\n\\n\\nalteration of expression regulation sequence of said DNA on\\nthe chromosome of the bacterium.” J.A. 5047 (emphasis\\nadded). 6 The examiner rejected the claim for lack of an ad-\\nequate written description and lack of enablement. As to\\nwritten description, the examiner explained that “[w]hile\\ngeneric expression regulation sequences are known in the\\nart, a particular, endogenous expression regulation se-\\nquence for the DNA that encodes [amino-acid] SEQ ID\\nNO:2, or related sequences, is not described.” J.A. 5371.\\n“Without description of the endogenous regulation se-\\nquence,” the examiner continued, “an endogenous regula-\\ntion sequence that has been altered to increase expression\\nof said protein also lacks adequate written description.” Id.\\nTurning to enablement, the examiner stated:\\n The specification, while being enabling for Esche-\\n richia strains wherein the native promoter for the\\n DNA encoding SEQ ID NO: 2 has been changed by\\n substitution with a more potent promoter, does not\\n reasonably provide enablement for the genus of an\\n L-amino acid producing bacterium wherein the ac-\\n tivity of proteins described by SEQ ID NO: 2 and\\n related sequences is increased due to specific alter-\\n ations within the chromosomal expression regula-\\n tion sequence for DNA encoding said proteins.\\n ....\\n\\n\\n\\n\\n 6 Although claims 9 and 15 issued from what were\\nnumbered as claims 12 and 24 when added during prose-\\ncution, the parties do not dispute that the amendments to\\noriginal claim 2 (which eventually was cancelled) are rele-\\nvant to construing issued claims 9 and 15. The same “re-\\nplacing the native promoter . . . with a more potent\\npromoter” language added to original claim 2 was eventu-\\nally added to claims 9 and 15.\\n\\n12 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\n The instant specification teaches how to select\\n Escherichia bacteria that have an increased pro-\\n duction of L-amino acids, and the art teaches how\\n to mutagenize chromosomal DNA and how to char-\\n acterize the mutations in the DNA. However, nei-\\n ther the specification nor the art contain any\\n examples of how to specifically change endogenous\\n Escherichia chromosomal expression regulation se-\\n quences for the DNA encoding proteins described\\n by SEQ ID NO: 2, or related sequences, such that\\n the activity of said proteins in the bacteria is in-\\n creased. The art and the specification provide en-\\n ablement for inserting a known promoter in the\\n chromosomal DNA to upregulate the expression of\\n the DNA encoding SEQ ID NO: 2; however, neither\\n the specification nor the art enable making specific\\n changes to expression regulation sequences for\\n DNA encoding SEQ ID NO:2 and related sequences\\n on the chromosome of Escherichia bacteria. The\\n art and specification lack a detailed description of\\n the structure of the instant endogenous expression\\n regulation sequences, and they lack any guidance\\n on how to alter such sequences such that DNA ex-\\n pression is increased; therefore, to make the in-\\n stant bacteria with altered expression regulation\\n sequences would be unpredictable.\\nJ.A. 5374–75.\\n In response to the rejections, the applicants amended\\nthe claim to recite “replacing the native promoter that pre-\\ncedes a DNA encoding said protein . . . with a more potent\\npromoter” instead of “by alteration of expression regulation\\nsequence of said DNA.” J.A. 5610. The applicants ex-\\nplained the amendment as follows: “Applicants have\\namended Claim 2 consistent with the Examiner’s recogni-\\ntion that the specification enables Escherichia strains\\nwherein the native promoter for the DNA encoding SEQ ID\\n\\nAJINOMOTO CO., INC. v. ITC 13\\n\\n\\n\\nNO: 2 has been changed by substitution with a more potent\\npromoter.” J.A. 5622.\\n Reading the written-description and enablement rejec-\\ntions together, we think that the most reasonable under-\\nstanding of the examiner’s comments is that the examiner\\nwas drawing a distinction between alterations to the pro-\\nmoter, which were sufficiently described and enabled be-\\ncause E. coli promoters were well understood in the art,\\nand alterations to the expression-regulation sequence more\\nbroadly, which were not adequately described or enabled.\\nTo be sure, the examiner’s statement that the art and the\\nspecification “lack any guidance on how to alter such se-\\nquences such that DNA expression is increased” might at\\nfirst suggest that the applicants had not described and en-\\nabled the full scope of “alteration.” But in context, this\\nstatement is best read as meaning that the applicants had\\nnot described and enabled the full scope of “expression reg-\\nulation sequence,” so that “alteration” of that sequence also\\nwas not adequately described or enabled, even though gen-\\neral techniques for altering DNA sequences were well\\nknown in the relevant art.\\n We need not determine the precise basis for the exam-\\niner’s rejections, however, as “there is no principle of patent\\nlaw that the scope of a surrender of subject matter during\\nprosecution is limited to what is absolutely necessary to\\navoid a prior art reference that was the basis for an exam-\\niner's rejection.” Norian Corp. v. Stryker Corp., 432 F.3d\\n1356, 1361 (Fed. Cir. 2005). Rather, patentees frequently\\n“surrender more through amendment than may have been\\nabsolutely necessary to avoid particular prior art.” Id.\\nThat principle logically extends to amendments made to\\novercome rejections under § 112. Cf. Biogen Idec, Inc. v.\\nGlaxoSmithKline LLC, 713 F.3d 1090, 1095–96 (Fed. Cir.\\n2013). Indeed, we have stated more generally that “[t]he\\nquestion is what a person of ordinary skill would under-\\nstand the patentee to have disclaimed during prosecution,\\nnot what a person of ordinary skill would think the\\n\\n14 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\npatentee needed to disclaim during prosecution.” Tech.\\nProps. Ltd. LLC v. Huawei Techs. Co., 849 F.3d 1349, 1359\\n(Fed. Cir. 2017). A patentee must “be held to what he de-\\nclares during the prosecution of his patent,” because a con-\\ntrary rule would undermine “[t]he public notice function of\\na patent.” Springs Window Fashions LP v. Novo Indus.,\\nL.P., 323 F.3d 989, 995 (Fed. Cir. 2003).\\n We conclude that this is a case where the applicants\\nsurrendered more than may have been necessary. As dis-\\ncussed above, the best reading of the prosecution history is\\nthat, to overcome the written-description and enablement\\nrejections, it might well have sufficed if the applicants had\\nnarrowed their claims from alterations to the overall ex-\\npression-regulation sequence to alterations to the pro-\\nmoter. But the applicants did not merely change\\n“expression regulation sequence” to “native promoter”;\\nthey also changed “alteration” to “replacing.” Just as\\n“when different words are used in separate claims, they are\\npresumed to have different meanings,” Aspex Eyewear, Inc.\\nv. Marchon Eyewear, Inc., 672 F.3d 1335, 1349 (Fed. Cir.\\n2012), when a word is changed during prosecution, the\\nchange tends to suggest that the new word differs in mean-\\ning in some way from the original word.\\n That inference is bolstered by the applicants’ remarks\\naccompanying the amendment. Those remarks effectively\\nequate “replacing the native promoter . . . with a more po-\\ntent promoter” in the amended claim with “chang[ing]” the\\nnative promoter “by substitution with a more potent pro-\\nmoter.” J.A. 5622. As we have already noted, Example 4,\\ndescribed as involving “substitution” of a promoter, in-\\nvolves removal of the entire native promoter and insertion\\nof a new promoter. ’665 patent, col. 11, line 5, through col.\\n12, line 46. The applicants’ remarks, understood in light of\\nthe word choices and the specification, thus reinforce the\\nCommission’s conclusion that the new claim language does\\nnot include mutagenesis of individual nucleotides.\\n\\nAJINOMOTO CO., INC. v. ITC 15\\n\\n\\n\\n For those reasons, we affirm the Commission’s claim\\nconstruction and its finding that CJ’s earlier strains do not\\ninfringe based on that claim construction.\\n III\\n CJ, in its cross-appeal, challenges the Commission’s\\ndeterminations that CJ’s second later strain met the pro-\\ntein limitation, that both of CJ’s later strains met the re-\\nsistance limitation, and that claim 20 is not invalid for lack\\nof an adequate written description. We affirm the Commis-\\nsion as to all three issues.\\n A determination of infringement or non-infringement,\\nwhether literal or under the doctrine of equivalents, is a\\nfinding of fact, reviewed here for substantial evidence.\\nKinik Co. v. Int’l Trade Comm’n, 362 F.3d 1359, 1361 (Fed.\\nCir. 2004). But a determination of the applicability or in-\\napplicability of prosecution history estoppel, which limits\\nthe availability of the doctrine of equivalents, is a matter\\nof law, reviewed de novo. Spectrum Pharm., Inc. v. Sandoz\\nInc., 802 F.3d 1326, 1337 (Fed. Cir. 2015). The determina-\\ntion that a patent claim did not lack adequate support in\\nthe written description is a factual finding, reviewed for\\nsubstantial evidence. Rivera v. Int’l Trade Comm’n, 857\\nF.3d 1315, 1319 (Fed. Cir. 2017). Ajinomoto had to prove\\ninfringement by a preponderance of the evidence, while CJ\\nhad to prove invalidity by clear and convincing evidence.\\nSee Motorola Mobility, LLC v. Int’l Trade Comm’n, 737\\nF.3d 1345, 1348 (Fed. Cir. 2013); Enercon GmbH v. Int’l\\nTrade Comm’n, 151 F.3d 1376, 1384 (Fed. Cir. 1998).\\n A\\n The Commission found that CJ’s second later strain in-\\nfringed claim 20, which covers two alternatives of rele-\\nvance in this case—the claim 9 alternative and the claim\\n15 alternative. The infringement finding for CJ’s second\\nlater strain does not rest on the claim 15 alternative,\\nwhich, in its protein limitation, requires a protein encoded\\n\\n16 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nby a nucleotide sequence that hybridizes with the comple-\\nment of SEQ ID NO:1 (the nucleotide sequence of the E.\\ncoli yddG gene). The Commission did not find, and\\nAjinomoto does not argue for, either literal or equivalents\\ninfringement based on claim 15. The Commission found\\ninfringement under the claim 9 alternative—specifically, it\\nfound that the YddG protein encoded by the codon-random-\\nized non-E. coli yddG gene of this strain is an equivalent of\\nSEQ ID NO:2 (the amino-acid sequence of the E. coli YddG\\nprotein), as required by the protein limitation of claim 9.\\n CJ challenges that finding on two grounds. Based on\\nan amendment to original claims made during prosecution,\\nCJ asserts that prosecution history estoppel bars\\nAjinomoto from relying on the doctrine of equivalents to\\nmeet the protein limitation. Separately, CJ asserts that\\nthe non-E. coli YddG protein of CJ’s second later strain\\ncannot reasonably be found to be an equivalent of the\\nclaimed E. coli YddG protein under the function-way-result\\ntest for equivalence. We address those arguments in turn.\\n 1\\n Under the doctrine of prosecution history estoppel, “[a]\\npatentee’s decision to narrow his claims through amend-\\nment may be presumed to be a general disclaimer of the\\nterritory between the original claim and the amended\\nclaim.” Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki\\nCo., 535 U.S. 722, 740 (2002). The Supreme Court has\\nspecified three ways the patentee can rebut that presump-\\ntion, each of which, if established, means that “the amend-\\nment cannot reasonably be viewed as surrendering a\\nparticular equivalent.” Id. First, “[t]he equivalent may\\nhave been unforeseeable at the time of the application.” Id.\\nSecond, “the rationale underlying the amendment may\\nbear no more than a tangential relation to the equivalent\\nin question.” Id. Third, “there may be some other reason\\nsuggesting that the patentee could not reasonably be\\n\\nAJINOMOTO CO., INC. v. ITC 17\\n\\n\\n\\nexpected to have described the insubstantial substitute in\\nquestion.” Id. at 740–41.\\n In this case, the relevant facts about what transpired\\nduring prosecution are as follows. Claim 1 as originally\\nfiled recited two alternative conditions for the claimed pro-\\ntein:\\n a protein as defined in the following (A) or (B) in a\\n cell of said bacterium:\\n (A) a protein which comprises the amino acid se-\\n quence shown in SEQ ID NO:2 in Sequence listing;\\n (B) a protein which comprises an amino acid se-\\n quence including deletion, substitution, insertion\\n or addition of one or several amino acids in the\\n amino acid sequence shown in SEQ ID NO:2 in Se-\\n quence listing.\\nJ.A. 5047. The examiner rejected that claim as anticipated\\nby a reference disclosing the E. coli “yfiK gene product” (i.e.,\\nthe E. coli YfiK protein)—which differed from SEQ ID\\nNO:2 by deletion, substitution, insertion, or addition of sev-\\neral amino acids and, therefore, did not come within the (A)\\nalternative but did come within the (B) alternative. J.A.\\n5378. In response, the applicants left the (A) alternative\\nalone but replaced the language following (B) with new lan-\\nguage: “a protein which comprises an amino acid sequence\\nthat is encoded by a nucleotide sequence that hybridizes\\nwith the nucleotide sequence of SEQ ID NO:1 under strin-\\ngent conditions.” J.A. 5609. 7\\n\\n\\n\\n\\n 7 As previously noted, claims 9 and 15 issued from\\nnew claims added at the same time as this amendment.\\nSee supra note 6. Claims 9 and 15 respectively contain the\\nsame language as the (A) and (B) limitations in claim 1 af-\\nter it was amended. Claim 20, the claim at issue, treats\\n\\n18 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\n As an initial matter, CJ’s argument for prosecution his-\\ntory estoppel in this case involves an unusual circum-\\nstance. The infringement determination does not rest on\\nfinding an equivalent of the new claim language—namely,\\nthe (nucleotide) SEQ ID NO:1 language now in claim 15.\\nRather, it rests on finding an equivalent of the (amino-acid)\\nSEQ ID NO:2 language now in claim 9, which was not itself\\naltered by the amendment at issue. That is, the original\\nclaim provided two alternatives; only the second was mod-\\nified by amendment; and only the first is asserted as the\\nbasis for infringement by CJ’s second later strain. But we\\nneed not reach Ajinomoto’s contention that, in this circum-\\nstance, prosecution history estoppel does not apply at all,\\ni.e., that there is not even a presumed (though rebuttable)\\nsurrender of the asserted equivalent. The Commission did\\nnot so rule, instead concluding that the “tangential rela-\\ntion” exception applied, so that Ajinomoto did not surren-\\nder the protein produced by the codon-randomized non-E.\\ncoli yddG gene of CJ’s second later strain. J.A. 41–44. We\\nagree with that conclusion. 8\\n In applying the “tangential relation” exception, we\\n“ask[] whether the reason for the narrowing amendment\\nwas peripheral, or not directly relevant, to the alleged\\nequivalent.” Festo Corp. v. Shoketsu Kinzoku Kogyo Ka-\\nbushiki Co., 344 F.3d 1359, 1369 (Fed. Cir. 2003). “[T]he\\ninquiry into whether a patentee can rebut the Festo\\n\\n\\n\\nclaims 9 and 15 as alternatives in the same way that orig-\\ninal and amended claim 1 treated (A) and (B).\\n 8 CJ contends that Ajinomoto forfeited invocation of\\nthe “tangential relation” exception because it did not in-\\nvoke the exception before the ALJ or in its request for re-\\nview by the full Commission. CJ cites no authority that\\nbarred the Commission from exercising discretion to raise\\nthe issue and give the parties an adequate opportunity to\\naddress it, as the Commission did here.\\n\\nAJINOMOTO CO., INC. v. ITC 19\\n\\n\\n\\npresumption under the ‘tangential’ criterion focuses on the\\npatentee’s objectively apparent reason for the narrowing\\namendment.” Id. Our cases require the patentee to show\\nthat the way in which the alleged equivalent departs from\\nwhat the claim limitation literally requires is tangential to\\nthe discernible objective reason for the narrowing amend-\\nment. In that situation, there is no surrender of the equiv-\\nalent by that amendment.\\n For instance, in Insituform Technologies, Inc. v. CAT\\nContracting, Inc., the patent claimed a method of using a\\nvacuum to impregnate a flexible tube with resin. 385 F.3d\\n1360, 1362–63 (Fed. Cir. 2004). The claims were originally\\nrejected over a prior-art reference that disclosed a single\\nvacuum source located far away from the resin source. Id.\\nat 1369. The applicant amended the claim at issue to re-\\nquire a single vacuum source placed near the resin source.\\nSee id. at 1368–70. The alleged equivalent used multiple\\nvacuum sources. Id. at 1369–70. We held that the “tan-\\ngential relation” exception applied, observing that the pur-\\npose of the narrowing amendment was to distinguish the\\ninvention from the prior art based on the location of the\\nvacuum source relative to the resin, not to limit the number\\nof vacuum sources. Id. at 1370.\\n Similarly, in Regents of the University of California v.\\nDakocytomation California, Inc., the patented method in-\\nvolved using DNA testing to detect chromosomal abnor-\\nmalities. 517 F.3d 1364, 1369–70 (Fed. Cir. 2008). The\\nclaim at issue originally recited “disabling the hybridiza-\\ntion capacity of repetitive sequences” generally. Id. at\\n1377. The examiner rejected the claim over several prior-\\nart references, one of which disclosed disabling hybridiza-\\ntion using unique sequence probes. Id. at 1378. In re-\\nsponse, the applicants amended the claim to recite a\\nparticular technique of disabling hybridization using block-\\ning nucleic acids. Id. The parties stipulated that the added\\n“blocking nucleic acid” limitation was limited to human nu-\\ncleic acids, but the alleged equivalent used synthetic\\n\\n20 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nnucleic acids. Id. at 1376. We concluded that the narrow-\\ning amendment was tangential to how the equivalent dif-\\nfered from the literal claim limitation: “[I]n narrowing the\\nclaim to overcome the prior art rejections, the focus of the\\npatentees’ arguments centered on the method of blocking—\\nnot on the particular type of nucleic acid that could be used\\nfor blocking.” Id. at 1378. Indeed, we noted, “the ‘nucleic\\nacid’ limitation was never narrowed during prosecution\\nand was not at issue in the office action rejecting the\\nclaims,” and “none of the cited references concerned the\\ntype of nucleic acid that could perform the blocking, or\\nmentioned the accused equivalent.” Id.\\n Our decision in Intervet Inc. v. Merial Ltd. is to similar\\neffect. In that case, the patent claimed DNA constructs en-\\ncoding a type of porcine circovirus. 617 F.3d 1282, 1284\\n(Fed. Cir. 2010). The claim at issue originally recited DNA\\nsequences from a group of thirteen open reading frames,\\nwhich are portions of a gene that encode a protein. Id. at\\n1285–86, 1291. The examiner rejected the claim over open\\nreading frames from another organism, noting that the\\nclaim as written could cover open reading frames from any\\norganism. Id. at 1291. The applicants then amended the\\nclaim to require that the open reading frames be “of porcine\\ncircovirus type II.” Id. The alleged equivalent was a nu-\\ncleotide sequence that was over 99% homologous to one of\\nthe claimed sequences. Id. at 1286. The “tangential rela-\\ntion” exception applied to that equivalent, we held, because\\n“[t]he rationale for the amendment was to narrow the\\nclaimed universe of [open reading frames] down to those of\\n[porcine circovirus type II], and bore only a tangential re-\\nlation to the question of which DNA sequences are and are\\nnot properly characterized as [porcine circovirus type II].”\\nId. at 1292.\\n This understanding of the “tangential relation” excep-\\ntion also underlies cases in which we have held that the\\npatentee failed to establish that a narrowing amendment\\nwas tangential to the equivalent at issue. For example, in\\n\\nAJINOMOTO CO., INC. v. ITC 21\\n\\n\\n\\nBiagro Western Sales, Inc. v. Grow More, Inc., the claims at\\nissue, which claimed buffered phosphorus fertilizers, were\\nrejected over a prior-art reference disclosing a fertilizer\\nthat was buffered only when diluted. 423 F.3d 1296, 1299,\\n1306 (Fed. Cir. 2005). In response, the applicant amended\\nthe claims by adding the limitation “wherein said phospho-\\nrous-containing acid or salt thereof is present in an amount\\nof about 30 to about 40 weight percent,” explaining that the\\nfertilizer must be concentrated and that the amendment\\nspecified a range for the concentration. Id. at 1305–06.\\nThe alleged equivalent contained phosphorus compounds\\nat a concentration of between 59% and 62%. Id. at 1305.\\nWe concluded that the “tangential relation” exception did\\nnot apply, reasoning that it was “clear from the prosecution\\nhistory that the reason for adding the range limitation was\\nto overcome a prior art fertilizer that was not concen-\\ntrated,” and “both the reason for the amendment and the\\nasserted equivalent relate to the concentration of the ferti-\\nlizer.” Id. at 1306.\\n Here, we conclude, the Commission correctly concluded\\nthat Ajinomoto had rebutted the Festo presumption be-\\ncause the amendment was tangential to the equivalent in\\nquestion. The objectively evident rationale for the amend-\\nment was to limit the set of proteins within the claim’s\\nscope so that it no longer included the prior-art E. coli YfiK\\nprotein and, more generally, no longer allowed as wide a\\nrange of amino acid alterations (hence changes in the pro-\\ntein) as original alternative (B), which had allowed “dele-\\ntion, substitution, insertion or addition of one or several\\namino acids in the amino acid sequence shown in SEQ ID\\nNO: 2.” J.A. 5047. The reason for the amendment had\\nnothing to do with choosing among several DNA sequences\\nin the redundant genetic code that correspond to the same\\nprotein. Indeed, it is undisputed that the non-E. coli YddG\\nprotein produced without codon randomization remains\\nwithin the literal claim scope even after the amendment\\nand that the non-E. coli YddG protein is identical whether\\n\\n22 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nproduced from the codon-randomized or the non-codon-ran-\\ndomized version of the non-E. coli yddG gene.\\n Accordingly, the reason for the narrowing amendment\\n—limiting the amino-acid makeup of the proteins included\\nin one of the alternatives covered by the claim—is unre-\\nlated to differences among the several DNA sequences that\\nencode a given protein. Under Festo’s express provision for\\na “tangential relation” exception to the presumption as to\\nthe scope of surrender by amendment during prosecution,\\nthis conclusion about the reason for the amendment at is-\\nsue does not “ignore[] how the patentee deliberately elected\\nto narrow the claims” (Dissent at 6); rather, it identifies\\nwhat was not within the “scope disclaimed” (id. at 7), so\\nthat it may be proved to infringe by satisfying the other\\nrequirements of the doctrine of equivalents. We therefore\\nreject CJ’s contention that prosecution history estoppel\\nprecludes the Commission’s finding of infringement under\\nthe doctrine of equivalents for the second later strain.\\n 2\\n CJ’s second challenge to the Commission’s finding re-\\ngarding the protein limitation and CJ’s second later strain\\nis that the non-E. coli YddG protein of CJ’s second later\\nstrain could not properly be found to be equivalent to the\\nclaimed E. coli YddG protein. Ajinomoto presented its\\nequivalence case within the function-way-result frame-\\nwork, under which a product or process that does not liter-\\nally satisfy a claim limitation may nevertheless infringe “if\\nit performs substantially the same function in substan-\\ntially the same way to obtain the same result.” Duncan\\nParking Techs., Inc. v. IPS Grp., Inc., 914 F.3d 1347, 1362\\n(Fed. Cir. 2019) (quoting Graver Tank & Mfg. Co. v. Linde\\nAir Prods. Co., 339 U.S. 605, 608 (1950)). We conclude that\\nsubstantial evidence supports the Commission’s finding of\\nequivalence under that test.\\n As to “function”: Ajinomoto’s expert, Dr. Gregory\\nStephanopoulos, testified that both E. coli and non-E. coli\\n\\nAJINOMOTO CO., INC. v. ITC 23\\n\\n\\n\\nYddG proteins function as “export protein[s] that actively\\nexport[] aromatic L-amino acids and aromatic L-amino acid\\nanalogs” out of the bacterial cell. J.A. 545–46. A 2007 ar-\\nticle by Doroshenko et al. similarly explains that both pro-\\nteins are involved in exporting aromatic compounds. See\\nJ.A. 9451. And Dr. So Young Kim, a CJ employee, testified\\nduring a deposition that both proteins would be expected\\nto have similar functions based on similarities in the or-\\nganisms from which they are derived. See J.A. 10641 (“Q.\\nBased on the similarity between E. coli and [the non-E. coli\\norganism], you would suspect that the protein coded by the\\n[non-E. coli] yddG gene would be useful for whatever it\\ndoes in E. coli, right? A. I think that way too.”). Thus, the\\nCommission’s finding that both proteins perform the same\\nfunction is supported by substantial evidence.\\n As to “way”: Dr. Stephanopoulos testified that the two\\nproteins are 85% to 95% identical in structure. J.A. 546.\\nThis range was corroborated by a 2002 article by Santivi-\\nago et al., which indicates an 85% structural identity, see\\nJ.A. 9444, and the Doroshenko article, which notes a 95%\\nidentity in amino-acid sequence, J.A. 9451. On the record\\nhere, substantial evidence supports a finding that the two\\nproteins perform the membrane-transport function in sub-\\nstantially the same way. See also Mylan Institutional LLC\\nv. Aurobindo Pharma Ltd., 857 F.3d 858, 868 (Fed. Cir.\\n2017) (noting that the “function” and “way” inquiries often\\noverlap or are synonymous).\\n As to “result”: Dr. Stephanopoulos testified that, by ex-\\nporting L-tryptophan out of the bacterial cell, both proteins\\nincrease the ability of bacteria to “produce and accumulate\\nL-tryptophan.” See J.A. 547. That statement is supported\\nby CJ’s fermentation data, which showed that strains con-\\ntaining the E. coli yddG gene but with a stronger promoter,\\nand strains containing the non-E. coli yddG gene with a\\nstrong promoter, both showed greater production of L-tryp-\\ntophan than did strains containing the E. coli yddG gene\\nwith the native promoter. See J.A. 7957; J.A. 10053. In\\n\\n24 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nother words, enhancing the expression of either the E. coli\\nor the non-E. coli yddG gene had the effect of increasing\\nproduction of L-tryptophan, which supports an inference\\nthat the proteins encoded by those genes both result in in-\\ncreased L-tryptophan production. The Commission’s find-\\nings regarding result are supported by substantial\\nevidence.\\n CJ argues that the two proteins do not perform the\\nsame function in the same way because the E. coli YddG\\nprotein exports aromatic L-amino acids such as L-trypto-\\nphan, whereas the non-E. coli YddG protein exports a dif-\\nferent compound—namely, paraquat (also known as\\nmethyl viologen). But a 2012 article by Liu et al. explains\\nthat YddG proteins can export both types of compounds.\\nSee J.A. 9751 (“YddG is classified as aromatic amino\\nacid/paraquat exporter . . . .”). And Dr. Stephanopoulos, re-\\nlying on the Santiviago article, testified that the non-E. coli\\nYddG protein must be coupled to the OmpD protein, which\\nis present in the non-E. coli organism but not E. coli, to\\nexport paraquat. J.A. 762 (citing J.A. 9439). The fact that\\nthe non-E. coli YddG protein may be involved in exporting\\ncompounds other than L-tryptophan in the non-E. coli or-\\nganism does not undermine the Commission’s well-sup-\\nported finding that the non-E. coli YddG protein is involved\\nin exporting L-tryptophan in the E. coli bacteria used by\\nCJ.\\n B\\n CJ challenges the Commission’s finding of infringe-\\nment of both later strains on one additional ground. The\\nCommission found that CJ’s later strains met the re-\\nsistance limitation. CJ argues that substantial evidence\\n\\nAJINOMOTO CO., INC. v. ITC 25\\n\\n\\n\\ndoes not exist to support that finding. 9 We reject that ar-\\ngument.\\n Several pieces of evidence indicate that, as a general\\nmatter, enhancing the activity of the YddG protein in-\\ncreases bacteria’s resistance to L-tryptophan. Table 1 of\\nthe ’655 patent shows that E. coli bacteria with multiple\\ncopies of the yddG gene introduced through plasmids\\ndemonstrated better growth on a tryptophan substrate,\\nand thus more resistance, than unmodified E. coli bacteria.\\nSee ’655 patent, col. 9, lines 50–65 (bottom row, compare\\ncolumn “pUC19” (-) with column “pYDDG1” (+)). Similarly,\\nthe Doroshenko article, mentioned above, describes an ex-\\nperiment in which E. coli bacteria with a stronger promoter\\npreceding the yddG gene demonstrated enhanced re-\\nsistance to L-tryptophan. See J.A. 9455 (row DV036, col-\\numn DL-5-f-Trp).\\n CJ’s fermentation data, mentioned above, also provides\\ndirect evidence that CJ’s later strains have increased re-\\nsistance to L-tryptophan. That data shows a greater vol-\\nume of tryptophan with both of CJ’s later strains than with\\nunmodified E. coli bacteria. See J.A. 7957 (first later\\nstrain: middle table, row F4, column “Volume produced”);\\nJ.A. 10053 (second later strain: row “Product (g)” toward\\nmiddle of table). Dr. Stephanopoulos indicated that a\\nstrain’s ability to overproduce L-tryptophan necessarily\\nmeant that the strain had increased resistance to L-trypto-\\nphan. See J.A. 1448 (“[I]f that product feedback inhibits its\\nown synthesis, clearly, this is not going to work.”); see also\\n\\n\\n 9 CJ does not challenge the finding that CJ’s first\\nlater strain meets the protein limitation of the claim 15 al-\\nternative of claim 20. Specifically, CJ’s first later strain\\nuses a non-E. coli yddG gene without codon randomization,\\nwhich hybridizes with the complement of SEQ ID NO:1\\n(i.e., the nucleotide sequence of the E. coli yddG gene), and\\nthus falls within the literal scope of claim 15.\\n\\n26 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nJ.A. 521 (stating that bacteria that “exhibit enhanced re-\\nsistance to an aromatic L-amino acid or an aromatic L-\\namino acid analog” also “overproduce the corresponding ar-\\nomatic L-amino acid analog”).\\n CJ’s objections to the sufficiency or even relevance of\\nthis evidence are unpersuasive. CJ points out that the bac-\\nteria used to generate the data in Table 1 of the ’655 patent\\ncontained plasmids with more than the two copies of the\\nyddG gene in CJ’s later strains. See J.A. 1229. The Doro-\\nshenko article, however, indicates that enhancing the ac-\\ntivity of even a single copy of the yddG gene can increase\\nresistance to L-tryptophan. CJ responds that the strain\\nstudied in Doroshenko used a strong λPL promoter, while\\nCJ’s later strains use relatively weaker non-E. coli native\\nyddG, rmf, and rhtB promoters. See J.A. 9454. But Dr.\\nStephanopoulos testified that at least the rmf promoter in\\nboth of CJ’s later strains also is more potent than the na-\\ntive E. coli yddG promoter. J.A. 554. Thus, even if CJ is\\ncorrect that its later strains do not contain tandem promot-\\ners, the Commission could reasonably infer that the pro-\\nmoters used in CJ’s later strains enhance the activity of the\\nyddG genes relative to unmodified E. coli bacteria and\\nthereby increase those strains’ resistance to L-tryptophan.\\n CJ also cites Ajinomoto’s 2002 Progress Report as evi-\\ndence that enhancing a single copy of the yddG gene is in-\\nsufficient to enhance resistance to L-tryptophan. That\\nreport states that an experiment using “only one copy” of\\nthe yddG gene with a PL promoter “does not correctly\\nmodel[]” an earlier experiment using a “moderate-copy-\\nnumber” plasmid with the yddG gene, which had shown a\\n“positive effect” of the yddG gene on tryptophan produc-\\ntion. J.A. 10268. No more need be inferred from the report\\nthan that enhancing a single copy of the yddG gene in-\\ncreases resistance to L-tryptophan less than using a\\ngreater number of copies. The Commission did not need to\\ninfer that enhancing a single copy as in CJ’s later strains\\ndoes not enhance resistance at all.\\n\\nAJINOMOTO CO., INC. v. ITC 27\\n\\n\\n\\n Further, CJ asserts that the increased production of L-\\ntryptophan, and thus the enhanced resistance to L-trypto-\\nphan, observed in its later strains could be attributable to\\nthe presence of other genetic mutations rather than to in-\\ncreased YddG protein activity alone. See J.A. 442. But the\\nclaims require only that the protein “has the activity to\\nmake the bacterium resistant” to L-tryptophan, not that\\nthe protein be the sole cause of the bacterium’s enhanced\\nresistance to L-tryptophan. See ’655 patent, col. 22, lines\\n57–59. Considering the already-mentioned evidence that\\nthe YddG protein generally has the effect of increasing re-\\nsistance to L-tryptophan, the Commission had substantial\\nevidence from which to find that it was more likely than\\nnot that increased activity of the YddG protein at least\\npartly contributed to the enhanced resistance of CJ’s later\\nstrains.\\n C\\n CJ’s final contention in its cross-appeal seeks reversal\\nof the Commission’s rejection of CJ’s invalidity challenge to\\nclaim 20. CJ argues that substantial evidence does not\\nsupport the Commission’s finding that CJ did not prove\\nlack of an adequate written description for the genus of\\n“more potent promoter[s]” recited in claims 9 and 15 and,\\nby incorporation, in claim 20. We reject CJ’s argument.\\n “[A] sufficient description of a genus . . . requires the\\ndisclosure of either a representative number of species fall-\\ning within the scope of the genus or structural features\\ncommon to the members of the genus so that one of skill in\\nthe art can ‘visualize or recognize’ the members of the ge-\\nnus.” Ariad Pharm., Inc. v. Eli Lilly & Co., 598 F.3d 1336,\\n1350 (Fed. Cir. 2010) (en banc). The Commission found\\nboth that the ’655 patent discloses a representative number\\nof species of more potent promoters and that there are\\nstructural features common to the genus of more potent\\npromoters. Both of those findings are supported by\\n\\n28 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nsubstantial evidence, and they suffice to uphold the Com-\\nmission’s rejection of CJ’s written-description challenge.\\n 1\\n As to a representative number of species, we have rec-\\nognized that the amount of disclosure necessary to satisfy\\nthe written-description requirement “will necessarily vary\\ndepending on the context,” considering such facts as “the\\nexisting knowledge in the particular field, the extent and\\ncontent of the prior art, the maturity of the science or tech-\\nnology,” and “the predictability of the aspect at issue.” Ar-\\niad, 598 F.3d at 1351 (quoting Capon v. Eshhar, 418 F.3d\\n1349, 1359 (Fed. Cir. 2005)). In some circumstances, we\\nhave added, “a patentee may rely on information that is\\n‘well-known in the art’ for purposes of meeting the written\\ndescription requirement,” because “the specification is\\nviewed from the perspective of one of skill” in the relevant\\nart. Bos. Sci. Corp. v. Johnson & Johnson, 647 F.3d 1353,\\n1366 (Fed. Cir. 2011).\\n The ’655 patent discloses four examples of “potent pro-\\nmoters”: “PL promoter of lambda phage,” the “lac pro-\\nmoter,” the “trp promoter,” and the “trc promoter.” ’655\\npatent, col. 6, lines 21–24. The patent also cites the 1986\\narticle by Deuschle et al. as disclosing “examples of potent\\npromoters” and “[m]ethods for [the] evaluation [of] the\\nstrength of promoter[s].” Id., col. 6, lines 16–21. That ar-\\nticle provides data about the relative strength of fourteen\\npromoters and describes a general methodology for deter-\\nmining promoter strength in E. coli bacteria. J.A. 6174–\\n77. This evidence supports the Commission’s finding that\\n“enhancing promoter activity was well-known” and that a\\nskilled artisan “would have been able to identify more po-\\ntent promoters by employing common tools for measuring\\nRNA transcription.” J.A. 46.\\n The ’655 patent also makes clear that its invention was\\n“identifying the yddG gene encoding a membrane protein”\\nand discovering that the gene “conferred on a\\n\\nAJINOMOTO CO., INC. v. ITC 29\\n\\n\\n\\nmicroorganism resistance to phenylalanine and several\\namino acid analogues” when the gene was amplified or its\\nexpression enhanced, see ’655 patent, col. 2, lines 46–57,\\nnot the well-known techniques for performing the amplifi-\\ncation or expression enhancement, see id., col. 5, line 57,\\nthrough col. 6, line 33. We have explained that the repre-\\nsentative-species inquiry is directed to whether the inven-\\ntor “has truly invented the genus” as opposed to “a research\\nplan, leaving it to others to explore the unknown contours\\nof the claimed genus.” AbbVie Deutschland GmbH & Co.,\\nKG v. Janssen Biotech, Inc., 759 F.3d 1285, 1300 (Fed. Cir.\\n2014). Here, the genus of more potent promoters was al-\\nready well explored in the relevant art by the time of the\\n’655 patent’s invention. In these circumstances, the Com-\\nmission permissibly found in the specification, read in light\\nof the background knowledge in the art, a representative\\nnumber of species for the genus of more potent promoters.\\n 2\\n As to a common structural feature, the Commission\\nfound that a skilled artisan could “identify more potent\\nyddG promoters given the well-known link between con-\\nsensus sequence and promoter strength,” i.e., that promot-\\ners having fewer departures from a “consensus sequence”\\nin a promoter are generally stronger than promoters with\\nmore departures from such a sequence. J.A. 46. 10 Substan-\\ntial evidence supports that finding. For instance, a 1983\\narticle by Hawley and McClure describes a study demon-\\nstrating that most “mutations that decrease initiation fre-\\nquency also decrease the homology of the promoter to the\\nconsensus sequence, while up-mutations increase the\\n\\n\\n 10 The consensus sequence is a specific nucleotide se-\\nquence that appears in the promoters associated with\\nmany different genes in the genome of a particular organ-\\nism. In E. coli, the consensus sequence has two parts:\\nTTGACA at the -35 region and TATAAT at the -10 region.\\n\\n30 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nhomology in” most instances. J.A. 6237. Similarly, a 1986\\narticle by Horwitz and Loeb explains that “mutations that\\nincrease transcription, ‘up mutations,’ usually increase ho-\\nmology with the consensus sequence and spacing,” while\\n“mutations that decrease transcription, ‘down mutations,’\\nusually decrease homology with the consensus sequence\\nand spacing.” J.A. 6251.\\n CJ disputes that similarity to the consensus sequence\\ndefines a common structural feature, citing several articles\\nas indicating that a promoter closer to the consensus se-\\nquence will not always be stronger than one farther from\\nthat sequence. For instance, a 1998 article by Jensen and\\nHammer reports that a pattern observed in another organ-\\nism—that “the relatively strong promoters were the perfect\\nones,” i.e., those closer to the consensus sequence—“did not\\nhold true for E. coli: here the promoters which had either\\nan error in the consensus sequence or a shorter spacer were\\nrelatively strong.” J.A. 9149. Moreover, a 1985 article by\\nAoyama and Takanami states that similarity to the con-\\nsensus sequence “is still not enough to predict the site and\\nstrength of promoter from a given sequence,” J.A. 6215,\\nand a 1999 book edited by Fernandez and Hoeffler notes\\nthat “the strongest promoters in E. coli do not necessarily\\nadhere to the consensus sequence,” J.A. 9113.\\n CJ’s argument both assumes too strict a legal standard\\nand reads too much into its cited references. Adequate\\nwritten description does not require a perfect correspond-\\nence between the members of the genus and the asserted\\ncommon structural feature; for a functionally defined ge-\\nnus like the one at issue here, we have spoken more mod-\\nestly of a “correlation between structure and function.”\\nAriad, 598 F.3d at 1350 (emphasis added). In any event,\\nCJ’s evidence at most establishes that, starting with the\\nconsensus sequence, deviations from that sequence do not\\nalways decrease promoter strength, at least in E. coli. But\\nthe genus at issue here is “more potent promoter[s]” than\\nthe native promoter, not less potent promoters than the\\n\\nAJINOMOTO CO., INC. v. ITC 31\\n\\n\\n\\nconsensus sequence. And the Commission had substantial\\nevidence from which to find that, starting from the native\\nE. coli yddG promoter, deviations toward the consensus se-\\nquence generally increase promoter strength.\\n The cases cited by CJ in which we have held genus\\nclaims to lack an adequate written description are inappo-\\nsite. In Boston Scientific, the specification contained “no\\nexamples of macrocyclic lactone analogs of rapamycin” (the\\nclaimed genus) and essentially “no guidance on how to\\nproperly determine whether a compound is a macrocyclic\\nlactone analog of rapamycin.” 647 F.3d at 1364. In AbbVie,\\nthere was “no evidence to show any described antibody to\\nbe structurally similar to, and thus representative of,” an\\nantibody accused of coming within the claim, nor was there\\n“evidence to show whether one of skill in the art could make\\npredictable changes to the described antibodies to arrive at\\nother types of antibodies such as” the accused antibody.\\n759 F.3d at 1301. And in Regents of the University of Cali-\\nfornia v. Eli Lilly & Co., the specification described “a pro-\\ncess for obtaining human insulin-encoding cDNA” (such\\ncDNA required by the claim at issue) but not any “sequence\\ninformation indicating which nucleotides constitute hu-\\nman cDNA” or “further information in the patent pertain-\\ning to that cDNA’s relevant structural or physical\\ncharacteristics.” 119 F.3d 1559, 1567 (Fed. Cir. 1997).\\nHere, by contrast, the ’655 patent expressly provides four\\nexamples of “more potent promoters,” and the Commission\\nsupportably found that a skilled artisan could make rela-\\ntively predictable changes to the native promoter to arrive\\nat a more potent promoter.\\n IV\\n For the foregoing reasons, we affirm the Commission’s\\ndecision.\\n No costs.\\n AFFIRMED\\n\\n United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\\n INC.,\\n Appellants\\n\\n v.\\n\\n INTERNATIONAL TRADE COMMISSION,\\n Appellee\\n\\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\\n CHEIJEDANG INDONESIA,\\n Intervenors\\n\\n--------------------------------------------\\n\\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\\n CHEIJEDANG INDONESIA,\\n Appellants\\n\\n v.\\n\\n INTERNATIONAL TRADE COMMISSION,\\n Appellee\\n\\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\\n INC.,\\n Intervenors\\n ______________________\\n\\n 2018-1590, 2018-1629\\n ______________________\\n\\n2 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\n Appeals from the United States International Trade\\nCommission in Investigation No. 337-TA-1005.\\n ______________________\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4423084/\", \"author_raw\": \"TARANTO, Circuit Judge.\"}, {\"author\": \"DYK, Circuit Judge, concurring\", \"type\": \"concurrence\", \"text\": \"DYK, Circuit Judge, concurring-in-part and dissenting-in-\\npart.\\n I join the majority as to parts I, II, III (B) (as it relates\\nto Strain A, corresponding to the “first later strain” in the\\nmajority) and (C). I respectfully dissent from the majority’s\\nconclusion that Ajinomoto successfully rebutted the pre-\\nsumption of prosecution history estoppel under the tangen-\\ntial exception as to respondent’s recombinant bacterial\\nStrain B, which corresponds to the “second later strain” re-\\nferred to by the majority, see Majority Op. at 15.\\n On appeal, the only asserted claim is claim 20 of U.S.\\nPatent No. 7,666,655 (’655 patent). It covers “[a] method for\\nproducing an [amino acid,] which comprises cultivating the\\nbacterium according to any one of claims 9[, or 15].” ’655\\npatent, col. 24, ll. 4–6. In relevant part, claim 9 covers a\\nrecombinant bacteria having a “protein consist[ing] of the\\namino acid sequence of SEQ ID NO: 2.” Id. col. 22, ll. 56–\\n57. This corresponds to the amino acid sequence of E. coli\\nYddG protein (a membrane-bound protein involved in the\\ncellular export of aromatic amino acids). Strain B does not\\nliterally infringe claim 9 because it produces a protein with\\nan amino acid sequence that differs from SEQ ID NO: 2.\\nSee J.A. 37. Instead, Ajinomoto asserts infringement under\\nthe doctrine of equivalents, arguing that Strain B’s non-E.\\ncoli YddG protein is equivalent to the E. coli YddG protein\\n(SEQ ID NO: 2) in claim 9.\\n The prosecution history shows that claim language was\\namended such that the accused equivalent is excluded. 1\\n\\n\\n\\n 1 Everyone agrees that the relevant prosecution his-\\ntory for the analysis focuses on the language in claim 1,\\n\\nAJINOMOTO CO., INC. v. ITC 3\\n\\n\\n\\nOriginally, the claim language covered variations of SEQ\\nID NO: 2, stating that it covered “deletion, substitution, in-\\nsertion or addition of one or several amino acids” of SEQ ID\\nNO: 2. J.A. 5609. During prosecution, the examiner re-\\njected the claim as anticipated by prior art (Livshits) that\\ndisclosed a recombinant E. coli bacteria producing YfiK\\nprotein, encoded by the yfiK gene, which had an amino acid\\nsequence different from the SEQ ID NO: 2 but still satis-\\nfied the claim limitations. J.A. 5378. Specifically, the ex-\\naminer stated “Livshits et al. anticipate claims 1-4 because\\nthe yfiK gene product can be considered a protein” meeting\\nthe claim limitation above. Id. In response to this rejection,\\nthe patentee narrowed the claim language (which now ap-\\npears in claim 15) to only cover protein variants differing\\nfrom SEQ ID NO: 2 when they are “encoded by a nucleotide\\nsequence that hybridizes with the nucleotide sequence of\\nSEQ ID NO: 1[, the E. coli yddG gene,] under stringent con-\\nditions comprising 60°C, 1 x SSC, 0.1% SDS.” J.A. 5609;\\nsee ’655 patent, col. 23, ll. 19–22. 2 The patentee stated that\\n“[i]n view of this amendment, Livshits et al no longer an-\\nticipates the claimed invention.” J.A. 5617.\\n The majority assumes that prosecution history estop-\\npel presumptively applies in this case. Majority Op. at 18.\\nBut the majority concludes that Ajinomoto is still not pre-\\ncluded from arguing infringement under the doctrine of\\nequivalents based on the tangential exception.\\n We have consistently described this exception as “very\\nnarrow.” Integrated Tech. Corp. v. Rudolph Techs., Inc.,\\n734 F.3d 1352, 1358 (Fed. Cir. 2013) (quoting Cross Med.\\n\\n\\nwhich was later utilized in claims 9 and 15 that were added\\nlater in prosecution.\\n 2 Strain B does not literally infringe claim 15 be-\\ncause the non-E. coli YddG protein’s encoding nucleotide\\nsequence does not hybridize with SEQ ID NO: 1 under the\\nclaimed conditions.\\n\\n4 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nProds., Inc. v. Medtronic Sofamor Danek, Inc., 480 F.3d\\n1335, 1342 (Fed. Cir. 2007)). Under this exception, the\\nquestion is “whether the reason for the narrowing amend-\\nment was peripheral, or not directly relevant, to the alleged\\nequivalent.” Festo Corp. v. Shoketsu Kinzoku Kogyo Ka-\\nbushiki Co., 344 F.3d 1359, 1369 (Fed. Cir. 2003) (en banc).\\nThe inquiry “focuses on the patentee’s objectively apparent\\nreason for the narrowing amendment,” which “should be\\ndiscernible from the prosecution history record.” Id. (em-\\nphasis added). In my view the “reason for the narrowing\\namendment” in this case is directly related to the equiva-\\nlent.\\n Originally, the claim covered proteins with amino acid\\nsequence variations from SEQ ID NO: 2, which would have\\nincluded the non-E. coli YddG protein at issue here. The\\nexaminer rejected the original claim based on anticipating\\nprior art, and the patentee responded with a narrowing\\namendment. Instead of continuing to define the covered\\nproteins in terms of amino acid sequence variations from\\nSEQ ID NO: 2, 3 the patentee deliberately chose to redefine\\nthe claimed proteins in terms of the ability of their encod-\\ning nucleotide sequences to hybridize with SEQ ID NO: 1\\nunder the claimed conditions. The amended claim lan-\\nguage excluded the prior art protein (Livshits) because it\\nwas made based on a nucleotide sequence that did not meet\\nthe newly added hybridization requirement. The accused\\nequivalent is similarly not covered by the amended claims\\nbecause it is produced based on an encoding nucleotide se-\\nquence that does not hybridize with SEQ ID NO: 1 under\\nthe claimed conditions. Thus, I do not see how the reason\\n\\n\\n\\n 3 The patentee later added claim language that cov-\\nered other, more limited, variations from the amino acid\\nsequence of SEQ ID NO: 2, by “one to five amino acids,” but\\nthat claim language is not at issue here. ’655 Patent, col.\\n21, l. 42.\\n\\nAJINOMOTO CO., INC. v. ITC 5\\n\\n\\n\\nfor the narrowing amendment is tangential to the accused\\nequivalent.\\n Ajinomoto argues that “[t]o the extent anything was\\ngiven up during prosecution, it was the YfiK protein [dis-\\nclosed in Livshits] . . . and, possibly, other non-YddG pro-\\nteins.” Ajinomoto Response & Reply Br. at 41 (emphasis\\nomitted). Ajinomoto’s argument that prosecution history\\nestoppel would only apply to the specific prior art protein\\n(or possibly other non-YddG proteins) is not only incon-\\nsistent with how the patentee amended the claims but also\\nour caselaw. Specifically, “[Ajinomoto’s] representations\\nconvey to the public that it was relying on [the claimed hy-\\nbridization requirement] to overcome the prior art. The\\npublic is entitled to rely on those representations.” Inte-\\ngrated Tech., 734 F.3d at 1359. “The fact that the inventors\\nmay have thought after the fact that they could have relied\\non other distinctions in order to defend their claims[, e.g.,\\nby limiting the claim to only YddG-type proteins,] is irrele-\\nvant and speculative . . . .” Schwarz Pharma, Inc. v. Pad-\\ndock Labs., Inc., 504 F.3d 1371, 1377 (Fed. Cir. 2007). “It is\\nnot relevant to the determination of the scope of the sur-\\nrender that the applicant did not need to amend the claims”\\nin the way that it chose to do so “in order to overcome the\\nprior art.” Lucent Techs., Inc. v. Gateway, 525 F.3d 1200,\\n1218 (Fed. Cir. 2008) (citing Norian Corp v. Stryker Corp.,\\n432 F.3d 1356, 1361–62 (Fed. Cir. 2005)).\\n The majority adopts a slightly different version of\\nAjinomoto’s untenable argument. The majority concludes\\nthat the “objectively evident rationale” for the narrowing\\namendment was “to limit the set of proteins within the\\nclaim’s scope so that it no longer included the prior-art\\n[protein], and, more generally, no longer allowed as wide a\\nrange of amino acid alterations.” Majority Op. at 21 (em-\\nphasis in original). The majority reasons that because\\nStrain A, which makes the same protein as Strain B but\\nwith a different nucleotide sequence, literally infringes\\nclaim 15, somehow Strain B should be found to infringe\\n\\n6 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nclaim 9 under the doctrine of equivalents. It theorizes that\\n“[t]he reason for the amendment had nothing to do with\\nchoosing among several DNA sequences in the redundant\\ngenetic code that correspond to the same protein” (i.e., the\\naccused equivalent). Id. at 21. In this way, the majority\\nconcludes that the reason for the narrowing amendment—\\nlimiting the range of proteins covered by the claim—is un-\\nrelated to the way in which the equivalent departs from the\\nliteral claim limitation—differences among the several\\nDNA sequences that encode a given protein.\\n The problem with the majority’s analysis is that it ig-\\nnores how the patentee deliberately elected to narrow the\\nclaims. The anticipating prior art disclosed E. coli YfiK pro-\\ntein, encoded by the yfiK gene, and this prior art was\\navoided by narrowing the claim to only cover certain encod-\\ning nucleotide sequences. That rationale is directly related\\nto the accused equivalent, which does not infringe because\\nit does not use a covered encoding nucleotide sequence. In\\nother words, the rationale for the narrowing amendment\\n(avoiding a prior art protein based on its encoding nucleo-\\ntide sequence that does not meet the newly claimed hybrid-\\nization requirement) directly relates to the accused\\nequivalent (a protein made by an encoding nucleotide se-\\nquence that does not meet the newly claimed hybridization\\nrequirement).\\n The cases cited by the majority also do not support its\\napproach. In Insituform Technologies, Inc. v. CAT Con-\\ntracting, Inc., 385 F.3d 1360 (Fed. Cir. 2004), and Regents\\nof the University of California v. Dakocytomation Califor-\\nnia, 517 F.3d 1364 (Fed. Cir. 2008), multiple limitations\\nwere added with a narrowing amendment but only one of\\nthose limitations related to what was taught in the prior\\nart cited by the examiner. We held that the equivalent to\\nthe other limitation was permitted under the tangential ex-\\nception. In Insituform, the rationale for the amendment\\nwas to limit the location of the vacuum source, not the num-\\nber of vacuum sources (the accused equivalent). 385 F.3d\\n\\nAJINOMOTO CO., INC. v. ITC 7\\n\\n\\n\\nat 1370. In Regents, the rationale for the amendment was\\nto limit the type of blocking method, not the particular\\ntypes of nucleic acids that could be used in that method (the\\naccused equivalent). 517 F.3d at 1378. These cases cannot\\nbe read as allowing the patentee to recapture scope dis-\\nclaimed in order to distinguish the prior art, which is ex-\\nactly what the patentee is attempting to do here. The\\nanticipating prior art cited by the examiner specifically\\ntaught a protein made by a particular gene, and the pa-\\ntentee narrowed the claim to avoid this prior art by limit-\\ning the claim to only cover proteins made by particular\\nnucleotide sequences (which neither the prior art nor\\nStrain B have).\\n Our decision in Intervet Inc. v. Merial Ltd., 617 F.3d\\n1282 (Fed. Cir. 2010), is also inapposite. There, the claims\\nwere “narrow[ed from] the claimed universe of [nucleotide\\nsequences] down to those of [porcine circovirus type II\\n(‘PCV-2’)],” but there remained the tangential “question of\\nwhich DNA sequences are and are not properly character-\\nized as PCV-2.” Id. at 1292 (emphasis added). In contrast,\\nthere is no question here of which nucleotide sequences are\\n“properly characterized” as being included under the claim\\nlanguage—only those that hybridize with SEQ ID NO: 1\\n“under stringent conditions comprising 60°C, 1 x SSC,\\n0.1% SDS” are covered. J.A. 5609. There is no dispute that\\nCJ’s bacterial strain does not satisfy this specific and un-\\nambiguous limitation.\\n In my view the tangential exception cannot apply. The\\nequivalent is directly related to the reason for the amend-\\nment—to exclude those proteins made by an encoding nu-\\ncleotide sequence that does not hybridize with SEQ ID\\nNO: 1 under the specified conditions. I respectfully dissent\\nfrom the majority’s contrary conclusion.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4423084/\", \"author_raw\": \"DYK, Circuit Judge, concurring\"}]}"},"judge1":{"kind":"string","value":"DYK"},"judge2":{"kind":"string","value":"MOORE"},"judge3":{"kind":"string","value":"TARANTO"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":2,"string":"2"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":1,"string":"1"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4645831/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: AJINOMOTO CO., INC., Ajinomoto Heartland Inc., Appellants v. INTERNATIONAL TRADE COMMISSION, Appellee CJ CheilJedang Corp., CJ America, Inc., PT CheilJedang Indonesia, Intervenors CJ CheilJedang Corp., CJ America, Inc., PT CheilJedang Indonesia, Appellants v. International Trade Commission, Appellee Ajinomoto Co., Inc., Ajinomoto Heartland Inc., Intervenors\nCourt: U.S. Court of Appeals for the Federal Circuit\nDate decided: 2019-08-06\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: Yes\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n for the Federal Circuit\n ______________________\n\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\n INC.,\n Appellants\n\n v.\n\n INTERNATIONAL TRADE COMMISSION,\n Appellee\n\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\n CHEILJEDANG INDONESIA,\n Intervenors\n\n--------------------------------------------\n\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\n CHEILJEDANG INDONESIA,\n Appellants\n\n v.\n\n INTERNATIONAL TRADE COMMISSION,\n Appellee\n\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\n INC.,\n Intervenors\n ______________________\n\n 2018-1590, 2018-1629\n ______________________\n\n2 AJINOMOTO CO., INC. v. ITC\n\n\n\n\n Appeals from the United States International Trade\nCommission in Investigation No. 337-TA-1005.\n ______________________\n\n Decided: August 6, 2019\n ______________________\n\n JOHN D. LIVINGSTONE, Finnegan, Henderson, Farabow,\nGarrett & Dunner, LLP, Atlanta, GA, argued for Ajinomoto\nCo., Inc., Ajinomoto Heartland Inc. Also represented by\nMARTIN DAVID WEINGARTEN; CHARLES E. LIPSEY, Reston,\nVA; MAREESA ARNITA FREDERICK, CORA RENAE HOLT,\nBARBARA RUDOLPH, Washington, DC.\n\n HOUDA MORAD, Office of General Counsel, United\nStates International Trade Commission, Washington, DC,\nargued for appellee. Also represented by SIDNEY A.\nROSENZWEIG, DOMINIC L. BIANCHI, WAYNE W. HERRINGTON.\n\n JAMES F. HALEY, JR., Haley Guiliano LLP, New York,\nNY, argued for CJ CheilJedang Corp., CJ America, Inc., PT\nCheiJedang Indonesia. Also represented by STEVEN PEPE,\nRopes & Gray LLP, New York, NY; MATTHEW RIZZOLO,\nWashington, DC.\n ______________________\n\n Before DYK, MOORE, and TARANTO, Circuit Judges.\n Opinion for the court filed by Circuit Judge TARANTO.\nOpinion concurring in part and dissenting in part filed by\n Circuit Judge DYK.\nTARANTO, Circuit Judge.\n Ajinomoto Co., Inc. and Ajinomoto Heartland Inc. (col-\nlectively, Ajinomoto) filed a complaint against CJ\nCheilJedang Corp., CJ America, Inc., and PT CheilJedang\nIndonesia (collectively, CJ) with the International Trade\nCommission, alleging that CJ was importing certain\n\nAJINOMOTO CO., INC. v. ITC 3\n\n\n\nproducts that infringed Ajinomoto’s U.S. Patent No.\n7,666,655. CJ used several strains of Escherichia coli bac-\nteria to produce L-tryptophan products, which it then im-\nported into the United States. The Commission\ndetermined that CJ’s earlier strains did not infringe but\nthat CJ’s two later strains did. The Commission also found\nthat the relevant claim of the ’655 patent is not invalid for\nlack of an adequate written description.\n Ajinomoto appeals the Commission’s claim construc-\ntion underlying the determination of no infringement by\nthe earlier strains. CJ cross-appeals aspects of the deter-\nmination of infringement by the later strains and the rejec-\ntion of the invalidity challenge. We affirm.\n I\n A\n The ’655 patent claims E. coli bacteria that have been\ngenetically engineered to increase their production of aro-\nmatic L-amino acids, such as L-tryptophan, during fermen-\ntation, as well as methods of producing aromatic L-amino\nacids using such bacteria. See ’655 patent, col. 2, lines 40–\n45. In particular, the ’655 patent identifies a specific gene\nin the E. coli genome, the yddG gene, that encodes a mem-\nbrane protein, the YddG protein. Id., col. 2, lines 46–48.\nThat protein transports aromatic L-amino acids out of the\nbacterial cell and into the surrounding culture medium,\nwhere they can be collected. See id., col. 7, lines 11–16.\nWhen yddG gene activity in bacteria is enhanced so that\nmore YddG protein is produced, the bacteria show in-\ncreased production of, and increased resistance to, aro-\nmatic L-amino acids. Id., col. 2, lines 49–57. 1\n\n\n\n 1 The specification defines a bacterium’s “resistance”\nto an amino acid as its ability “to grow on a minimal me-\ndium containing” the amino acid on “which unmodified or\n\n4 AJINOMOTO CO., INC. v. ITC\n\n\n\n\n The ’655 patent describes three ways to enhance the\nactivity of the yddG gene. First, plasmids containing addi-\ntional copies of the yddG gene can be introduced into the\nbacterium. Id., col. 2, lines 50–52; id., col. 5, line 62,\nthrough col. 6, line 2. Second, additional copies of the yddG\ngene can be inserted into the bacterial chromosome. Id.,\ncol. 2, lines 52–54; id., col. 6, lines 3–6. Third, a stronger\n“promoter” than the one native to the E. coli yddG gene can\nbe used. Id., col. 2, lines 54–57; id., col. 6, lines 12–15. 2\n Claim 20, the only claim of the ’655 patent still asserted\nwhen the Commission issued its decision, claims “[a]\nmethod for producing an aromatic L-amino acid, which\ncomprises cultivating the bacterium according to any\none of claims 9–12, 13, 14, 15–18, or 19.” Id., col. 24, lines\n\n\n\nthe wild type, or the parental strain of the bacterium can-\nnot grow,” or its ability “to grow faster” on such a medium\n“than unmodified or the wild type, or the parental strain of\nthe bacterium.” ’655 patent, col. 4, lines 49–56.\n 2 A promoter is a nucleotide sequence within a DNA\nmolecule, located adjacent to the nucleotide sequence that\nconstitutes the gene to be expressed. The Lewin textbook\ncited by Ajinomoto shows a “typical promoter” around 41\nnucleotides long. J.A. 6043; see also J.A. 6177 (article by\nDeuschle et al., cited at ’655 patent, col. 6, lines 18–21,\nshowing longer promoters). The promoter is the binding\nsite for RNA polymerase, which initiates transcription (the\nfirst step in gene expression) by separating the two strands\nof DNA. The ’655 patent’s specification defines “[s]trength\nof promoter” with reference to the “frequency of acts of the\nRNA synthesis initiation.” ’655 patent, col. 6, lines 15–16.\n The promoter is only one part of a gene’s “expression\nregulation sequence,” which controls expression of the\ngene. See id., col. 3, line 14; id., col. 5, line 2. Besides pro-\nmoters, the “expression regulation sequence” can include,\ne.g., operators, enhancers, terminators, and silencers.\n\nAJINOMOTO CO., INC. v. ITC 5\n\n\n\n4–6 (emphasis added). Of the claims in that list, claims 9\nand 15 are the independent claims, and they are the two\nalternatives, under claim 20, of importance in this case.\n Claim 9 recites:\n 9. A recombinant Escherichia coli bacterium,\n which has the ability to accumulate aromatic L-\n amino acid in a medium, wherein the aromatic L-\n amino acid production by said bacterium is en-\n hanced by enhancing activity of a protein in a cell\n of said bacterium beyond the levels observed in a\n wild-type of said bacterium,\n [1] and in which said protein consists of the\n amino acid sequence of SEQ ID NO: 2\n [2] and said protein has the activity to make\n the bacterium resistant to L-phenylalanine, fluoro-\n phenylalanine or 5[-]fluoro-DL-tryptophan,\n [3] wherein the activity of the protein is en-\n hanced by [3a] transformation of the bacterium\n with a DNA encoding the protein to express the\n protein in the bacterium, [3b] by replacing the na-\n tive promoter which precedes the DNA on the chro-\n mosome of the bacterium with a more potent\n promoter, [3c] or by introduction of multiple copies\n of the DNA encoding said protein into the chromo-\n some of said bacterium to express the protein in\n said bacterium.\nId., col. 22, lines 51–67 (paragraph breaks and bold num-\nbering added). The Commission referred to limitation [1]\nas the “protein limitation,” limitation [2] as the “resistance\nlimitation,” and limitation [3] as the “enhancement limita-\ntion.” Claim 15 is materially identical to claim 9, except for\nthe protein limitation. Whereas claim 9 identifies the\nclaimed protein by a specific amino-acid sequence, claim 15\nidentifies it by reference to a corresponding DNA se-\nquence—a protein “encoded by the nucleotide sequence\n\n6 AJINOMOTO CO., INC. v. ITC\n\n\n\n\nwhich hybridizes with the complement of the nucleotide se-\nquence of SEQ ID NO: 1 under” certain conditions. See id.,\ncol. 23, lines 14–32.\n B\n In May 2016, Ajinomoto filed a complaint against CJ\nwith the Commission under 19 U.S.C. § 1337. Ajinomoto\nalleged that CJ violated § 1337(a)(1)(B)(ii) by importing an-\nimal-feed-grade L-tryptophan products produced by a pro-\ncess covered by the ’655 patent. 3 The Commission\ninstituted an investigation based on Ajinomoto’s com-\nplaint.\n The parties before us, including the Commission, agree\nthat whether the accused products were produced by a pro-\ncess covered by the patent is a question of infringement.\nThe proceeding focused on three groups of E. coli strains\nthat CJ has used to produce tryptophan. First, CJ’s “ear-\nlier strains” contained both the native E. coli yddG gene\nand the native E. coli yddG promoter, except that the first\nnucleotide of the promoter was changed through chemical\nmutagenesis, resulting in a stronger promoter. Second, in\nNovember 2016, several months after Ajinomoto filed its\ncomplaint, CJ began using its first “later strain,” which\ncontained two copies of a yddG gene: (1) the native E. coli\nyddG gene with the native E. coli yddG promoter; and (2) a\nnon-E. coli yddG gene with two promoters—(2a) a native\nnon-E. coli yddG promoter and (2b) an rmf promoter. 4\nThird, in December 2016, CJ started using its second “later\n\n\n\n 3 Ajinomoto also alleged that CJ infringed U.S. Pa-\ntent No. 6,180,373, which similarly claims methods of pro-\nducing tryptophan using genetically engineered bacteria.\nThe ’373 patent expired on January 30, 2018, and is not at\nissue in this court.\n 4 The rmf and rhtB promoters are promoters associ-\nated with other genes in the E. coli genome.\n\nAJINOMOTO CO., INC. v. ITC 7\n\n\n\nstrain,” which also contained two copies of a yddG gene:\n(1) the native E. coli yddG gene with the native E. coli\nyddG promoter; and (2) a codon-randomized non-E. coli\nyddG gene with two promoters—(2a) an rmf promoter and\n(2b) an rhtB promoter. 5\n In August 2017, the administrative law judge (ALJ) is-\nsued a final initial determination. The ALJ construed “re-\nplacing the native promoter . . . with a more potent\npromoter” in the enhancement limitation to mean “remov-\ning the native upstream region of the yddG gene and in-\nserting one of a class of promoters that controls expression\nof a different gene.” J.A. 90–91. Using that construction,\nthe ALJ found that CJ’s earlier strains did not infringe; he\nfound that they failed to meet the enhancement limitation\nbecause CJ created the more potent promoter in those\nstrains by mutagenesis of a single nucleotide rather than\nremoval of the entire native promoter and insertion of a\nnew promoter. As to CJ’s later strains, the ALJ found that\n(a) the first later strain did not infringe because Ajinomoto\nhad failed to prove that it met the resistance limitation,\nand (b) the second later strain also did not infringe because\n\n\n\n 5 Each particular codon (three nucleotides in a row\non a DNA molecule) that encodes for an amino acid always\nencodes for the same amino acid, but many of the 20 amino\nacids are encoded by more than one of the 64 codons. See\nAmgen, Inc. v. Chugai Pharm. Co., 927 F.2d 1200, 1208 n.4\n(Fed. Cir. 1991) (discussing “redundancy” of genetic code).\nFor instance, the DNA sequences TTA and TTG both code\nfor the amino acid leucine. “Codon randomization” refers\nto creation of DNA molecules that use different codons (e.g.,\nTTA or TTG) to code for the same amino acid (e.g., leucine)\nin building the same protein. See Mycogen Plant Sci. v.\nMonsanto Co., 243 F.3d 1316, 1323 (Fed. Cir. 2001) (“[O]ne\ncodon can be substituted for another in the gene without\nchanging the amino acid and resulting protein.”).\n\n8 AJINOMOTO CO., INC. v. ITC\n\n\n\n\nits non-E. coli YddG protein was not equivalent to the\nclaimed E. coli YddG protein under the doctrine of equiva-\nlents. Finally, the ALJ found that claim 20 of the ’655 pa-\ntent is invalid for lack of an adequate written description\nof the “more potent promoter” limitation incorporated into\nthat claim.\n In October 2017, the full Commission decided to review\nthe ALJ’s final initial determination in its entirety, and in\nDecember 2017, the Commission issued its decision. It af-\nfirmed the ALJ’s construction of “replacing the native pro-\nmoter . . . with a more potent promoter” and accordingly\naffirmed the ALJ’s finding that CJ’s earlier strains did not\ninfringe. But the Commission reversed several of the ALJ’s\nother findings. Specifically, it determined that both of CJ’s\nlater strains met all disputed claim limitations and thus\ninfringed claim 20 and that claim 20 was not proved to lack\nan adequate written description. The Commission accord-\ningly entered a limited exclusion order against CJ’s infring-\ning products, i.e., those made by both of CJ’s later strains\nbut not its earlier strains. The Commission also issued a\ncease-and-desist order against CJ America, which held in-\nventory of the infringing products.\n Ajinomoto and CJ both timely appealed. We have ju-\nrisdiction under 28 U.S.C. § 1295(a)(6).\n II\n We begin with Ajinomoto’s appeal of the Commission’s\nfinding of no infringement by the earlier strains.\nAjinomoto challenges that finding solely by arguing that\nthe Commission erred in its claim construction of “replac-\ning the native promoter . . . with a more potent promoter.”\nAjinomoto argues that, properly construed, the phrase is\nnot limited to removing the entire native promoter and in-\nserting a new promoter, as the Commission concluded, but\nencompasses mutagenesis of individual nucleotides within\nthe native promoter. We review the Commission’s claim\nconstruction de novo, as the Commission relied on only\n\nAJINOMOTO CO., INC. v. ITC 9\n\n\n\nintrinsic evidence and made no factual findings based on\nextrinsic evidence. Teva Pharm. USA, Inc. v. Sandoz, Inc.,\n135 S. Ct. 831, 841 (2015); see Cont’l Circuits LLC v. Intel\nCorp., 915 F.3d 788, 795 (Fed. Cir. 2019). We agree with\nthe Commission’s claim construction and therefore affirm\nthe non-infringement finding.\n The ordinary and customary meaning of the claim lan-\nguage provides support for the Commission’s claim con-\nstruction. The language of “replacing the native promoter\n. . . with a more potent promoter” suggests, in ordinary par-\nlance, an operation at the level of the entire promoter as a\nunit, not at the level of a single nucleotide that is just one\nsmall component of the promoter. To say that one is “re-\nplacing” an object (e.g., a laptop computer, a bicycle, a sail-\nboat, a blender) suggests that one is doing more than\naltering one small part of it. That suggestion is bolstered\nwhen one also uses language (here, a “more potent pro-\nmoter”) referring to the replacement at the level of the\noverall object. The suggestion is further reinforced by the\nmost apt of the dictionary definitions of “replace” intro-\nduced before the Commission—“to provide a substitute\nfor.” J.A. 10361; see also J.A. 5622 (patent applicants ex-\nplaining “replacing” as “substitut[ing]”). In many contexts,\none would not refer to swapping out one small component\nof a larger unit as “replacing” the unit or as providing a\n“substitute” for the unit, even though the net result is a\ndifferently constituted larger unit. Context matters, of\ncourse, but here, Ajinomoto has not shown a contrary com-\nmon understanding (or even one of several common under-\nstandings) among relevant artisans in the specific context\nof replacing a promoter with a more potent promoter. Ac-\ncordingly, the claim language, though hardly establishing\na plain meaning, supports the Commission’s construction.\n The specification offers additional support, though it\ntoo is hardly plain insofar as it bears on the particular con-\nstruction issue. The specification states that “the enhance-\nment of gene expression can be achieved by locating the\n\n10 AJINOMOTO CO., INC. v. ITC\n\n\n\n\nDNA of the present invention under control of more potent\npromoter instead of the native promoter.” ’655 patent,\ncol. 6, lines 12–15. That statement speaks of a promoter as\na unit, but it does not use the language of “replacing.” In-\ndeed, the specification nowhere uses that language. But it\ndoes discuss “substituting” promoters, using a term that,\nas indicated above, is an apt definition of “replacing” here.\nThe specification describes “[t]he present inventions” as in-\ncluding “[t]he bacterium according to the above bacterium,\nwherein native promoter of said DNA is substituted with\nmore potent promoter.” Id., col. 3, lines 19–21. The term\nis then used in Example 4, which is titled “Substitution of\nthe Native Upstream Region of yddG Gene by the Hybrid\nRegulatory Element Carrying the PL Promoter and SDlacZ\nin E. coli Chromosome,” and which involves removing the\nentire native promoter and inserting a new promoter. See\nid., col. 11, line 5, through col. 12, line 46. The sole specifi-\ncation example of “substitution” thus fits the Commission’s\nclaim construction. And while the specification discusses\nmutagenesis, it does so only in the context of the protein-\ncoding region of the yddG gene, not the promoter. See id.,\ncol. 5, lines 18–30.\n We turn finally to the prosecution history—on which\nthe parties to this case have focused most of their compet-\ning analyses. We conclude that the best understanding of\nwhat transpired before the examiner further supports the\nCommission’s construction. Because the prosecution his-\ntory reinforces what is already suggested by the claim lan-\nguage and specification, this case provides no occasion,\ncontrary to Ajinomoto’s contention (Ajinomoto Br. 34), for\nrequiring clear and unmistakable disavowal or disclaimer\nto justify a claim construction contrary to a meaning evi-\ndent from the claim language and specification.\n What was claim 2 of the original application recited\n“[t]he bacterium according to claim 1, wherein said activi-\nties of proteins . . . is enhanced by transformation of said\nbacterium with DNA coding for the protein . . . or by\n\nAJINOMOTO CO., INC. v. ITC 11\n\n\n\nalteration of expression regulation sequence of said DNA on\nthe chromosome of the bacterium.” J.A. 5047 (emphasis\nadded). 6 The examiner rejected the claim for lack of an ad-\nequate written description and lack of enablement. As to\nwritten description, the examiner explained that “[w]hile\ngeneric expression regulation sequences are known in the\nart, a particular, endogenous expression regulation se-\nquence for the DNA that encodes [amino-acid] SEQ ID\nNO:2, or related sequences, is not described.” J.A. 5371.\n“Without description of the endogenous regulation se-\nquence,” the examiner continued, “an endogenous regula-\ntion sequence that has been altered to increase expression\nof said protein also lacks adequate written description.” Id.\nTurning to enablement, the examiner stated:\n The specification, while being enabling for Esche-\n richia strains wherein the native promoter for the\n DNA encoding SEQ ID NO: 2 has been changed by\n substitution with a more potent promoter, does not\n reasonably provide enablement for the genus of an\n L-amino acid producing bacterium wherein the ac-\n tivity of proteins described by SEQ ID NO: 2 and\n related sequences is increased due to specific alter-\n ations within the chromosomal expression regula-\n tion sequence for DNA encoding said proteins.\n ....\n\n\n\n\n 6 Although claims 9 and 15 issued from what were\nnumbered as claims 12 and 24 when added during prose-\ncution, the parties do not dispute that the amendments to\noriginal claim 2 (which eventually was cancelled) are rele-\nvant to construing issued claims 9 and 15. The same “re-\nplacing the native promoter . . . with a more potent\npromoter” language added to original claim 2 was eventu-\nally added to claims 9 and 15.\n\n12 AJINOMOTO CO., INC. v. ITC\n\n\n\n\n The instant specification teaches how to select\n Escherichia bacteria that have an increased pro-\n duction of L-amino acids, and the art teaches how\n to mutagenize chromosomal DNA and how to char-\n acterize the mutations in the DNA. However, nei-\n ther the specification nor the art contain any\n examples of how to specifically change endogenous\n Escherichia chromosomal expression regulation se-\n quences for the DNA encoding proteins described\n by SEQ ID NO: 2, or related sequences, such that\n the activity of said proteins in the bacteria is in-\n creased. The art and the specification provide en-\n ablement for inserting a known promoter in the\n chromosomal DNA to upregulate the expression of\n the DNA encoding SEQ ID NO: 2; however, neither\n the specification nor the art enable making specific\n changes to expression regulation sequences for\n DNA encoding SEQ ID NO:2 and related sequences\n on the chromosome of Escherichia bacteria. The\n art and specification lack a detailed description of\n the structure of the instant endogenous expression\n regulation sequences, and they lack any guidance\n on how to alter such sequences such that DNA ex-\n pression is increased; therefore, to make the in-\n stant bacteria with altered expression regulation\n sequences would be unpredictable.\nJ.A. 5374–75.\n In response to the rejections, the applicants amended\nthe claim to recite “replacing the native promoter that pre-\ncedes a DNA encoding said protein . . . with a more potent\npromoter” instead of “by alteration of expression regulation\nsequence of said DNA.” J.A. 5610. The applicants ex-\nplained the amendment as follows: “Applicants have\namended Claim 2 consistent with the Examiner’s recogni-\ntion that the specification enables Escherichia strains\nwherein the native promoter for the DNA encoding SEQ ID\n\nAJINOMOTO CO., INC. v. ITC 13\n\n\n\nNO: 2 has been changed by substitution with a more potent\npromoter.” J.A. 5622.\n Reading the written-description and enablement rejec-\ntions together, we think that the most reasonable under-\nstanding of the examiner’s comments is that the examiner\nwas drawing a distinction between alterations to the pro-\nmoter, which were sufficiently described and enabled be-\ncause E. coli promoters were well understood in the art,\nand alterations to the expression-regulation sequence more\nbroadly, which were not adequately described or enabled.\nTo be sure, the examiner’s statement that the art and the\nspecification “lack any guidance on how to alter such se-\nquences such that DNA expression is increased” might at\nfirst suggest that the applicants had not described and en-\nabled the full scope of “alteration.” But in context, this\nstatement is best read as meaning that the applicants had\nnot described and enabled the full scope of “expression reg-\nulation sequence,” so that “alteration” of that sequence also\nwas not adequately described or enabled, even though gen-\neral techniques for altering DNA sequences were well\nknown in the relevant art.\n We need not determine the precise basis for the exam-\niner’s rejections, however, as “there is no principle of patent\nlaw that the scope of a surrender of subject matter during\nprosecution is limited to what is absolutely necessary to\navoid a prior art reference that was the basis for an exam-\niner's rejection.” Norian Corp. v. Stryker Corp., 432 F.3d\n1356, 1361 (Fed. Cir. 2005). Rather, patentees frequently\n“surrender more through amendment than may have been\nabsolutely necessary to avoid particular prior art.” Id.\nThat principle logically extends to amendments made to\novercome rejections under § 112. Cf. Biogen Idec, Inc. v.\nGlaxoSmithKline LLC, 713 F.3d 1090, 1095–96 (Fed. Cir.\n2013). Indeed, we have stated more generally that “[t]he\nquestion is what a person of ordinary skill would under-\nstand the patentee to have disclaimed during prosecution,\nnot what a person of ordinary skill would think the\n\n14 AJINOMOTO CO., INC. v. ITC\n\n\n\n\npatentee needed to disclaim during prosecution.” Tech.\nProps. Ltd. LLC v. Huawei Techs. Co., 849 F.3d 1349, 1359\n(Fed. Cir. 2017). A patentee must “be held to what he de-\nclares during the prosecution of his patent,” because a con-\ntrary rule would undermine “[t]he public notice function of\na patent.” Springs Window Fashions LP v. Novo Indus.,\nL.P., 323 F.3d 989, 995 (Fed. Cir. 2003).\n We conclude that this is a case where the applicants\nsurrendered more than may have been necessary. As dis-\ncussed above, the best reading of the prosecution history is\nthat, to overcome the written-description and enablement\nrejections, it might well have sufficed if the applicants had\nnarrowed their claims from alterations to the overall ex-\npression-regulation sequence to alterations to the pro-\nmoter. But the applicants did not merely change\n“expression regulation sequence” to “native promoter”;\nthey also changed “alteration” to “replacing.” Just as\n“when different words are used in separate claims, they are\npresumed to have different meanings,” Aspex Eyewear, Inc.\nv. Marchon Eyewear, Inc., 672 F.3d 1335, 1349 (Fed. Cir.\n2012), when a word is changed during prosecution, the\nchange tends to suggest that the new word differs in mean-\ning in some way from the original word.\n That inference is bolstered by the applicants’ remarks\naccompanying the amendment. Those remarks effectively\nequate “replacing the native promoter . . . with a more po-\ntent promoter” in the amended claim with “chang[ing]” the\nnative promoter “by substitution with a more potent pro-\nmoter.” J.A. 5622. As we have already noted, Example 4,\ndescribed as involving “substitution” of a promoter, in-\nvolves removal of the entire native promoter and insertion\nof a new promoter. ’665 patent, col. 11, line 5, through col.\n12, line 46. The applicants’ remarks, understood in light of\nthe word choices and the specification, thus reinforce the\nCommission’s conclusion that the new claim language does\nnot include mutagenesis of individual nucleotides.\n\nAJINOMOTO CO., INC. v. ITC 15\n\n\n\n For those reasons, we affirm the Commission’s claim\nconstruction and its finding that CJ’s earlier strains do not\ninfringe based on that claim construction.\n III\n CJ, in its cross-appeal, challenges the Commission’s\ndeterminations that CJ’s second later strain met the pro-\ntein limitation, that both of CJ’s later strains met the re-\nsistance limitation, and that claim 20 is not invalid for lack\nof an adequate written description. We affirm the Commis-\nsion as to all three issues.\n A determination of infringement or non-infringement,\nwhether literal or under the doctrine of equivalents, is a\nfinding of fact, reviewed here for substantial evidence.\nKinik Co. v. Int’l Trade Comm’n, 362 F.3d 1359, 1361 (Fed.\nCir. 2004). But a determination of the applicability or in-\napplicability of prosecution history estoppel, which limits\nthe availability of the doctrine of equivalents, is a matter\nof law, reviewed de novo. Spectrum Pharm., Inc. v. Sandoz\nInc., 802 F.3d 1326, 1337 (Fed. Cir. 2015). The determina-\ntion that a patent claim did not lack adequate support in\nthe written description is a factual finding, reviewed for\nsubstantial evidence. Rivera v. Int’l Trade Comm’n, 857\nF.3d 1315, 1319 (Fed. Cir. 2017). Ajinomoto had to prove\ninfringement by a preponderance of the evidence, while CJ\nhad to prove invalidity by clear and convincing evidence.\nSee Motorola Mobility, LLC v. Int’l Trade Comm’n, 737\nF.3d 1345, 1348 (Fed. Cir. 2013); Enercon GmbH v. Int’l\nTrade Comm’n, 151 F.3d 1376, 1384 (Fed. Cir. 1998).\n A\n The Commission found that CJ’s second later strain in-\nfringed claim 20, which covers two alternatives of rele-\nvance in this case—the claim 9 alternative and the claim\n15 alternative. The infringement finding for CJ’s second\nlater strain does not rest on the claim 15 alternative,\nwhich, in its protein limitation, requires a protein encoded\n\n16 AJINOMOTO CO., INC. v. ITC\n\n\n\n\nby a nucleotide sequence that hybridizes with the comple-\nment of SEQ ID NO:1 (the nucleotide sequence of the E.\ncoli yddG gene). The Commission did not find, and\nAjinomoto does not argue for, either literal or equivalents\ninfringement based on claim 15. The Commission found\ninfringement under the claim 9 alternative—specifically, it\nfound that the YddG protein encoded by the codon-random-\nized non-E. coli yddG gene of this strain is an equivalent of\nSEQ ID NO:2 (the amino-acid sequence of the E. coli YddG\nprotein), as required by the protein limitation of claim 9.\n CJ challenges that finding on two grounds. Based on\nan amendment to original claims made during prosecution,\nCJ asserts that prosecution history estoppel bars\nAjinomoto from relying on the doctrine of equivalents to\nmeet the protein limitation. Separately, CJ asserts that\nthe non-E. coli YddG protein of CJ’s second later strain\ncannot reasonably be found to be an equivalent of the\nclaimed E. coli YddG protein under the function-way-result\ntest for equivalence. We address those arguments in turn.\n 1\n Under the doctrine of prosecution history estoppel, “[a]\npatentee’s decision to narrow his claims through amend-\nment may be presumed to be a general disclaimer of the\nterritory between the original claim and the amended\nclaim.” Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki\nCo., 535 U.S. 722, 740 (2002). The Supreme Court has\nspecified three ways the patentee can rebut that presump-\ntion, each of which, if established, means that “the amend-\nment cannot reasonably be viewed as surrendering a\nparticular equivalent.” Id. First, “[t]he equivalent may\nhave been unforeseeable at the time of the application.” Id.\nSecond, “the rationale underlying the amendment may\nbear no more than a tangential relation to the equivalent\nin question.” Id. Third, “there may be some other reason\nsuggesting that the patentee could not reasonably be\n\nAJINOMOTO CO., INC. v. ITC 17\n\n\n\nexpected to have described the insubstantial substitute in\nquestion.” Id. at 740–41.\n In this case, the relevant facts about what transpired\nduring prosecution are as follows. Claim 1 as originally\nfiled recited two alternative conditions for the claimed pro-\ntein:\n a protein as defined in the following (A) or (B) in a\n cell of said bacterium:\n (A) a protein which comprises the amino acid se-\n quence shown in SEQ ID NO:2 in Sequence listing;\n (B) a protein which comprises an amino acid se-\n quence including deletion, substitution, insertion\n or addition of one or several amino acids in the\n amino acid sequence shown in SEQ ID NO:2 in Se-\n quence listing.\nJ.A. 5047. The examiner rejected that claim as anticipated\nby a reference disclosing the E. coli “yfiK gene product” (i.e.,\nthe E. coli YfiK protein)—which differed from SEQ ID\nNO:2 by deletion, substitution, insertion, or addition of sev-\neral amino acids and, therefore, did not come within the (A)\nalternative but did come within the (B) alternative. J.A.\n5378. In response, the applicants left the (A) alternative\nalone but replaced the language following (B) with new lan-\nguage: “a protein which comprises an amino acid sequence\nthat is encoded by a nucleotide sequence that hybridizes\nwith the nucleotide sequence of SEQ ID NO:1 under strin-\ngent conditions.” J.A. 5609. 7\n\n\n\n\n 7 As previously noted, claims 9 and 15 issued from\nnew claims added at the same time as this amendment.\nSee supra note 6. Claims 9 and 15 respectively contain the\nsame language as the (A) and (B) limitations in claim 1 af-\nter it was amended. Claim 20, the claim at issue, treats\n\n18 AJINOMOTO CO., INC. v. ITC\n\n\n\n\n As an initial matter, CJ’s argument for prosecution his-\ntory estoppel in this case involves an unusual circum-\nstance. The infringement determination does not rest on\nfinding an equivalent of the new claim language—namely,\nthe (nucleotide) SEQ ID NO:1 language now in claim 15.\nRather, it rests on finding an equivalent of the (amino-acid)\nSEQ ID NO:2 language now in claim 9, which was not itself\naltered by the amendment at issue. That is, the original\nclaim provided two alternatives; only the second was mod-\nified by amendment; and only the first is asserted as the\nbasis for infringement by CJ’s second later strain. But we\nneed not reach Ajinomoto’s contention that, in this circum-\nstance, prosecution history estoppel does not apply at all,\ni.e., that there is not even a presumed (though rebuttable)\nsurrender of the asserted equivalent. The Commission did\nnot so rule, instead concluding that the “tangential rela-\ntion” exception applied, so that Ajinomoto did not surren-\nder the protein produced by the codon-randomized non-E.\ncoli yddG gene of CJ’s second later strain. J.A. 41–44. We\nagree with that conclusion. 8\n In applying the “tangential relation” exception, we\n“ask[] whether the reason for the narrowing amendment\nwas peripheral, or not directly relevant, to the alleged\nequivalent.” Festo Corp. v. Shoketsu Kinzoku Kogyo Ka-\nbushiki Co., 344 F.3d 1359, 1369 (Fed. Cir. 2003). “[T]he\ninquiry into whether a patentee can rebut the Festo\n\n\n\nclaims 9 and 15 as alternatives in the same way that orig-\ninal and amended claim 1 treated (A) and (B).\n 8 CJ contends that Ajinomoto forfeited invocation of\nthe “tangential relation” exception because it did not in-\nvoke the exception before the ALJ or in its request for re-\nview by the full Commission. CJ cites no authority that\nbarred the Commission from exercising discretion to raise\nthe issue and give the parties an adequate opportunity to\naddress it, as the Commission did here.\n\nAJINOMOTO CO., INC. v. ITC 19\n\n\n\npresumption under the ‘tangential’ criterion focuses on the\npatentee’s objectively apparent reason for the narrowing\namendment.” Id. Our cases require the patentee to show\nthat the way in which the alleged equivalent departs from\nwhat the claim limitation literally requires is tangential to\nthe discernible objective reason for the narrowing amend-\nment. In that situation, there is no surrender of the equiv-\nalent by that amendment.\n For instance, in Insituform Technologies, Inc. v. CAT\nContracting, Inc., the patent claimed a method of using a\nvacuum to impregnate a flexible tube with resin. 385 F.3d\n1360, 1362–63 (Fed. Cir. 2004). The claims were originally\nrejected over a prior-art reference that disclosed a single\nvacuum source located far away from the resin source. Id.\nat 1369. The applicant amended the claim at issue to re-\nquire a single vacuum source placed near the resin source.\nSee id. at 1368–70. The alleged equivalent used multiple\nvacuum sources. Id. at 1369–70. We held that the “tan-\ngential relation” exception applied, observing that the pur-\npose of the narrowing amendment was to distinguish the\ninvention from the prior art based on the location of the\nvacuum source relative to the resin, not to limit the number\nof vacuum sources. Id. at 1370.\n Similarly, in Regents of the University of California v.\nDakocytomation California, Inc., the patented method in-\nvolved using DNA testing to detect chromosomal abnor-\nmalities. 517 F.3d 1364, 1369–70 (Fed. Cir. 2008). The\nclaim at issue originally recited “disabling the hybridiza-\ntion capacity of repetitive sequences” generally. Id. at\n1377. The examiner rejected the claim over several prior-\nart references, one of which disclosed disabling hybridiza-\ntion using unique sequence probes. Id. at 1378. In re-\nsponse, the applicants amended the claim to recite a\nparticular technique of disabling hybridization using block-\ning nucleic acids. Id. The parties stipulated that the added\n“blocking nucleic acid” limitation was limited to human nu-\ncleic acids, but the alleged equivalent used synthetic\n\n20 AJINOMOTO CO., INC. v. ITC\n\n\n\n\nnucleic acids. Id. at 1376. We concluded that the narrow-\ning amendment was tangential to how the equivalent dif-\nfered from the literal claim limitation: “[I]n narrowing the\nclaim to overcome the prior art rejections, the focus of the\npatentees’ arguments centered on the method of blocking—\nnot on the particular type of nucleic acid that could be used\nfor blocking.” Id. at 1378. Indeed, we noted, “the ‘nucleic\nacid’ limitation was never narrowed during prosecution\nand was not at issue in the office action rejecting the\nclaims,” and “none of the cited references concerned the\ntype of nucleic acid that could perform the blocking, or\nmentioned the accused equivalent.” Id.\n Our decision in Intervet Inc. v. Merial Ltd. is to similar\neffect. In that case, the patent claimed DNA constructs en-\ncoding a type of porcine circovirus. 617 F.3d 1282, 1284\n(Fed. Cir. 2010). The claim at issue originally recited DNA\nsequences from a group of thirteen open reading frames,\nwhich are portions of a gene that encode a protein. Id. at\n1285–86, 1291. The examiner rejected the claim over open\nreading frames from another organism, noting that the\nclaim as written could cover open reading frames from any\norganism. Id. at 1291. The applicants then amended the\nclaim to require that the open reading frames be “of porcine\ncircovirus type II.” Id. The alleged equivalent was a nu-\ncleotide sequence that was over 99% homologous to one of\nthe claimed sequences. Id. at 1286. The “tangential rela-\ntion” exception applied to that equivalent, we held, because\n“[t]he rationale for the amendment was to narrow the\nclaimed universe of [open reading frames] down to those of\n[porcine circovirus type II], and bore only a tangential re-\nlation to the question of which DNA sequences are and are\nnot properly characterized as [porcine circovirus type II].”\nId. at 1292.\n This understanding of the “tangential relation” excep-\ntion also underlies cases in which we have held that the\npatentee failed to establish that a narrowing amendment\nwas tangential to the equivalent at issue. For example, in\n\nAJINOMOTO CO., INC. v. ITC 21\n\n\n\nBiagro Western Sales, Inc. v. Grow More, Inc., the claims at\nissue, which claimed buffered phosphorus fertilizers, were\nrejected over a prior-art reference disclosing a fertilizer\nthat was buffered only when diluted. 423 F.3d 1296, 1299,\n1306 (Fed. Cir. 2005). In response, the applicant amended\nthe claims by adding the limitation “wherein said phospho-\nrous-containing acid or salt thereof is present in an amount\nof about 30 to about 40 weight percent,” explaining that the\nfertilizer must be concentrated and that the amendment\nspecified a range for the concentration. Id. at 1305–06.\nThe alleged equivalent contained phosphorus compounds\nat a concentration of between 59% and 62%. Id. at 1305.\nWe concluded that the “tangential relation” exception did\nnot apply, reasoning that it was “clear from the prosecution\nhistory that the reason for adding the range limitation was\nto overcome a prior art fertilizer that was not concen-\ntrated,” and “both the reason for the amendment and the\nasserted equivalent relate to the concentration of the ferti-\nlizer.” Id. at 1306.\n Here, we conclude, the Commission correctly concluded\nthat Ajinomoto had rebutted the Festo presumption be-\ncause the amendment was tangential to the equivalent in\nquestion. The objectively evident rationale for the amend-\nment was to limit the set of proteins within the claim’s\nscope so that it no longer included the prior-art E. coli YfiK\nprotein and, more generally, no longer allowed as wide a\nrange of amino acid alterations (hence changes in the pro-\ntein) as original alternative (B), which had allowed “dele-\ntion, substitution, insertion or addition of one or several\namino acids in the amino acid sequence shown in SEQ ID\nNO: 2.” J.A. 5047. The reason for the amendment had\nnothing to do with choosing among several DNA sequences\nin the redundant genetic code that correspond to the same\nprotein. Indeed, it is undisputed that the non-E. coli YddG\nprotein produced without codon randomization remains\nwithin the literal claim scope even after the amendment\nand that the non-E. coli YddG protein is identical whether\n\n22 AJINOMOTO CO., INC. v. ITC\n\n\n\n\nproduced from the codon-randomized or the non-codon-ran-\ndomized version of the non-E. coli yddG gene.\n Accordingly, the reason for the narrowing amendment\n—limiting the amino-acid makeup of the proteins included\nin one of the alternatives covered by the claim—is unre-\nlated to differences among the several DNA sequences that\nencode a given protein. Under Festo’s express provision for\na “tangential relation” exception to the presumption as to\nthe scope of surrender by amendment during prosecution,\nthis conclusion about the reason for the amendment at is-\nsue does not “ignore[] how the patentee deliberately elected\nto narrow the claims” (Dissent at 6); rather, it identifies\nwhat was not within the “scope disclaimed” (id. at 7), so\nthat it may be proved to infringe by satisfying the other\nrequirements of the doctrine of equivalents. We therefore\nreject CJ’s contention that prosecution history estoppel\nprecludes the Commission’s finding of infringement under\nthe doctrine of equivalents for the second later strain.\n 2\n CJ’s second challenge to the Commission’s finding re-\ngarding the protein limitation and CJ’s second later strain\nis that the non-E. coli YddG protein of CJ’s second later\nstrain could not properly be found to be equivalent to the\nclaimed E. coli YddG protein. Ajinomoto presented its\nequivalence case within the function-way-result frame-\nwork, under which a product or process that does not liter-\nally satisfy a claim limitation may nevertheless infringe “if\nit performs substantially the same function in substan-\ntially the same way to obtain the same result.” Duncan\nParking Techs., Inc. v. IPS Grp., Inc., 914 F.3d 1347, 1362\n(Fed. Cir. 2019) (quoting Graver Tank & Mfg. Co. v. Linde\nAir Prods. Co., 339 U.S. 605, 608 (1950)). We conclude that\nsubstantial evidence supports the Commission’s finding of\nequivalence under that test.\n As to “function”: Ajinomoto’s expert, Dr. Gregory\nStephanopoulos, testified that both E. coli and non-E. coli\n\nAJINOMOTO CO., INC. v. ITC 23\n\n\n\nYddG proteins function as “export protein[s] that actively\nexport[] aromatic L-amino acids and aromatic L-amino acid\nanalogs” out of the bacterial cell. J.A. 545–46. A 2007 ar-\nticle by Doroshenko et al. similarly explains that both pro-\nteins are involved in exporting aromatic compounds. See\nJ.A. 9451. And Dr. So Young Kim, a CJ employee, testified\nduring a deposition that both proteins would be expected\nto have similar functions based on similarities in the or-\nganisms from which they are derived. See J.A. 10641 (“Q.\nBased on the similarity between E. coli and [the non-E. coli\norganism], you would suspect that the protein coded by the\n[non-E. coli] yddG gene would be useful for whatever it\ndoes in E. coli, right? A. I think that way too.”). Thus, the\nCommission’s finding that both proteins perform the same\nfunction is supported by substantial evidence.\n As to “way”: Dr. Stephanopoulos testified that the two\nproteins are 85% to 95% identical in structure. J.A. 546.\nThis range was corroborated by a 2002 article by Santivi-\nago et al., which indicates an 85% structural identity, see\nJ.A. 9444, and the Doroshenko article, which notes a 95%\nidentity in amino-acid sequence, J.A. 9451. On the record\nhere, substantial evidence supports a finding that the two\nproteins perform the membrane-transport function in sub-\nstantially the same way. See also Mylan Institutional LLC\nv. Aurobindo Pharma Ltd., 857 F.3d 858, 868 (Fed. Cir.\n2017) (noting that the “function” and “way” inquiries often\noverlap or are synonymous).\n As to “result”: Dr. Stephanopoulos testified that, by ex-\nporting L-tryptophan out of the bacterial cell, both proteins\nincrease the ability of bacteria to “produce and accumulate\nL-tryptophan.” See J.A. 547. That statement is supported\nby CJ’s fermentation data, which showed that strains con-\ntaining the E. coli yddG gene but with a stronger promoter,\nand strains containing the non-E. coli yddG gene with a\nstrong promoter, both showed greater production of L-tryp-\ntophan than did strains containing the E. coli yddG gene\nwith the native promoter. See J.A. 7957; J.A. 10053. In\n\n24 AJINOMOTO CO., INC. v. ITC\n\n\n\n\nother words, enhancing the expression of either the E. coli\nor the non-E. coli yddG gene had the effect of increasing\nproduction of L-tryptophan, which supports an inference\nthat the proteins encoded by those genes both result in in-\ncreased L-tryptophan production. The Commission’s find-\nings regarding result are supported by substantial\nevidence.\n CJ argues that the two proteins do not perform the\nsame function in the same way because the E. coli YddG\nprotein exports aromatic L-amino acids such as L-trypto-\nphan, whereas the non-E. coli YddG protein exports a dif-\nferent compound—namely, paraquat (also known as\nmethyl viologen). But a 2012 article by Liu et al. explains\nthat YddG proteins can export both types of compounds.\nSee J.A. 9751 (“YddG is classified as aromatic amino\nacid/paraquat exporter . . . .”). And Dr. Stephanopoulos, re-\nlying on the Santiviago article, testified that the non-E. coli\nYddG protein must be coupled to the OmpD protein, which\nis present in the non-E. coli organism but not E. coli, to\nexport paraquat. J.A. 762 (citing J.A. 9439). The fact that\nthe non-E. coli YddG protein may be involved in exporting\ncompounds other than L-tryptophan in the non-E. coli or-\nganism does not undermine the Commission’s well-sup-\nported finding that the non-E. coli YddG protein is involved\nin exporting L-tryptophan in the E. coli bacteria used by\nCJ.\n B\n CJ challenges the Commission’s finding of infringe-\nment of both later strains on one additional ground. The\nCommission found that CJ’s later strains met the re-\nsistance limitation. CJ argues that substantial evidence\n\nAJINOMOTO CO., INC. v. ITC 25\n\n\n\ndoes not exist to support that finding. 9 We reject that ar-\ngument.\n Several pieces of evidence indicate that, as a general\nmatter, enhancing the activity of the YddG protein in-\ncreases bacteria’s resistance to L-tryptophan. Table 1 of\nthe ’655 patent shows that E. coli bacteria with multiple\ncopies of the yddG gene introduced through plasmids\ndemonstrated better growth on a tryptophan substrate,\nand thus more resistance, than unmodified E. coli bacteria.\nSee ’655 patent, col. 9, lines 50–65 (bottom row, compare\ncolumn “pUC19” (-) with column “pYDDG1” (+)). Similarly,\nthe Doroshenko article, mentioned above, describes an ex-\nperiment in which E. coli bacteria with a stronger promoter\npreceding the yddG gene demonstrated enhanced re-\nsistance to L-tryptophan. See J.A. 9455 (row DV036, col-\numn DL-5-f-Trp).\n CJ’s fermentation data, mentioned above, also provides\ndirect evidence that CJ’s later strains have increased re-\nsistance to L-tryptophan. That data shows a greater vol-\nume of tryptophan with both of CJ’s later strains than with\nunmodified E. coli bacteria. See J.A. 7957 (first later\nstrain: middle table, row F4, column “Volume produced”);\nJ.A. 10053 (second later strain: row “Product (g)” toward\nmiddle of table). Dr. Stephanopoulos indicated that a\nstrain’s ability to overproduce L-tryptophan necessarily\nmeant that the strain had increased resistance to L-trypto-\nphan. See J.A. 1448 (“[I]f that product feedback inhibits its\nown synthesis, clearly, this is not going to work.”); see also\n\n\n 9 CJ does not challenge the finding that CJ’s first\nlater strain meets the protein limitation of the claim 15 al-\nternative of claim 20. Specifically, CJ’s first later strain\nuses a non-E. coli yddG gene without codon randomization,\nwhich hybridizes with the complement of SEQ ID NO:1\n(i.e., the nucleotide sequence of the E. coli yddG gene), and\nthus falls within the literal scope of claim 15.\n\n26 AJINOMOTO CO., INC. v. ITC\n\n\n\n\nJ.A. 521 (stating that bacteria that “exhibit enhanced re-\nsistance to an aromatic L-amino acid or an aromatic L-\namino acid analog” also “overproduce the corresponding ar-\nomatic L-amino acid analog”).\n CJ’s objections to the sufficiency or even relevance of\nthis evidence are unpersuasive. CJ points out that the bac-\nteria used to generate the data in Table 1 of the ’655 patent\ncontained plasmids with more than the two copies of the\nyddG gene in CJ’s later strains. See J.A. 1229. The Doro-\nshenko article, however, indicates that enhancing the ac-\ntivity of even a single copy of the yddG gene can increase\nresistance to L-tryptophan. CJ responds that the strain\nstudied in Doroshenko used a strong λPL promoter, while\nCJ’s later strains use relatively weaker non-E. coli native\nyddG, rmf, and rhtB promoters. See J.A. 9454. But Dr.\nStephanopoulos testified that at least the rmf promoter in\nboth of CJ’s later strains also is more potent than the na-\ntive E. coli yddG promoter. J.A. 554. Thus, even if CJ is\ncorrect that its later strains do not contain tandem promot-\ners, the Commission could reasonably infer that the pro-\nmoters used in CJ’s later strains enhance the activity of the\nyddG genes relative to unmodified E. coli bacteria and\nthereby increase those strains’ resistance to L-tryptophan.\n CJ also cites Ajinomoto’s 2002 Progress Report as evi-\ndence that enhancing a single copy of the yddG gene is in-\nsufficient to enhance resistance to L-tryptophan. That\nreport states that an experiment using “only one copy” of\nthe yddG gene with a PL promoter “does not correctly\nmodel[]” an earlier experiment using a “moderate-copy-\nnumber” plasmid with the yddG gene, which had shown a\n“positive effect” of the yddG gene on tryptophan produc-\ntion. J.A. 10268. No more need be inferred from the report\nthan that enhancing a single copy of the yddG gene in-\ncreases resistance to L-tryptophan less than using a\ngreater number of copies. The Commission did not need to\ninfer that enhancing a single copy as in CJ’s later strains\ndoes not enhance resistance at all.\n\nAJINOMOTO CO., INC. v. ITC 27\n\n\n\n Further, CJ asserts that the increased production of L-\ntryptophan, and thus the enhanced resistance to L-trypto-\nphan, observed in its later strains could be attributable to\nthe presence of other genetic mutations rather than to in-\ncreased YddG protein activity alone. See J.A. 442. But the\nclaims require only that the protein “has the activity to\nmake the bacterium resistant” to L-tryptophan, not that\nthe protein be the sole cause of the bacterium’s enhanced\nresistance to L-tryptophan. See ’655 patent, col. 22, lines\n57–59. Considering the already-mentioned evidence that\nthe YddG protein generally has the effect of increasing re-\nsistance to L-tryptophan, the Commission had substantial\nevidence from which to find that it was more likely than\nnot that increased activity of the YddG protein at least\npartly contributed to the enhanced resistance of CJ’s later\nstrains.\n C\n CJ’s final contention in its cross-appeal seeks reversal\nof the Commission’s rejection of CJ’s invalidity challenge to\nclaim 20. CJ argues that substantial evidence does not\nsupport the Commission’s finding that CJ did not prove\nlack of an adequate written description for the genus of\n“more potent promoter[s]” recited in claims 9 and 15 and,\nby incorporation, in claim 20. We reject CJ’s argument.\n “[A] sufficient description of a genus . . . requires the\ndisclosure of either a representative number of species fall-\ning within the scope of the genus or structural features\ncommon to the members of the genus so that one of skill in\nthe art can ‘visualize or recognize’ the members of the ge-\nnus.” Ariad Pharm., Inc. v. Eli Lilly & Co., 598 F.3d 1336,\n1350 (Fed. Cir. 2010) (en banc). The Commission found\nboth that the ’655 patent discloses a representative number\nof species of more potent promoters and that there are\nstructural features common to the genus of more potent\npromoters. Both of those findings are supported by\n\n28 AJINOMOTO CO., INC. v. ITC\n\n\n\n\nsubstantial evidence, and they suffice to uphold the Com-\nmission’s rejection of CJ’s written-description challenge.\n 1\n As to a representative number of species, we have rec-\nognized that the amount of disclosure necessary to satisfy\nthe written-description requirement “will necessarily vary\ndepending on the context,” considering such facts as “the\nexisting knowledge in the particular field, the extent and\ncontent of the prior art, the maturity of the science or tech-\nnology,” and “the predictability of the aspect at issue.” Ar-\niad, 598 F.3d at 1351 (quoting Capon v. Eshhar, 418 F.3d\n1349, 1359 (Fed. Cir. 2005)). In some circumstances, we\nhave added, “a patentee may rely on information that is\n‘well-known in the art’ for purposes of meeting the written\ndescription requirement,” because “the specification is\nviewed from the perspective of one of skill” in the relevant\nart. Bos. Sci. Corp. v. Johnson & Johnson, 647 F.3d 1353,\n1366 (Fed. Cir. 2011).\n The ’655 patent discloses four examples of “potent pro-\nmoters”: “PL promoter of lambda phage,” the “lac pro-\nmoter,” the “trp promoter,” and the “trc promoter.” ’655\npatent, col. 6, lines 21–24. The patent also cites the 1986\narticle by Deuschle et al. as disclosing “examples of potent\npromoters” and “[m]ethods for [the] evaluation [of] the\nstrength of promoter[s].” Id., col. 6, lines 16–21. That ar-\nticle provides data about the relative strength of fourteen\npromoters and describes a general methodology for deter-\nmining promoter strength in E. coli bacteria. J.A. 6174–\n77. This evidence supports the Commission’s finding that\n“enhancing promoter activity was well-known” and that a\nskilled artisan “would have been able to identify more po-\ntent promoters by employing common tools for measuring\nRNA transcription.” J.A. 46.\n The ’655 patent also makes clear that its invention was\n“identifying the yddG gene encoding a membrane protein”\nand discovering that the gene “conferred on a\n\nAJINOMOTO CO., INC. v. ITC 29\n\n\n\nmicroorganism resistance to phenylalanine and several\namino acid analogues” when the gene was amplified or its\nexpression enhanced, see ’655 patent, col. 2, lines 46–57,\nnot the well-known techniques for performing the amplifi-\ncation or expression enhancement, see id., col. 5, line 57,\nthrough col. 6, line 33. We have explained that the repre-\nsentative-species inquiry is directed to whether the inven-\ntor “has truly invented the genus” as opposed to “a research\nplan, leaving it to others to explore the unknown contours\nof the claimed genus.” AbbVie Deutschland GmbH & Co.,\nKG v. Janssen Biotech, Inc., 759 F.3d 1285, 1300 (Fed. Cir.\n2014). Here, the genus of more potent promoters was al-\nready well explored in the relevant art by the time of the\n’655 patent’s invention. In these circumstances, the Com-\nmission permissibly found in the specification, read in light\nof the background knowledge in the art, a representative\nnumber of species for the genus of more potent promoters.\n 2\n As to a common structural feature, the Commission\nfound that a skilled artisan could “identify more potent\nyddG promoters given the well-known link between con-\nsensus sequence and promoter strength,” i.e., that promot-\ners having fewer departures from a “consensus sequence”\nin a promoter are generally stronger than promoters with\nmore departures from such a sequence. J.A. 46. 10 Substan-\ntial evidence supports that finding. For instance, a 1983\narticle by Hawley and McClure describes a study demon-\nstrating that most “mutations that decrease initiation fre-\nquency also decrease the homology of the promoter to the\nconsensus sequence, while up-mutations increase the\n\n\n 10 The consensus sequence is a specific nucleotide se-\nquence that appears in the promoters associated with\nmany different genes in the genome of a particular organ-\nism. In E. coli, the consensus sequence has two parts:\nTTGACA at the -35 region and TATAAT at the -10 region.\n\n30 AJINOMOTO CO., INC. v. ITC\n\n\n\n\nhomology in” most instances. J.A. 6237. Similarly, a 1986\narticle by Horwitz and Loeb explains that “mutations that\nincrease transcription, ‘up mutations,’ usually increase ho-\nmology with the consensus sequence and spacing,” while\n“mutations that decrease transcription, ‘down mutations,’\nusually decrease homology with the consensus sequence\nand spacing.” J.A. 6251.\n CJ disputes that similarity to the consensus sequence\ndefines a common structural feature, citing several articles\nas indicating that a promoter closer to the consensus se-\nquence will not always be stronger than one farther from\nthat sequence. For instance, a 1998 article by Jensen and\nHammer reports that a pattern observed in another organ-\nism—that “the relatively strong promoters were the perfect\nones,” i.e., those closer to the consensus sequence—“did not\nhold true for E. coli: here the promoters which had either\nan error in the consensus sequence or a shorter spacer were\nrelatively strong.” J.A. 9149. Moreover, a 1985 article by\nAoyama and Takanami states that similarity to the con-\nsensus sequence “is still not enough to predict the site and\nstrength of promoter from a given sequence,” J.A. 6215,\nand a 1999 book edited by Fernandez and Hoeffler notes\nthat “the strongest promoters in E. coli do not necessarily\nadhere to the consensus sequence,” J.A. 9113.\n CJ’s argument both assumes too strict a legal standard\nand reads too much into its cited references. Adequate\nwritten description does not require a perfect correspond-\nence between the members of the genus and the asserted\ncommon structural feature; for a functionally defined ge-\nnus like the one at issue here, we have spoken more mod-\nestly of a “correlation between structure and function.”\nAriad, 598 F.3d at 1350 (emphasis added). In any event,\nCJ’s evidence at most establishes that, starting with the\nconsensus sequence, deviations from that sequence do not\nalways decrease promoter strength, at least in E. coli. But\nthe genus at issue here is “more potent promoter[s]” than\nthe native promoter, not less potent promoters than the\n\nAJINOMOTO CO., INC. v. ITC 31\n\n\n\nconsensus sequence. And the Commission had substantial\nevidence from which to find that, starting from the native\nE. coli yddG promoter, deviations toward the consensus se-\nquence generally increase promoter strength.\n The cases cited by CJ in which we have held genus\nclaims to lack an adequate written description are inappo-\nsite. In Boston Scientific, the specification contained “no\nexamples of macrocyclic lactone analogs of rapamycin” (the\nclaimed genus) and essentially “no guidance on how to\nproperly determine whether a compound is a macrocyclic\nlactone analog of rapamycin.” 647 F.3d at 1364. In AbbVie,\nthere was “no evidence to show any described antibody to\nbe structurally similar to, and thus representative of,” an\nantibody accused of coming within the claim, nor was there\n“evidence to show whether one of skill in the art could make\npredictable changes to the described antibodies to arrive at\nother types of antibodies such as” the accused antibody.\n759 F.3d at 1301. And in Regents of the University of Cali-\nfornia v. Eli Lilly & Co., the specification described “a pro-\ncess for obtaining human insulin-encoding cDNA” (such\ncDNA required by the claim at issue) but not any “sequence\ninformation indicating which nucleotides constitute hu-\nman cDNA” or “further information in the patent pertain-\ning to that cDNA’s relevant structural or physical\ncharacteristics.” 119 F.3d 1559, 1567 (Fed. Cir. 1997).\nHere, by contrast, the ’655 patent expressly provides four\nexamples of “more potent promoters,” and the Commission\nsupportably found that a skilled artisan could make rela-\ntively predictable changes to the native promoter to arrive\nat a more potent promoter.\n IV\n For the foregoing reasons, we affirm the Commission’s\ndecision.\n No costs.\n AFFIRMED\n\n United States Court of Appeals\n for the Federal Circuit\n ______________________\n\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\n INC.,\n Appellants\n\n v.\n\n INTERNATIONAL TRADE COMMISSION,\n Appellee\n\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\n CHEIJEDANG INDONESIA,\n Intervenors\n\n--------------------------------------------\n\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\n CHEIJEDANG INDONESIA,\n Appellants\n\n v.\n\n INTERNATIONAL TRADE COMMISSION,\n Appellee\n\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\n INC.,\n Intervenors\n ______________________\n\n 2018-1590, 2018-1629\n ______________________\n\n2 AJINOMOTO CO., INC. v. ITC\n\n\n\n\n Appeals from the United States International Trade\nCommission in Investigation No. 337-TA-1005.\n ______________________\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: AJINOMOTO CO., INC., Ajinomoto Heartland Inc., Appellants v. INTERNATIONAL TRADE COMMISSION, Appellee CJ CheilJedang Corp., CJ America, Inc., PT CheilJedang Indonesia, Intervenors CJ CheilJedang Corp., CJ America, Inc., PT CheilJedang Indonesia, Appellants v. International Trade Commission, Appellee Ajinomoto Co., Inc., Ajinomoto Heartland Inc., Intervenors\\nCourt: U.S. Court of Appeals for the Federal Circuit\\nDate decided: 2019-08-06\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: Yes\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\\n INC.,\\n Appellants\\n\\n v.\\n\\n INTERNATIONAL TRADE COMMISSION,\\n Appellee\\n\\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\\n CHEILJEDANG INDONESIA,\\n Intervenors\\n\\n--------------------------------------------\\n\\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\\n CHEILJEDANG INDONESIA,\\n Appellants\\n\\n v.\\n\\n INTERNATIONAL TRADE COMMISSION,\\n Appellee\\n\\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\\n INC.,\\n Intervenors\\n ______________________\\n\\n 2018-1590, 2018-1629\\n ______________________\\n\\n2 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\n Appeals from the United States International Trade\\nCommission in Investigation No. 337-TA-1005.\\n ______________________\\n\\n Decided: August 6, 2019\\n ______________________\\n\\n JOHN D. LIVINGSTONE, Finnegan, Henderson, Farabow,\\nGarrett & Dunner, LLP, Atlanta, GA, argued for Ajinomoto\\nCo., Inc., Ajinomoto Heartland Inc. Also represented by\\nMARTIN DAVID WEINGARTEN; CHARLES E. LIPSEY, Reston,\\nVA; MAREESA ARNITA FREDERICK, CORA RENAE HOLT,\\nBARBARA RUDOLPH, Washington, DC.\\n\\n HOUDA MORAD, Office of General Counsel, United\\nStates International Trade Commission, Washington, DC,\\nargued for appellee. Also represented by SIDNEY A.\\nROSENZWEIG, DOMINIC L. BIANCHI, WAYNE W. HERRINGTON.\\n\\n JAMES F. HALEY, JR., Haley Guiliano LLP, New York,\\nNY, argued for CJ CheilJedang Corp., CJ America, Inc., PT\\nCheiJedang Indonesia. Also represented by STEVEN PEPE,\\nRopes & Gray LLP, New York, NY; MATTHEW RIZZOLO,\\nWashington, DC.\\n ______________________\\n\\n Before DYK, MOORE, and TARANTO, Circuit Judges.\\n Opinion for the court filed by Circuit Judge TARANTO.\\nOpinion concurring in part and dissenting in part filed by\\n Circuit Judge DYK.\\nTARANTO, Circuit Judge.\\n Ajinomoto Co., Inc. and Ajinomoto Heartland Inc. (col-\\nlectively, Ajinomoto) filed a complaint against CJ\\nCheilJedang Corp., CJ America, Inc., and PT CheilJedang\\nIndonesia (collectively, CJ) with the International Trade\\nCommission, alleging that CJ was importing certain\\n\\nAJINOMOTO CO., INC. v. ITC 3\\n\\n\\n\\nproducts that infringed Ajinomoto’s U.S. Patent No.\\n7,666,655. CJ used several strains of Escherichia coli bac-\\nteria to produce L-tryptophan products, which it then im-\\nported into the United States. The Commission\\ndetermined that CJ’s earlier strains did not infringe but\\nthat CJ’s two later strains did. The Commission also found\\nthat the relevant claim of the ’655 patent is not invalid for\\nlack of an adequate written description.\\n Ajinomoto appeals the Commission’s claim construc-\\ntion underlying the determination of no infringement by\\nthe earlier strains. CJ cross-appeals aspects of the deter-\\nmination of infringement by the later strains and the rejec-\\ntion of the invalidity challenge. We affirm.\\n I\\n A\\n The ’655 patent claims E. coli bacteria that have been\\ngenetically engineered to increase their production of aro-\\nmatic L-amino acids, such as L-tryptophan, during fermen-\\ntation, as well as methods of producing aromatic L-amino\\nacids using such bacteria. See ’655 patent, col. 2, lines 40–\\n45. In particular, the ’655 patent identifies a specific gene\\nin the E. coli genome, the yddG gene, that encodes a mem-\\nbrane protein, the YddG protein. Id., col. 2, lines 46–48.\\nThat protein transports aromatic L-amino acids out of the\\nbacterial cell and into the surrounding culture medium,\\nwhere they can be collected. See id., col. 7, lines 11–16.\\nWhen yddG gene activity in bacteria is enhanced so that\\nmore YddG protein is produced, the bacteria show in-\\ncreased production of, and increased resistance to, aro-\\nmatic L-amino acids. Id., col. 2, lines 49–57. 1\\n\\n\\n\\n 1 The specification defines a bacterium’s “resistance”\\nto an amino acid as its ability “to grow on a minimal me-\\ndium containing” the amino acid on “which unmodified or\\n\\n4 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\n The ’655 patent describes three ways to enhance the\\nactivity of the yddG gene. First, plasmids containing addi-\\ntional copies of the yddG gene can be introduced into the\\nbacterium. Id., col. 2, lines 50–52; id., col. 5, line 62,\\nthrough col. 6, line 2. Second, additional copies of the yddG\\ngene can be inserted into the bacterial chromosome. Id.,\\ncol. 2, lines 52–54; id., col. 6, lines 3–6. Third, a stronger\\n“promoter” than the one native to the E. coli yddG gene can\\nbe used. Id., col. 2, lines 54–57; id., col. 6, lines 12–15. 2\\n Claim 20, the only claim of the ’655 patent still asserted\\nwhen the Commission issued its decision, claims “[a]\\nmethod for producing an aromatic L-amino acid, which\\ncomprises cultivating the bacterium according to any\\none of claims 9–12, 13, 14, 15–18, or 19.” Id., col. 24, lines\\n\\n\\n\\nthe wild type, or the parental strain of the bacterium can-\\nnot grow,” or its ability “to grow faster” on such a medium\\n“than unmodified or the wild type, or the parental strain of\\nthe bacterium.” ’655 patent, col. 4, lines 49–56.\\n 2 A promoter is a nucleotide sequence within a DNA\\nmolecule, located adjacent to the nucleotide sequence that\\nconstitutes the gene to be expressed. The Lewin textbook\\ncited by Ajinomoto shows a “typical promoter” around 41\\nnucleotides long. J.A. 6043; see also J.A. 6177 (article by\\nDeuschle et al., cited at ’655 patent, col. 6, lines 18–21,\\nshowing longer promoters). The promoter is the binding\\nsite for RNA polymerase, which initiates transcription (the\\nfirst step in gene expression) by separating the two strands\\nof DNA. The ’655 patent’s specification defines “[s]trength\\nof promoter” with reference to the “frequency of acts of the\\nRNA synthesis initiation.” ’655 patent, col. 6, lines 15–16.\\n The promoter is only one part of a gene’s “expression\\nregulation sequence,” which controls expression of the\\ngene. See id., col. 3, line 14; id., col. 5, line 2. Besides pro-\\nmoters, the “expression regulation sequence” can include,\\ne.g., operators, enhancers, terminators, and silencers.\\n\\nAJINOMOTO CO., INC. v. ITC 5\\n\\n\\n\\n4–6 (emphasis added). Of the claims in that list, claims 9\\nand 15 are the independent claims, and they are the two\\nalternatives, under claim 20, of importance in this case.\\n Claim 9 recites:\\n 9. A recombinant Escherichia coli bacterium,\\n which has the ability to accumulate aromatic L-\\n amino acid in a medium, wherein the aromatic L-\\n amino acid production by said bacterium is en-\\n hanced by enhancing activity of a protein in a cell\\n of said bacterium beyond the levels observed in a\\n wild-type of said bacterium,\\n [1] and in which said protein consists of the\\n amino acid sequence of SEQ ID NO: 2\\n [2] and said protein has the activity to make\\n the bacterium resistant to L-phenylalanine, fluoro-\\n phenylalanine or 5[-]fluoro-DL-tryptophan,\\n [3] wherein the activity of the protein is en-\\n hanced by [3a] transformation of the bacterium\\n with a DNA encoding the protein to express the\\n protein in the bacterium, [3b] by replacing the na-\\n tive promoter which precedes the DNA on the chro-\\n mosome of the bacterium with a more potent\\n promoter, [3c] or by introduction of multiple copies\\n of the DNA encoding said protein into the chromo-\\n some of said bacterium to express the protein in\\n said bacterium.\\nId., col. 22, lines 51–67 (paragraph breaks and bold num-\\nbering added). The Commission referred to limitation [1]\\nas the “protein limitation,” limitation [2] as the “resistance\\nlimitation,” and limitation [3] as the “enhancement limita-\\ntion.” Claim 15 is materially identical to claim 9, except for\\nthe protein limitation. Whereas claim 9 identifies the\\nclaimed protein by a specific amino-acid sequence, claim 15\\nidentifies it by reference to a corresponding DNA se-\\nquence—a protein “encoded by the nucleotide sequence\\n\\n6 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nwhich hybridizes with the complement of the nucleotide se-\\nquence of SEQ ID NO: 1 under” certain conditions. See id.,\\ncol. 23, lines 14–32.\\n B\\n In May 2016, Ajinomoto filed a complaint against CJ\\nwith the Commission under 19 U.S.C. § 1337. Ajinomoto\\nalleged that CJ violated § 1337(a)(1)(B)(ii) by importing an-\\nimal-feed-grade L-tryptophan products produced by a pro-\\ncess covered by the ’655 patent. 3 The Commission\\ninstituted an investigation based on Ajinomoto’s com-\\nplaint.\\n The parties before us, including the Commission, agree\\nthat whether the accused products were produced by a pro-\\ncess covered by the patent is a question of infringement.\\nThe proceeding focused on three groups of E. coli strains\\nthat CJ has used to produce tryptophan. First, CJ’s “ear-\\nlier strains” contained both the native E. coli yddG gene\\nand the native E. coli yddG promoter, except that the first\\nnucleotide of the promoter was changed through chemical\\nmutagenesis, resulting in a stronger promoter. Second, in\\nNovember 2016, several months after Ajinomoto filed its\\ncomplaint, CJ began using its first “later strain,” which\\ncontained two copies of a yddG gene: (1) the native E. coli\\nyddG gene with the native E. coli yddG promoter; and (2) a\\nnon-E. coli yddG gene with two promoters—(2a) a native\\nnon-E. coli yddG promoter and (2b) an rmf promoter. 4\\nThird, in December 2016, CJ started using its second “later\\n\\n\\n\\n 3 Ajinomoto also alleged that CJ infringed U.S. Pa-\\ntent No. 6,180,373, which similarly claims methods of pro-\\nducing tryptophan using genetically engineered bacteria.\\nThe ’373 patent expired on January 30, 2018, and is not at\\nissue in this court.\\n 4 The rmf and rhtB promoters are promoters associ-\\nated with other genes in the E. coli genome.\\n\\nAJINOMOTO CO., INC. v. ITC 7\\n\\n\\n\\nstrain,” which also contained two copies of a yddG gene:\\n(1) the native E. coli yddG gene with the native E. coli\\nyddG promoter; and (2) a codon-randomized non-E. coli\\nyddG gene with two promoters—(2a) an rmf promoter and\\n(2b) an rhtB promoter. 5\\n In August 2017, the administrative law judge (ALJ) is-\\nsued a final initial determination. The ALJ construed “re-\\nplacing the native promoter . . . with a more potent\\npromoter” in the enhancement limitation to mean “remov-\\ning the native upstream region of the yddG gene and in-\\nserting one of a class of promoters that controls expression\\nof a different gene.” J.A. 90–91. Using that construction,\\nthe ALJ found that CJ’s earlier strains did not infringe; he\\nfound that they failed to meet the enhancement limitation\\nbecause CJ created the more potent promoter in those\\nstrains by mutagenesis of a single nucleotide rather than\\nremoval of the entire native promoter and insertion of a\\nnew promoter. As to CJ’s later strains, the ALJ found that\\n(a) the first later strain did not infringe because Ajinomoto\\nhad failed to prove that it met the resistance limitation,\\nand (b) the second later strain also did not infringe because\\n\\n\\n\\n 5 Each particular codon (three nucleotides in a row\\non a DNA molecule) that encodes for an amino acid always\\nencodes for the same amino acid, but many of the 20 amino\\nacids are encoded by more than one of the 64 codons. See\\nAmgen, Inc. v. Chugai Pharm. Co., 927 F.2d 1200, 1208 n.4\\n(Fed. Cir. 1991) (discussing “redundancy” of genetic code).\\nFor instance, the DNA sequences TTA and TTG both code\\nfor the amino acid leucine. “Codon randomization” refers\\nto creation of DNA molecules that use different codons (e.g.,\\nTTA or TTG) to code for the same amino acid (e.g., leucine)\\nin building the same protein. See Mycogen Plant Sci. v.\\nMonsanto Co., 243 F.3d 1316, 1323 (Fed. Cir. 2001) (“[O]ne\\ncodon can be substituted for another in the gene without\\nchanging the amino acid and resulting protein.”).\\n\\n8 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nits non-E. coli YddG protein was not equivalent to the\\nclaimed E. coli YddG protein under the doctrine of equiva-\\nlents. Finally, the ALJ found that claim 20 of the ’655 pa-\\ntent is invalid for lack of an adequate written description\\nof the “more potent promoter” limitation incorporated into\\nthat claim.\\n In October 2017, the full Commission decided to review\\nthe ALJ’s final initial determination in its entirety, and in\\nDecember 2017, the Commission issued its decision. It af-\\nfirmed the ALJ’s construction of “replacing the native pro-\\nmoter . . . with a more potent promoter” and accordingly\\naffirmed the ALJ’s finding that CJ’s earlier strains did not\\ninfringe. But the Commission reversed several of the ALJ’s\\nother findings. Specifically, it determined that both of CJ’s\\nlater strains met all disputed claim limitations and thus\\ninfringed claim 20 and that claim 20 was not proved to lack\\nan adequate written description. The Commission accord-\\ningly entered a limited exclusion order against CJ’s infring-\\ning products, i.e., those made by both of CJ’s later strains\\nbut not its earlier strains. The Commission also issued a\\ncease-and-desist order against CJ America, which held in-\\nventory of the infringing products.\\n Ajinomoto and CJ both timely appealed. We have ju-\\nrisdiction under 28 U.S.C. § 1295(a)(6).\\n II\\n We begin with Ajinomoto’s appeal of the Commission’s\\nfinding of no infringement by the earlier strains.\\nAjinomoto challenges that finding solely by arguing that\\nthe Commission erred in its claim construction of “replac-\\ning the native promoter . . . with a more potent promoter.”\\nAjinomoto argues that, properly construed, the phrase is\\nnot limited to removing the entire native promoter and in-\\nserting a new promoter, as the Commission concluded, but\\nencompasses mutagenesis of individual nucleotides within\\nthe native promoter. We review the Commission’s claim\\nconstruction de novo, as the Commission relied on only\\n\\nAJINOMOTO CO., INC. v. ITC 9\\n\\n\\n\\nintrinsic evidence and made no factual findings based on\\nextrinsic evidence. Teva Pharm. USA, Inc. v. Sandoz, Inc.,\\n135 S. Ct. 831, 841 (2015); see Cont’l Circuits LLC v. Intel\\nCorp., 915 F.3d 788, 795 (Fed. Cir. 2019). We agree with\\nthe Commission’s claim construction and therefore affirm\\nthe non-infringement finding.\\n The ordinary and customary meaning of the claim lan-\\nguage provides support for the Commission’s claim con-\\nstruction. The language of “replacing the native promoter\\n. . . with a more potent promoter” suggests, in ordinary par-\\nlance, an operation at the level of the entire promoter as a\\nunit, not at the level of a single nucleotide that is just one\\nsmall component of the promoter. To say that one is “re-\\nplacing” an object (e.g., a laptop computer, a bicycle, a sail-\\nboat, a blender) suggests that one is doing more than\\naltering one small part of it. That suggestion is bolstered\\nwhen one also uses language (here, a “more potent pro-\\nmoter”) referring to the replacement at the level of the\\noverall object. The suggestion is further reinforced by the\\nmost apt of the dictionary definitions of “replace” intro-\\nduced before the Commission—“to provide a substitute\\nfor.” J.A. 10361; see also J.A. 5622 (patent applicants ex-\\nplaining “replacing” as “substitut[ing]”). In many contexts,\\none would not refer to swapping out one small component\\nof a larger unit as “replacing” the unit or as providing a\\n“substitute” for the unit, even though the net result is a\\ndifferently constituted larger unit. Context matters, of\\ncourse, but here, Ajinomoto has not shown a contrary com-\\nmon understanding (or even one of several common under-\\nstandings) among relevant artisans in the specific context\\nof replacing a promoter with a more potent promoter. Ac-\\ncordingly, the claim language, though hardly establishing\\na plain meaning, supports the Commission’s construction.\\n The specification offers additional support, though it\\ntoo is hardly plain insofar as it bears on the particular con-\\nstruction issue. The specification states that “the enhance-\\nment of gene expression can be achieved by locating the\\n\\n10 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nDNA of the present invention under control of more potent\\npromoter instead of the native promoter.” ’655 patent,\\ncol. 6, lines 12–15. That statement speaks of a promoter as\\na unit, but it does not use the language of “replacing.” In-\\ndeed, the specification nowhere uses that language. But it\\ndoes discuss “substituting” promoters, using a term that,\\nas indicated above, is an apt definition of “replacing” here.\\nThe specification describes “[t]he present inventions” as in-\\ncluding “[t]he bacterium according to the above bacterium,\\nwherein native promoter of said DNA is substituted with\\nmore potent promoter.” Id., col. 3, lines 19–21. The term\\nis then used in Example 4, which is titled “Substitution of\\nthe Native Upstream Region of yddG Gene by the Hybrid\\nRegulatory Element Carrying the PL Promoter and SDlacZ\\nin E. coli Chromosome,” and which involves removing the\\nentire native promoter and inserting a new promoter. See\\nid., col. 11, line 5, through col. 12, line 46. The sole specifi-\\ncation example of “substitution” thus fits the Commission’s\\nclaim construction. And while the specification discusses\\nmutagenesis, it does so only in the context of the protein-\\ncoding region of the yddG gene, not the promoter. See id.,\\ncol. 5, lines 18–30.\\n We turn finally to the prosecution history—on which\\nthe parties to this case have focused most of their compet-\\ning analyses. We conclude that the best understanding of\\nwhat transpired before the examiner further supports the\\nCommission’s construction. Because the prosecution his-\\ntory reinforces what is already suggested by the claim lan-\\nguage and specification, this case provides no occasion,\\ncontrary to Ajinomoto’s contention (Ajinomoto Br. 34), for\\nrequiring clear and unmistakable disavowal or disclaimer\\nto justify a claim construction contrary to a meaning evi-\\ndent from the claim language and specification.\\n What was claim 2 of the original application recited\\n“[t]he bacterium according to claim 1, wherein said activi-\\nties of proteins . . . is enhanced by transformation of said\\nbacterium with DNA coding for the protein . . . or by\\n\\nAJINOMOTO CO., INC. v. ITC 11\\n\\n\\n\\nalteration of expression regulation sequence of said DNA on\\nthe chromosome of the bacterium.” J.A. 5047 (emphasis\\nadded). 6 The examiner rejected the claim for lack of an ad-\\nequate written description and lack of enablement. As to\\nwritten description, the examiner explained that “[w]hile\\ngeneric expression regulation sequences are known in the\\nart, a particular, endogenous expression regulation se-\\nquence for the DNA that encodes [amino-acid] SEQ ID\\nNO:2, or related sequences, is not described.” J.A. 5371.\\n“Without description of the endogenous regulation se-\\nquence,” the examiner continued, “an endogenous regula-\\ntion sequence that has been altered to increase expression\\nof said protein also lacks adequate written description.” Id.\\nTurning to enablement, the examiner stated:\\n The specification, while being enabling for Esche-\\n richia strains wherein the native promoter for the\\n DNA encoding SEQ ID NO: 2 has been changed by\\n substitution with a more potent promoter, does not\\n reasonably provide enablement for the genus of an\\n L-amino acid producing bacterium wherein the ac-\\n tivity of proteins described by SEQ ID NO: 2 and\\n related sequences is increased due to specific alter-\\n ations within the chromosomal expression regula-\\n tion sequence for DNA encoding said proteins.\\n ....\\n\\n\\n\\n\\n 6 Although claims 9 and 15 issued from what were\\nnumbered as claims 12 and 24 when added during prose-\\ncution, the parties do not dispute that the amendments to\\noriginal claim 2 (which eventually was cancelled) are rele-\\nvant to construing issued claims 9 and 15. The same “re-\\nplacing the native promoter . . . with a more potent\\npromoter” language added to original claim 2 was eventu-\\nally added to claims 9 and 15.\\n\\n12 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\n The instant specification teaches how to select\\n Escherichia bacteria that have an increased pro-\\n duction of L-amino acids, and the art teaches how\\n to mutagenize chromosomal DNA and how to char-\\n acterize the mutations in the DNA. However, nei-\\n ther the specification nor the art contain any\\n examples of how to specifically change endogenous\\n Escherichia chromosomal expression regulation se-\\n quences for the DNA encoding proteins described\\n by SEQ ID NO: 2, or related sequences, such that\\n the activity of said proteins in the bacteria is in-\\n creased. The art and the specification provide en-\\n ablement for inserting a known promoter in the\\n chromosomal DNA to upregulate the expression of\\n the DNA encoding SEQ ID NO: 2; however, neither\\n the specification nor the art enable making specific\\n changes to expression regulation sequences for\\n DNA encoding SEQ ID NO:2 and related sequences\\n on the chromosome of Escherichia bacteria. The\\n art and specification lack a detailed description of\\n the structure of the instant endogenous expression\\n regulation sequences, and they lack any guidance\\n on how to alter such sequences such that DNA ex-\\n pression is increased; therefore, to make the in-\\n stant bacteria with altered expression regulation\\n sequences would be unpredictable.\\nJ.A. 5374–75.\\n In response to the rejections, the applicants amended\\nthe claim to recite “replacing the native promoter that pre-\\ncedes a DNA encoding said protein . . . with a more potent\\npromoter” instead of “by alteration of expression regulation\\nsequence of said DNA.” J.A. 5610. The applicants ex-\\nplained the amendment as follows: “Applicants have\\namended Claim 2 consistent with the Examiner’s recogni-\\ntion that the specification enables Escherichia strains\\nwherein the native promoter for the DNA encoding SEQ ID\\n\\nAJINOMOTO CO., INC. v. ITC 13\\n\\n\\n\\nNO: 2 has been changed by substitution with a more potent\\npromoter.” J.A. 5622.\\n Reading the written-description and enablement rejec-\\ntions together, we think that the most reasonable under-\\nstanding of the examiner’s comments is that the examiner\\nwas drawing a distinction between alterations to the pro-\\nmoter, which were sufficiently described and enabled be-\\ncause E. coli promoters were well understood in the art,\\nand alterations to the expression-regulation sequence more\\nbroadly, which were not adequately described or enabled.\\nTo be sure, the examiner’s statement that the art and the\\nspecification “lack any guidance on how to alter such se-\\nquences such that DNA expression is increased” might at\\nfirst suggest that the applicants had not described and en-\\nabled the full scope of “alteration.” But in context, this\\nstatement is best read as meaning that the applicants had\\nnot described and enabled the full scope of “expression reg-\\nulation sequence,” so that “alteration” of that sequence also\\nwas not adequately described or enabled, even though gen-\\neral techniques for altering DNA sequences were well\\nknown in the relevant art.\\n We need not determine the precise basis for the exam-\\niner’s rejections, however, as “there is no principle of patent\\nlaw that the scope of a surrender of subject matter during\\nprosecution is limited to what is absolutely necessary to\\navoid a prior art reference that was the basis for an exam-\\niner's rejection.” Norian Corp. v. Stryker Corp., 432 F.3d\\n1356, 1361 (Fed. Cir. 2005). Rather, patentees frequently\\n“surrender more through amendment than may have been\\nabsolutely necessary to avoid particular prior art.” Id.\\nThat principle logically extends to amendments made to\\novercome rejections under § 112. Cf. Biogen Idec, Inc. v.\\nGlaxoSmithKline LLC, 713 F.3d 1090, 1095–96 (Fed. Cir.\\n2013). Indeed, we have stated more generally that “[t]he\\nquestion is what a person of ordinary skill would under-\\nstand the patentee to have disclaimed during prosecution,\\nnot what a person of ordinary skill would think the\\n\\n14 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\npatentee needed to disclaim during prosecution.” Tech.\\nProps. Ltd. LLC v. Huawei Techs. Co., 849 F.3d 1349, 1359\\n(Fed. Cir. 2017). A patentee must “be held to what he de-\\nclares during the prosecution of his patent,” because a con-\\ntrary rule would undermine “[t]he public notice function of\\na patent.” Springs Window Fashions LP v. Novo Indus.,\\nL.P., 323 F.3d 989, 995 (Fed. Cir. 2003).\\n We conclude that this is a case where the applicants\\nsurrendered more than may have been necessary. As dis-\\ncussed above, the best reading of the prosecution history is\\nthat, to overcome the written-description and enablement\\nrejections, it might well have sufficed if the applicants had\\nnarrowed their claims from alterations to the overall ex-\\npression-regulation sequence to alterations to the pro-\\nmoter. But the applicants did not merely change\\n“expression regulation sequence” to “native promoter”;\\nthey also changed “alteration” to “replacing.” Just as\\n“when different words are used in separate claims, they are\\npresumed to have different meanings,” Aspex Eyewear, Inc.\\nv. Marchon Eyewear, Inc., 672 F.3d 1335, 1349 (Fed. Cir.\\n2012), when a word is changed during prosecution, the\\nchange tends to suggest that the new word differs in mean-\\ning in some way from the original word.\\n That inference is bolstered by the applicants’ remarks\\naccompanying the amendment. Those remarks effectively\\nequate “replacing the native promoter . . . with a more po-\\ntent promoter” in the amended claim with “chang[ing]” the\\nnative promoter “by substitution with a more potent pro-\\nmoter.” J.A. 5622. As we have already noted, Example 4,\\ndescribed as involving “substitution” of a promoter, in-\\nvolves removal of the entire native promoter and insertion\\nof a new promoter. ’665 patent, col. 11, line 5, through col.\\n12, line 46. The applicants’ remarks, understood in light of\\nthe word choices and the specification, thus reinforce the\\nCommission’s conclusion that the new claim language does\\nnot include mutagenesis of individual nucleotides.\\n\\nAJINOMOTO CO., INC. v. ITC 15\\n\\n\\n\\n For those reasons, we affirm the Commission’s claim\\nconstruction and its finding that CJ’s earlier strains do not\\ninfringe based on that claim construction.\\n III\\n CJ, in its cross-appeal, challenges the Commission’s\\ndeterminations that CJ’s second later strain met the pro-\\ntein limitation, that both of CJ’s later strains met the re-\\nsistance limitation, and that claim 20 is not invalid for lack\\nof an adequate written description. We affirm the Commis-\\nsion as to all three issues.\\n A determination of infringement or non-infringement,\\nwhether literal or under the doctrine of equivalents, is a\\nfinding of fact, reviewed here for substantial evidence.\\nKinik Co. v. Int’l Trade Comm’n, 362 F.3d 1359, 1361 (Fed.\\nCir. 2004). But a determination of the applicability or in-\\napplicability of prosecution history estoppel, which limits\\nthe availability of the doctrine of equivalents, is a matter\\nof law, reviewed de novo. Spectrum Pharm., Inc. v. Sandoz\\nInc., 802 F.3d 1326, 1337 (Fed. Cir. 2015). The determina-\\ntion that a patent claim did not lack adequate support in\\nthe written description is a factual finding, reviewed for\\nsubstantial evidence. Rivera v. Int’l Trade Comm’n, 857\\nF.3d 1315, 1319 (Fed. Cir. 2017). Ajinomoto had to prove\\ninfringement by a preponderance of the evidence, while CJ\\nhad to prove invalidity by clear and convincing evidence.\\nSee Motorola Mobility, LLC v. Int’l Trade Comm’n, 737\\nF.3d 1345, 1348 (Fed. Cir. 2013); Enercon GmbH v. Int’l\\nTrade Comm’n, 151 F.3d 1376, 1384 (Fed. Cir. 1998).\\n A\\n The Commission found that CJ’s second later strain in-\\nfringed claim 20, which covers two alternatives of rele-\\nvance in this case—the claim 9 alternative and the claim\\n15 alternative. The infringement finding for CJ’s second\\nlater strain does not rest on the claim 15 alternative,\\nwhich, in its protein limitation, requires a protein encoded\\n\\n16 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nby a nucleotide sequence that hybridizes with the comple-\\nment of SEQ ID NO:1 (the nucleotide sequence of the E.\\ncoli yddG gene). The Commission did not find, and\\nAjinomoto does not argue for, either literal or equivalents\\ninfringement based on claim 15. The Commission found\\ninfringement under the claim 9 alternative—specifically, it\\nfound that the YddG protein encoded by the codon-random-\\nized non-E. coli yddG gene of this strain is an equivalent of\\nSEQ ID NO:2 (the amino-acid sequence of the E. coli YddG\\nprotein), as required by the protein limitation of claim 9.\\n CJ challenges that finding on two grounds. Based on\\nan amendment to original claims made during prosecution,\\nCJ asserts that prosecution history estoppel bars\\nAjinomoto from relying on the doctrine of equivalents to\\nmeet the protein limitation. Separately, CJ asserts that\\nthe non-E. coli YddG protein of CJ’s second later strain\\ncannot reasonably be found to be an equivalent of the\\nclaimed E. coli YddG protein under the function-way-result\\ntest for equivalence. We address those arguments in turn.\\n 1\\n Under the doctrine of prosecution history estoppel, “[a]\\npatentee’s decision to narrow his claims through amend-\\nment may be presumed to be a general disclaimer of the\\nterritory between the original claim and the amended\\nclaim.” Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki\\nCo., 535 U.S. 722, 740 (2002). The Supreme Court has\\nspecified three ways the patentee can rebut that presump-\\ntion, each of which, if established, means that “the amend-\\nment cannot reasonably be viewed as surrendering a\\nparticular equivalent.” Id. First, “[t]he equivalent may\\nhave been unforeseeable at the time of the application.” Id.\\nSecond, “the rationale underlying the amendment may\\nbear no more than a tangential relation to the equivalent\\nin question.” Id. Third, “there may be some other reason\\nsuggesting that the patentee could not reasonably be\\n\\nAJINOMOTO CO., INC. v. ITC 17\\n\\n\\n\\nexpected to have described the insubstantial substitute in\\nquestion.” Id. at 740–41.\\n In this case, the relevant facts about what transpired\\nduring prosecution are as follows. Claim 1 as originally\\nfiled recited two alternative conditions for the claimed pro-\\ntein:\\n a protein as defined in the following (A) or (B) in a\\n cell of said bacterium:\\n (A) a protein which comprises the amino acid se-\\n quence shown in SEQ ID NO:2 in Sequence listing;\\n (B) a protein which comprises an amino acid se-\\n quence including deletion, substitution, insertion\\n or addition of one or several amino acids in the\\n amino acid sequence shown in SEQ ID NO:2 in Se-\\n quence listing.\\nJ.A. 5047. The examiner rejected that claim as anticipated\\nby a reference disclosing the E. coli “yfiK gene product” (i.e.,\\nthe E. coli YfiK protein)—which differed from SEQ ID\\nNO:2 by deletion, substitution, insertion, or addition of sev-\\neral amino acids and, therefore, did not come within the (A)\\nalternative but did come within the (B) alternative. J.A.\\n5378. In response, the applicants left the (A) alternative\\nalone but replaced the language following (B) with new lan-\\nguage: “a protein which comprises an amino acid sequence\\nthat is encoded by a nucleotide sequence that hybridizes\\nwith the nucleotide sequence of SEQ ID NO:1 under strin-\\ngent conditions.” J.A. 5609. 7\\n\\n\\n\\n\\n 7 As previously noted, claims 9 and 15 issued from\\nnew claims added at the same time as this amendment.\\nSee supra note 6. Claims 9 and 15 respectively contain the\\nsame language as the (A) and (B) limitations in claim 1 af-\\nter it was amended. Claim 20, the claim at issue, treats\\n\\n18 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\n As an initial matter, CJ’s argument for prosecution his-\\ntory estoppel in this case involves an unusual circum-\\nstance. The infringement determination does not rest on\\nfinding an equivalent of the new claim language—namely,\\nthe (nucleotide) SEQ ID NO:1 language now in claim 15.\\nRather, it rests on finding an equivalent of the (amino-acid)\\nSEQ ID NO:2 language now in claim 9, which was not itself\\naltered by the amendment at issue. That is, the original\\nclaim provided two alternatives; only the second was mod-\\nified by amendment; and only the first is asserted as the\\nbasis for infringement by CJ’s second later strain. But we\\nneed not reach Ajinomoto’s contention that, in this circum-\\nstance, prosecution history estoppel does not apply at all,\\ni.e., that there is not even a presumed (though rebuttable)\\nsurrender of the asserted equivalent. The Commission did\\nnot so rule, instead concluding that the “tangential rela-\\ntion” exception applied, so that Ajinomoto did not surren-\\nder the protein produced by the codon-randomized non-E.\\ncoli yddG gene of CJ’s second later strain. J.A. 41–44. We\\nagree with that conclusion. 8\\n In applying the “tangential relation” exception, we\\n“ask[] whether the reason for the narrowing amendment\\nwas peripheral, or not directly relevant, to the alleged\\nequivalent.” Festo Corp. v. Shoketsu Kinzoku Kogyo Ka-\\nbushiki Co., 344 F.3d 1359, 1369 (Fed. Cir. 2003). “[T]he\\ninquiry into whether a patentee can rebut the Festo\\n\\n\\n\\nclaims 9 and 15 as alternatives in the same way that orig-\\ninal and amended claim 1 treated (A) and (B).\\n 8 CJ contends that Ajinomoto forfeited invocation of\\nthe “tangential relation” exception because it did not in-\\nvoke the exception before the ALJ or in its request for re-\\nview by the full Commission. CJ cites no authority that\\nbarred the Commission from exercising discretion to raise\\nthe issue and give the parties an adequate opportunity to\\naddress it, as the Commission did here.\\n\\nAJINOMOTO CO., INC. v. ITC 19\\n\\n\\n\\npresumption under the ‘tangential’ criterion focuses on the\\npatentee’s objectively apparent reason for the narrowing\\namendment.” Id. Our cases require the patentee to show\\nthat the way in which the alleged equivalent departs from\\nwhat the claim limitation literally requires is tangential to\\nthe discernible objective reason for the narrowing amend-\\nment. In that situation, there is no surrender of the equiv-\\nalent by that amendment.\\n For instance, in Insituform Technologies, Inc. v. CAT\\nContracting, Inc., the patent claimed a method of using a\\nvacuum to impregnate a flexible tube with resin. 385 F.3d\\n1360, 1362–63 (Fed. Cir. 2004). The claims were originally\\nrejected over a prior-art reference that disclosed a single\\nvacuum source located far away from the resin source. Id.\\nat 1369. The applicant amended the claim at issue to re-\\nquire a single vacuum source placed near the resin source.\\nSee id. at 1368–70. The alleged equivalent used multiple\\nvacuum sources. Id. at 1369–70. We held that the “tan-\\ngential relation” exception applied, observing that the pur-\\npose of the narrowing amendment was to distinguish the\\ninvention from the prior art based on the location of the\\nvacuum source relative to the resin, not to limit the number\\nof vacuum sources. Id. at 1370.\\n Similarly, in Regents of the University of California v.\\nDakocytomation California, Inc., the patented method in-\\nvolved using DNA testing to detect chromosomal abnor-\\nmalities. 517 F.3d 1364, 1369–70 (Fed. Cir. 2008). The\\nclaim at issue originally recited “disabling the hybridiza-\\ntion capacity of repetitive sequences” generally. Id. at\\n1377. The examiner rejected the claim over several prior-\\nart references, one of which disclosed disabling hybridiza-\\ntion using unique sequence probes. Id. at 1378. In re-\\nsponse, the applicants amended the claim to recite a\\nparticular technique of disabling hybridization using block-\\ning nucleic acids. Id. The parties stipulated that the added\\n“blocking nucleic acid” limitation was limited to human nu-\\ncleic acids, but the alleged equivalent used synthetic\\n\\n20 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nnucleic acids. Id. at 1376. We concluded that the narrow-\\ning amendment was tangential to how the equivalent dif-\\nfered from the literal claim limitation: “[I]n narrowing the\\nclaim to overcome the prior art rejections, the focus of the\\npatentees’ arguments centered on the method of blocking—\\nnot on the particular type of nucleic acid that could be used\\nfor blocking.” Id. at 1378. Indeed, we noted, “the ‘nucleic\\nacid’ limitation was never narrowed during prosecution\\nand was not at issue in the office action rejecting the\\nclaims,” and “none of the cited references concerned the\\ntype of nucleic acid that could perform the blocking, or\\nmentioned the accused equivalent.” Id.\\n Our decision in Intervet Inc. v. Merial Ltd. is to similar\\neffect. In that case, the patent claimed DNA constructs en-\\ncoding a type of porcine circovirus. 617 F.3d 1282, 1284\\n(Fed. Cir. 2010). The claim at issue originally recited DNA\\nsequences from a group of thirteen open reading frames,\\nwhich are portions of a gene that encode a protein. Id. at\\n1285–86, 1291. The examiner rejected the claim over open\\nreading frames from another organism, noting that the\\nclaim as written could cover open reading frames from any\\norganism. Id. at 1291. The applicants then amended the\\nclaim to require that the open reading frames be “of porcine\\ncircovirus type II.” Id. The alleged equivalent was a nu-\\ncleotide sequence that was over 99% homologous to one of\\nthe claimed sequences. Id. at 1286. The “tangential rela-\\ntion” exception applied to that equivalent, we held, because\\n“[t]he rationale for the amendment was to narrow the\\nclaimed universe of [open reading frames] down to those of\\n[porcine circovirus type II], and bore only a tangential re-\\nlation to the question of which DNA sequences are and are\\nnot properly characterized as [porcine circovirus type II].”\\nId. at 1292.\\n This understanding of the “tangential relation” excep-\\ntion also underlies cases in which we have held that the\\npatentee failed to establish that a narrowing amendment\\nwas tangential to the equivalent at issue. For example, in\\n\\nAJINOMOTO CO., INC. v. ITC 21\\n\\n\\n\\nBiagro Western Sales, Inc. v. Grow More, Inc., the claims at\\nissue, which claimed buffered phosphorus fertilizers, were\\nrejected over a prior-art reference disclosing a fertilizer\\nthat was buffered only when diluted. 423 F.3d 1296, 1299,\\n1306 (Fed. Cir. 2005). In response, the applicant amended\\nthe claims by adding the limitation “wherein said phospho-\\nrous-containing acid or salt thereof is present in an amount\\nof about 30 to about 40 weight percent,” explaining that the\\nfertilizer must be concentrated and that the amendment\\nspecified a range for the concentration. Id. at 1305–06.\\nThe alleged equivalent contained phosphorus compounds\\nat a concentration of between 59% and 62%. Id. at 1305.\\nWe concluded that the “tangential relation” exception did\\nnot apply, reasoning that it was “clear from the prosecution\\nhistory that the reason for adding the range limitation was\\nto overcome a prior art fertilizer that was not concen-\\ntrated,” and “both the reason for the amendment and the\\nasserted equivalent relate to the concentration of the ferti-\\nlizer.” Id. at 1306.\\n Here, we conclude, the Commission correctly concluded\\nthat Ajinomoto had rebutted the Festo presumption be-\\ncause the amendment was tangential to the equivalent in\\nquestion. The objectively evident rationale for the amend-\\nment was to limit the set of proteins within the claim’s\\nscope so that it no longer included the prior-art E. coli YfiK\\nprotein and, more generally, no longer allowed as wide a\\nrange of amino acid alterations (hence changes in the pro-\\ntein) as original alternative (B), which had allowed “dele-\\ntion, substitution, insertion or addition of one or several\\namino acids in the amino acid sequence shown in SEQ ID\\nNO: 2.” J.A. 5047. The reason for the amendment had\\nnothing to do with choosing among several DNA sequences\\nin the redundant genetic code that correspond to the same\\nprotein. Indeed, it is undisputed that the non-E. coli YddG\\nprotein produced without codon randomization remains\\nwithin the literal claim scope even after the amendment\\nand that the non-E. coli YddG protein is identical whether\\n\\n22 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nproduced from the codon-randomized or the non-codon-ran-\\ndomized version of the non-E. coli yddG gene.\\n Accordingly, the reason for the narrowing amendment\\n—limiting the amino-acid makeup of the proteins included\\nin one of the alternatives covered by the claim—is unre-\\nlated to differences among the several DNA sequences that\\nencode a given protein. Under Festo’s express provision for\\na “tangential relation” exception to the presumption as to\\nthe scope of surrender by amendment during prosecution,\\nthis conclusion about the reason for the amendment at is-\\nsue does not “ignore[] how the patentee deliberately elected\\nto narrow the claims” (Dissent at 6); rather, it identifies\\nwhat was not within the “scope disclaimed” (id. at 7), so\\nthat it may be proved to infringe by satisfying the other\\nrequirements of the doctrine of equivalents. We therefore\\nreject CJ’s contention that prosecution history estoppel\\nprecludes the Commission’s finding of infringement under\\nthe doctrine of equivalents for the second later strain.\\n 2\\n CJ’s second challenge to the Commission’s finding re-\\ngarding the protein limitation and CJ’s second later strain\\nis that the non-E. coli YddG protein of CJ’s second later\\nstrain could not properly be found to be equivalent to the\\nclaimed E. coli YddG protein. Ajinomoto presented its\\nequivalence case within the function-way-result frame-\\nwork, under which a product or process that does not liter-\\nally satisfy a claim limitation may nevertheless infringe “if\\nit performs substantially the same function in substan-\\ntially the same way to obtain the same result.” Duncan\\nParking Techs., Inc. v. IPS Grp., Inc., 914 F.3d 1347, 1362\\n(Fed. Cir. 2019) (quoting Graver Tank & Mfg. Co. v. Linde\\nAir Prods. Co., 339 U.S. 605, 608 (1950)). We conclude that\\nsubstantial evidence supports the Commission’s finding of\\nequivalence under that test.\\n As to “function”: Ajinomoto’s expert, Dr. Gregory\\nStephanopoulos, testified that both E. coli and non-E. coli\\n\\nAJINOMOTO CO., INC. v. ITC 23\\n\\n\\n\\nYddG proteins function as “export protein[s] that actively\\nexport[] aromatic L-amino acids and aromatic L-amino acid\\nanalogs” out of the bacterial cell. J.A. 545–46. A 2007 ar-\\nticle by Doroshenko et al. similarly explains that both pro-\\nteins are involved in exporting aromatic compounds. See\\nJ.A. 9451. And Dr. So Young Kim, a CJ employee, testified\\nduring a deposition that both proteins would be expected\\nto have similar functions based on similarities in the or-\\nganisms from which they are derived. See J.A. 10641 (“Q.\\nBased on the similarity between E. coli and [the non-E. coli\\norganism], you would suspect that the protein coded by the\\n[non-E. coli] yddG gene would be useful for whatever it\\ndoes in E. coli, right? A. I think that way too.”). Thus, the\\nCommission’s finding that both proteins perform the same\\nfunction is supported by substantial evidence.\\n As to “way”: Dr. Stephanopoulos testified that the two\\nproteins are 85% to 95% identical in structure. J.A. 546.\\nThis range was corroborated by a 2002 article by Santivi-\\nago et al., which indicates an 85% structural identity, see\\nJ.A. 9444, and the Doroshenko article, which notes a 95%\\nidentity in amino-acid sequence, J.A. 9451. On the record\\nhere, substantial evidence supports a finding that the two\\nproteins perform the membrane-transport function in sub-\\nstantially the same way. See also Mylan Institutional LLC\\nv. Aurobindo Pharma Ltd., 857 F.3d 858, 868 (Fed. Cir.\\n2017) (noting that the “function” and “way” inquiries often\\noverlap or are synonymous).\\n As to “result”: Dr. Stephanopoulos testified that, by ex-\\nporting L-tryptophan out of the bacterial cell, both proteins\\nincrease the ability of bacteria to “produce and accumulate\\nL-tryptophan.” See J.A. 547. That statement is supported\\nby CJ’s fermentation data, which showed that strains con-\\ntaining the E. coli yddG gene but with a stronger promoter,\\nand strains containing the non-E. coli yddG gene with a\\nstrong promoter, both showed greater production of L-tryp-\\ntophan than did strains containing the E. coli yddG gene\\nwith the native promoter. See J.A. 7957; J.A. 10053. In\\n\\n24 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nother words, enhancing the expression of either the E. coli\\nor the non-E. coli yddG gene had the effect of increasing\\nproduction of L-tryptophan, which supports an inference\\nthat the proteins encoded by those genes both result in in-\\ncreased L-tryptophan production. The Commission’s find-\\nings regarding result are supported by substantial\\nevidence.\\n CJ argues that the two proteins do not perform the\\nsame function in the same way because the E. coli YddG\\nprotein exports aromatic L-amino acids such as L-trypto-\\nphan, whereas the non-E. coli YddG protein exports a dif-\\nferent compound—namely, paraquat (also known as\\nmethyl viologen). But a 2012 article by Liu et al. explains\\nthat YddG proteins can export both types of compounds.\\nSee J.A. 9751 (“YddG is classified as aromatic amino\\nacid/paraquat exporter . . . .”). And Dr. Stephanopoulos, re-\\nlying on the Santiviago article, testified that the non-E. coli\\nYddG protein must be coupled to the OmpD protein, which\\nis present in the non-E. coli organism but not E. coli, to\\nexport paraquat. J.A. 762 (citing J.A. 9439). The fact that\\nthe non-E. coli YddG protein may be involved in exporting\\ncompounds other than L-tryptophan in the non-E. coli or-\\nganism does not undermine the Commission’s well-sup-\\nported finding that the non-E. coli YddG protein is involved\\nin exporting L-tryptophan in the E. coli bacteria used by\\nCJ.\\n B\\n CJ challenges the Commission’s finding of infringe-\\nment of both later strains on one additional ground. The\\nCommission found that CJ’s later strains met the re-\\nsistance limitation. CJ argues that substantial evidence\\n\\nAJINOMOTO CO., INC. v. ITC 25\\n\\n\\n\\ndoes not exist to support that finding. 9 We reject that ar-\\ngument.\\n Several pieces of evidence indicate that, as a general\\nmatter, enhancing the activity of the YddG protein in-\\ncreases bacteria’s resistance to L-tryptophan. Table 1 of\\nthe ’655 patent shows that E. coli bacteria with multiple\\ncopies of the yddG gene introduced through plasmids\\ndemonstrated better growth on a tryptophan substrate,\\nand thus more resistance, than unmodified E. coli bacteria.\\nSee ’655 patent, col. 9, lines 50–65 (bottom row, compare\\ncolumn “pUC19” (-) with column “pYDDG1” (+)). Similarly,\\nthe Doroshenko article, mentioned above, describes an ex-\\nperiment in which E. coli bacteria with a stronger promoter\\npreceding the yddG gene demonstrated enhanced re-\\nsistance to L-tryptophan. See J.A. 9455 (row DV036, col-\\numn DL-5-f-Trp).\\n CJ’s fermentation data, mentioned above, also provides\\ndirect evidence that CJ’s later strains have increased re-\\nsistance to L-tryptophan. That data shows a greater vol-\\nume of tryptophan with both of CJ’s later strains than with\\nunmodified E. coli bacteria. See J.A. 7957 (first later\\nstrain: middle table, row F4, column “Volume produced”);\\nJ.A. 10053 (second later strain: row “Product (g)” toward\\nmiddle of table). Dr. Stephanopoulos indicated that a\\nstrain’s ability to overproduce L-tryptophan necessarily\\nmeant that the strain had increased resistance to L-trypto-\\nphan. See J.A. 1448 (“[I]f that product feedback inhibits its\\nown synthesis, clearly, this is not going to work.”); see also\\n\\n\\n 9 CJ does not challenge the finding that CJ’s first\\nlater strain meets the protein limitation of the claim 15 al-\\nternative of claim 20. Specifically, CJ’s first later strain\\nuses a non-E. coli yddG gene without codon randomization,\\nwhich hybridizes with the complement of SEQ ID NO:1\\n(i.e., the nucleotide sequence of the E. coli yddG gene), and\\nthus falls within the literal scope of claim 15.\\n\\n26 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nJ.A. 521 (stating that bacteria that “exhibit enhanced re-\\nsistance to an aromatic L-amino acid or an aromatic L-\\namino acid analog” also “overproduce the corresponding ar-\\nomatic L-amino acid analog”).\\n CJ’s objections to the sufficiency or even relevance of\\nthis evidence are unpersuasive. CJ points out that the bac-\\nteria used to generate the data in Table 1 of the ’655 patent\\ncontained plasmids with more than the two copies of the\\nyddG gene in CJ’s later strains. See J.A. 1229. The Doro-\\nshenko article, however, indicates that enhancing the ac-\\ntivity of even a single copy of the yddG gene can increase\\nresistance to L-tryptophan. CJ responds that the strain\\nstudied in Doroshenko used a strong λPL promoter, while\\nCJ’s later strains use relatively weaker non-E. coli native\\nyddG, rmf, and rhtB promoters. See J.A. 9454. But Dr.\\nStephanopoulos testified that at least the rmf promoter in\\nboth of CJ’s later strains also is more potent than the na-\\ntive E. coli yddG promoter. J.A. 554. Thus, even if CJ is\\ncorrect that its later strains do not contain tandem promot-\\ners, the Commission could reasonably infer that the pro-\\nmoters used in CJ’s later strains enhance the activity of the\\nyddG genes relative to unmodified E. coli bacteria and\\nthereby increase those strains’ resistance to L-tryptophan.\\n CJ also cites Ajinomoto’s 2002 Progress Report as evi-\\ndence that enhancing a single copy of the yddG gene is in-\\nsufficient to enhance resistance to L-tryptophan. That\\nreport states that an experiment using “only one copy” of\\nthe yddG gene with a PL promoter “does not correctly\\nmodel[]” an earlier experiment using a “moderate-copy-\\nnumber” plasmid with the yddG gene, which had shown a\\n“positive effect” of the yddG gene on tryptophan produc-\\ntion. J.A. 10268. No more need be inferred from the report\\nthan that enhancing a single copy of the yddG gene in-\\ncreases resistance to L-tryptophan less than using a\\ngreater number of copies. The Commission did not need to\\ninfer that enhancing a single copy as in CJ’s later strains\\ndoes not enhance resistance at all.\\n\\nAJINOMOTO CO., INC. v. ITC 27\\n\\n\\n\\n Further, CJ asserts that the increased production of L-\\ntryptophan, and thus the enhanced resistance to L-trypto-\\nphan, observed in its later strains could be attributable to\\nthe presence of other genetic mutations rather than to in-\\ncreased YddG protein activity alone. See J.A. 442. But the\\nclaims require only that the protein “has the activity to\\nmake the bacterium resistant” to L-tryptophan, not that\\nthe protein be the sole cause of the bacterium’s enhanced\\nresistance to L-tryptophan. See ’655 patent, col. 22, lines\\n57–59. Considering the already-mentioned evidence that\\nthe YddG protein generally has the effect of increasing re-\\nsistance to L-tryptophan, the Commission had substantial\\nevidence from which to find that it was more likely than\\nnot that increased activity of the YddG protein at least\\npartly contributed to the enhanced resistance of CJ’s later\\nstrains.\\n C\\n CJ’s final contention in its cross-appeal seeks reversal\\nof the Commission’s rejection of CJ’s invalidity challenge to\\nclaim 20. CJ argues that substantial evidence does not\\nsupport the Commission’s finding that CJ did not prove\\nlack of an adequate written description for the genus of\\n“more potent promoter[s]” recited in claims 9 and 15 and,\\nby incorporation, in claim 20. We reject CJ’s argument.\\n “[A] sufficient description of a genus . . . requires the\\ndisclosure of either a representative number of species fall-\\ning within the scope of the genus or structural features\\ncommon to the members of the genus so that one of skill in\\nthe art can ‘visualize or recognize’ the members of the ge-\\nnus.” Ariad Pharm., Inc. v. Eli Lilly & Co., 598 F.3d 1336,\\n1350 (Fed. Cir. 2010) (en banc). The Commission found\\nboth that the ’655 patent discloses a representative number\\nof species of more potent promoters and that there are\\nstructural features common to the genus of more potent\\npromoters. Both of those findings are supported by\\n\\n28 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nsubstantial evidence, and they suffice to uphold the Com-\\nmission’s rejection of CJ’s written-description challenge.\\n 1\\n As to a representative number of species, we have rec-\\nognized that the amount of disclosure necessary to satisfy\\nthe written-description requirement “will necessarily vary\\ndepending on the context,” considering such facts as “the\\nexisting knowledge in the particular field, the extent and\\ncontent of the prior art, the maturity of the science or tech-\\nnology,” and “the predictability of the aspect at issue.” Ar-\\niad, 598 F.3d at 1351 (quoting Capon v. Eshhar, 418 F.3d\\n1349, 1359 (Fed. Cir. 2005)). In some circumstances, we\\nhave added, “a patentee may rely on information that is\\n‘well-known in the art’ for purposes of meeting the written\\ndescription requirement,” because “the specification is\\nviewed from the perspective of one of skill” in the relevant\\nart. Bos. Sci. Corp. v. Johnson & Johnson, 647 F.3d 1353,\\n1366 (Fed. Cir. 2011).\\n The ’655 patent discloses four examples of “potent pro-\\nmoters”: “PL promoter of lambda phage,” the “lac pro-\\nmoter,” the “trp promoter,” and the “trc promoter.” ’655\\npatent, col. 6, lines 21–24. The patent also cites the 1986\\narticle by Deuschle et al. as disclosing “examples of potent\\npromoters” and “[m]ethods for [the] evaluation [of] the\\nstrength of promoter[s].” Id., col. 6, lines 16–21. That ar-\\nticle provides data about the relative strength of fourteen\\npromoters and describes a general methodology for deter-\\nmining promoter strength in E. coli bacteria. J.A. 6174–\\n77. This evidence supports the Commission’s finding that\\n“enhancing promoter activity was well-known” and that a\\nskilled artisan “would have been able to identify more po-\\ntent promoters by employing common tools for measuring\\nRNA transcription.” J.A. 46.\\n The ’655 patent also makes clear that its invention was\\n“identifying the yddG gene encoding a membrane protein”\\nand discovering that the gene “conferred on a\\n\\nAJINOMOTO CO., INC. v. ITC 29\\n\\n\\n\\nmicroorganism resistance to phenylalanine and several\\namino acid analogues” when the gene was amplified or its\\nexpression enhanced, see ’655 patent, col. 2, lines 46–57,\\nnot the well-known techniques for performing the amplifi-\\ncation or expression enhancement, see id., col. 5, line 57,\\nthrough col. 6, line 33. We have explained that the repre-\\nsentative-species inquiry is directed to whether the inven-\\ntor “has truly invented the genus” as opposed to “a research\\nplan, leaving it to others to explore the unknown contours\\nof the claimed genus.” AbbVie Deutschland GmbH & Co.,\\nKG v. Janssen Biotech, Inc., 759 F.3d 1285, 1300 (Fed. Cir.\\n2014). Here, the genus of more potent promoters was al-\\nready well explored in the relevant art by the time of the\\n’655 patent’s invention. In these circumstances, the Com-\\nmission permissibly found in the specification, read in light\\nof the background knowledge in the art, a representative\\nnumber of species for the genus of more potent promoters.\\n 2\\n As to a common structural feature, the Commission\\nfound that a skilled artisan could “identify more potent\\nyddG promoters given the well-known link between con-\\nsensus sequence and promoter strength,” i.e., that promot-\\ners having fewer departures from a “consensus sequence”\\nin a promoter are generally stronger than promoters with\\nmore departures from such a sequence. J.A. 46. 10 Substan-\\ntial evidence supports that finding. For instance, a 1983\\narticle by Hawley and McClure describes a study demon-\\nstrating that most “mutations that decrease initiation fre-\\nquency also decrease the homology of the promoter to the\\nconsensus sequence, while up-mutations increase the\\n\\n\\n 10 The consensus sequence is a specific nucleotide se-\\nquence that appears in the promoters associated with\\nmany different genes in the genome of a particular organ-\\nism. In E. coli, the consensus sequence has two parts:\\nTTGACA at the -35 region and TATAAT at the -10 region.\\n\\n30 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\nhomology in” most instances. J.A. 6237. Similarly, a 1986\\narticle by Horwitz and Loeb explains that “mutations that\\nincrease transcription, ‘up mutations,’ usually increase ho-\\nmology with the consensus sequence and spacing,” while\\n“mutations that decrease transcription, ‘down mutations,’\\nusually decrease homology with the consensus sequence\\nand spacing.” J.A. 6251.\\n CJ disputes that similarity to the consensus sequence\\ndefines a common structural feature, citing several articles\\nas indicating that a promoter closer to the consensus se-\\nquence will not always be stronger than one farther from\\nthat sequence. For instance, a 1998 article by Jensen and\\nHammer reports that a pattern observed in another organ-\\nism—that “the relatively strong promoters were the perfect\\nones,” i.e., those closer to the consensus sequence—“did not\\nhold true for E. coli: here the promoters which had either\\nan error in the consensus sequence or a shorter spacer were\\nrelatively strong.” J.A. 9149. Moreover, a 1985 article by\\nAoyama and Takanami states that similarity to the con-\\nsensus sequence “is still not enough to predict the site and\\nstrength of promoter from a given sequence,” J.A. 6215,\\nand a 1999 book edited by Fernandez and Hoeffler notes\\nthat “the strongest promoters in E. coli do not necessarily\\nadhere to the consensus sequence,” J.A. 9113.\\n CJ’s argument both assumes too strict a legal standard\\nand reads too much into its cited references. Adequate\\nwritten description does not require a perfect correspond-\\nence between the members of the genus and the asserted\\ncommon structural feature; for a functionally defined ge-\\nnus like the one at issue here, we have spoken more mod-\\nestly of a “correlation between structure and function.”\\nAriad, 598 F.3d at 1350 (emphasis added). In any event,\\nCJ’s evidence at most establishes that, starting with the\\nconsensus sequence, deviations from that sequence do not\\nalways decrease promoter strength, at least in E. coli. But\\nthe genus at issue here is “more potent promoter[s]” than\\nthe native promoter, not less potent promoters than the\\n\\nAJINOMOTO CO., INC. v. ITC 31\\n\\n\\n\\nconsensus sequence. And the Commission had substantial\\nevidence from which to find that, starting from the native\\nE. coli yddG promoter, deviations toward the consensus se-\\nquence generally increase promoter strength.\\n The cases cited by CJ in which we have held genus\\nclaims to lack an adequate written description are inappo-\\nsite. In Boston Scientific, the specification contained “no\\nexamples of macrocyclic lactone analogs of rapamycin” (the\\nclaimed genus) and essentially “no guidance on how to\\nproperly determine whether a compound is a macrocyclic\\nlactone analog of rapamycin.” 647 F.3d at 1364. In AbbVie,\\nthere was “no evidence to show any described antibody to\\nbe structurally similar to, and thus representative of,” an\\nantibody accused of coming within the claim, nor was there\\n“evidence to show whether one of skill in the art could make\\npredictable changes to the described antibodies to arrive at\\nother types of antibodies such as” the accused antibody.\\n759 F.3d at 1301. And in Regents of the University of Cali-\\nfornia v. Eli Lilly & Co., the specification described “a pro-\\ncess for obtaining human insulin-encoding cDNA” (such\\ncDNA required by the claim at issue) but not any “sequence\\ninformation indicating which nucleotides constitute hu-\\nman cDNA” or “further information in the patent pertain-\\ning to that cDNA’s relevant structural or physical\\ncharacteristics.” 119 F.3d 1559, 1567 (Fed. Cir. 1997).\\nHere, by contrast, the ’655 patent expressly provides four\\nexamples of “more potent promoters,” and the Commission\\nsupportably found that a skilled artisan could make rela-\\ntively predictable changes to the native promoter to arrive\\nat a more potent promoter.\\n IV\\n For the foregoing reasons, we affirm the Commission’s\\ndecision.\\n No costs.\\n AFFIRMED\\n\\n United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\\n INC.,\\n Appellants\\n\\n v.\\n\\n INTERNATIONAL TRADE COMMISSION,\\n Appellee\\n\\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\\n CHEIJEDANG INDONESIA,\\n Intervenors\\n\\n--------------------------------------------\\n\\nCJ CHEILJEDANG CORP., CJ AMERICA, INC., PT\\n CHEIJEDANG INDONESIA,\\n Appellants\\n\\n v.\\n\\n INTERNATIONAL TRADE COMMISSION,\\n Appellee\\n\\nAJINOMOTO CO., INC., AJINOMOTO HEARTLAND\\n INC.,\\n Intervenors\\n ______________________\\n\\n 2018-1590, 2018-1629\\n ______________________\\n\\n2 AJINOMOTO CO., INC. v. ITC\\n\\n\\n\\n\\n Appeals from the United States International Trade\\nCommission in Investigation No. 337-TA-1005.\\n ______________________\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4196,"cells":{"id":{"kind":"number","value":4646661,"string":"4,646,661"},"name":{"kind":"string","value":"AMERICAN BANKERS ASSOCIATION, Washington Federal, N.A., Individually and on Behalf of All Others Similarly Situated, Plaintiffs-Appellants v. UNITED STATES, Defendant-Appellee"},"name_abbreviation":{"kind":"string","value":"American Bankers Association v. United States"},"decision_date":{"kind":"string","value":"2019-08-08"},"docket_number":{"kind":"string","value":"2018-1341"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Federal Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Wallach, Chen, Hughes\", \"parties\": \"\", \"opinions\": [{\"author\": \"HUGHES, Circuit Judge.\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n AMERICAN BANKERS ASSOCIATION,\\n WASHINGTON FEDERAL, N.A., INDIVIDUALLY\\n AND ON BEHALF OF ALL OTHERS SIMILARLY\\n SITUATED,\\n Plaintiffs-Appellants\\n\\n v.\\n\\n UNITED STATES,\\n Defendant-Appellee\\n ______________________\\n\\n 2018-1341\\n ______________________\\n\\n Appeal from the United States Court of Federal Claims\\nin No. 1:17-cv-0194-SGB, Senior Judge Susan G. Braden.\\n ______________________\\n\\n Decided: August 8, 2019\\n ______________________\\n\\n STEPHEN JOSEPH OBERMEIER, Wiley Rein LLP, Wash-\\nington, DC, argued for plaintiffs-appellants. Also repre-\\nsented by CLAIRE J. EVANS, MICHAEL E. TONER.\\n\\n ERIC PETER BRUSKIN, Commercial Litigation Branch,\\nCivil Division, United States Department of Justice, Wash-\\nington, DC, argued for defendant-appellee. Also repre-\\nsented by JOSEPH H. HUNT, ROBERT E. KIRSCHMAN, JR.,\\nKENNETH M. DINTZER, CLAUDIA BURKE; KATHERINE H.\\n\\f2 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\nWHEATLEY, Board of Governors of the Federal Reserve Sys-\\ntem, Washington, DC.\\n ______________________\\n\\n Before WALLACH, CHEN, and HUGHES, Circuit Judges.\\nHUGHES, Circuit Judge.\\n This case arises out of legislation amending the statu-\\ntory rate for dividend payments on Federal Reserve Bank\\nstock. The Federal Reserve Act of 1913 set the dividend\\nrate at six percent per year, which remained in effect until\\nCongress amended the dividend provision in 2016. The\\namendment effectively reduced the dividend rate for cer-\\ntain stockholder banks from the fixed six percent rate to a\\nlower variable rate. American Bankers Association and\\nWashington Federal, N.A. sued the United States in the\\nCourt of Federal Claims, arguing that banks who sub-\\nscribed to Reserve Bank stock before the amendment are\\nentitled to dividends at the six percent rate. The complaint\\nalleged that, by paying dividends at the amended statutory\\nrate, the United States breached a contractual duty or, in\\nthe alternative, effected a Fifth Amendment taking. The\\ntrial court dismissed the complaint under Rules of the U.S.\\nCourt of Federal Claims 12(b)(6) for failure to state a claim.\\nAmerican Bankers and Washington Federal now appeal.\\nBecause the complaint does not allege facts establishing\\nthe existence of a contract or an unconstitutional taking,\\nwe affirm.\\n I\\n A.\\n We begin with a brief overview of the Federal Reserve\\nSystem and its statutory origins. The Federal Reserve Act\\n\\fAMERICAN BANKERS v. UNITED STATES 3\\n\\n\\n\\nof 1913, Pub. L. No. 63−43, ch. 6, 38 Stat. 251 (1913), 1 es-\\ntablished a system to oversee banking operations and pro-\\nmote greater economic stability. The Federal Reserve\\nSystem includes the Federal Reserve Board of Governors,\\nsee id., §§ 10−11, 38 Stat. 260–63, and twelve regional Re-\\nserve Banks, see id. § 2, 38 Stat. 251–52. The Board exer-\\ncises broad regulatory supervision over the Reserve Banks,\\nwhich serve as banks to the U.S. government and to com-\\nmercial banks who are members of the Federal Reserve\\nSystem.\\n The Act sets forth the conditions under which commer-\\ncial banks may join the Federal Reserve System. One of\\nthe conditions of membership is that member banks must\\n“subscribe” to the stock of their regional Reserve Bank in\\nan amount “equal to six per centum of the paid-up capital\\nstock and surplus of [the] applicant bank . . . .” § 5, 38 Stat.\\n257. Every national bank 2 is required to join the system\\nand subscribe to Reserve Bank stock. § 2, 38 Stat 252.\\nOther financial institutions, such as state banks, are per-\\nmitted but not required to apply for membership and sub-\\nscribe to stock. § 9, 38 Stat. 259.\\n Reserve Bank stock is “divided into shares of $100,”\\nwhich “shall not be transferred or hypothecated.” § 5, 38\\nStat. 257. From 1913 to 2015, the Act provided that “the\\nstockholders of the [Reserve] bank shall be entitled to\\n\\n\\n\\n\\n 1 The Federal Reserve Act is codified as amended in\\nscattered sections of Chapter 3 of Title 12 of the United\\nStates Code. See 12 U.S.C. §§ 221−522. This opinion cites\\nto the original 1913 Act, which is the same as the current\\nversion except where otherwise noted.\\n 2 A national bank refers to a commercial bank char-\\ntered by the federal government under the National Bank\\nAct. See 12 U.S.C. § 21 et seq.\\n\\f4 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\nreceive an annual dividend of six per centum on the paid-\\nin capital stock . . . .” § 7, 38 Stat 258.\\n On December 4, 2015, Congress passed the Fixing\\nAmerica’s Surface Transportation Act (FAST Act), which\\nauthorized substantial appropriations for surface trans-\\nportation infrastructure. See Pub. L. No. 114–94, 129 Stat.\\n1312. The FAST Act included an amendment to the statu-\\ntory dividend rate for Reserve Bank stock owned by mem-\\nber banks with consolidated assets of more than $10\\nbillion. Under the amended dividend provision, these\\nbanks would receive a variable dividend rate equal to the\\nlesser of: (1) the rate of the 10-year Treasury note or (2) six\\npercent. See § 32203, 129 Stat. 1739 (codified as amended\\nat 12 U.S.C. § 289(a)(1)).\\n B.\\n Prior to 2013, Washington Federal operated as a feder-\\nally chartered savings and loan association. On May 29,\\n2013, Washington Federal received approval from the Of-\\nfice of the Comptroller of the Currency to convert to a na-\\ntional bank, contingent on, inter alia, Washington Federal\\napplying for membership in the Federal Reserve System.\\n On July 8, 2013, Washington Federal submitted an ap-\\nplication for Reserve Bank stock to the Reserve Bank of\\nSan Francisco (BSF). A letter from BSF, dated July 17,\\n2013, informed Washington Federal that its application\\nand payment for stock had been processed and enclosed an\\nAdvice of Holdings for 479,610 shares of BSF stock. The\\nletter further noted that “[d]ividends are paid at the statu-\\ntory rate of 6 percent per annum, or $1.50 per share semi-\\nannually.” J.A. 65.\\n From 2013 to 2015, Washington Federal received divi-\\ndend payments on its stock at a rate of six percent per year.\\nAfter the FAST Act took effect on January 1, 2016, Wash-\\nington Federal received dividends at the rate of the 10-year\\nTreasury note. In 2016, Washington Federal received\\n\\fAMERICAN BANKERS v. UNITED STATES 5\\n\\n\\n\\ndividends totaling $502,471.53, reflecting an annual rate of\\napproximately two percent.\\n C.\\n Washington Federal and American Bankers Associa-\\ntion 3 filed a complaint against the United States in the\\nCourt of Federal Claims on February 9, 2017. 4 The com-\\nplaint alleged that, by paying dividends at a rate lower\\nthan six percent in 2016, the government breached a con-\\ntractual duty to member banks that subscribed to Reserve\\nBank stock before December 4, 2015. The complaint also\\nasserted, in the alternative, that the government’s conduct\\neffected a Fifth Amendment taking.\\n The government filed a motion to dismiss for lack of\\nstanding under RCFC 12(b)(1) and failure to state a claim\\nunder RCFC 12(b)(6). The Court of Federal Claims deter-\\nmined that American Bankers failed to meet the require-\\nments for associational standing because the damages\\nrequested would require individualized proof for each asso-\\nciation member. The court found that Washington Federal\\nhad standing but dismissed all counts of the complaint un-\\nder RCFC 12(b)(6) for failure to state a claim. Washington\\nFederal and American Bankers now appeal the court’s dis-\\nmissal of the claims and its standing determination. We\\nhave jurisdiction under 28 U.S.C. § 1295(a)(3).\\n\\n\\n\\n\\n 3 American Bankers Association is a national trade\\nassociation for the banking industry. Its members include\\nWashington Federal, as well as other banks affected by the\\namendment to the dividend rate, i.e., member banks with\\nmore than $10 billion in consolidated assets.\\n 4 The complaint was subsequently amended on April\\n14, 2017. This opinion refers to the amended complaint\\nunless otherwise stated.\\n\\f6 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\n II\\n We review de novo whether the Court of Federal\\nClaims properly dismissed a complaint for failure to state\\na claim upon which relief may be granted. Frankel v.\\nUnited States, 842 F.3d 1246, 1249 (Fed. Cir. 2016). To\\navoid dismissal under RCFC 12(b)(6), a plaintiff “must al-\\nlege facts ‘plausibly suggesting (not merely consistent\\nwith)’ a showing of entitlement to relief.” Acceptance Ins.\\nCos., Inc. v. United States, 583 F.3d 849, 853 (Fed. Cir.\\n2009) (quoting Bell Atl. Corp. v. Twombly, 550 U.S. 544,\\n557 (2007)). “A claim has facial plausibility when the plain-\\ntiff pleads factual content that allows the court to draw the\\nreasonable inference that the defendant is liable for the\\nmisconduct alleged.” Ashcroft v. Iqbal, 556 U.S. 662, 678\\n(2009). In reviewing a motion to dismiss, we accept as true\\nthe complaint’s well-pled factual allegations; however, we\\nare not required to accept the asserted legal conclusions.\\nId.\\n For the reasons set forth below, we conclude that the\\ntrial court did not err in dismissing Washington Federal’s\\nbreach of contract and takings claims under RCFC\\n12(b)(6). 5\\n A.\\n First, we address Washington Federal’s breach of con-\\ntract claim. The complaint asserts that the government\\nbreached an implied-in-fact or express contract with Wash-\\nington Federal by paying dividends at a rate lower than six\\npercent in 2016. Washington Federal alleges that an im-\\nplied-in-fact contract exists because the Federal Reserve\\nAct constitutes an offer by the government, which\\n\\n\\n 5 We need not reach American Bankers’ standing ar-\\ngument because American Bankers’ allegations are the\\nsame as Washington Federal’s and would thus share the\\nsame flaws regardless of our outcome.\\n\\fAMERICAN BANKERS v. UNITED STATES 7\\n\\n\\n\\nWashington Federal accepted by submitting its application\\nand payment for Reserve Bank stock. Alternatively, Wash-\\nington Federal contends that an express contract was\\nformed based on its application for stock, which was a con-\\ntractual offer that the government accepted by approving\\nthe application and issuing stock.\\n There are four requirements to form a contract binding\\nupon the government: “(1) mutuality of intent to contract;\\n(2) lack of ambiguity in offer and acceptance; (3) consider-\\nation; and (4) a government representative having actual\\nauthority to bind the United States in contract.” Anderson\\nv. United States, 344 F.3d 1343, 1353 (Fed. Cir. 2003).\\nThese requirements apply to both express and implied-in-\\nfact contracts. Id. at 1353 n.3. “To satisfy its burden to\\nprove such a mutuality of intent, a plaintiff must show, by\\nobjective evidence, the existence of an offer and a reciprocal\\nacceptance.” Id.\\n For both its implied-in-fact and express contract theo-\\nries, Washington Federal relies largely on the Federal Re-\\nserve Act as evidence of the government’s intent to\\ncontract. Under its implied-in-fact contract theory, the Act\\nwas an offer to contract; under its express contract theory,\\nthe Act was an invitation to receive offers to contract. Un-\\nder either theory of contract formation, Washington Fed-\\neral argues that the Act contemplates a contractual\\nagreement between member banks and the government.\\nBecause Washington Federal failed to allege facts estab-\\nlishing the existence of a contract with the government, we\\ndetermine that the trial court did not err in dismissing this\\nclaim.\\n 1.\\n “[A]bsent some clear indication that the legislature in-\\ntends to bind itself contractually, the presumption is that\\n‘a law is not intended to create private contractual or\\nvested rights but merely declares a policy to be pursued\\nuntil the legislature shall ordain otherwise.’” Nat’l R.R.\\n\\f8 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\nPassenger Corp. v. Atchison Topeka & Santa Fe Ry. Co.,\\n470 U.S. 451, 465–66 (1985) (quoting with alterations\\nDodge v. Bd. of Educ., 302 U.S. 74, 79 (1937)). This “well-\\nestablished presumption” reflects a recognition that “the\\nprincipal function of a legislature is not to make contracts,\\nbut to make laws that establish the policy of the state.” Id.\\nAnd “[p]olicies, unlike contracts, are inherently subject to\\nrevision and repeal . . . .” Id.\\n To overcome the presumption, there must be a “clear\\nindication” that the legislature intended to create contrac-\\ntual rights enforceable against the government. Id. at\\n465−66. The Supreme Court has recognized evidence of an\\nintent to contract where a statute “provide[s] for the execu-\\ntion of a written contract on behalf of the United States” or\\n“speak[s] of a contract” with the United States. Id. at 467\\n(emphasis in original); see also Dodge, 302 U.S. at 78; Indi-\\nana ex rel. Anderson v. Brand, 303 U.S. 95, 105 (1938). For\\nexample, in Hall v. Wisconsin, 103 U.S. 5 (1880), the stat-\\nute provided for a geological, mineralogical, and agricul-\\ntural survey to be carried out by commissioners appointed\\nby the governor. Id. at 5–6. The statutory text directed the\\ngovernor to “make a written contract with each of the com-\\nmissioners . . . expressly stipulating and setting forth the\\nnature and extent of the services to be rendered by each,\\nand the compensation therefor . . . .” Id. at 8–9. Likewise,\\nin Indiana ex rel. Anderson, the Court found that Indiana’s\\nTeachers’ Tenure Law contemplated contracts binding on\\nthe state, noting that “[t]he title of the act is couched in\\nterms of contract” and the text “speaks of the making and\\ncanceling of indefinite contracts” between teachers and\\nschool districts. 303 U.S. at 105.\\n In finding that a statute or regulation constitutes an\\noffer to enter into a unilateral contract, courts have also\\nrelied on explicit references to contractual undertakings.\\nFor example, in Radium Mines, Inc. v. United States, 153\\nF. Supp. 403 (Ct. Cl. 1957), the regulation at issue included\\na section entitled “Purchase Contract,” which stated that,\\n\\fAMERICAN BANKERS v. UNITED STATES 9\\n\\n\\n\\nif a sample of uranium delivered to the Commission\\n“meet[s] the conditions of this section, the Commission will\\nforward to the person making the offer a form of contract\\ncontaining applicable terms and conditions ready for his\\nacceptance.” Id. at 405. Similarly, the statutory provision\\nin Grav v. United States, 14 Cl. Ct. 390 (1988), aff’d, 886\\nF.2d 1305 (Fed. Cir. 1989), provided that “[t]he Secretary\\nshall offer to enter into a contract” with milk producers. Id.\\nat 392.\\n In contrast, the Supreme Court determined in Dodge\\nthat the Miller Law did not clearly express the govern-\\nment’s intent to contract. 302 U.S. at 80. As originally en-\\nacted, the Miller Law established a compulsory retirement\\nage for public school teachers and provided for the payment\\nof annuities to retired teachers. Id. at 76. The law stated\\nthat teachers “who served in the public schools of such city\\nfor twenty or more years prior to such retirement, shall be\\npaid the sum of fifteen hundred dollars ($1,500.00) annu-\\nally and for life from the date of such retirement . . . .” Id.\\nNearly ten years after it was passed, the Miller Law was\\namended to reduce annuity payments to $500 for all retired\\nteachers, including those who had retired prior to the\\namendment. Id. at 77. The teachers who filed suit against\\nthe Board of Education argued that they were contractu-\\nally entitled to annuity payments at $1,500 because the\\nMiller Law constituted an offer to contract, which they had\\naccepted by remaining in service for at least twenty years.\\nId. at 77. The Supreme Court rejected this argument, con-\\ncluding that neither the statutory language nor the circum-\\nstances of enactment indicated a legislative intent to create\\nbinding contractual obligations. Id. at 79−81.\\n 2.\\n To determine whether a statute gives rise to a contrac-\\ntual obligation, we first look to the language of the statute.\\nSee Dodge, 302 U.S. at 78; Nat’l R.R., 470 U.S. at 466. The\\nlanguage of the Federal Reserve Act is devoid of the\\n\\f10 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\ntraditional indicia of a contractual undertaking. The Act\\ndoes not “speak of a contract” between Reserve Banks and\\nmember banks; nor does it “provide for the execution of a\\nwritten contract on behalf of the United States.” See Nat’l\\nR.R., 470 U.S. at 467. Rather, the Act sets forth a regula-\\ntory system, in which member banks are granted certain\\n“powers and privileges” and are subject to specified “duties,\\nliabilities, and regulations.” § 8, 38 Stat. 259. Among the\\nduties, member banks are “required . . . to subscribe to the\\ncapital stock” of their regional Reserve Bank. § 2, 38 Stat.\\n252. Among the privileges, banks “shall be entitled to re-\\nceive an annual dividend of six per centum on the paid-in\\ncapital stock. . . .” § 7, 38 Stat. 258.\\n Washington Federal urges us to discern contractual in-\\ntent from the terms “subscribe” and “subscription,” which\\nit contends are “contractual terms of art in the context of\\nstock offerings. . . .” Appellant’s Op. Br. 31; see also id. at\\n32−33. But we must interpret the language in the context\\nin which it is written. In the context of a regulatory stat-\\nute, we will not infer a contractual undertaking “absent ‘an\\nadequate expression of an actual intent’ of the State to bind\\nitself. . . .” Nat’l R.R., 470 U.S. at 466–67 (quoting Wis. &\\nMich. Ry. Co. v. Powers, 191 U.S. 379, 386–87 (1903)). And\\nthe use of terminology that carries contractual connota-\\ntions when used in the private sector does not, on its own,\\nestablish such intent. For example, in Dodge, the Supreme\\nCourt rejected appellants’ argument that “annuity” is “ter-\\nminology based on contract” that reflected the legislature’s\\nintent to establish contractual rights. 302 U.S. at 81. Like-\\nwise, we find that the subscription language in the Federal\\nReserve Act does not unequivocally express the govern-\\nment’s intent to bind itself in contract.\\n Washington Federal further argues that the govern-\\nment’s intent to contract is evident from the exchange of\\nobligations between member banks and Reserve Banks.\\nAccording to Washington Federal, the Act contemplates an\\nagreement that Reserve Banks will pay member banks an\\n\\fAMERICAN BANKERS v. UNITED STATES 11\\n\\n\\n\\nannual dividend of six percent in exchange for member\\nbanks’ subscription to Reserve Bank stock. The language\\nand structure of the Act, however, do not reflect a bar-\\ngained-for quid pro quo between two parties. The dividend\\nrate is set forth in an entirely different section than the\\nprovisions governing stock subscription. Compare § 7, 38\\nStat. 258, with § 2, 38 Stat. 251−52, and § 5, 38 Stat. 257–\\n58. Moreover, the Act does not frame dividend payments\\nas a contractual obligation of the Reserve Banks. Rather,\\nthe six percent dividend is described as a benefit that mem-\\nber banks “shall be entitled to receive.” § 7, 38 Stat. 258.\\nAnd a statute does not create contractual obligations\\nmerely by setting forth “benefits to those who comply with\\nits conditions.” Wis. & Mich, Ry. Co., 191 U.S. at 387. In-\\ndeed, in Dodge, the statutory language provided that teach-\\ners who retired after serving twenty or more years in the\\npublic schools “shall be paid the sum of fifteen hundred dol-\\nlars ($1,500.00) annually and for life from the date of such\\nretirement . . . .” 302 U.S. at 76 (emphasis added). Yet the\\nSupreme Court declined to find that the teachers’ rights to\\n$1,500 annuity payments vested upon meeting the statu-\\ntory conditions. Id. at 77, 79−81. Similarly, we conclude\\nthat the Federal Reserve Act does not demonstrate the gov-\\nernment’s clear intent to confer vested contractual rights\\non each member bank who complies with the stock sub-\\nscription requirement.\\n The circumstances surrounding the passage of the Fed-\\neral Reserve Act and its legislative history offer further\\nsupport. See Nat’l R.R., 470 U.S. at 468. The Federal Re-\\nserve Act was passed in 1913 in response to ongoing insta-\\nbility within the financial sector, which had given rise to a\\nseries of banking crises in the preceding decades. See H.R.\\nRep. No. 63-69, at 3−5 (1913). One of the deficiencies of the\\nbanking system at the time was that it “fail[ed] to afford\\nany safeguard against panics and commercial stringencies\\nor any means of alleviating them.” Id. at 6.\\n\\f12 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\n Congress created the Federal Reserve System to pre-\\nvent and contain the financial disruption caused by bank\\nfailures. The Reserve Banks were intended to serve as\\nlenders of last resort by maintaining a reserve of liquid cap-\\nital “ready for use in protecting the banks of any section of\\nthe country and for enabling them to go on meeting their\\nobligations instead of suspending payments, as so often in\\nthe past.” Id. at 11; see also id. at 19−22. To provide the\\nfunds for this reserve, Congress established the require-\\nment that member banks subscribe to Reserve Bank stock.\\nId. at 16−17, 20−21. Thus, the origins of the subscription\\nrequirement reflect a regulatory effort to promote stability\\nin the banking system through collaboration, rather than\\na collection of private contractual undertakings.\\n Statements in the legislative history bear this out. For\\nexample, the House Report accompanying the bill later\\npassed as the Federal Reserve Act, expressed the view that\\n“banking institutions which desire to be known by the\\nname ‘national’ should be required, and can well afford, to\\ntake upon themselves the responsibilities involved in joint\\nor federated organization.” Id. at 16. Likewise, the Senate\\nReport stated that the Reserve Banks were “not intended\\nto be merely money-making banks,” but “guardians of the\\npublic welfare, primarily safeguarding the member banks,\\nprotecting their reserves, safeguarding their credit, [and]\\nprotecting them from panic or financial stringency. . . .” S.\\nRep. No. 63-133, at 10 (1913). Although the proposed leg-\\nislation was to provide member banks with a return on the\\nuse of their funds, the Senate Report noted that “the sta-\\nbility of the business of the bank, and the peace of mind it\\nwill give to the bankers in having freedom from constant\\nanxiety, would more than compensate them, even if the fi-\\nnancial advantages did not do so.” Id. at 12.\\n Accordingly, we discern no “clear indication” of the gov-\\nernment’s intent to contract in either the language of the\\nFederal Reserve Act or the circumstances under which it\\nwas passed.\\n\\fAMERICAN BANKERS v. UNITED STATES 13\\n\\n\\n\\n The additional evidence on which Washington Federal\\nrelies, primarily for its express contract theory, does noth-\\ning to remedy this deficiency. For example, the July 17,\\n2013 letter from the BSF merely states that Washington\\nFederal’s application and payment have been processed\\nand informs Washington Federal of some of the obligations\\nand benefits associated with membership in the Federal\\nReserve System. Washington Federal points to the letter’s\\nstatement that “[d]ividends are paid at the statutory rate\\nof 6 percent per annum, or $1.50 per share semi-annually.”\\nJ.A. 65. But this is simply a statement of policy based on\\nthe statutory dividend rate in effect at the time, not the\\nlanguage of a promise or contractual undertaking. See\\nChattler v. United States, 632 F.3d 1324, 1330 (Fed. Cir.\\n2011) (“[T]he obligation of the government, if it is to be held\\nliable, must be stated in the form of an undertaking, not as\\na mere prediction or statement of opinion or intention.\\nLikewise statements of information or definition are not\\nstatements of obligation.” (internal citations and quotation\\nmarks omitted)).\\n Because Washington Federal did not plead facts suffi-\\ncient to establish the government’s intent to contract, the\\ncomplaint fails to state a plausible claim for breach of con-\\ntract. Accordingly, this claim was properly dismissed. 6\\n\\n\\n\\n\\n 6 Washington Federal’s claim for breach of implied\\nduty of good faith and fair dealing likewise depends on the\\nexistence of a valid contract. See Centex Corp. v. United\\nStates, 395 F.3d 1283, 1304 (Fed. Cir. 2005) (“The covenant\\nof good faith and fair dealing is an implied duty that each\\nparty to a contract owes to its contracting partner.”).\\nTherefore, this claim was also properly dismissed.\\n\\f14 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\n B.\\n We now turn to Washington Federal’s Fifth Amend-\\nment takings claim. The complaint sets forth two takings\\ntheories: (1) by enacting the FAST Act, the government de-\\nprived Washington Federal of its “property interest[] in the\\npromised six percent dividend,” J.A. 57 ¶ 92; and (2) by\\npaying dividends at a rate lower than six percent, the gov-\\nernment effected “a taking of [Washington Federal’s] capi-\\ntal investment[] in Federal Reserve Bank stock without\\njust compensation in the form of a market return on the\\ninvested capital,” J.A. 57 ¶ 93. Under either theory, Wash-\\nington Federal failed to state a plausible takings claim.\\n To state a claim for a taking under the Fifth Amend-\\nment, a plaintiff must identify a legally cognizable property\\ninterest. Tex. State Bank v. United States, 423 F.3d 1370,\\n1378 (Fed. Cir. 2005). “The Constitution neither creates\\nnor defines the scope of property interests compensable un-\\nder the Fifth Amendment.” Conti v. United States, 291\\nF.3d 1334, 1340 (Fed. Cir. 2002) (citing Bd. of Regents of\\nState Colls. v. Roth, 408 U.S. 564, 577 (1972)). Property\\ninterests arise from “existing rules and understandings\\nand background principles derived from an independent\\nsource, such as state, federal, or common law. . . .” Air Peg-\\nasus of D.C., Inc. v. United States, 424 F.3d 1206, 1213\\n(Fed. Cir. 2005) (internal quotation marks omitted). To\\nsupport a takings claim, a property interest must be more\\nthan a “mere unilateral expectation or an abstract need.”\\nWebb’s Fabulous Pharmacies, Inc. v. Beckwith, 449 U.S.\\n155, 161 (1980).\\n Under its first takings theory, Washington Federal as-\\nserts a property interest in its “contractual and statutory\\nrights to receive a six percent dividend on Federal Reserve\\nBank stock . . . .” J.A. 57 ¶ 90. While contract rights are a\\nform of property that may be compensable under the Fifth\\nAmendment, see Cienega Gardens v. United States, 331\\nF.3d 1319, 1329–30 (Fed. Cir. 2003), the complaint does not\\n\\fAMERICAN BANKERS v. UNITED STATES 15\\n\\n\\n\\nestablish that Washington Federal had a contractual right\\nto a six percent dividend, see supra Section II.A.2. Thus,\\nthe trial court properly dismissed the contract-based tak-\\nings claim.\\n Likewise, Washington Federal has not alleged a legally\\ncognizable property interest arising from its “statutory\\nrights” under the Federal Reserve Act. Absent independ-\\nent evidence of a contractual undertaking, a statutory en-\\ntitlement “creates no vested right.” Dodge, 302 U.S. at 79.\\nBecause “Congress at all times retains the ability to amend\\nstatutes, a power which inheres in its authority to legislate,\\nCongress at all times retains the right to revoke legisla-\\ntively created entitlements.” Members of Peanut Quota\\nHolders Ass’n, Inc. v. United States, 421 F.3d 1323, 1335\\n(Fed. Cir. 2005). Indeed, in this case, Congress “expressly\\nreserved” its “right to amend, alter, or repeal” any provi-\\nsion of the Federal Reserve Act. See § 30, 38 Stat. 275\\n(1913), renumbered § 31, Pub. L. No. 95−630, title I, § 101,\\n92 Stat. 3641 (1978). Washington Federal emphasizes\\nthat, prior to the FAST Act, the six percent dividend rate\\nhad remained unchanged for over 100 years. But Wash-\\nington Federal’s “unilateral expectation” that Congress\\nwould not exercise its right to amend the dividend provi-\\nsion going forward does not give rise to a compensable\\nproperty interest under the Fifth Amendment. See Webb’s,\\n449 U.S. 161; Peanut Quota Holders, 421 F.3d at 1334\\n(“[Appellants] have no legally protected right against the\\ngovernment’s making changes in the underlying [regula-\\ntory] program and no right to compensation for the loss in\\nvalue resulting from the changes.”).\\n Washington Federal’s alternative takings theory con-\\ntemplates a taking of its underlying capital investment in\\nReserve Bank stock without just compensation. 7 This\\n\\n\\n 7 The trial court did not directly address this alter-\\nnative takings theory. Washington Federal contends that\\n\\f16 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\nclaim also fails. Washington Federal’s initial subscription\\nof stock was part of its voluntary participation in a regula-\\ntory scheme, and we have held that “enforceable rights suf-\\nficient to support a taking claim against the United States\\ncannot arise in an area voluntarily entered into and one\\nwhich, from the start, is subject to pervasive Government\\ncontrol.” Mitchell Arms, Inc. v. United States, 7 F.3d 212,\\n216 (Fed. Cir. 1993) (internal quotation marks omitted); see\\nalso Commonwealth Edison Co. v. United States, 271 F.3d\\n1327, 1339 (Fed. Cir. 2001) (“[R]egulatory actions requiring\\nthe payment of money are not takings.”). Furthermore, un-\\nder the Federal Reserve Act, Washington Federal can sur-\\nrender its stock and obtain a refund of its paid-in capital.\\nSee 12 U.S.C. §§ 287, 321. Thus, the requirement that\\nmember banks subscribe to reserve bank stock under the\\nFederal Reserve Act does not constitute a regulatory tak-\\ning.\\n\\n\\n\\n\\nremand is necessary for the trial court to rule on this issue\\nin the first instance. We conclude, however, that principles\\nof judicial economy counsel against remand, and whether\\nWashington Federal has adequately stated a claim under\\nits alternative takings theory is an issue amenable to reso-\\nlution for the first time on appeal. See Glaxo Grp. Ltd. v.\\nTorPharm, Inc., 153 F.3d 1366, 1371 (Fed. Cir. 1998) (not-\\ning that “an appellate court may choose to decide [an] issue\\neven if not passed on by the trial court” where the issue “is\\none of law” and “has been fully vetted by the parties on ap-\\npeal”); Singleton v. Wulff, 428 U.S. 106, 121 (1976) (“The\\nmatter of what questions may be taken up and resolved for\\nthe first time on appeal is one left primarily to the discre-\\ntion of the courts of appeals, to be exercised on the facts of\\nindividual cases.”).\\n\\fAMERICAN BANKERS v. UNITED STATES 17\\n\\n\\n\\n Because the complaint fails to allege facts sufficient to\\nsupport a taking under the Fifth Amendment, the trial\\ncourt properly dismissed this claim under RCFC 12(b)(6).\\n III\\n For the foregoing reasons, we conclude that the com-\\nplaint fails to state a claim upon which relief may be\\ngranted. Thus, we affirm the trial court’s grant of the gov-\\nernment’s motion to dismiss Washington Federal’s claims\\nunder RCFC 12(b)(6).\\n AFFIRMED\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4423914/\", \"author_raw\": \"HUGHES, Circuit Judge.\"}]}"},"judge1":{"kind":"string","value":"WALLACH"},"judge2":{"kind":"string","value":"CHEN"},"judge3":{"kind":"string","value":"HUGHES"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4646661/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: AMERICAN BANKERS ASSOCIATION, Washington Federal, N.A., Individually and on Behalf of All Others Similarly Situated, Plaintiffs-Appellants v. UNITED STATES, Defendant-Appellee\nCourt: U.S. Court of Appeals for the Federal Circuit\nDate decided: 2019-08-08\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n for the Federal Circuit\n ______________________\n\n AMERICAN BANKERS ASSOCIATION,\n WASHINGTON FEDERAL, N.A., INDIVIDUALLY\n AND ON BEHALF OF ALL OTHERS SIMILARLY\n SITUATED,\n Plaintiffs-Appellants\n\n v.\n\n UNITED STATES,\n Defendant-Appellee\n ______________________\n\n 2018-1341\n ______________________\n\n Appeal from the United States Court of Federal Claims\nin No. 1:17-cv-0194-SGB, Senior Judge Susan G. Braden.\n ______________________\n\n Decided: August 8, 2019\n ______________________\n\n STEPHEN JOSEPH OBERMEIER, Wiley Rein LLP, Wash-\nington, DC, argued for plaintiffs-appellants. Also repre-\nsented by CLAIRE J. EVANS, MICHAEL E. TONER.\n\n ERIC PETER BRUSKIN, Commercial Litigation Branch,\nCivil Division, United States Department of Justice, Wash-\nington, DC, argued for defendant-appellee. Also repre-\nsented by JOSEPH H. HUNT, ROBERT E. KIRSCHMAN, JR.,\nKENNETH M. DINTZER, CLAUDIA BURKE; KATHERINE H.\n\f2 AMERICAN BANKERS v. UNITED STATES\n\n\n\nWHEATLEY, Board of Governors of the Federal Reserve Sys-\ntem, Washington, DC.\n ______________________\n\n Before WALLACH, CHEN, and HUGHES, Circuit Judges.\nHUGHES, Circuit Judge.\n This case arises out of legislation amending the statu-\ntory rate for dividend payments on Federal Reserve Bank\nstock. The Federal Reserve Act of 1913 set the dividend\nrate at six percent per year, which remained in effect until\nCongress amended the dividend provision in 2016. The\namendment effectively reduced the dividend rate for cer-\ntain stockholder banks from the fixed six percent rate to a\nlower variable rate. American Bankers Association and\nWashington Federal, N.A. sued the United States in the\nCourt of Federal Claims, arguing that banks who sub-\nscribed to Reserve Bank stock before the amendment are\nentitled to dividends at the six percent rate. The complaint\nalleged that, by paying dividends at the amended statutory\nrate, the United States breached a contractual duty or, in\nthe alternative, effected a Fifth Amendment taking. The\ntrial court dismissed the complaint under Rules of the U.S.\nCourt of Federal Claims 12(b)(6) for failure to state a claim.\nAmerican Bankers and Washington Federal now appeal.\nBecause the complaint does not allege facts establishing\nthe existence of a contract or an unconstitutional taking,\nwe affirm.\n I\n A.\n We begin with a brief overview of the Federal Reserve\nSystem and its statutory origins. The Federal Reserve Act\n\fAMERICAN BANKERS v. UNITED STATES 3\n\n\n\nof 1913, Pub. L. No. 63−43, ch. 6, 38 Stat. 251 (1913), 1 es-\ntablished a system to oversee banking operations and pro-\nmote greater economic stability. The Federal Reserve\nSystem includes the Federal Reserve Board of Governors,\nsee id., §§ 10−11, 38 Stat. 260–63, and twelve regional Re-\nserve Banks, see id. § 2, 38 Stat. 251–52. The Board exer-\ncises broad regulatory supervision over the Reserve Banks,\nwhich serve as banks to the U.S. government and to com-\nmercial banks who are members of the Federal Reserve\nSystem.\n The Act sets forth the conditions under which commer-\ncial banks may join the Federal Reserve System. One of\nthe conditions of membership is that member banks must\n“subscribe” to the stock of their regional Reserve Bank in\nan amount “equal to six per centum of the paid-up capital\nstock and surplus of [the] applicant bank . . . .” § 5, 38 Stat.\n257. Every national bank 2 is required to join the system\nand subscribe to Reserve Bank stock. § 2, 38 Stat 252.\nOther financial institutions, such as state banks, are per-\nmitted but not required to apply for membership and sub-\nscribe to stock. § 9, 38 Stat. 259.\n Reserve Bank stock is “divided into shares of $100,”\nwhich “shall not be transferred or hypothecated.” § 5, 38\nStat. 257. From 1913 to 2015, the Act provided that “the\nstockholders of the [Reserve] bank shall be entitled to\n\n\n\n\n 1 The Federal Reserve Act is codified as amended in\nscattered sections of Chapter 3 of Title 12 of the United\nStates Code. See 12 U.S.C. §§ 221−522. This opinion cites\nto the original 1913 Act, which is the same as the current\nversion except where otherwise noted.\n 2 A national bank refers to a commercial bank char-\ntered by the federal government under the National Bank\nAct. See 12 U.S.C. § 21 et seq.\n\f4 AMERICAN BANKERS v. UNITED STATES\n\n\n\nreceive an annual dividend of six per centum on the paid-\nin capital stock . . . .” § 7, 38 Stat 258.\n On December 4, 2015, Congress passed the Fixing\nAmerica’s Surface Transportation Act (FAST Act), which\nauthorized substantial appropriations for surface trans-\nportation infrastructure. See Pub. L. No. 114–94, 129 Stat.\n1312. The FAST Act included an amendment to the statu-\ntory dividend rate for Reserve Bank stock owned by mem-\nber banks with consolidated assets of more than $10\nbillion. Under the amended dividend provision, these\nbanks would receive a variable dividend rate equal to the\nlesser of: (1) the rate of the 10-year Treasury note or (2) six\npercent. See § 32203, 129 Stat. 1739 (codified as amended\nat 12 U.S.C. § 289(a)(1)).\n B.\n Prior to 2013, Washington Federal operated as a feder-\nally chartered savings and loan association. On May 29,\n2013, Washington Federal received approval from the Of-\nfice of the Comptroller of the Currency to convert to a na-\ntional bank, contingent on, inter alia, Washington Federal\napplying for membership in the Federal Reserve System.\n On July 8, 2013, Washington Federal submitted an ap-\nplication for Reserve Bank stock to the Reserve Bank of\nSan Francisco (BSF). A letter from BSF, dated July 17,\n2013, informed Washington Federal that its application\nand payment for stock had been processed and enclosed an\nAdvice of Holdings for 479,610 shares of BSF stock. The\nletter further noted that “[d]ividends are paid at the statu-\ntory rate of 6 percent per annum, or $1.50 per share semi-\nannually.” J.A. 65.\n From 2013 to 2015, Washington Federal received divi-\ndend payments on its stock at a rate of six percent per year.\nAfter the FAST Act took effect on January 1, 2016, Wash-\nington Federal received dividends at the rate of the 10-year\nTreasury note. In 2016, Washington Federal received\n\fAMERICAN BANKERS v. UNITED STATES 5\n\n\n\ndividends totaling $502,471.53, reflecting an annual rate of\napproximately two percent.\n C.\n Washington Federal and American Bankers Associa-\ntion 3 filed a complaint against the United States in the\nCourt of Federal Claims on February 9, 2017. 4 The com-\nplaint alleged that, by paying dividends at a rate lower\nthan six percent in 2016, the government breached a con-\ntractual duty to member banks that subscribed to Reserve\nBank stock before December 4, 2015. The complaint also\nasserted, in the alternative, that the government’s conduct\neffected a Fifth Amendment taking.\n The government filed a motion to dismiss for lack of\nstanding under RCFC 12(b)(1) and failure to state a claim\nunder RCFC 12(b)(6). The Court of Federal Claims deter-\nmined that American Bankers failed to meet the require-\nments for associational standing because the damages\nrequested would require individualized proof for each asso-\nciation member. The court found that Washington Federal\nhad standing but dismissed all counts of the complaint un-\nder RCFC 12(b)(6) for failure to state a claim. Washington\nFederal and American Bankers now appeal the court’s dis-\nmissal of the claims and its standing determination. We\nhave jurisdiction under 28 U.S.C. § 1295(a)(3).\n\n\n\n\n 3 American Bankers Association is a national trade\nassociation for the banking industry. Its members include\nWashington Federal, as well as other banks affected by the\namendment to the dividend rate, i.e., member banks with\nmore than $10 billion in consolidated assets.\n 4 The complaint was subsequently amended on April\n14, 2017. This opinion refers to the amended complaint\nunless otherwise stated.\n\f6 AMERICAN BANKERS v. UNITED STATES\n\n\n\n II\n We review de novo whether the Court of Federal\nClaims properly dismissed a complaint for failure to state\na claim upon which relief may be granted. Frankel v.\nUnited States, 842 F.3d 1246, 1249 (Fed. Cir. 2016). To\navoid dismissal under RCFC 12(b)(6), a plaintiff “must al-\nlege facts ‘plausibly suggesting (not merely consistent\nwith)’ a showing of entitlement to relief.” Acceptance Ins.\nCos., Inc. v. United States, 583 F.3d 849, 853 (Fed. Cir.\n2009) (quoting Bell Atl. Corp. v. Twombly, 550 U.S. 544,\n557 (2007)). “A claim has facial plausibility when the plain-\ntiff pleads factual content that allows the court to draw the\nreasonable inference that the defendant is liable for the\nmisconduct alleged.” Ashcroft v. Iqbal, 556 U.S. 662, 678\n(2009). In reviewing a motion to dismiss, we accept as true\nthe complaint’s well-pled factual allegations; however, we\nare not required to accept the asserted legal conclusions.\nId.\n For the reasons set forth below, we conclude that the\ntrial court did not err in dismissing Washington Federal’s\nbreach of contract and takings claims under RCFC\n12(b)(6). 5\n A.\n First, we address Washington Federal’s breach of con-\ntract claim. The complaint asserts that the government\nbreached an implied-in-fact or express contract with Wash-\nington Federal by paying dividends at a rate lower than six\npercent in 2016. Washington Federal alleges that an im-\nplied-in-fact contract exists because the Federal Reserve\nAct constitutes an offer by the government, which\n\n\n 5 We need not reach American Bankers’ standing ar-\ngument because American Bankers’ allegations are the\nsame as Washington Federal’s and would thus share the\nsame flaws regardless of our outcome.\n\fAMERICAN BANKERS v. UNITED STATES 7\n\n\n\nWashington Federal accepted by submitting its application\nand payment for Reserve Bank stock. Alternatively, Wash-\nington Federal contends that an express contract was\nformed based on its application for stock, which was a con-\ntractual offer that the government accepted by approving\nthe application and issuing stock.\n There are four requirements to form a contract binding\nupon the government: “(1) mutuality of intent to contract;\n(2) lack of ambiguity in offer and acceptance; (3) consider-\nation; and (4) a government representative having actual\nauthority to bind the United States in contract.” Anderson\nv. United States, 344 F.3d 1343, 1353 (Fed. Cir. 2003).\nThese requirements apply to both express and implied-in-\nfact contracts. Id. at 1353 n.3. “To satisfy its burden to\nprove such a mutuality of intent, a plaintiff must show, by\nobjective evidence, the existence of an offer and a reciprocal\nacceptance.” Id.\n For both its implied-in-fact and express contract theo-\nries, Washington Federal relies largely on the Federal Re-\nserve Act as evidence of the government’s intent to\ncontract. Under its implied-in-fact contract theory, the Act\nwas an offer to contract; under its express contract theory,\nthe Act was an invitation to receive offers to contract. Un-\nder either theory of contract formation, Washington Fed-\neral argues that the Act contemplates a contractual\nagreement between member banks and the government.\nBecause Washington Federal failed to allege facts estab-\nlishing the existence of a contract with the government, we\ndetermine that the trial court did not err in dismissing this\nclaim.\n 1.\n “[A]bsent some clear indication that the legislature in-\ntends to bind itself contractually, the presumption is that\n‘a law is not intended to create private contractual or\nvested rights but merely declares a policy to be pursued\nuntil the legislature shall ordain otherwise.’” Nat’l R.R.\n\f8 AMERICAN BANKERS v. UNITED STATES\n\n\n\nPassenger Corp. v. Atchison Topeka & Santa Fe Ry. Co.,\n470 U.S. 451, 465–66 (1985) (quoting with alterations\nDodge v. Bd. of Educ., 302 U.S. 74, 79 (1937)). This “well-\nestablished presumption” reflects a recognition that “the\nprincipal function of a legislature is not to make contracts,\nbut to make laws that establish the policy of the state.” Id.\nAnd “[p]olicies, unlike contracts, are inherently subject to\nrevision and repeal . . . .” Id.\n To overcome the presumption, there must be a “clear\nindication” that the legislature intended to create contrac-\ntual rights enforceable against the government. Id. at\n465−66. The Supreme Court has recognized evidence of an\nintent to contract where a statute “provide[s] for the execu-\ntion of a written contract on behalf of the United States” or\n“speak[s] of a contract” with the United States. Id. at 467\n(emphasis in original); see also Dodge, 302 U.S. at 78; Indi-\nana ex rel. Anderson v. Brand, 303 U.S. 95, 105 (1938). For\nexample, in Hall v. Wisconsin, 103 U.S. 5 (1880), the stat-\nute provided for a geological, mineralogical, and agricul-\ntural survey to be carried out by commissioners appointed\nby the governor. Id. at 5–6. The statutory text directed the\ngovernor to “make a written contract with each of the com-\nmissioners . . . expressly stipulating and setting forth the\nnature and extent of the services to be rendered by each,\nand the compensation therefor . . . .” Id. at 8–9. Likewise,\nin Indiana ex rel. Anderson, the Court found that Indiana’s\nTeachers’ Tenure Law contemplated contracts binding on\nthe state, noting that “[t]he title of the act is couched in\nterms of contract” and the text “speaks of the making and\ncanceling of indefinite contracts” between teachers and\nschool districts. 303 U.S. at 105.\n In finding that a statute or regulation constitutes an\noffer to enter into a unilateral contract, courts have also\nrelied on explicit references to contractual undertakings.\nFor example, in Radium Mines, Inc. v. United States, 153\nF. Supp. 403 (Ct. Cl. 1957), the regulation at issue included\na section entitled “Purchase Contract,” which stated that,\n\fAMERICAN BANKERS v. UNITED STATES 9\n\n\n\nif a sample of uranium delivered to the Commission\n“meet[s] the conditions of this section, the Commission will\nforward to the person making the offer a form of contract\ncontaining applicable terms and conditions ready for his\nacceptance.” Id. at 405. Similarly, the statutory provision\nin Grav v. United States, 14 Cl. Ct. 390 (1988), aff’d, 886\nF.2d 1305 (Fed. Cir. 1989), provided that “[t]he Secretary\nshall offer to enter into a contract” with milk producers. Id.\nat 392.\n In contrast, the Supreme Court determined in Dodge\nthat the Miller Law did not clearly express the govern-\nment’s intent to contract. 302 U.S. at 80. As originally en-\nacted, the Miller Law established a compulsory retirement\nage for public school teachers and provided for the payment\nof annuities to retired teachers. Id. at 76. The law stated\nthat teachers “who served in the public schools of such city\nfor twenty or more years prior to such retirement, shall be\npaid the sum of fifteen hundred dollars ($1,500.00) annu-\nally and for life from the date of such retirement . . . .” Id.\nNearly ten years after it was passed, the Miller Law was\namended to reduce annuity payments to $500 for all retired\nteachers, including those who had retired prior to the\namendment. Id. at 77. The teachers who filed suit against\nthe Board of Education argued that they were contractu-\nally entitled to annuity payments at $1,500 because the\nMiller Law constituted an offer to contract, which they had\naccepted by remaining in service for at least twenty years.\nId. at 77. The Supreme Court rejected this argument, con-\ncluding that neither the statutory language nor the circum-\nstances of enactment indicated a legislative intent to create\nbinding contractual obligations. Id. at 79−81.\n 2.\n To determine whether a statute gives rise to a contrac-\ntual obligation, we first look to the language of the statute.\nSee Dodge, 302 U.S. at 78; Nat’l R.R., 470 U.S. at 466. The\nlanguage of the Federal Reserve Act is devoid of the\n\f10 AMERICAN BANKERS v. UNITED STATES\n\n\n\ntraditional indicia of a contractual undertaking. The Act\ndoes not “speak of a contract” between Reserve Banks and\nmember banks; nor does it “provide for the execution of a\nwritten contract on behalf of the United States.” See Nat’l\nR.R., 470 U.S. at 467. Rather, the Act sets forth a regula-\ntory system, in which member banks are granted certain\n“powers and privileges” and are subject to specified “duties,\nliabilities, and regulations.” § 8, 38 Stat. 259. Among the\nduties, member banks are “required . . . to subscribe to the\ncapital stock” of their regional Reserve Bank. § 2, 38 Stat.\n252. Among the privileges, banks “shall be entitled to re-\nceive an annual dividend of six per centum on the paid-in\ncapital stock. . . .” § 7, 38 Stat. 258.\n Washington Federal urges us to discern contractual in-\ntent from the terms “subscribe” and “subscription,” which\nit contends are “contractual terms of art in the context of\nstock offerings. . . .” Appellant’s Op. Br. 31; see also id. at\n32−33. But we must interpret the language in the context\nin which it is written. In the context of a regulatory stat-\nute, we will not infer a contractual undertaking “absent ‘an\nadequate expression of an actual intent’ of the State to bind\nitself. . . .” Nat’l R.R., 470 U.S. at 466–67 (quoting Wis. &\nMich. Ry. Co. v. Powers, 191 U.S. 379, 386–87 (1903)). And\nthe use of terminology that carries contractual connota-\ntions when used in the private sector does not, on its own,\nestablish such intent. For example, in Dodge, the Supreme\nCourt rejected appellants’ argument that “annuity” is “ter-\nminology based on contract” that reflected the legislature’s\nintent to establish contractual rights. 302 U.S. at 81. Like-\nwise, we find that the subscription language in the Federal\nReserve Act does not unequivocally express the govern-\nment’s intent to bind itself in contract.\n Washington Federal further argues that the govern-\nment’s intent to contract is evident from the exchange of\nobligations between member banks and Reserve Banks.\nAccording to Washington Federal, the Act contemplates an\nagreement that Reserve Banks will pay member banks an\n\fAMERICAN BANKERS v. UNITED STATES 11\n\n\n\nannual dividend of six percent in exchange for member\nbanks’ subscription to Reserve Bank stock. The language\nand structure of the Act, however, do not reflect a bar-\ngained-for quid pro quo between two parties. The dividend\nrate is set forth in an entirely different section than the\nprovisions governing stock subscription. Compare § 7, 38\nStat. 258, with § 2, 38 Stat. 251−52, and § 5, 38 Stat. 257–\n58. Moreover, the Act does not frame dividend payments\nas a contractual obligation of the Reserve Banks. Rather,\nthe six percent dividend is described as a benefit that mem-\nber banks “shall be entitled to receive.” § 7, 38 Stat. 258.\nAnd a statute does not create contractual obligations\nmerely by setting forth “benefits to those who comply with\nits conditions.” Wis. & Mich, Ry. Co., 191 U.S. at 387. In-\ndeed, in Dodge, the statutory language provided that teach-\ners who retired after serving twenty or more years in the\npublic schools “shall be paid the sum of fifteen hundred dol-\nlars ($1,500.00) annually and for life from the date of such\nretirement . . . .” 302 U.S. at 76 (emphasis added). Yet the\nSupreme Court declined to find that the teachers’ rights to\n$1,500 annuity payments vested upon meeting the statu-\ntory conditions. Id. at 77, 79−81. Similarly, we conclude\nthat the Federal Reserve Act does not demonstrate the gov-\nernment’s clear intent to confer vested contractual rights\non each member bank who complies with the stock sub-\nscription requirement.\n The circumstances surrounding the passage of the Fed-\neral Reserve Act and its legislative history offer further\nsupport. See Nat’l R.R., 470 U.S. at 468. The Federal Re-\nserve Act was passed in 1913 in response to ongoing insta-\nbility within the financial sector, which had given rise to a\nseries of banking crises in the preceding decades. See H.R.\nRep. No. 63-69, at 3−5 (1913). One of the deficiencies of the\nbanking system at the time was that it “fail[ed] to afford\nany safeguard against panics and commercial stringencies\nor any means of alleviating them.” Id. at 6.\n\f12 AMERICAN BANKERS v. UNITED STATES\n\n\n\n Congress created the Federal Reserve System to pre-\nvent and contain the financial disruption caused by bank\nfailures. The Reserve Banks were intended to serve as\nlenders of last resort by maintaining a reserve of liquid cap-\nital “ready for use in protecting the banks of any section of\nthe country and for enabling them to go on meeting their\nobligations instead of suspending payments, as so often in\nthe past.” Id. at 11; see also id. at 19−22. To provide the\nfunds for this reserve, Congress established the require-\nment that member banks subscribe to Reserve Bank stock.\nId. at 16−17, 20−21. Thus, the origins of the subscription\nrequirement reflect a regulatory effort to promote stability\nin the banking system through collaboration, rather than\na collection of private contractual undertakings.\n Statements in the legislative history bear this out. For\nexample, the House Report accompanying the bill later\npassed as the Federal Reserve Act, expressed the view that\n“banking institutions which desire to be known by the\nname ‘national’ should be required, and can well afford, to\ntake upon themselves the responsibilities involved in joint\nor federated organization.” Id. at 16. Likewise, the Senate\nReport stated that the Reserve Banks were “not intended\nto be merely money-making banks,” but “guardians of the\npublic welfare, primarily safeguarding the member banks,\nprotecting their reserves, safeguarding their credit, [and]\nprotecting them from panic or financial stringency. . . .” S.\nRep. No. 63-133, at 10 (1913). Although the proposed leg-\nislation was to provide member banks with a return on the\nuse of their funds, the Senate Report noted that “the sta-\nbility of the business of the bank, and the peace of mind it\nwill give to the bankers in having freedom from constant\nanxiety, would more than compensate them, even if the fi-\nnancial advantages did not do so.” Id. at 12.\n Accordingly, we discern no “clear indication” of the gov-\nernment’s intent to contract in either the language of the\nFederal Reserve Act or the circumstances under which it\nwas passed.\n\fAMERICAN BANKERS v. UNITED STATES 13\n\n\n\n The additional evidence on which Washington Federal\nrelies, primarily for its express contract theory, does noth-\ning to remedy this deficiency. For example, the July 17,\n2013 letter from the BSF merely states that Washington\nFederal’s application and payment have been processed\nand informs Washington Federal of some of the obligations\nand benefits associated with membership in the Federal\nReserve System. Washington Federal points to the letter’s\nstatement that “[d]ividends are paid at the statutory rate\nof 6 percent per annum, or $1.50 per share semi-annually.”\nJ.A. 65. But this is simply a statement of policy based on\nthe statutory dividend rate in effect at the time, not the\nlanguage of a promise or contractual undertaking. See\nChattler v. United States, 632 F.3d 1324, 1330 (Fed. Cir.\n2011) (“[T]he obligation of the government, if it is to be held\nliable, must be stated in the form of an undertaking, not as\na mere prediction or statement of opinion or intention.\nLikewise statements of information or definition are not\nstatements of obligation.” (internal citations and quotation\nmarks omitted)).\n Because Washington Federal did not plead facts suffi-\ncient to establish the government’s intent to contract, the\ncomplaint fails to state a plausible claim for breach of con-\ntract. Accordingly, this claim was properly dismissed. 6\n\n\n\n\n 6 Washington Federal’s claim for breach of implied\nduty of good faith and fair dealing likewise depends on the\nexistence of a valid contract. See Centex Corp. v. United\nStates, 395 F.3d 1283, 1304 (Fed. Cir. 2005) (“The covenant\nof good faith and fair dealing is an implied duty that each\nparty to a contract owes to its contracting partner.”).\nTherefore, this claim was also properly dismissed.\n\f14 AMERICAN BANKERS v. UNITED STATES\n\n\n\n B.\n We now turn to Washington Federal’s Fifth Amend-\nment takings claim. The complaint sets forth two takings\ntheories: (1) by enacting the FAST Act, the government de-\nprived Washington Federal of its “property interest[] in the\npromised six percent dividend,” J.A. 57 ¶ 92; and (2) by\npaying dividends at a rate lower than six percent, the gov-\nernment effected “a taking of [Washington Federal’s] capi-\ntal investment[] in Federal Reserve Bank stock without\njust compensation in the form of a market return on the\ninvested capital,” J.A. 57 ¶ 93. Under either theory, Wash-\nington Federal failed to state a plausible takings claim.\n To state a claim for a taking under the Fifth Amend-\nment, a plaintiff must identify a legally cognizable property\ninterest. Tex. State Bank v. United States, 423 F.3d 1370,\n1378 (Fed. Cir. 2005). “The Constitution neither creates\nnor defines the scope of property interests compensable un-\nder the Fifth Amendment.” Conti v. United States, 291\nF.3d 1334, 1340 (Fed. Cir. 2002) (citing Bd. of Regents of\nState Colls. v. Roth, 408 U.S. 564, 577 (1972)). Property\ninterests arise from “existing rules and understandings\nand background principles derived from an independent\nsource, such as state, federal, or common law. . . .” Air Peg-\nasus of D.C., Inc. v. United States, 424 F.3d 1206, 1213\n(Fed. Cir. 2005) (internal quotation marks omitted). To\nsupport a takings claim, a property interest must be more\nthan a “mere unilateral expectation or an abstract need.”\nWebb’s Fabulous Pharmacies, Inc. v. Beckwith, 449 U.S.\n155, 161 (1980).\n Under its first takings theory, Washington Federal as-\nserts a property interest in its “contractual and statutory\nrights to receive a six percent dividend on Federal Reserve\nBank stock . . . .” J.A. 57 ¶ 90. While contract rights are a\nform of property that may be compensable under the Fifth\nAmendment, see Cienega Gardens v. United States, 331\nF.3d 1319, 1329–30 (Fed. Cir. 2003), the complaint does not\n\fAMERICAN BANKERS v. UNITED STATES 15\n\n\n\nestablish that Washington Federal had a contractual right\nto a six percent dividend, see supra Section II.A.2. Thus,\nthe trial court properly dismissed the contract-based tak-\nings claim.\n Likewise, Washington Federal has not alleged a legally\ncognizable property interest arising from its “statutory\nrights” under the Federal Reserve Act. Absent independ-\nent evidence of a contractual undertaking, a statutory en-\ntitlement “creates no vested right.” Dodge, 302 U.S. at 79.\nBecause “Congress at all times retains the ability to amend\nstatutes, a power which inheres in its authority to legislate,\nCongress at all times retains the right to revoke legisla-\ntively created entitlements.” Members of Peanut Quota\nHolders Ass’n, Inc. v. United States, 421 F.3d 1323, 1335\n(Fed. Cir. 2005). Indeed, in this case, Congress “expressly\nreserved” its “right to amend, alter, or repeal” any provi-\nsion of the Federal Reserve Act. See § 30, 38 Stat. 275\n(1913), renumbered § 31, Pub. L. No. 95−630, title I, § 101,\n92 Stat. 3641 (1978). Washington Federal emphasizes\nthat, prior to the FAST Act, the six percent dividend rate\nhad remained unchanged for over 100 years. But Wash-\nington Federal’s “unilateral expectation” that Congress\nwould not exercise its right to amend the dividend provi-\nsion going forward does not give rise to a compensable\nproperty interest under the Fifth Amendment. See Webb’s,\n449 U.S. 161; Peanut Quota Holders, 421 F.3d at 1334\n(“[Appellants] have no legally protected right against the\ngovernment’s making changes in the underlying [regula-\ntory] program and no right to compensation for the loss in\nvalue resulting from the changes.”).\n Washington Federal’s alternative takings theory con-\ntemplates a taking of its underlying capital investment in\nReserve Bank stock without just compensation. 7 This\n\n\n 7 The trial court did not directly address this alter-\nnative takings theory. Washington Federal contends that\n\f16 AMERICAN BANKERS v. UNITED STATES\n\n\n\nclaim also fails. Washington Federal’s initial subscription\nof stock was part of its voluntary participation in a regula-\ntory scheme, and we have held that “enforceable rights suf-\nficient to support a taking claim against the United States\ncannot arise in an area voluntarily entered into and one\nwhich, from the start, is subject to pervasive Government\ncontrol.” Mitchell Arms, Inc. v. United States, 7 F.3d 212,\n216 (Fed. Cir. 1993) (internal quotation marks omitted); see\nalso Commonwealth Edison Co. v. United States, 271 F.3d\n1327, 1339 (Fed. Cir. 2001) (“[R]egulatory actions requiring\nthe payment of money are not takings.”). Furthermore, un-\nder the Federal Reserve Act, Washington Federal can sur-\nrender its stock and obtain a refund of its paid-in capital.\nSee 12 U.S.C. §§ 287, 321. Thus, the requirement that\nmember banks subscribe to reserve bank stock under the\nFederal Reserve Act does not constitute a regulatory tak-\ning.\n\n\n\n\nremand is necessary for the trial court to rule on this issue\nin the first instance. We conclude, however, that principles\nof judicial economy counsel against remand, and whether\nWashington Federal has adequately stated a claim under\nits alternative takings theory is an issue amenable to reso-\nlution for the first time on appeal. See Glaxo Grp. Ltd. v.\nTorPharm, Inc., 153 F.3d 1366, 1371 (Fed. Cir. 1998) (not-\ning that “an appellate court may choose to decide [an] issue\neven if not passed on by the trial court” where the issue “is\none of law” and “has been fully vetted by the parties on ap-\npeal”); Singleton v. Wulff, 428 U.S. 106, 121 (1976) (“The\nmatter of what questions may be taken up and resolved for\nthe first time on appeal is one left primarily to the discre-\ntion of the courts of appeals, to be exercised on the facts of\nindividual cases.”).\n\fAMERICAN BANKERS v. UNITED STATES 17\n\n\n\n Because the complaint fails to allege facts sufficient to\nsupport a taking under the Fifth Amendment, the trial\ncourt properly dismissed this claim under RCFC 12(b)(6).\n III\n For the foregoing reasons, we conclude that the com-\nplaint fails to state a claim upon which relief may be\ngranted. Thus, we affirm the trial court’s grant of the gov-\nernment’s motion to dismiss Washington Federal’s claims\nunder RCFC 12(b)(6).\n AFFIRMED\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: AMERICAN BANKERS ASSOCIATION, Washington Federal, N.A., Individually and on Behalf of All Others Similarly Situated, Plaintiffs-Appellants v. UNITED STATES, Defendant-Appellee\\nCourt: U.S. Court of Appeals for the Federal Circuit\\nDate decided: 2019-08-08\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n AMERICAN BANKERS ASSOCIATION,\\n WASHINGTON FEDERAL, N.A., INDIVIDUALLY\\n AND ON BEHALF OF ALL OTHERS SIMILARLY\\n SITUATED,\\n Plaintiffs-Appellants\\n\\n v.\\n\\n UNITED STATES,\\n Defendant-Appellee\\n ______________________\\n\\n 2018-1341\\n ______________________\\n\\n Appeal from the United States Court of Federal Claims\\nin No. 1:17-cv-0194-SGB, Senior Judge Susan G. Braden.\\n ______________________\\n\\n Decided: August 8, 2019\\n ______________________\\n\\n STEPHEN JOSEPH OBERMEIER, Wiley Rein LLP, Wash-\\nington, DC, argued for plaintiffs-appellants. Also repre-\\nsented by CLAIRE J. EVANS, MICHAEL E. TONER.\\n\\n ERIC PETER BRUSKIN, Commercial Litigation Branch,\\nCivil Division, United States Department of Justice, Wash-\\nington, DC, argued for defendant-appellee. Also repre-\\nsented by JOSEPH H. HUNT, ROBERT E. KIRSCHMAN, JR.,\\nKENNETH M. DINTZER, CLAUDIA BURKE; KATHERINE H.\\n\\f2 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\nWHEATLEY, Board of Governors of the Federal Reserve Sys-\\ntem, Washington, DC.\\n ______________________\\n\\n Before WALLACH, CHEN, and HUGHES, Circuit Judges.\\nHUGHES, Circuit Judge.\\n This case arises out of legislation amending the statu-\\ntory rate for dividend payments on Federal Reserve Bank\\nstock. The Federal Reserve Act of 1913 set the dividend\\nrate at six percent per year, which remained in effect until\\nCongress amended the dividend provision in 2016. The\\namendment effectively reduced the dividend rate for cer-\\ntain stockholder banks from the fixed six percent rate to a\\nlower variable rate. American Bankers Association and\\nWashington Federal, N.A. sued the United States in the\\nCourt of Federal Claims, arguing that banks who sub-\\nscribed to Reserve Bank stock before the amendment are\\nentitled to dividends at the six percent rate. The complaint\\nalleged that, by paying dividends at the amended statutory\\nrate, the United States breached a contractual duty or, in\\nthe alternative, effected a Fifth Amendment taking. The\\ntrial court dismissed the complaint under Rules of the U.S.\\nCourt of Federal Claims 12(b)(6) for failure to state a claim.\\nAmerican Bankers and Washington Federal now appeal.\\nBecause the complaint does not allege facts establishing\\nthe existence of a contract or an unconstitutional taking,\\nwe affirm.\\n I\\n A.\\n We begin with a brief overview of the Federal Reserve\\nSystem and its statutory origins. The Federal Reserve Act\\n\\fAMERICAN BANKERS v. UNITED STATES 3\\n\\n\\n\\nof 1913, Pub. L. No. 63−43, ch. 6, 38 Stat. 251 (1913), 1 es-\\ntablished a system to oversee banking operations and pro-\\nmote greater economic stability. The Federal Reserve\\nSystem includes the Federal Reserve Board of Governors,\\nsee id., §§ 10−11, 38 Stat. 260–63, and twelve regional Re-\\nserve Banks, see id. § 2, 38 Stat. 251–52. The Board exer-\\ncises broad regulatory supervision over the Reserve Banks,\\nwhich serve as banks to the U.S. government and to com-\\nmercial banks who are members of the Federal Reserve\\nSystem.\\n The Act sets forth the conditions under which commer-\\ncial banks may join the Federal Reserve System. One of\\nthe conditions of membership is that member banks must\\n“subscribe” to the stock of their regional Reserve Bank in\\nan amount “equal to six per centum of the paid-up capital\\nstock and surplus of [the] applicant bank . . . .” § 5, 38 Stat.\\n257. Every national bank 2 is required to join the system\\nand subscribe to Reserve Bank stock. § 2, 38 Stat 252.\\nOther financial institutions, such as state banks, are per-\\nmitted but not required to apply for membership and sub-\\nscribe to stock. § 9, 38 Stat. 259.\\n Reserve Bank stock is “divided into shares of $100,”\\nwhich “shall not be transferred or hypothecated.” § 5, 38\\nStat. 257. From 1913 to 2015, the Act provided that “the\\nstockholders of the [Reserve] bank shall be entitled to\\n\\n\\n\\n\\n 1 The Federal Reserve Act is codified as amended in\\nscattered sections of Chapter 3 of Title 12 of the United\\nStates Code. See 12 U.S.C. §§ 221−522. This opinion cites\\nto the original 1913 Act, which is the same as the current\\nversion except where otherwise noted.\\n 2 A national bank refers to a commercial bank char-\\ntered by the federal government under the National Bank\\nAct. See 12 U.S.C. § 21 et seq.\\n\\f4 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\nreceive an annual dividend of six per centum on the paid-\\nin capital stock . . . .” § 7, 38 Stat 258.\\n On December 4, 2015, Congress passed the Fixing\\nAmerica’s Surface Transportation Act (FAST Act), which\\nauthorized substantial appropriations for surface trans-\\nportation infrastructure. See Pub. L. No. 114–94, 129 Stat.\\n1312. The FAST Act included an amendment to the statu-\\ntory dividend rate for Reserve Bank stock owned by mem-\\nber banks with consolidated assets of more than $10\\nbillion. Under the amended dividend provision, these\\nbanks would receive a variable dividend rate equal to the\\nlesser of: (1) the rate of the 10-year Treasury note or (2) six\\npercent. See § 32203, 129 Stat. 1739 (codified as amended\\nat 12 U.S.C. § 289(a)(1)).\\n B.\\n Prior to 2013, Washington Federal operated as a feder-\\nally chartered savings and loan association. On May 29,\\n2013, Washington Federal received approval from the Of-\\nfice of the Comptroller of the Currency to convert to a na-\\ntional bank, contingent on, inter alia, Washington Federal\\napplying for membership in the Federal Reserve System.\\n On July 8, 2013, Washington Federal submitted an ap-\\nplication for Reserve Bank stock to the Reserve Bank of\\nSan Francisco (BSF). A letter from BSF, dated July 17,\\n2013, informed Washington Federal that its application\\nand payment for stock had been processed and enclosed an\\nAdvice of Holdings for 479,610 shares of BSF stock. The\\nletter further noted that “[d]ividends are paid at the statu-\\ntory rate of 6 percent per annum, or $1.50 per share semi-\\nannually.” J.A. 65.\\n From 2013 to 2015, Washington Federal received divi-\\ndend payments on its stock at a rate of six percent per year.\\nAfter the FAST Act took effect on January 1, 2016, Wash-\\nington Federal received dividends at the rate of the 10-year\\nTreasury note. In 2016, Washington Federal received\\n\\fAMERICAN BANKERS v. UNITED STATES 5\\n\\n\\n\\ndividends totaling $502,471.53, reflecting an annual rate of\\napproximately two percent.\\n C.\\n Washington Federal and American Bankers Associa-\\ntion 3 filed a complaint against the United States in the\\nCourt of Federal Claims on February 9, 2017. 4 The com-\\nplaint alleged that, by paying dividends at a rate lower\\nthan six percent in 2016, the government breached a con-\\ntractual duty to member banks that subscribed to Reserve\\nBank stock before December 4, 2015. The complaint also\\nasserted, in the alternative, that the government’s conduct\\neffected a Fifth Amendment taking.\\n The government filed a motion to dismiss for lack of\\nstanding under RCFC 12(b)(1) and failure to state a claim\\nunder RCFC 12(b)(6). The Court of Federal Claims deter-\\nmined that American Bankers failed to meet the require-\\nments for associational standing because the damages\\nrequested would require individualized proof for each asso-\\nciation member. The court found that Washington Federal\\nhad standing but dismissed all counts of the complaint un-\\nder RCFC 12(b)(6) for failure to state a claim. Washington\\nFederal and American Bankers now appeal the court’s dis-\\nmissal of the claims and its standing determination. We\\nhave jurisdiction under 28 U.S.C. § 1295(a)(3).\\n\\n\\n\\n\\n 3 American Bankers Association is a national trade\\nassociation for the banking industry. Its members include\\nWashington Federal, as well as other banks affected by the\\namendment to the dividend rate, i.e., member banks with\\nmore than $10 billion in consolidated assets.\\n 4 The complaint was subsequently amended on April\\n14, 2017. This opinion refers to the amended complaint\\nunless otherwise stated.\\n\\f6 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\n II\\n We review de novo whether the Court of Federal\\nClaims properly dismissed a complaint for failure to state\\na claim upon which relief may be granted. Frankel v.\\nUnited States, 842 F.3d 1246, 1249 (Fed. Cir. 2016). To\\navoid dismissal under RCFC 12(b)(6), a plaintiff “must al-\\nlege facts ‘plausibly suggesting (not merely consistent\\nwith)’ a showing of entitlement to relief.” Acceptance Ins.\\nCos., Inc. v. United States, 583 F.3d 849, 853 (Fed. Cir.\\n2009) (quoting Bell Atl. Corp. v. Twombly, 550 U.S. 544,\\n557 (2007)). “A claim has facial plausibility when the plain-\\ntiff pleads factual content that allows the court to draw the\\nreasonable inference that the defendant is liable for the\\nmisconduct alleged.” Ashcroft v. Iqbal, 556 U.S. 662, 678\\n(2009). In reviewing a motion to dismiss, we accept as true\\nthe complaint’s well-pled factual allegations; however, we\\nare not required to accept the asserted legal conclusions.\\nId.\\n For the reasons set forth below, we conclude that the\\ntrial court did not err in dismissing Washington Federal’s\\nbreach of contract and takings claims under RCFC\\n12(b)(6). 5\\n A.\\n First, we address Washington Federal’s breach of con-\\ntract claim. The complaint asserts that the government\\nbreached an implied-in-fact or express contract with Wash-\\nington Federal by paying dividends at a rate lower than six\\npercent in 2016. Washington Federal alleges that an im-\\nplied-in-fact contract exists because the Federal Reserve\\nAct constitutes an offer by the government, which\\n\\n\\n 5 We need not reach American Bankers’ standing ar-\\ngument because American Bankers’ allegations are the\\nsame as Washington Federal’s and would thus share the\\nsame flaws regardless of our outcome.\\n\\fAMERICAN BANKERS v. UNITED STATES 7\\n\\n\\n\\nWashington Federal accepted by submitting its application\\nand payment for Reserve Bank stock. Alternatively, Wash-\\nington Federal contends that an express contract was\\nformed based on its application for stock, which was a con-\\ntractual offer that the government accepted by approving\\nthe application and issuing stock.\\n There are four requirements to form a contract binding\\nupon the government: “(1) mutuality of intent to contract;\\n(2) lack of ambiguity in offer and acceptance; (3) consider-\\nation; and (4) a government representative having actual\\nauthority to bind the United States in contract.” Anderson\\nv. United States, 344 F.3d 1343, 1353 (Fed. Cir. 2003).\\nThese requirements apply to both express and implied-in-\\nfact contracts. Id. at 1353 n.3. “To satisfy its burden to\\nprove such a mutuality of intent, a plaintiff must show, by\\nobjective evidence, the existence of an offer and a reciprocal\\nacceptance.” Id.\\n For both its implied-in-fact and express contract theo-\\nries, Washington Federal relies largely on the Federal Re-\\nserve Act as evidence of the government’s intent to\\ncontract. Under its implied-in-fact contract theory, the Act\\nwas an offer to contract; under its express contract theory,\\nthe Act was an invitation to receive offers to contract. Un-\\nder either theory of contract formation, Washington Fed-\\neral argues that the Act contemplates a contractual\\nagreement between member banks and the government.\\nBecause Washington Federal failed to allege facts estab-\\nlishing the existence of a contract with the government, we\\ndetermine that the trial court did not err in dismissing this\\nclaim.\\n 1.\\n “[A]bsent some clear indication that the legislature in-\\ntends to bind itself contractually, the presumption is that\\n‘a law is not intended to create private contractual or\\nvested rights but merely declares a policy to be pursued\\nuntil the legislature shall ordain otherwise.’” Nat’l R.R.\\n\\f8 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\nPassenger Corp. v. Atchison Topeka & Santa Fe Ry. Co.,\\n470 U.S. 451, 465–66 (1985) (quoting with alterations\\nDodge v. Bd. of Educ., 302 U.S. 74, 79 (1937)). This “well-\\nestablished presumption” reflects a recognition that “the\\nprincipal function of a legislature is not to make contracts,\\nbut to make laws that establish the policy of the state.” Id.\\nAnd “[p]olicies, unlike contracts, are inherently subject to\\nrevision and repeal . . . .” Id.\\n To overcome the presumption, there must be a “clear\\nindication” that the legislature intended to create contrac-\\ntual rights enforceable against the government. Id. at\\n465−66. The Supreme Court has recognized evidence of an\\nintent to contract where a statute “provide[s] for the execu-\\ntion of a written contract on behalf of the United States” or\\n“speak[s] of a contract” with the United States. Id. at 467\\n(emphasis in original); see also Dodge, 302 U.S. at 78; Indi-\\nana ex rel. Anderson v. Brand, 303 U.S. 95, 105 (1938). For\\nexample, in Hall v. Wisconsin, 103 U.S. 5 (1880), the stat-\\nute provided for a geological, mineralogical, and agricul-\\ntural survey to be carried out by commissioners appointed\\nby the governor. Id. at 5–6. The statutory text directed the\\ngovernor to “make a written contract with each of the com-\\nmissioners . . . expressly stipulating and setting forth the\\nnature and extent of the services to be rendered by each,\\nand the compensation therefor . . . .” Id. at 8–9. Likewise,\\nin Indiana ex rel. Anderson, the Court found that Indiana’s\\nTeachers’ Tenure Law contemplated contracts binding on\\nthe state, noting that “[t]he title of the act is couched in\\nterms of contract” and the text “speaks of the making and\\ncanceling of indefinite contracts” between teachers and\\nschool districts. 303 U.S. at 105.\\n In finding that a statute or regulation constitutes an\\noffer to enter into a unilateral contract, courts have also\\nrelied on explicit references to contractual undertakings.\\nFor example, in Radium Mines, Inc. v. United States, 153\\nF. Supp. 403 (Ct. Cl. 1957), the regulation at issue included\\na section entitled “Purchase Contract,” which stated that,\\n\\fAMERICAN BANKERS v. UNITED STATES 9\\n\\n\\n\\nif a sample of uranium delivered to the Commission\\n“meet[s] the conditions of this section, the Commission will\\nforward to the person making the offer a form of contract\\ncontaining applicable terms and conditions ready for his\\nacceptance.” Id. at 405. Similarly, the statutory provision\\nin Grav v. United States, 14 Cl. Ct. 390 (1988), aff’d, 886\\nF.2d 1305 (Fed. Cir. 1989), provided that “[t]he Secretary\\nshall offer to enter into a contract” with milk producers. Id.\\nat 392.\\n In contrast, the Supreme Court determined in Dodge\\nthat the Miller Law did not clearly express the govern-\\nment’s intent to contract. 302 U.S. at 80. As originally en-\\nacted, the Miller Law established a compulsory retirement\\nage for public school teachers and provided for the payment\\nof annuities to retired teachers. Id. at 76. The law stated\\nthat teachers “who served in the public schools of such city\\nfor twenty or more years prior to such retirement, shall be\\npaid the sum of fifteen hundred dollars ($1,500.00) annu-\\nally and for life from the date of such retirement . . . .” Id.\\nNearly ten years after it was passed, the Miller Law was\\namended to reduce annuity payments to $500 for all retired\\nteachers, including those who had retired prior to the\\namendment. Id. at 77. The teachers who filed suit against\\nthe Board of Education argued that they were contractu-\\nally entitled to annuity payments at $1,500 because the\\nMiller Law constituted an offer to contract, which they had\\naccepted by remaining in service for at least twenty years.\\nId. at 77. The Supreme Court rejected this argument, con-\\ncluding that neither the statutory language nor the circum-\\nstances of enactment indicated a legislative intent to create\\nbinding contractual obligations. Id. at 79−81.\\n 2.\\n To determine whether a statute gives rise to a contrac-\\ntual obligation, we first look to the language of the statute.\\nSee Dodge, 302 U.S. at 78; Nat’l R.R., 470 U.S. at 466. The\\nlanguage of the Federal Reserve Act is devoid of the\\n\\f10 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\ntraditional indicia of a contractual undertaking. The Act\\ndoes not “speak of a contract” between Reserve Banks and\\nmember banks; nor does it “provide for the execution of a\\nwritten contract on behalf of the United States.” See Nat’l\\nR.R., 470 U.S. at 467. Rather, the Act sets forth a regula-\\ntory system, in which member banks are granted certain\\n“powers and privileges” and are subject to specified “duties,\\nliabilities, and regulations.” § 8, 38 Stat. 259. Among the\\nduties, member banks are “required . . . to subscribe to the\\ncapital stock” of their regional Reserve Bank. § 2, 38 Stat.\\n252. Among the privileges, banks “shall be entitled to re-\\nceive an annual dividend of six per centum on the paid-in\\ncapital stock. . . .” § 7, 38 Stat. 258.\\n Washington Federal urges us to discern contractual in-\\ntent from the terms “subscribe” and “subscription,” which\\nit contends are “contractual terms of art in the context of\\nstock offerings. . . .” Appellant’s Op. Br. 31; see also id. at\\n32−33. But we must interpret the language in the context\\nin which it is written. In the context of a regulatory stat-\\nute, we will not infer a contractual undertaking “absent ‘an\\nadequate expression of an actual intent’ of the State to bind\\nitself. . . .” Nat’l R.R., 470 U.S. at 466–67 (quoting Wis. &\\nMich. Ry. Co. v. Powers, 191 U.S. 379, 386–87 (1903)). And\\nthe use of terminology that carries contractual connota-\\ntions when used in the private sector does not, on its own,\\nestablish such intent. For example, in Dodge, the Supreme\\nCourt rejected appellants’ argument that “annuity” is “ter-\\nminology based on contract” that reflected the legislature’s\\nintent to establish contractual rights. 302 U.S. at 81. Like-\\nwise, we find that the subscription language in the Federal\\nReserve Act does not unequivocally express the govern-\\nment’s intent to bind itself in contract.\\n Washington Federal further argues that the govern-\\nment’s intent to contract is evident from the exchange of\\nobligations between member banks and Reserve Banks.\\nAccording to Washington Federal, the Act contemplates an\\nagreement that Reserve Banks will pay member banks an\\n\\fAMERICAN BANKERS v. UNITED STATES 11\\n\\n\\n\\nannual dividend of six percent in exchange for member\\nbanks’ subscription to Reserve Bank stock. The language\\nand structure of the Act, however, do not reflect a bar-\\ngained-for quid pro quo between two parties. The dividend\\nrate is set forth in an entirely different section than the\\nprovisions governing stock subscription. Compare § 7, 38\\nStat. 258, with § 2, 38 Stat. 251−52, and § 5, 38 Stat. 257–\\n58. Moreover, the Act does not frame dividend payments\\nas a contractual obligation of the Reserve Banks. Rather,\\nthe six percent dividend is described as a benefit that mem-\\nber banks “shall be entitled to receive.” § 7, 38 Stat. 258.\\nAnd a statute does not create contractual obligations\\nmerely by setting forth “benefits to those who comply with\\nits conditions.” Wis. & Mich, Ry. Co., 191 U.S. at 387. In-\\ndeed, in Dodge, the statutory language provided that teach-\\ners who retired after serving twenty or more years in the\\npublic schools “shall be paid the sum of fifteen hundred dol-\\nlars ($1,500.00) annually and for life from the date of such\\nretirement . . . .” 302 U.S. at 76 (emphasis added). Yet the\\nSupreme Court declined to find that the teachers’ rights to\\n$1,500 annuity payments vested upon meeting the statu-\\ntory conditions. Id. at 77, 79−81. Similarly, we conclude\\nthat the Federal Reserve Act does not demonstrate the gov-\\nernment’s clear intent to confer vested contractual rights\\non each member bank who complies with the stock sub-\\nscription requirement.\\n The circumstances surrounding the passage of the Fed-\\neral Reserve Act and its legislative history offer further\\nsupport. See Nat’l R.R., 470 U.S. at 468. The Federal Re-\\nserve Act was passed in 1913 in response to ongoing insta-\\nbility within the financial sector, which had given rise to a\\nseries of banking crises in the preceding decades. See H.R.\\nRep. No. 63-69, at 3−5 (1913). One of the deficiencies of the\\nbanking system at the time was that it “fail[ed] to afford\\nany safeguard against panics and commercial stringencies\\nor any means of alleviating them.” Id. at 6.\\n\\f12 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\n Congress created the Federal Reserve System to pre-\\nvent and contain the financial disruption caused by bank\\nfailures. The Reserve Banks were intended to serve as\\nlenders of last resort by maintaining a reserve of liquid cap-\\nital “ready for use in protecting the banks of any section of\\nthe country and for enabling them to go on meeting their\\nobligations instead of suspending payments, as so often in\\nthe past.” Id. at 11; see also id. at 19−22. To provide the\\nfunds for this reserve, Congress established the require-\\nment that member banks subscribe to Reserve Bank stock.\\nId. at 16−17, 20−21. Thus, the origins of the subscription\\nrequirement reflect a regulatory effort to promote stability\\nin the banking system through collaboration, rather than\\na collection of private contractual undertakings.\\n Statements in the legislative history bear this out. For\\nexample, the House Report accompanying the bill later\\npassed as the Federal Reserve Act, expressed the view that\\n“banking institutions which desire to be known by the\\nname ‘national’ should be required, and can well afford, to\\ntake upon themselves the responsibilities involved in joint\\nor federated organization.” Id. at 16. Likewise, the Senate\\nReport stated that the Reserve Banks were “not intended\\nto be merely money-making banks,” but “guardians of the\\npublic welfare, primarily safeguarding the member banks,\\nprotecting their reserves, safeguarding their credit, [and]\\nprotecting them from panic or financial stringency. . . .” S.\\nRep. No. 63-133, at 10 (1913). Although the proposed leg-\\nislation was to provide member banks with a return on the\\nuse of their funds, the Senate Report noted that “the sta-\\nbility of the business of the bank, and the peace of mind it\\nwill give to the bankers in having freedom from constant\\nanxiety, would more than compensate them, even if the fi-\\nnancial advantages did not do so.” Id. at 12.\\n Accordingly, we discern no “clear indication” of the gov-\\nernment’s intent to contract in either the language of the\\nFederal Reserve Act or the circumstances under which it\\nwas passed.\\n\\fAMERICAN BANKERS v. UNITED STATES 13\\n\\n\\n\\n The additional evidence on which Washington Federal\\nrelies, primarily for its express contract theory, does noth-\\ning to remedy this deficiency. For example, the July 17,\\n2013 letter from the BSF merely states that Washington\\nFederal’s application and payment have been processed\\nand informs Washington Federal of some of the obligations\\nand benefits associated with membership in the Federal\\nReserve System. Washington Federal points to the letter’s\\nstatement that “[d]ividends are paid at the statutory rate\\nof 6 percent per annum, or $1.50 per share semi-annually.”\\nJ.A. 65. But this is simply a statement of policy based on\\nthe statutory dividend rate in effect at the time, not the\\nlanguage of a promise or contractual undertaking. See\\nChattler v. United States, 632 F.3d 1324, 1330 (Fed. Cir.\\n2011) (“[T]he obligation of the government, if it is to be held\\nliable, must be stated in the form of an undertaking, not as\\na mere prediction or statement of opinion or intention.\\nLikewise statements of information or definition are not\\nstatements of obligation.” (internal citations and quotation\\nmarks omitted)).\\n Because Washington Federal did not plead facts suffi-\\ncient to establish the government’s intent to contract, the\\ncomplaint fails to state a plausible claim for breach of con-\\ntract. Accordingly, this claim was properly dismissed. 6\\n\\n\\n\\n\\n 6 Washington Federal’s claim for breach of implied\\nduty of good faith and fair dealing likewise depends on the\\nexistence of a valid contract. See Centex Corp. v. United\\nStates, 395 F.3d 1283, 1304 (Fed. Cir. 2005) (“The covenant\\nof good faith and fair dealing is an implied duty that each\\nparty to a contract owes to its contracting partner.”).\\nTherefore, this claim was also properly dismissed.\\n\\f14 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\n B.\\n We now turn to Washington Federal’s Fifth Amend-\\nment takings claim. The complaint sets forth two takings\\ntheories: (1) by enacting the FAST Act, the government de-\\nprived Washington Federal of its “property interest[] in the\\npromised six percent dividend,” J.A. 57 ¶ 92; and (2) by\\npaying dividends at a rate lower than six percent, the gov-\\nernment effected “a taking of [Washington Federal’s] capi-\\ntal investment[] in Federal Reserve Bank stock without\\njust compensation in the form of a market return on the\\ninvested capital,” J.A. 57 ¶ 93. Under either theory, Wash-\\nington Federal failed to state a plausible takings claim.\\n To state a claim for a taking under the Fifth Amend-\\nment, a plaintiff must identify a legally cognizable property\\ninterest. Tex. State Bank v. United States, 423 F.3d 1370,\\n1378 (Fed. Cir. 2005). “The Constitution neither creates\\nnor defines the scope of property interests compensable un-\\nder the Fifth Amendment.” Conti v. United States, 291\\nF.3d 1334, 1340 (Fed. Cir. 2002) (citing Bd. of Regents of\\nState Colls. v. Roth, 408 U.S. 564, 577 (1972)). Property\\ninterests arise from “existing rules and understandings\\nand background principles derived from an independent\\nsource, such as state, federal, or common law. . . .” Air Peg-\\nasus of D.C., Inc. v. United States, 424 F.3d 1206, 1213\\n(Fed. Cir. 2005) (internal quotation marks omitted). To\\nsupport a takings claim, a property interest must be more\\nthan a “mere unilateral expectation or an abstract need.”\\nWebb’s Fabulous Pharmacies, Inc. v. Beckwith, 449 U.S.\\n155, 161 (1980).\\n Under its first takings theory, Washington Federal as-\\nserts a property interest in its “contractual and statutory\\nrights to receive a six percent dividend on Federal Reserve\\nBank stock . . . .” J.A. 57 ¶ 90. While contract rights are a\\nform of property that may be compensable under the Fifth\\nAmendment, see Cienega Gardens v. United States, 331\\nF.3d 1319, 1329–30 (Fed. Cir. 2003), the complaint does not\\n\\fAMERICAN BANKERS v. UNITED STATES 15\\n\\n\\n\\nestablish that Washington Federal had a contractual right\\nto a six percent dividend, see supra Section II.A.2. Thus,\\nthe trial court properly dismissed the contract-based tak-\\nings claim.\\n Likewise, Washington Federal has not alleged a legally\\ncognizable property interest arising from its “statutory\\nrights” under the Federal Reserve Act. Absent independ-\\nent evidence of a contractual undertaking, a statutory en-\\ntitlement “creates no vested right.” Dodge, 302 U.S. at 79.\\nBecause “Congress at all times retains the ability to amend\\nstatutes, a power which inheres in its authority to legislate,\\nCongress at all times retains the right to revoke legisla-\\ntively created entitlements.” Members of Peanut Quota\\nHolders Ass’n, Inc. v. United States, 421 F.3d 1323, 1335\\n(Fed. Cir. 2005). Indeed, in this case, Congress “expressly\\nreserved” its “right to amend, alter, or repeal” any provi-\\nsion of the Federal Reserve Act. See § 30, 38 Stat. 275\\n(1913), renumbered § 31, Pub. L. No. 95−630, title I, § 101,\\n92 Stat. 3641 (1978). Washington Federal emphasizes\\nthat, prior to the FAST Act, the six percent dividend rate\\nhad remained unchanged for over 100 years. But Wash-\\nington Federal’s “unilateral expectation” that Congress\\nwould not exercise its right to amend the dividend provi-\\nsion going forward does not give rise to a compensable\\nproperty interest under the Fifth Amendment. See Webb’s,\\n449 U.S. 161; Peanut Quota Holders, 421 F.3d at 1334\\n(“[Appellants] have no legally protected right against the\\ngovernment’s making changes in the underlying [regula-\\ntory] program and no right to compensation for the loss in\\nvalue resulting from the changes.”).\\n Washington Federal’s alternative takings theory con-\\ntemplates a taking of its underlying capital investment in\\nReserve Bank stock without just compensation. 7 This\\n\\n\\n 7 The trial court did not directly address this alter-\\nnative takings theory. Washington Federal contends that\\n\\f16 AMERICAN BANKERS v. UNITED STATES\\n\\n\\n\\nclaim also fails. Washington Federal’s initial subscription\\nof stock was part of its voluntary participation in a regula-\\ntory scheme, and we have held that “enforceable rights suf-\\nficient to support a taking claim against the United States\\ncannot arise in an area voluntarily entered into and one\\nwhich, from the start, is subject to pervasive Government\\ncontrol.” Mitchell Arms, Inc. v. United States, 7 F.3d 212,\\n216 (Fed. Cir. 1993) (internal quotation marks omitted); see\\nalso Commonwealth Edison Co. v. United States, 271 F.3d\\n1327, 1339 (Fed. Cir. 2001) (“[R]egulatory actions requiring\\nthe payment of money are not takings.”). Furthermore, un-\\nder the Federal Reserve Act, Washington Federal can sur-\\nrender its stock and obtain a refund of its paid-in capital.\\nSee 12 U.S.C. §§ 287, 321. Thus, the requirement that\\nmember banks subscribe to reserve bank stock under the\\nFederal Reserve Act does not constitute a regulatory tak-\\ning.\\n\\n\\n\\n\\nremand is necessary for the trial court to rule on this issue\\nin the first instance. We conclude, however, that principles\\nof judicial economy counsel against remand, and whether\\nWashington Federal has adequately stated a claim under\\nits alternative takings theory is an issue amenable to reso-\\nlution for the first time on appeal. See Glaxo Grp. Ltd. v.\\nTorPharm, Inc., 153 F.3d 1366, 1371 (Fed. Cir. 1998) (not-\\ning that “an appellate court may choose to decide [an] issue\\neven if not passed on by the trial court” where the issue “is\\none of law” and “has been fully vetted by the parties on ap-\\npeal”); Singleton v. Wulff, 428 U.S. 106, 121 (1976) (“The\\nmatter of what questions may be taken up and resolved for\\nthe first time on appeal is one left primarily to the discre-\\ntion of the courts of appeals, to be exercised on the facts of\\nindividual cases.”).\\n\\fAMERICAN BANKERS v. UNITED STATES 17\\n\\n\\n\\n Because the complaint fails to allege facts sufficient to\\nsupport a taking under the Fifth Amendment, the trial\\ncourt properly dismissed this claim under RCFC 12(b)(6).\\n III\\n For the foregoing reasons, we conclude that the com-\\nplaint fails to state a claim upon which relief may be\\ngranted. Thus, we affirm the trial court’s grant of the gov-\\nernment’s motion to dismiss Washington Federal’s claims\\nunder RCFC 12(b)(6).\\n AFFIRMED\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4197,"cells":{"id":{"kind":"number","value":4647095,"string":"4,647,095"},"name":{"kind":"string","value":"ELI LILLY AND COMPANY, Plaintiff-Appellee v. HOSPIRA, INC., Defendant-Appellant Eli Lilly and Company, Plaintiff-Appellee v. Dr. Reddy's Laboratories, Ltd., Dr. Reddy's Laboratories, Inc., Defendants-Appellants"},"name_abbreviation":{"kind":"string","value":"Eli Lilly and Company v. Hospira, Inc."},"decision_date":{"kind":"string","value":"2019-08-09"},"docket_number":{"kind":"string","value":"2018-2126; 2018-2127; 2018-2128"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Federal Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Lourie, Moore, Taranto\", \"parties\": \"\", \"opinions\": [{\"author\": \"LOURIE, Circuit Judge.\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n ELI LILLY AND COMPANY,\\n Plaintiff-Appellee\\n\\n v.\\n\\n HOSPIRA, INC.,\\n Defendant-Appellant\\n ______________________\\n\\n 2018-2126, 2018-2127\\n ______________________\\n\\n Appeals from the United States District Court for the\\nSouthern District of Indiana in No. 1:16-cv-03460-TWP-\\nMPB, Judge Tanya Walton Pratt.\\n --------------------------------------------------\\n ELI LILLY AND COMPANY,\\n Plaintiff-Appellee\\n\\n v.\\n\\n DR. REDDY'S LABORATORIES, LTD., DR.\\n REDDY'S LABORATORIES, INC.,\\n Defendants-Appellants\\n ______________________\\n\\n 2018-2128\\n ______________________\\n\\f2 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\n Appeal from the United States District Court for the\\nSouthern District of Indiana in No. 1:16-cv-00308-TWP-\\nMPB, Judge Tanya Walton Pratt.\\n ______________________\\n\\n Decided: August 9, 2019\\n ______________________\\n\\n ADAM LAWRENCE PERLMAN, Williams & Connolly LLP,\\nWashington, DC, argued for plaintiff-appellee in 2018-2126\\nand 2018-2128. Also represented by GALINA I. FOMENKOVA,\\nDOV PHILIP GROSSMAN, DAVID M. KRINSKY, ANDREW P.\\nLEMENS, CHARLES MCCLOUD; JAMES PATRICK LEEDS, Eli\\nLilly and Company, Indianapolis, IN.\\n\\n BRADFORD PETER LYERLA, Jenner & Block LLP, Chi-\\ncago, IL, argued for defendant-appellant in 2018-2126.\\nAlso represented by YUSUF ESAT, SARA TONNIES HORTON;\\nADAM G. UNIKOWSKY, Washington, DC.\\n\\n JOHN C. O'QUINN, Kirkland & Ellis LLP, Washington,\\nDC, argued for defendants-appellants in 2018-2128. Also\\nrepresented by WILLIAM H. BURGESS, CALVIN ALEXANDER\\nSHANK; JEFFERY B. ARNOLD, Holland & Knight LLP, At-\\nlanta, GA; MERRI C. MOKEN, CHARLES A. WEISS, ERIC H.\\nYECIES, New York, NY.\\n\\n BRIAN TIMOTHY BURGESS, Goodwin Procter LLP, Wash-\\nington, DC, for amicus curiae Actavis LLC in 2018-2128.\\nAlso represented by EDWINA CLARKE, EMILY L. RAPALINO,\\nDARYL L. WIESEN, Boston, MA; LINNEA P. CIPRIANO, New\\nYork, NY.\\n ______________________\\n\\n\\n Before LOURIE, MOORE, and TARANTO, Circuit Judges.\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 3\\n\\n\\n\\nLOURIE, Circuit Judge.\\n Hospira Inc. (“Hospira”), Dr. Reddy’s Laboratories\\nLtd., and Dr. Reddy’s Laboratories Inc. (collectively,\\n“DRL”) appeal from two judgments of the United States\\nDistrict Court for the Southern District of Indiana in two\\ninfringement suits brought by Eli Lilly & Company\\n(“Lilly”) under the Hatch-Waxman Act, 21 U.S.C. § 355.\\nThe district court held in each case that the defendant’s\\nsubmission of a New Drug Application pursuant to 21\\nU.S.C. § 355(b)(2) infringed U.S. Patent 7,772,209 (the\\n“’209 patent”) under 35 U.S.C. § 271(e)(2). See Eli Lilly &\\nCo. v. Hospira, Inc., No. 1:16-cv-03460-TWP-MPB, 2018\\nWL 3008570 (S.D. Ind. June 15, 2018) (“Hospira Decision”);\\nEli Lilly & Co. v. Dr. Reddy’s Labs., Ltd., 323 F. Supp. 3d\\n1042 (S.D. Ind. 2018) (“DRL Decision”); see also Eli Lilly &\\nCo. v. Dr. Reddy’s Labs., Ltd., No. 1:16-cv-00308-TWP-\\nMPB, 2017 WL 6387316 (S.D. Ind. Dec. 14, 2017) (“DRL\\nSummary Judgment Decision”). Accordingly, the district\\ncourt entered orders under 35 U.S.C. § 271(e)(4)(A) prohib-\\niting FDA approval of the products at issue until the expi-\\nration of the ’209 patent. Eli Lilly & Co. v. Hospira, Inc.,\\nNo. 1:16-cv-03460-TWP-MPB (S.D. Ind. June 27, 2018),\\nECF No. 94; Eli Lilly & Co. v. Dr. Reddy’s Labs., Ltd., No.\\n1:16-cv-00308-TWP-MPB, 2018 WL 3616715 (S.D. Ind.\\nJuly 27, 2018). We decide these appeals together in this\\ncombined opinion. 1\\n We reverse the district court’s finding of literal in-\\nfringement in the Hospira Decision as clearly erroneous in\\nlight of the court’s claim construction of “administration of\\npemetrexed disodium.” Because the district court did not\\nerr in its application of the doctrine of equivalents in either\\n\\n\\n 1 We refer to the joint appendices in these appeals by\\nreference to each appellant. Lilly’s brief in the Hospira ap-\\npeal is referred to as “Lilly Br. I” and its brief in the DRL\\nappeal as “Lilly Br. II.”\\n\\f4 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\ndecision, we affirm both judgments of infringement. Thus,\\nthe Hospira Decision is affirmed-in-part and reversed-in-\\npart, and the DRL Decision is affirmed.\\n BACKGROUND\\n Lilly markets the compound pemetrexed in the form of\\na disodium salt as Alimta®, which is indicated, both alone\\nand in combination with other active agents, for treating\\ncertain types of non-small cell lung cancer and mesotheli-\\noma. Pemetrexed is an antifolate, a class of molecules\\nwhich, at the time of the invention in 2001, was “one of the\\nmost thoroughly studied classes of antineoplastic agents.”\\n’209 patent col. 1 ll. 19–20. Antifolates are structurally\\nsimilar to folic acid and work by competitively binding to\\ncertain enzymes that use folic acid metabolites as cofactors\\nin several steps of de novo nucleotide synthesis. Id. col. 1\\nll. 40–41. Unlike folic acid, antifolates do not enable these\\nsynthetic steps, but instead inhibit them. Pemetrexed in-\\nhibits several of these enzymes, including thymidylate syn-\\nthase, which methylates deoxyuridine in the final step of\\ndeoxythymidine synthesis. Id. col. 1 ll. 59–61. By inhibit-\\ning the creation of these nucleotides, antifolates slow down\\nDNA and RNA synthesis, and with it, cell growth and divi-\\nsion. Cancer cells tend to grow rapidly, so antifolate ther-\\napy affects them disproportionately, but healthy cells can\\nalso be damaged.\\n Pemetrexed had been known for at least a decade in\\n2001. Lilly’s U.S. Patent 5,344,932 (“Taylor”) disclosed\\nthat certain glutamic acid derivatives with pyrrolo[2,3-\\nd]pyrimidine heterocyclic ring structures, exemplified by\\npemetrexed, are “particularly active . . . inhibitors of thy-\\nmidylate synth[ase],” Taylor col. 1 ll. 59–60; see also id. col.\\n19 l. 37–col. 20 l. 25 (disclosing data indicating that\\npemetrexed inhibits thymidylate synthase activity in vitro\\nin human cell lines and in vivo in mice). The Taylor patent\\nalso disclosed that its compounds could be employed as\\n“pharmaceutically acceptable salt[s],” id. col. 2 l. 35, and\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 5\\n\\n\\n\\nthat the disodium salt form was particularly advantageous,\\nid. col. 2 ll. 47–48. U.S. Patent 4,997,838 (“Akimoto”), to\\nwhich Lilly took a license, disclosed a large genus of com-\\npounds containing pyrrolo[2,3-d]pyrimidine heterocyclic\\nring structures and a glutamic acid functional group, and\\nthat encompassed pemetrexed. The Akimoto patent dis-\\ncloses nearly fifty exemplary compounds, col. 14 l. 61–col.\\n16 l. 48, none of which is pemetrexed. Akimoto further dis-\\ncloses that its compounds may be prepared as salts of\\n“pharmaceutically acceptable bases,” such as “alkali met-\\nals, alkali earth metals, non-toxic metals, ammonium, and\\nsubstituted ammonium.” Id. col. 14 ll. 44–47.\\n By 2001, Lilly had also published the results of several\\nclinical trials investigating the use of pemetrexed disodium\\nas a treatment for different types of cancer. See, e.g., W.\\nJohn et al., “Activity of Multitargeted Antifolate\\n(Pemetrexed Disodium, LY231514) in Patients with Ad-\\nvanced Colorectal Carcinoma: Results from a Phase II\\nStudy,” Cancer, 88(8):1807–13 (2000). In the course of con-\\nducting these studies, Lilly discovered that pemetrexed\\ndisodium caused severe hematologic and immunologic side\\neffects, resulting in infections, nausea, rashes, and even\\nsome deaths. See id.; see also Neptune Generics, LLC v. Eli\\nLilly & Co., 921 F.3d 1372, 1377–78 (Fed. Cir. 2019) (dis-\\ncussing Lilly’s response to adverse clinical data), and Nep-\\ntune Generics, LLC v. Eli Lilly & Co., No. IPR2016-00240,\\n2017 WL 4466557, at *28–30 (P.T.A.B. Oct. 5, 2017) (same).\\nAs the ’209 patent teaches, such side effects are not uncom-\\nmon among antifolates. See ’209 patent col. 1 ll. 11–14.\\nSome researchers hypothesized that folic acid deficiency\\ncaused these side effects and suggested supplementing\\npemetrexed disodium treatment with folic acid. DRL J.A.\\n7870 (citing J.F. Worzalla et al., “Role of Folic Acid in Mod-\\nulating the Toxicity and Efficacy of the Multitargeted An-\\ntifolate, LY231514,” Anticancer Research, 18:3235–40\\n(1998)).\\n\\f6 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\n The invention of the ’209 patent is an improved method\\nof treatment with antifolates, particularly pemetrexed\\ndisodium, through supplementation with a methylmalonic\\nacid lowering agent and folic acid. Doing so, according to\\nthe patent, lessens antifolate toxicity without sacrificing ef-\\nficacy. See ’209 patent col. 10 ll. 17–53 (reporting that pre-\\nsupplementation regimen of vitamin B12 and folic acid in\\nclinical studies substantially reduced pemetrexed-induced\\ntoxicity and deaths while delivering a superior chemother-\\napeutic response rate). The ’209 patent lists preferred an-\\ntifolates, including some then-existing antifolate therapies,\\nas well as “derivatives described in” several patents includ-\\ning the Akimoto patent, and “most preferred, Pemetrexed\\nDisodium.” Id. col. 4 ll. 28–43. Each of the claims of the\\n’209 patent requires administration of pemetrexed diso-\\ndium following administration of folic acid and a\\nmethylmalonic acid lowering agent, specified in some\\nclaims, as well as the Alimta® label, as vitamin B12. Claim\\n12 is representative2:\\n 12. An improved method for administering\\n pemetrexed disodium to a patient in need of chemo-\\n therapeutic treatment, wherein the improvement\\n comprises:\\n a) administration of between about 350 µg\\n and about 1000 µg of folic acid prior to the\\n first administration of pemetrexed diso-\\n dium;\\n\\n\\n\\n\\n 2 The district court treated claim 12 as representa-\\ntive, DRL Summary Judgment Decision, 2017 WL\\n6387316, at *1–2; Hospira Decision, 2018 WL 3008570, at\\n*2, and no party has disputed that determination on ap-\\npeal. See, e.g., DRL Opening Br. 8–9; Hospira Opening Br.\\n23.\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 7\\n\\n\\n\\n b) administration of about 500 µg to about\\n 1500 µg of vitamin B12, prior to the first\\n administration of pemetrexed disodium;\\n and\\n c) administration of pemetrexed disodium.\\n In a parent application, Application 10/297,821 (the\\n“’821 application”), Lilly originally sought broad claims to\\nmethods of administering an antifolate in conjunction with\\na methylmalonic acid lowering agent, with or without folic\\nacid. The original independent claims 2 and 5 read:\\n 2. (Original) A method of reducing the toxicity as-\\n sociated with the administration of an antifolate to\\n a mammal comprising\\n administering to said mammal an effective\\n amount of said antifolate in combination\\n with a methylmalonic acid lowering agent.\\n 5. (Original) A method of reducing the toxicity as-\\n sociated with the administration of an antifolate to\\n a mammal comprising\\n administering to said mammal an effective\\n amount of said antifolate in combination\\n with a methylmalonic acid lowering agent\\n and FBP binding agent.\\nDRL J.A. 7860. A dependent claim further limited the an-\\ntifolate to pemetrexed disodium. Id. at 7861.\\n Claim 2 was rejected as anticipated by F.G. Arsenyan\\net al., “Influence of Methylcobalamin on the Antineoplastic\\nActivity of Methotrexate,” Onkol. Nauchn., 12(10):1299-\\n1303 (1978), which disclosed experiments treating mice\\nwith various tumors with a combination of methotrexate,\\nan antifolate, and methylcobalamin, a vitamin B12 deriva-\\ntive. The rest of the pending claims, including Claim 5,\\nwere rejected as obvious over a collection of references: U.S.\\nPatent 5,431,925 (“Ohmori”)—which taught treatment of\\n\\f8 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\nchemotherapeutically-induced immunosuppression with a\\ncombination of vitamins that could include folic acid and\\nvitamin B12—Worzalla, John, and Arsenyan. ’821 appli-\\ncation, Sept. 27, 2004, Office Action; DRL J.A. 7868–72.\\n In response, Lilly amended both claims to narrow “an-\\ntifolate” to “pemetrexed disodium” and cancelled its de-\\npendent claim limited to pemetrexed disodium. ’821\\napplication, Jan. 25, 2005, Response to Office Action; DRL\\nJ.A. 7877–84. In its remarks, Lilly asserted that the\\namendment to claim 2 overcame the anticipation rejection\\nbecause Arsenyan does not disclose pemetrexed disodium.\\nId. To overcome the obviousness rejection of claim 5 and\\nits dependents, Lilly generally argued that, while John dis-\\ncloses hematologic and immunologic toxicities from admin-\\nistration of pemetrexed disodium, it never suggests\\nvitamin supplementation, and none of the other references\\n“teach the use of [vitamin B12] to reduce toxicities associ-\\nated with an antifolate.” Id. The examiner then withdrew\\nthe anticipation rejection and later withdrew the obvious-\\nness rejection. The ’821 application issued as U.S. Patent\\n7,053,065, and the ’209 patent later issued from a continu-\\nation application.\\n These appeals were taken from cases which are among\\nthe latest in a series of patent disputes about Alimta® that\\nreaches back more than a decade. 3 In this most recent\\nchapter, DRL, Hospira, and Actavis 4 submitted New Drug\\n\\n\\n 3 This is the fourth appeal we have decided concern-\\ning Alimta® and the third specifically concerning the ’209\\npatent. See Neptune Generics, 921 F.3d 1372; Eli Lilly &\\nCo. v. Teva Parenteral Meds., Inc., 845 F.3d 1357 (Fed. Cir.\\n2017); Eli Lilly & Co. v. Teva Parenteral Meds., Inc., 689\\nF.3d 1368 (Fed. Cir. 2012).\\n 4 Lilly also sued Actavis LLC (“Actavis”) for infringe-\\nment of the ’209 patent, Eli Lilly & Co. v. Actavis LLC, No.\\n1:17-cv-00982-TWP-MPB (S.D. Ind. Mar. 30, 2017), ECF\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 9\\n\\n\\n\\nApplications under § 505(b)(2) of the Federal Food, Drug,\\nand Cosmetic Act, 21 U.S.C. § 355(b)(2), relying on Lilly’s\\nclinical data for pemetrexed disodium. But each applicant\\nseeks to market different pemetrexed salts—in DRL’s and\\nHospira’s applications, pemetrexed ditromethamine. Both\\nDRL and Hospira represented to the FDA that their choice\\nof the tromethamine cation was immaterial because\\npemetrexed dissociates from its counterion in solution,\\nDRL J.A. 8555–57; Hospira J.A. 124, and tromethamine\\nwas known to be safe for pharmaceutical use, DRL J.A.\\n8555, 8557.\\n Lilly then asserted the ’209 patent against each of\\nthese NDA applicants in the United States District Court\\nfor the Southern District of Indiana. In the DRL case, the\\ndistrict court construed the phrase “administration of\\npemetrexed disodium” to mean “liquid administration of\\npemetrexed disodium,” which “is accomplished by dissolv-\\ning the solid compound pemetrexed disodium into solu-\\ntion.” DRL Summary Judgment Decision, 2017 WL\\n6387316, at *4. The district court denied DRL’s motion for\\nsummary judgment of noninfringement, holding that pros-\\necution history estoppel does not bar Lilly from asserting\\nthat DRL’s proposed pemetrexed ditromethamine product\\nwould infringe through the doctrine of equivalents because\\nthe reason for Lilly’s amendment was to distinguish other\\nantifolates and was therefore only tangential to\\npemetrexed ditromethamine. Id. at *6–7. The district\\ncourt also rejected DRL’s argument that Lilly dedicated\\n\\n\\n\\nNo. 1, but the parties stipulated to be bound by the district\\ncourt’s decision in the DRL case that neither prosecution\\nhistory estoppel nor the disclosure-dedication rule bars\\nLilly’s assertion of infringement through the doctrine of\\nequivalents. Actavis Br. 2. Actavis filed a brief in the DRL\\nappeal as amicus curiae requesting reversal of that portion\\nof the district court’s decision.\\n\\f10 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\npemetrexed ditromethamine to the public under the disclo-\\nsure-dedication rule through its reference to Akimoto’s an-\\ntifolate compounds because Akimoto is not incorporated by\\nreference into the ’209 patent and in any event discloses\\npemetrexed ditromethamine only within a genus of thou-\\nsands of compounds, which the district court held does not\\nconstitute the requisite disclosure of an identifiable alter-\\nnative under this court’s precedent. Id. at *7–8; see, e.g.,\\nSanDisk Corp. v. Kingston Tech. Co., 695 F.3d 1348, 1363\\n(Fed. Cir. 2012).\\n Following a bench trial, the district court’s opinion\\nlargely followed its rationale in the DRL Summary Judg-\\nment Decision with respect to the applicability of prosecu-\\ntion history estoppel and the disclosure-dedication rule.\\nDRL Decision, 323 F. Supp. 3d at 1046–48. In addition, the\\ncourt found that DRL’s proposed product would be admin-\\nistered in a manner that would meet the “administration\\nof pemetrexed disodium” step of the asserted claims under\\nthe doctrine of equivalents, id. at 1049, regardless of the\\n“differences in chemical properties between pemetrexed\\ndisodium and pemetrexed ditromethamine,” id. at 1050.\\n In the Hospira case, the parties similarly disputed the\\ndoctrine of equivalents, but Lilly also asserted literal in-\\nfringement because Hospira’s proposed product label al-\\nlows reconstitution of its pemetrexed ditromethamine salt\\nin saline. Hospira Decision, 2018 WL 3008570, at *2–3;\\nHospira J.A. 229. After the district court issued the DRL\\nSummary Judgment Decision, Hospira conceded, contin-\\ngent upon its right to appeal, that its product would in-\\nfringe under the claim construction of “administration of\\npemetrexed disodium” set forth in that opinion and that its\\ndoctrine of equivalents arguments were likewise fore-\\nclosed. Hospira Br. 18. The district court, “rel[ying] heav-\\nily” on the DRL Summary Judgment Decision, granted\\nLilly’s motion for summary judgment of infringement, both\\nliterally and under the doctrine of equivalents. Hospira\\nDecision, 2018 WL 3008570, at *1 n.2, *6.\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 11\\n\\n\\n\\n These appeals followed. We have jurisdiction under 28\\nU.S.C. § 1295(a)(1).\\n DISCUSSION\\n We review a district court’s grant of summary judg-\\nment according to the law of the regional circuit. Kaneka\\nCorp. v. Xiamen Kingdomway Grp. Co., 790 F.3d 1298,\\n1303 (Fed. Cir. 2015) (citing Halo Elecs., Inc. v. Pulse El-\\necs., Inc., 769 F.3d 1371, 1377 (Fed. Cir. 2014)). In the Sev-\\nenth Circuit, summary judgment is reviewed de novo, con-\\nstruing all facts and drawing all inferences in favor of the\\nnon-movant. Wis. Alumni Research Found. v. Apple Inc.,\\n905 F.3d 1341, 1352 (Fed. Cir. 2018) (citing Austin v.\\nWalgreen Co., 885 F.3d 1085, 1087 (7th Cir. 2018)). On ap-\\npeal from a bench trial, we review a district court’s conclu-\\nsions of law de novo and its findings of fact for clear error.\\nBraintree Labs., Inc. v. Novel Labs., Inc., 749 F.3d 1349,\\n1358 (Fed. Cir. 2014) (citing Brown & Williamson Tobacco\\nCorp. v. Philip Morris Inc., 229 F.3d 1120, 1123 (Fed. Cir.\\n2000)). A factual finding is clearly erroneous if, despite\\nsome supporting evidence, we are left with the definite and\\nfirm conviction that a mistake has been made. United\\nStates v. U.S. Gypsum Co., 333 U.S. 364, 395 (1948).\\n Claim construction is ultimately an issue of law, which\\nwe review de novo. Shire Dev., LLC v. Watson Pharm., Inc.,\\n787 F.3d 1359, 1364 (Fed. Cir. 2015). We review de novo\\nthe district court’s findings of fact on evidence “intrinsic to\\nthe patent (the patent claims and specification[], along\\nwith the patent’s prosecution history),” and review for clear\\nerror extrinsic findings of fact. Teva Pharm. USA, Inc. v.\\nSandoz, Inc., 135 S. Ct. 831, 841 (2015). While infringe-\\nment is a question of fact, Lucent Techs., Inc. v. Gateway,\\nInc., 580 F.3d 1301, 1309 (Fed. Cir. 2009), we review de\\nnovo the district court’s grant of summary judgment of non-\\ninfringement, Unwired Planet, LLC v. Apple Inc., 829 F.3d\\n1353, 1356 (Fed. Cir. 2016). To prove infringement, a pa-\\ntentee “must supply sufficient evidence to prove that the\\n\\f12 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\naccused product or process contains, either literally or un-\\nder the doctrine of equivalents, every limitation of the\\nproperly construed claim.” Seal-Flex, Inc. v. Athletic Track\\n& Court Const., 172 F.3d 836, 842 (Fed. Cir. 1999). The\\npatentee has the burden of proving infringement by a pre-\\nponderance of the evidence. SmithKline Diagnostics, Inc.\\nv. Helena Labs. Corp., 859 F.2d 878, 889 (Fed. Cir. 1988).\\n Hospira requests reversal of the district court’s finding\\nthat its submission of a § 505(b)(2) NDA for its pemetrexed\\nproduct literally infringed the claims of the ’209 patent.\\nDRL and Hospira both argue, as does the amicus curiae\\nActavis, that the district court erred as a matter of law by\\nrefusing to apply prosecution history estoppel to bar Lilly’s\\ndoctrine of equivalents claim, and DRL further contends\\nthat the disclosure-dedication rule precludes Lilly’s equiv-\\nalents claim. Finally, DRL disputes the district court’s\\nfinding that administration of pemetrexed ditromethamine\\nis equivalent to the claim element “administration of\\npemetrexed disodium.” We address each argument in turn.\\n A. Literal Infringement\\n Hospira argues that it cannot literally infringe the\\nclaims of the ’209 patent because intravenous administra-\\ntion of pemetrexed ditromethamine dissolved in saline—a\\nsolution which contains pemetrexed and chloride anions\\nalongside sodium and tromethamine cations—is not “ad-\\nministration of pemetrexed disodium.” Hospira also notes\\nthat such a solution will, in any case, contain far more than\\ntwo sodium cations per pemetrexed anion. Finally, Hos-\\npira appears to make a perfunctory argument that, in the\\nalternative, we should reverse the district court’s construc-\\ntion and hold that the term encompasses any route of ad-\\nministering pemetrexed disodium, not just liquid, as the\\ndistrict court’s construction requires.\\n Lilly counters that Hospira’s view improperly imposes\\na “source limitation,” requiring that the pemetrexed diso-\\ndium salt exist in solid form before administration, even\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 13\\n\\n\\n\\nthough Hospira’s proposed product label, like that of Ali-\\nmta®, calls for administration of a solution containing\\npemetrexed anions and sodium cations. Lilly also contends\\nthat Hospira’s claim construction arguments are irrelevant\\nbecause Hospira’s proposed product will be administered\\nintravenously anyway.\\n We agree with Hospira. It was clearly erroneous for\\nthe district court to hold that the “administration of\\npemetrexed disodium” step was met because Hospira’s\\npemetrexed ditromethamine product will be dissolved in\\nsaline before administration. A solution of pemetrexed and\\nchloride anions and tromethamine and sodium cations can-\\nnot be deemed pemetrexed disodium simply because some\\nassortment of the ions in the solution consists of\\npemetrexed and two sodium cations. As Lilly acknowl-\\nedges throughout its brief, pemetrexed disodium is a salt.\\nSee, e.g., Lilly Br. I 12 (pemetrexed toxicity is caused “by\\npemetrexed itself once dissociated in solution,” not\\npemetrexed disodium); see also Hospira J.A. 1596 (October\\n2017 Alimta® Label referring to the drug substance as the\\n“disodium salt” of pemetrexed). Once diluted, the salt’s\\ncrystalline structure dissolves, and the individual ions dis-\\nsociate. See Hospira J.A. 2820 (declaration of Lilly’s ex-\\npert). In other words, pemetrexed disodium no longer\\nexists once dissolved in solution, and, as a corollary, a dif-\\nferent salt of pemetrexed dissolved in saline is not\\npemetrexed disodium.\\n We conclude that to literally practice the “administra-\\ntion of pemetrexed disodium” step under the district court’s\\nclaim construction, the pemetrexed disodium salt must be\\nitself administered. See DRL Summary Judgment Deci-\\nsion, 2017 WL 6387316, at *4 (“‘[A]dministration of\\npemetrexed disodium’ . . . refer[s] to a liquid administration\\nof pemetrexed disodium. . . ., accomplished by dissolving\\nthe solid compound pemetrexed disodium into solu-\\ntion . . . .”); see also Tex. Instruments Inc. v. Cypress Semi-\\nconductor Corp., 90 F.3d 1558, 1563 (Fed. Cir. 1996) (“To\\n\\f14 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\nliterally infringe, the accused . . . process must contain\\nevery limitation of the asserted claim.” (citing Laitram\\nCorp. v. Rexnord, Inc., 939 F.2d 1533, 1535 (Fed. Cir.\\n1991))). There is no dispute that Hospira has only sought\\napproval to market pemetrexed ditromethamine, Lilly Br.\\nI 4, and that neither its proposed product nor methods of\\nadministering it will constitute administering the\\npemetrexed disodium salt. Accordingly, Hospira will not\\npractice the step of “administration of pemetrexed diso-\\ndium,” and the district court’s finding of literal infringe-\\nment must be reversed.\\n B. Doctrine of Equivalents\\n Few propositions of patent law have been so consist-\\nently sustained by the Supreme Court as the doctrine of\\nequivalents. See Festo Corp. v. Shoketsu Kinzoku Kogyo\\nKabushki Co., 535 U.S. 722, 733 (2002) (“Festo VIII”)\\n(“[E]quivalents remain a firmly entrenched part of the set-\\ntled rights protected by the patent.”); Warner-Jenkinson\\nCo. v. Hilton Davis Chem. Co., 520 U.S. 17, 40 (1997) (“[W]e\\nadhere to the doctrine of equivalents.”); Graver Tank &\\nMfg. Co. v. Linde Air Prods. Co., 339 U.S. 605, 608 (1950)\\n(“Originating almost a century ago in the case of Winans v.\\nDenmead, [56 U.S. 330 (1853)] . . . [the doctrine of equiva-\\nlents] has been consistently applied by this Court and the\\nlower federal courts, and continues today ready and avail-\\nable for utilization when the proper circumstances for its\\napplication arise.”). It is settled that a patentee is entitled\\n“in all cases to invoke to some extent the doctrine of equiv-\\nalents,” Seymour v. Osborne, 78 U.S. 516, 555 (1870), with-\\nout a “judicial exploration of the equities of a case”\\nbeforehand. See Warner-Jenkinson, 520 U.S. at 34.\\n Yet the Supreme Court has also acknowledged that the\\ndoctrine of equivalents, “when applied broadly, conflicts\\nwith the definitional and public-notice functions of the stat-\\nutory claiming requirement,” Warner-Jenkinson, 520 U.S.\\nat 29, and that, without the proper balance between these\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 15\\n\\n\\n\\ntwo imperatives, the doctrine may “take[] on a life of its\\nown, unbounded by the patent claims.” See id. at 28–29.\\nWe have emphasized, moreover, that the doctrine of equiv-\\nalents is “the exception, however, not the rule,” and not\\nmerely “the second prong of every infringement charge,\\nregularly available to extend protection beyond the scope\\nof the claims.” London v. Carson Pirie Scott & Co., 946 F.2d\\n1534, 1538 (Fed. Cir. 1991). Patent infringement is princi-\\npally determined by examining whether the accused sub-\\nject matter falls within the scope of the claims.\\n To that end, courts have placed important limitations\\non a patentee’s ability to assert infringement under the\\ndoctrine of equivalents. See, e.g., Festo VIII, 535 U.S. at\\n737–41 (prosecution history estoppel); Warner-Jenkinson,\\n520 U.S. at 39 n.8 (“[A] theory of equivalence [cannot] en-\\ntirely vitiate a particular claim element . . . .”); Graver\\nTank, 339 U.S. at 608 (accused equivalent cannot differ\\nsubstantially from the claimed invention); Johnson &\\nJohnston Assocs. Inc. v. R.E. Serv. Co., 285 F.3d 1046, 1054\\n(Fed. Cir. 2002) (en banc) (subject matter disclosed but not\\nclaimed is dedicated to the public) (citing Maxwell v. J.\\nBaker, Inc., 86 F.3d 1098 (Fed. Cir. 1996)); Wilson Sporting\\nGoods Co. v. David Geoffrey & Assocs., 904 F.2d 677, 683\\n(Fed. Cir. 1990) (“[T]he asserted scope of equivalency [can-\\nnot] encompass the prior art . . . .” (Rich, J.) (citations omit-\\nted)). These appeals implicate several of these limitations.\\n 1. Prosecution History Estoppel\\n The main dispute in these appeals is whether Lilly has\\nrebutted the presumption of prosecution history estoppel\\nthat attached to its amendment in the ’821 application.\\nProsecution history estoppel arises when a patent appli-\\ncant narrows the scope of his claims during prosecution for\\na reason “substantial[ly] relating to patentability.” See\\ngenerally Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki\\nCo., 344 F.3d 1359, 1366–67 (Fed. Cir. 2003) (en banc)\\n(“Festo X”). Such a narrowing amendment is presumed to\\n\\f16 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\nbe a surrender of all equivalents within “the territory be-\\ntween the original claim and the amended claim,” but the\\npresumption is overcome if the patentee can show the ap-\\nplicability of one of the few exceptions identified by the Su-\\npreme Court. Festo VIII, 535 U.S. at 740–41 (citing Exhibit\\nSupply Co. v. Ace Patents Corp., 315 U.S. 126, 136–37\\n(1942)). Whether prosecution history estoppel applies to\\nbar a doctrine of equivalents claim is a question of law, re-\\nviewed de novo. See Regents of Univ. of Cal. v. Dakocyto-\\nmation Cal., Inc., 517 F.3d 1364, 1371 (Fed. Cir. 2008)\\n(citing Pharmacia & Upjohn Co. v. Mylan Pharm., Inc., 170\\nF.3d 1373, 1376 (Fed. Cir. 1999)).\\n Lilly does not dispute that the amendment in question\\nwas both narrowing and made for a substantial reason re-\\nlating to patentability. Lilly Br. II 21. Furthermore, Lilly\\nrelies on only one exception to giving effect to the presump-\\ntion as to the scope of surrender: that the rationale of its\\namendment “[bore] no more than a tangential relation to\\nthe equivalent in question.” Festo VIII, 535 U.S. at 740. As\\na result, the parties’ dispute about whether prosecution\\nhistory estoppel applies is confined to whether Lilly’s\\namendment narrowing “an antifolate” to “pemetrexed diso-\\ndium” was only tangential to pemetrexed ditromethamine,\\nwhich is the accused compound. Whether the tangential\\nexception applies is a question of law, Integrated Tech.\\nCorp. v. Rudolph Techs., Inc., 734 F.3d 1352, 1356 (Fed.\\nCir. 2013), and a patentee seeking to use the exception\\n“must base his arguments solely upon the public record of\\nthe patent’s prosecution.” Festo X, 344 F.3d at 1369–70 (ci-\\ntation omitted).\\n The Appellants argue that Lilly failed to explain why\\nit did not pursue a narrower amendment literally encom-\\npassing pemetrexed ditromethamine, and they emphasize\\nour statement that the tangential exception is “very nar-\\nrow.” Integrated, 734 F.3d at 1358 (quoting Cross Med.\\nProds., Inc. v. Medtronic Sofamor Danek, Inc., 480 F.3d\\n1335, 1342 (Fed. Cir. 2007)). The Appellants further point\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 17\\n\\n\\n\\nout that Lilly cannot be said to have “lacked the words to\\ndescribe” pemetrexed ditromethamine, see Festo VIII, 535\\nU.S. at 734, because Lilly’s previous patents, as well as the\\nEuropean companion to the ’209 patent, claimed\\npemetrexed salts generally and pemetrexed disodium in a\\ndependent claim. They also assert that the district court\\nerred by focusing on whether Lilly actually needed to relin-\\nquish pemetrexed ditromethamine to overcome the Arsen-\\nyan anticipation rejection because “the tangential\\nexception is not a patentee’s-buyer’s-remorse exception.”\\nDRL Br. 39.\\n In response, Lilly argues that the district court\\nproperly held that the reason for its amendment was to dis-\\ntinguish pemetrexed from antifolates generally and that\\nthe different salt type is a merely tangential change with\\nno consequence for pemetrexed’s administration or mecha-\\nnism of action within the body. Lilly also contends that it\\nis not barred from asserting the tangential exception\\nsimply because pemetrexed ditromethamine is within “the\\nterritory between the original claim and the amended\\nclaim.” Festo VIII, 535 U.S. at 740. Finally, Lilly argues\\nthat Appellants’ view that courts must “consider hypothet-\\nical alternative amendments” that would literally encom-\\npass the alleged equivalent “would eviscerate the\\ntangentiality exception.” Lilly Br. II 44.\\n We agree with Lilly. As a general matter, we find Ap-\\npellants’ view of prosecution history estoppel, and the tan-\\ngential exception in particular, too rigid. Tangential\\nmeans “touching lightly or in the most tenuous way.” Web-\\nster’s Third New International Dictionary (2002). The rea-\\nson for Lilly’s amendment, as the district court concluded,\\nwas to narrow original claim 2 to avoid Arsenyan, which\\nonly discloses treatments using methotrexate, a different\\nantifolate. See DRL J.A. 7879–80 (overcoming the Arsen-\\nyan anticipation rejection by arguing that it “does not dis-\\nclose pemetrexed disodium”). To overcome a clear\\nanticipation, Lilly opted to narrow its original claim 2 and\\n\\f18 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\nits dependents to more accurately define what it actually\\ninvented, an improved method of administering\\npemetrexed. In other words, the particular type of salt to\\nwhich pemetrexed is complexed relates only tenuously to\\nthe reason for the narrowing amendment, which was to\\navoid Arsenyan. We therefore hold that Lilly’s amendment\\nwas merely tangential to pemetrexed ditromethamine be-\\ncause the prosecution history, in view of the ’209 patent it-\\nself, strongly indicates that the reason for the amendment\\nwas not to cede other, functionally identical, pemetrexed\\nsalts.\\n The prosecution record confirms our understanding.\\nOriginal claim 5, which, like all the current claims of the\\n’209 patent, required supplementation with both vitamin\\nB12 and folic acid, was never rejected as anticipated over\\nArsenyan. Instead, the art cited against original claim 5\\nand its dependent claims in the obviousness ground of re-\\njection was replete with information about pemetrexed\\ndisodium; John disclosed clinical trials using pemetrexed\\ndisodium, reporting both its efficacy and its toxic side ef-\\nfects, and in response, DRL J.A. 7869–70, Worzalla sug-\\ngested folic acid supplementation to counteract these side\\neffects, DRL J.A. 7870–71. The prosecution record implies\\nthat Lilly’s amendment, inartful though it might have\\nbeen, was prudential in nature and did not need or intend\\nto cede other pemetrexed salts.\\n Hospira argues that the amendment was made to over-\\ncome the obviousness rejection over Ohmori and John and\\nthat Lilly has provided no reason for the amendment rela-\\ntive to that rejection. Like Lilly, we find this argument\\nmakes little sense. John discloses the results of a clinical\\ntrial of pemetrexed disodium and explicitly suggests the\\ntoxicities caused by pemetrexed; as we concluded above,\\nnarrowing “antifolate” to “pemetrexed disodium” could not\\npossibly distinguish the art cited in the obviousness ground\\nof rejection.\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 19\\n\\n\\n\\n DRL also insists that we have held that an applicant’s\\nremorse at ceding more claim scope than necessary is not a\\nreason for the tangential exception to apply. See, e.g., Lu-\\ncent Techs., Inc. v. Gateway, Inc., 525 F.3d 1200, 1218 (Fed.\\nCir. 2008); Schwarz Pharma, Inc. v. Paddock Labs., Inc.,\\n504 F.3d 1371, 1377 (Fed. Cir. 2007). This is generally\\ntrue, but DRL overreads the holdings of these cases. After\\nall, the tangential exception only exists because applicants\\nover-narrow their claims during prosecution. Amend-\\nments are not construed to cede only that which is neces-\\nsary to overcome the prior art, see Schwarz, 504 F.3d at\\n1377, nor will the court “speculat[e]” whether an amend-\\nment was necessary, see Kinzenbaw v. Deere & Co., 741\\nF.2d 383, 389 (Fed. Cir. 1984). But the reason for an\\namendment, where the tangential exception is invoked,\\ncannot be determined without reference to the context in\\nwhich it was made, including the prior art that might have\\ngiven rise to the amendment in the first place. See Festo X,\\n344 F.3d at 1370. Here, it is unlikely that a competitor\\nwould have been “justified in assuming that if he [made an\\nequivalent pemetrexed salt], he would not infringe [the\\n’209 patent].” Kinzenbaw, 741 F.2d at 389; cf. Festo VIII,\\n535 U.S. at 738 (“There is no reason why a narrowing\\namendment should be deemed to relinquish equivalents\\n. . . beyond a fair interpretation of what was surrendered.”).\\n Furthermore, Appellants’ suggestion that Lilly must\\nprove that it could not have drafted a claim that literally\\nencompassed pemetrexed ditromethamine is unsupported\\nby our precedent on prosecution history estoppel, not to\\nmention excessive. We do not demand perfection from pa-\\ntent prosecutors, and neither does the Supreme Court. See\\nFesto VIII, 535 U.S. at 738 (“It does not follow . . . that [an]\\namended claim becomes so perfect in its description that\\nno one could devise an equivalent.”). Lilly’s burden was to\\nshow that pemetrexed ditromethamine was “peripheral, or\\nnot directly relevant,” to its amendment, Festo X, 344 F.3d\\nat 1369. And as we concluded above, Lilly has done so.\\n\\f20 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\n In addition, the Appellants maintain that when a pa-\\ntentee submits an amendment adding two claim limita-\\ntions, it cannot later argue that the reason for the\\namendment was tangential to an accused equivalent con-\\ntaining only one of the added limitations simply because\\nthe second limitation was unnecessary to overcome the\\nprior art. They offer Felix v. American Honda Motor Co.,\\n562 F.3d 1167 (Fed. Cir. 2009), as an illustration of this\\nprinciple. 5 In that case, we held that prosecution history\\nestoppel applied to a claim directed to a vehicle bed storage\\nsystem—limited in response to a rejection to having a\\nchannel with a flange and a gasket mounted on that\\nflange—barring assertion of equivalence with respect to a\\nproduct that met the channel aspect, but not the gasket as-\\npect, of the limitation. Id. at 1184–85.\\n But as Lilly points out, this holding was determined by\\nthat patent’s prosecution history, Felix, 562 F.3d at 1184,\\nand we have also held that prosecution history estoppel\\ndoes not apply in similar circumstances, where the\\n\\n\\n\\n 5 The parties argue at length about which of our\\ncases are properly analogous to the facts presented in these\\nappeals. Here, in applying the Supreme Court’s frame-\\nwork, we find the analogies to other cases less helpful than\\na direct consideration of the specific record of this case and\\nwhat it shows about the reason for amendment and the re-\\nlation of that reason to the asserted equivalent. This case-\\nspecific focus, within the governing framework, comports\\nwith the equitable nature of prosecution history estoppel.\\nSee Festo VIII, 535 U.S. at 738 (“[The Supreme Court has]\\nconsistently applied the doctrine in a flexible way, not a\\nrigid one.”); cf. Heckler v. Cmty. Health Servs. of Crawford\\nCty., Inc., 467 U.S. 51, 59 (1984) (“Estoppel is an equitable\\ndoctrine invoked to avoid injustice in particular cases. . . .\\n[and] a hallmark of the doctrine is its flexible applica-\\ntion . . . .”).\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 21\\n\\n\\n\\nprosecution record differed. See, e.g., Regents, 517 F.3d at\\n1376–78 (amendment narrowing “disabling hybridization\\ncapacity of [nucleic acid] sequences” to methods using a\\n“blocking nucleic acid” was merely tangential to unclaimed\\nrepetitive sequence nucleic acids); Insituform Techs., Inc.\\nv. CAT Contracting, Inc., 385 F.3d 1360, 1368 (Fed. Cir.\\n2004) (amendment narrowing method of inserting resin\\ninto tube using a vacuum to one using “a cup” to do so was\\nmerely tangential to a multiple cup embodiment because\\nthe number of cups bore no relationship to the cited prior\\nart or the rationale behind the narrowing amendment).\\nThus, our cases demonstrate that prosecution history es-\\ntoppel is resistant to the rigid legal formulae that Appel-\\nlants seek to extract from them. See Intervet Inc. v. Merial\\nLtd., 617 F.3d 1282, 1291 (Fed. Cir. 2010) (“[T]here is no\\nhard-and-fast test for what is and what is not a tangential\\nrelation . . . .”).\\n Finally, DRL also contends that our precedent squarely\\nforecloses Lilly’s tangentiality argument, and it invites us\\nto read those cases to hold that “where the reason for the\\namendment and the equivalent in question both relate to\\nthe same claim element, the tangential exception does not\\napply.” DRL Br. 47. We decline this invitation because\\nsuch a bright-line rule is both contrary to the equitable na-\\nture of prosecution history estoppel, as articulated in Festo\\nVIII, 535 U.S. at 738, and inconsistent with the equitable\\nspirit that animates the doctrine of equivalents, see Graver\\nTank, 339 U.S. at 608–09 (the doctrine is one of “whole-\\nsome realism”). Instead, we reaffirm that whether an\\namendment was merely tangential to an equivalent must\\nbe decided in the context of the invention disclosed in the\\npatent and the prosecution history. Festo X, 344 F.3d at\\n1370.\\n DRL’s intuition—that an amendment that narrows an\\nexisting claim element evinces an intention to relinquish\\nthat claim scope—is often correct. Indeed, as we have\\nfound in previous cases, it is a powerful indication that an\\n\\f22 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\namendment was not merely tangential. See, e.g., Honey-\\nwell Int’l, Inc. v. Hamilton Sundstrand Corp., 523 F.3d\\n1304, 1315–16 (Fed. Cir. 2008); Biagro W. Sales, Inc. v.\\nGrow More, Inc., 423 F.3d 1296, 1306 (Fed. Cir. 2005). But\\nhere, we conclude that this consideration is not dispositive\\nbecause the rest of the prosecution history, and the ’209\\npatent itself, show that it is implausible that the reason for\\nLilly’s amendment was to surrender other pemetrexed\\nsalts. Indeed, such a relinquishment would effectively ded-\\nicate the entirety of Lilly’s invention to the public and\\nthereby render the ’209 patent worthless, and it would\\nhave been irrelevant for distinguishing the prior art.\\nAgain, the prosecution history strongly indicates a less\\nsweeping and more sensible reason for Lilly’s amendment:\\nto surrender antifolates other than pemetrexed. Thus, we\\nconclude on this prosecution record that Lilly’s amendment\\nwas merely tangential to pemetrexed ditromethamine.\\n 2. Disclosure-Dedication Rule\\n DRL next argues that the disclosure-dedication rule\\nbars Lilly from asserting infringement under the doctrine\\nof equivalents. The ’209 patent sets forth its invention as\\nan improved method of administering antifolates, ’209 pa-\\ntent col. 2 ll. 47–58, and teaches that the derivatives de-\\nscribed in the Akimoto patent are preferred examples of\\nantifolates, id. col. 4 ll. 34–40. DRL contends that one of\\nthese derivatives is pemetrexed ditromethamine and that\\nit was dedicated to the public when Lilly declined to claim\\nit. DRL asserts that the district court erred because it both\\nrequired express incorporation of Akimoto by reference\\ninto the ’209 patent and concluded that Akimoto does not\\nspecifically disclose pemetrexed ditromethamine.\\n Lilly counters that the disclosure-dedication rule re-\\nquires express disclosure of the subject matter in question\\nin the specification except in narrow circumstances, such\\nas when that subject matter is disclosed in a priority appli-\\ncation, see Abbott Labs. v. Sandoz, Inc., 566 F.3d 1282,\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 23\\n\\n\\n\\n1297 (Fed. Cir. 2009), or prior art expressly incorporated\\nby reference, SanDisk, 695 F.3d at 1366. Lilly also argues\\nthat the district court correctly determined that the rele-\\nvant portion of Akimoto discloses only a generic formula\\nfrom which a skilled artisan would not be able to recognize\\npemetrexed ditromethamine.\\n We agree with Lilly and hold that the disclosure-dedi-\\ncation rule is inapplicable to this case because the ’209 pa-\\ntent does not disclose methods of treatment using\\npemetrexed ditromethamine, and, as a result, Lilly could\\nnot have dedicated such a method to the public.\\n Under the disclosure-dedication rule, subject matter\\ndisclosed by a patentee, but not claimed, is considered ded-\\nicated to the public. See Johnson & Johnston, 285 F.3d at\\n1054. The reason for the doctrine is that members of the\\npublic reading a disclosure of particular subject matter are\\nentitled, absent a claim to it, to assume that it is not pa-\\ntented and therefore dedicated to the public (unless, for ex-\\nample, claimed in a continuation or other application based\\non the disclosure). Cf. Maxwell, 86 F.3d at 1107 (failure to\\nclaim inventive subject matter “is clearly contrary to 35\\nU.S.C. § 112, which requires that a patent applicant ‘par-\\nticularly point[] out and distinctly claim[] the subject mat-\\nter which the applicant regards as his invention’”). Subject\\nmatter is considered disclosed when a skilled artisan “can\\nunderstand the unclaimed disclosed teaching upon reading\\nthe written description,” but not “any generic reference . . .\\nnecessarily dedicates all members of that particular ge-\\nnus.” PSC Comput. Prod., Inc. v. Foxconn Int’l, Inc., 355\\nF.3d 1353, 1360 (Fed. Cir. 2004).\\n DRL further contends that the disclosure-dedication\\nrule does not impose a § 112 requirement for sufficiency of\\ndisclosure, see Toro Co. v. White Consol. Indus., Inc., 383\\nF.3d 1326, 1334 (Fed. Cir. 2004), and that a skilled artisan\\nreading the ’209 patent would both look for a disclosure of\\npemetrexed in Akimoto, and also seek to use a well-known\\n\\f24 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\ncation like tromethamine, which it maintains is generically\\ndisclosed in Akimoto in the form of “substituted ammo-\\nnium” base salts.\\n We are unpersuaded by DRL’s arguments. As the dis-\\ntrict court noted, Akimoto’s formula, col. 1 l. 49–col. 2 l. 3,\\nincludes seven functional group variables and encompasses\\nthousands of compounds, and while Akimoto discloses\\nabout fifty exemplary compounds, none of them is\\npemetrexed. Moreover, Akimoto does not even disclose tro-\\nmethamine expressly but only generically among dozens of\\nother salts. At most, Akimoto discloses ammonium salts\\ngenerally, which is far from a description of tromethamine.\\nIn similar circumstances, we have held that “sufficient de-\\nscription of a genus” requires that a skilled artisan be able\\nto “‘visualize or recognize’ the members of the genus.” See\\nAriad Pharm., Inc. v. Eli Lilly & Co., 598 F.3d 1336, 1350\\n(Fed. Cir. 2010) (quoting Regents of the Univ. of Cal. v. Eli\\nLilly & Co., 119 F.3d 1559, 1568–69 (Fed. Cir. 1997)).\\nAkimoto does not so describe pemetrexed ditromethamine,\\nand we see no reason why a skilled artisan would set out\\non DRL’s winding path to cobble together pemetrexed dit-\\nromethamine. While the ’209 patent teaches that\\npemetrexed disodium is the “most preferred” antifolate,\\nthat knowledge would not change the skilled artisan’s un-\\nderstanding of what Akimoto discloses.\\n Because Akimoto contains only a “generic reference” to\\npemetrexed ditromethamine, PSC Comput., 355 F.3d at\\n1360, we conclude that it was not dedicated to the public.\\n 3. Merits\\n A component in an accused product or process may be\\nequivalent to a claim element if the two are insubstantially\\ndifferent with respect to the “role played by [the] element\\nin the context of the specific patent claim.” Warner-Jen-\\nkinson, 520 U.S. at 39–40. Relevant differences can in-\\nclude the function each serves, the way in which each\\nworks, and the result each obtains, id. at 39, and, especially\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 25\\n\\n\\n\\nin biochemical cases, structural or pharmacological charac-\\nteristics, Mylan Inst. LLC v. Aurobindo Pharm. Ltd., 857\\nF.3d 858, 869 (Fed. Cir. 2017). “The determination of\\nequivalency vel non is a question of fact,” Canton Bio Med.,\\nInc. v. Integrated Liner Techs., Inc., 216 F.3d 1367, 1369\\n(Fed. Cir. 2000) (citing Pall Corp. v. Micron Separations,\\nInc., 66 F.3d 1211, 1218 (Fed. Cir. 1995)), which we review\\nfor clear error in an appeal from a bench trial, Pfizer, Inc.\\nv. Apotex, Inc., 480 F.3d 1348, 1359 (Fed. Cir. 2007).\\n DRL argues that the district court erred in finding that\\nits proposed pemetrexed ditromethamine product will be\\nadministered in an insubstantially different way from the\\nclaimed method. DRL maintains that the district court fo-\\ncused on the fact that each product treats the same dis-\\neases by delivering pemetrexed intravenously, when the\\nrelevant context is the manner of administration. In DRL’s\\nview, the chemical differences between sodium and tro-\\nmethamine—e.g., pH, buffering capacity, or solubility—\\nDRL Br. 20–21, render the methods in which each is ad-\\nministered to a patient substantially different.\\n Lilly responds that the relevant context is treatment of\\na patient “in need of chemotherapeutic treatment.” ’209\\npatent claim 12. Lilly agrees with the district court that\\nthe chemical differences between sodium and trometham-\\nine are clinically irrelevant because each undisputedly\\nlacks therapeutic activity.\\n We see no clear error in the district court’s findings. As\\nthe district court found, DRL’s product will accomplish an\\nidentical aim, furnishing the same amount of pemetrexed\\nto active sites in the body; in exactly the same way, by di-\\nluting a pemetrexed salt in an aqueous solution for intra-\\nvenous administration. Indeed, after dilution and\\nimmediately before administration, DRL’s product is func-\\ntionally identical to Lilly’s in that it contains the same\\namount of diluted pemetrexed anion. DRL J.A. 8557. And\\nDRL declines to identify the relevance of any of the\\n\\f26 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\nchemical differences it identifies. See UCB, Inc. v. Watson\\nLabs. Inc., 927 F.3d 1272, 1284–86 (Fed. Cir. 2019) (chem-\\nical differences may not be relevant if the equivalent has\\nknown interchangeability in the context of the claimed\\ncomposition). We find DRL’s arguments unconvincing and\\ntherefore affirm the district court’s findings.\\n In summary, these cases are eminently suitable for ap-\\nplication of the doctrine of equivalents, and we conclude\\nthat neither prosecution history estoppel nor the disclo-\\nsure-dedication rule bars Lilly from asserting infringement\\nthrough equivalence.\\n CONCLUSION\\n We have fully considered each party’s further argu-\\nments but find them unpersuasive. For the foregoing rea-\\nsons, we reverse the district court’s finding of literal\\ninfringement in the Hospira Decision but affirm its judg-\\nment of infringement under the doctrine of equivalents.\\nThe judgment of infringement under the doctrine of equiv-\\nalents in the DRL Decision is likewise affirmed.\\nAFFIRMED-IN-PART AND REVERSED-IN-PART IN\\n APPEAL NOS. 2018-2126, 2018-2127\\n AFFIRMED IN APPEAL NO. 2018-2128\\n COSTS\\n Each party shall bear its own costs.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4424348/\", \"author_raw\": \"LOURIE, Circuit Judge.\"}]}"},"judge1":{"kind":"string","value":"LOURIE"},"judge2":{"kind":"string","value":"MOORE"},"judge3":{"kind":"string","value":"TARANTO"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4647095/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: ELI LILLY AND COMPANY, Plaintiff-Appellee v. HOSPIRA, INC., Defendant-Appellant Eli Lilly and Company, Plaintiff-Appellee v. Dr. Reddy's Laboratories, Ltd., Dr. Reddy's Laboratories, Inc., Defendants-Appellants\nCourt: U.S. Court of Appeals for the Federal Circuit\nDate decided: 2019-08-09\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n for the Federal Circuit\n ______________________\n\n ELI LILLY AND COMPANY,\n Plaintiff-Appellee\n\n v.\n\n HOSPIRA, INC.,\n Defendant-Appellant\n ______________________\n\n 2018-2126, 2018-2127\n ______________________\n\n Appeals from the United States District Court for the\nSouthern District of Indiana in No. 1:16-cv-03460-TWP-\nMPB, Judge Tanya Walton Pratt.\n --------------------------------------------------\n ELI LILLY AND COMPANY,\n Plaintiff-Appellee\n\n v.\n\n DR. REDDY'S LABORATORIES, LTD., DR.\n REDDY'S LABORATORIES, INC.,\n Defendants-Appellants\n ______________________\n\n 2018-2128\n ______________________\n\f2 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\n Appeal from the United States District Court for the\nSouthern District of Indiana in No. 1:16-cv-00308-TWP-\nMPB, Judge Tanya Walton Pratt.\n ______________________\n\n Decided: August 9, 2019\n ______________________\n\n ADAM LAWRENCE PERLMAN, Williams & Connolly LLP,\nWashington, DC, argued for plaintiff-appellee in 2018-2126\nand 2018-2128. Also represented by GALINA I. FOMENKOVA,\nDOV PHILIP GROSSMAN, DAVID M. KRINSKY, ANDREW P.\nLEMENS, CHARLES MCCLOUD; JAMES PATRICK LEEDS, Eli\nLilly and Company, Indianapolis, IN.\n\n BRADFORD PETER LYERLA, Jenner & Block LLP, Chi-\ncago, IL, argued for defendant-appellant in 2018-2126.\nAlso represented by YUSUF ESAT, SARA TONNIES HORTON;\nADAM G. UNIKOWSKY, Washington, DC.\n\n JOHN C. O'QUINN, Kirkland & Ellis LLP, Washington,\nDC, argued for defendants-appellants in 2018-2128. Also\nrepresented by WILLIAM H. BURGESS, CALVIN ALEXANDER\nSHANK; JEFFERY B. ARNOLD, Holland & Knight LLP, At-\nlanta, GA; MERRI C. MOKEN, CHARLES A. WEISS, ERIC H.\nYECIES, New York, NY.\n\n BRIAN TIMOTHY BURGESS, Goodwin Procter LLP, Wash-\nington, DC, for amicus curiae Actavis LLC in 2018-2128.\nAlso represented by EDWINA CLARKE, EMILY L. RAPALINO,\nDARYL L. WIESEN, Boston, MA; LINNEA P. CIPRIANO, New\nYork, NY.\n ______________________\n\n\n Before LOURIE, MOORE, and TARANTO, Circuit Judges.\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 3\n\n\n\nLOURIE, Circuit Judge.\n Hospira Inc. (“Hospira”), Dr. Reddy’s Laboratories\nLtd., and Dr. Reddy’s Laboratories Inc. (collectively,\n“DRL”) appeal from two judgments of the United States\nDistrict Court for the Southern District of Indiana in two\ninfringement suits brought by Eli Lilly & Company\n(“Lilly”) under the Hatch-Waxman Act, 21 U.S.C. § 355.\nThe district court held in each case that the defendant’s\nsubmission of a New Drug Application pursuant to 21\nU.S.C. § 355(b)(2) infringed U.S. Patent 7,772,209 (the\n“’209 patent”) under 35 U.S.C. § 271(e)(2). See Eli Lilly &\nCo. v. Hospira, Inc., No. 1:16-cv-03460-TWP-MPB, 2018\nWL 3008570 (S.D. Ind. June 15, 2018) (“Hospira Decision”);\nEli Lilly & Co. v. Dr. Reddy’s Labs., Ltd., 323 F. Supp. 3d\n1042 (S.D. Ind. 2018) (“DRL Decision”); see also Eli Lilly &\nCo. v. Dr. Reddy’s Labs., Ltd., No. 1:16-cv-00308-TWP-\nMPB, 2017 WL 6387316 (S.D. Ind. Dec. 14, 2017) (“DRL\nSummary Judgment Decision”). Accordingly, the district\ncourt entered orders under 35 U.S.C. § 271(e)(4)(A) prohib-\niting FDA approval of the products at issue until the expi-\nration of the ’209 patent. Eli Lilly & Co. v. Hospira, Inc.,\nNo. 1:16-cv-03460-TWP-MPB (S.D. Ind. June 27, 2018),\nECF No. 94; Eli Lilly & Co. v. Dr. Reddy’s Labs., Ltd., No.\n1:16-cv-00308-TWP-MPB, 2018 WL 3616715 (S.D. Ind.\nJuly 27, 2018). We decide these appeals together in this\ncombined opinion. 1\n We reverse the district court’s finding of literal in-\nfringement in the Hospira Decision as clearly erroneous in\nlight of the court’s claim construction of “administration of\npemetrexed disodium.” Because the district court did not\nerr in its application of the doctrine of equivalents in either\n\n\n 1 We refer to the joint appendices in these appeals by\nreference to each appellant. Lilly’s brief in the Hospira ap-\npeal is referred to as “Lilly Br. I” and its brief in the DRL\nappeal as “Lilly Br. II.”\n\f4 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\ndecision, we affirm both judgments of infringement. Thus,\nthe Hospira Decision is affirmed-in-part and reversed-in-\npart, and the DRL Decision is affirmed.\n BACKGROUND\n Lilly markets the compound pemetrexed in the form of\na disodium salt as Alimta®, which is indicated, both alone\nand in combination with other active agents, for treating\ncertain types of non-small cell lung cancer and mesotheli-\noma. Pemetrexed is an antifolate, a class of molecules\nwhich, at the time of the invention in 2001, was “one of the\nmost thoroughly studied classes of antineoplastic agents.”\n’209 patent col. 1 ll. 19–20. Antifolates are structurally\nsimilar to folic acid and work by competitively binding to\ncertain enzymes that use folic acid metabolites as cofactors\nin several steps of de novo nucleotide synthesis. Id. col. 1\nll. 40–41. Unlike folic acid, antifolates do not enable these\nsynthetic steps, but instead inhibit them. Pemetrexed in-\nhibits several of these enzymes, including thymidylate syn-\nthase, which methylates deoxyuridine in the final step of\ndeoxythymidine synthesis. Id. col. 1 ll. 59–61. By inhibit-\ning the creation of these nucleotides, antifolates slow down\nDNA and RNA synthesis, and with it, cell growth and divi-\nsion. Cancer cells tend to grow rapidly, so antifolate ther-\napy affects them disproportionately, but healthy cells can\nalso be damaged.\n Pemetrexed had been known for at least a decade in\n2001. Lilly’s U.S. Patent 5,344,932 (“Taylor”) disclosed\nthat certain glutamic acid derivatives with pyrrolo[2,3-\nd]pyrimidine heterocyclic ring structures, exemplified by\npemetrexed, are “particularly active . . . inhibitors of thy-\nmidylate synth[ase],” Taylor col. 1 ll. 59–60; see also id. col.\n19 l. 37–col. 20 l. 25 (disclosing data indicating that\npemetrexed inhibits thymidylate synthase activity in vitro\nin human cell lines and in vivo in mice). The Taylor patent\nalso disclosed that its compounds could be employed as\n“pharmaceutically acceptable salt[s],” id. col. 2 l. 35, and\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 5\n\n\n\nthat the disodium salt form was particularly advantageous,\nid. col. 2 ll. 47–48. U.S. Patent 4,997,838 (“Akimoto”), to\nwhich Lilly took a license, disclosed a large genus of com-\npounds containing pyrrolo[2,3-d]pyrimidine heterocyclic\nring structures and a glutamic acid functional group, and\nthat encompassed pemetrexed. The Akimoto patent dis-\ncloses nearly fifty exemplary compounds, col. 14 l. 61–col.\n16 l. 48, none of which is pemetrexed. Akimoto further dis-\ncloses that its compounds may be prepared as salts of\n“pharmaceutically acceptable bases,” such as “alkali met-\nals, alkali earth metals, non-toxic metals, ammonium, and\nsubstituted ammonium.” Id. col. 14 ll. 44–47.\n By 2001, Lilly had also published the results of several\nclinical trials investigating the use of pemetrexed disodium\nas a treatment for different types of cancer. See, e.g., W.\nJohn et al., “Activity of Multitargeted Antifolate\n(Pemetrexed Disodium, LY231514) in Patients with Ad-\nvanced Colorectal Carcinoma: Results from a Phase II\nStudy,” Cancer, 88(8):1807–13 (2000). In the course of con-\nducting these studies, Lilly discovered that pemetrexed\ndisodium caused severe hematologic and immunologic side\neffects, resulting in infections, nausea, rashes, and even\nsome deaths. See id.; see also Neptune Generics, LLC v. Eli\nLilly & Co., 921 F.3d 1372, 1377–78 (Fed. Cir. 2019) (dis-\ncussing Lilly’s response to adverse clinical data), and Nep-\ntune Generics, LLC v. Eli Lilly & Co., No. IPR2016-00240,\n2017 WL 4466557, at *28–30 (P.T.A.B. Oct. 5, 2017) (same).\nAs the ’209 patent teaches, such side effects are not uncom-\nmon among antifolates. See ’209 patent col. 1 ll. 11–14.\nSome researchers hypothesized that folic acid deficiency\ncaused these side effects and suggested supplementing\npemetrexed disodium treatment with folic acid. DRL J.A.\n7870 (citing J.F. Worzalla et al., “Role of Folic Acid in Mod-\nulating the Toxicity and Efficacy of the Multitargeted An-\ntifolate, LY231514,” Anticancer Research, 18:3235–40\n(1998)).\n\f6 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\n The invention of the ’209 patent is an improved method\nof treatment with antifolates, particularly pemetrexed\ndisodium, through supplementation with a methylmalonic\nacid lowering agent and folic acid. Doing so, according to\nthe patent, lessens antifolate toxicity without sacrificing ef-\nficacy. See ’209 patent col. 10 ll. 17–53 (reporting that pre-\nsupplementation regimen of vitamin B12 and folic acid in\nclinical studies substantially reduced pemetrexed-induced\ntoxicity and deaths while delivering a superior chemother-\napeutic response rate). The ’209 patent lists preferred an-\ntifolates, including some then-existing antifolate therapies,\nas well as “derivatives described in” several patents includ-\ning the Akimoto patent, and “most preferred, Pemetrexed\nDisodium.” Id. col. 4 ll. 28–43. Each of the claims of the\n’209 patent requires administration of pemetrexed diso-\ndium following administration of folic acid and a\nmethylmalonic acid lowering agent, specified in some\nclaims, as well as the Alimta® label, as vitamin B12. Claim\n12 is representative2:\n 12. An improved method for administering\n pemetrexed disodium to a patient in need of chemo-\n therapeutic treatment, wherein the improvement\n comprises:\n a) administration of between about 350 µg\n and about 1000 µg of folic acid prior to the\n first administration of pemetrexed diso-\n dium;\n\n\n\n\n 2 The district court treated claim 12 as representa-\ntive, DRL Summary Judgment Decision, 2017 WL\n6387316, at *1–2; Hospira Decision, 2018 WL 3008570, at\n*2, and no party has disputed that determination on ap-\npeal. See, e.g., DRL Opening Br. 8–9; Hospira Opening Br.\n23.\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 7\n\n\n\n b) administration of about 500 µg to about\n 1500 µg of vitamin B12, prior to the first\n administration of pemetrexed disodium;\n and\n c) administration of pemetrexed disodium.\n In a parent application, Application 10/297,821 (the\n“’821 application”), Lilly originally sought broad claims to\nmethods of administering an antifolate in conjunction with\na methylmalonic acid lowering agent, with or without folic\nacid. The original independent claims 2 and 5 read:\n 2. (Original) A method of reducing the toxicity as-\n sociated with the administration of an antifolate to\n a mammal comprising\n administering to said mammal an effective\n amount of said antifolate in combination\n with a methylmalonic acid lowering agent.\n 5. (Original) A method of reducing the toxicity as-\n sociated with the administration of an antifolate to\n a mammal comprising\n administering to said mammal an effective\n amount of said antifolate in combination\n with a methylmalonic acid lowering agent\n and FBP binding agent.\nDRL J.A. 7860. A dependent claim further limited the an-\ntifolate to pemetrexed disodium. Id. at 7861.\n Claim 2 was rejected as anticipated by F.G. Arsenyan\net al., “Influence of Methylcobalamin on the Antineoplastic\nActivity of Methotrexate,” Onkol. Nauchn., 12(10):1299-\n1303 (1978), which disclosed experiments treating mice\nwith various tumors with a combination of methotrexate,\nan antifolate, and methylcobalamin, a vitamin B12 deriva-\ntive. The rest of the pending claims, including Claim 5,\nwere rejected as obvious over a collection of references: U.S.\nPatent 5,431,925 (“Ohmori”)—which taught treatment of\n\f8 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\nchemotherapeutically-induced immunosuppression with a\ncombination of vitamins that could include folic acid and\nvitamin B12—Worzalla, John, and Arsenyan. ’821 appli-\ncation, Sept. 27, 2004, Office Action; DRL J.A. 7868–72.\n In response, Lilly amended both claims to narrow “an-\ntifolate” to “pemetrexed disodium” and cancelled its de-\npendent claim limited to pemetrexed disodium. ’821\napplication, Jan. 25, 2005, Response to Office Action; DRL\nJ.A. 7877–84. In its remarks, Lilly asserted that the\namendment to claim 2 overcame the anticipation rejection\nbecause Arsenyan does not disclose pemetrexed disodium.\nId. To overcome the obviousness rejection of claim 5 and\nits dependents, Lilly generally argued that, while John dis-\ncloses hematologic and immunologic toxicities from admin-\nistration of pemetrexed disodium, it never suggests\nvitamin supplementation, and none of the other references\n“teach the use of [vitamin B12] to reduce toxicities associ-\nated with an antifolate.” Id. The examiner then withdrew\nthe anticipation rejection and later withdrew the obvious-\nness rejection. The ’821 application issued as U.S. Patent\n7,053,065, and the ’209 patent later issued from a continu-\nation application.\n These appeals were taken from cases which are among\nthe latest in a series of patent disputes about Alimta® that\nreaches back more than a decade. 3 In this most recent\nchapter, DRL, Hospira, and Actavis 4 submitted New Drug\n\n\n 3 This is the fourth appeal we have decided concern-\ning Alimta® and the third specifically concerning the ’209\npatent. See Neptune Generics, 921 F.3d 1372; Eli Lilly &\nCo. v. Teva Parenteral Meds., Inc., 845 F.3d 1357 (Fed. Cir.\n2017); Eli Lilly & Co. v. Teva Parenteral Meds., Inc., 689\nF.3d 1368 (Fed. Cir. 2012).\n 4 Lilly also sued Actavis LLC (“Actavis”) for infringe-\nment of the ’209 patent, Eli Lilly & Co. v. Actavis LLC, No.\n1:17-cv-00982-TWP-MPB (S.D. Ind. Mar. 30, 2017), ECF\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 9\n\n\n\nApplications under § 505(b)(2) of the Federal Food, Drug,\nand Cosmetic Act, 21 U.S.C. § 355(b)(2), relying on Lilly’s\nclinical data for pemetrexed disodium. But each applicant\nseeks to market different pemetrexed salts—in DRL’s and\nHospira’s applications, pemetrexed ditromethamine. Both\nDRL and Hospira represented to the FDA that their choice\nof the tromethamine cation was immaterial because\npemetrexed dissociates from its counterion in solution,\nDRL J.A. 8555–57; Hospira J.A. 124, and tromethamine\nwas known to be safe for pharmaceutical use, DRL J.A.\n8555, 8557.\n Lilly then asserted the ’209 patent against each of\nthese NDA applicants in the United States District Court\nfor the Southern District of Indiana. In the DRL case, the\ndistrict court construed the phrase “administration of\npemetrexed disodium” to mean “liquid administration of\npemetrexed disodium,” which “is accomplished by dissolv-\ning the solid compound pemetrexed disodium into solu-\ntion.” DRL Summary Judgment Decision, 2017 WL\n6387316, at *4. The district court denied DRL’s motion for\nsummary judgment of noninfringement, holding that pros-\necution history estoppel does not bar Lilly from asserting\nthat DRL’s proposed pemetrexed ditromethamine product\nwould infringe through the doctrine of equivalents because\nthe reason for Lilly’s amendment was to distinguish other\nantifolates and was therefore only tangential to\npemetrexed ditromethamine. Id. at *6–7. The district\ncourt also rejected DRL’s argument that Lilly dedicated\n\n\n\nNo. 1, but the parties stipulated to be bound by the district\ncourt’s decision in the DRL case that neither prosecution\nhistory estoppel nor the disclosure-dedication rule bars\nLilly’s assertion of infringement through the doctrine of\nequivalents. Actavis Br. 2. Actavis filed a brief in the DRL\nappeal as amicus curiae requesting reversal of that portion\nof the district court’s decision.\n\f10 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\npemetrexed ditromethamine to the public under the disclo-\nsure-dedication rule through its reference to Akimoto’s an-\ntifolate compounds because Akimoto is not incorporated by\nreference into the ’209 patent and in any event discloses\npemetrexed ditromethamine only within a genus of thou-\nsands of compounds, which the district court held does not\nconstitute the requisite disclosure of an identifiable alter-\nnative under this court’s precedent. Id. at *7–8; see, e.g.,\nSanDisk Corp. v. Kingston Tech. Co., 695 F.3d 1348, 1363\n(Fed. Cir. 2012).\n Following a bench trial, the district court’s opinion\nlargely followed its rationale in the DRL Summary Judg-\nment Decision with respect to the applicability of prosecu-\ntion history estoppel and the disclosure-dedication rule.\nDRL Decision, 323 F. Supp. 3d at 1046–48. In addition, the\ncourt found that DRL’s proposed product would be admin-\nistered in a manner that would meet the “administration\nof pemetrexed disodium” step of the asserted claims under\nthe doctrine of equivalents, id. at 1049, regardless of the\n“differences in chemical properties between pemetrexed\ndisodium and pemetrexed ditromethamine,” id. at 1050.\n In the Hospira case, the parties similarly disputed the\ndoctrine of equivalents, but Lilly also asserted literal in-\nfringement because Hospira’s proposed product label al-\nlows reconstitution of its pemetrexed ditromethamine salt\nin saline. Hospira Decision, 2018 WL 3008570, at *2–3;\nHospira J.A. 229. After the district court issued the DRL\nSummary Judgment Decision, Hospira conceded, contin-\ngent upon its right to appeal, that its product would in-\nfringe under the claim construction of “administration of\npemetrexed disodium” set forth in that opinion and that its\ndoctrine of equivalents arguments were likewise fore-\nclosed. Hospira Br. 18. The district court, “rel[ying] heav-\nily” on the DRL Summary Judgment Decision, granted\nLilly’s motion for summary judgment of infringement, both\nliterally and under the doctrine of equivalents. Hospira\nDecision, 2018 WL 3008570, at *1 n.2, *6.\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 11\n\n\n\n These appeals followed. We have jurisdiction under 28\nU.S.C. § 1295(a)(1).\n DISCUSSION\n We review a district court’s grant of summary judg-\nment according to the law of the regional circuit. Kaneka\nCorp. v. Xiamen Kingdomway Grp. Co., 790 F.3d 1298,\n1303 (Fed. Cir. 2015) (citing Halo Elecs., Inc. v. Pulse El-\necs., Inc., 769 F.3d 1371, 1377 (Fed. Cir. 2014)). In the Sev-\nenth Circuit, summary judgment is reviewed de novo, con-\nstruing all facts and drawing all inferences in favor of the\nnon-movant. Wis. Alumni Research Found. v. Apple Inc.,\n905 F.3d 1341, 1352 (Fed. Cir. 2018) (citing Austin v.\nWalgreen Co., 885 F.3d 1085, 1087 (7th Cir. 2018)). On ap-\npeal from a bench trial, we review a district court’s conclu-\nsions of law de novo and its findings of fact for clear error.\nBraintree Labs., Inc. v. Novel Labs., Inc., 749 F.3d 1349,\n1358 (Fed. Cir. 2014) (citing Brown & Williamson Tobacco\nCorp. v. Philip Morris Inc., 229 F.3d 1120, 1123 (Fed. Cir.\n2000)). A factual finding is clearly erroneous if, despite\nsome supporting evidence, we are left with the definite and\nfirm conviction that a mistake has been made. United\nStates v. U.S. Gypsum Co., 333 U.S. 364, 395 (1948).\n Claim construction is ultimately an issue of law, which\nwe review de novo. Shire Dev., LLC v. Watson Pharm., Inc.,\n787 F.3d 1359, 1364 (Fed. Cir. 2015). We review de novo\nthe district court’s findings of fact on evidence “intrinsic to\nthe patent (the patent claims and specification[], along\nwith the patent’s prosecution history),” and review for clear\nerror extrinsic findings of fact. Teva Pharm. USA, Inc. v.\nSandoz, Inc., 135 S. Ct. 831, 841 (2015). While infringe-\nment is a question of fact, Lucent Techs., Inc. v. Gateway,\nInc., 580 F.3d 1301, 1309 (Fed. Cir. 2009), we review de\nnovo the district court’s grant of summary judgment of non-\ninfringement, Unwired Planet, LLC v. Apple Inc., 829 F.3d\n1353, 1356 (Fed. Cir. 2016). To prove infringement, a pa-\ntentee “must supply sufficient evidence to prove that the\n\f12 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\naccused product or process contains, either literally or un-\nder the doctrine of equivalents, every limitation of the\nproperly construed claim.” Seal-Flex, Inc. v. Athletic Track\n& Court Const., 172 F.3d 836, 842 (Fed. Cir. 1999). The\npatentee has the burden of proving infringement by a pre-\nponderance of the evidence. SmithKline Diagnostics, Inc.\nv. Helena Labs. Corp., 859 F.2d 878, 889 (Fed. Cir. 1988).\n Hospira requests reversal of the district court’s finding\nthat its submission of a § 505(b)(2) NDA for its pemetrexed\nproduct literally infringed the claims of the ’209 patent.\nDRL and Hospira both argue, as does the amicus curiae\nActavis, that the district court erred as a matter of law by\nrefusing to apply prosecution history estoppel to bar Lilly’s\ndoctrine of equivalents claim, and DRL further contends\nthat the disclosure-dedication rule precludes Lilly’s equiv-\nalents claim. Finally, DRL disputes the district court’s\nfinding that administration of pemetrexed ditromethamine\nis equivalent to the claim element “administration of\npemetrexed disodium.” We address each argument in turn.\n A. Literal Infringement\n Hospira argues that it cannot literally infringe the\nclaims of the ’209 patent because intravenous administra-\ntion of pemetrexed ditromethamine dissolved in saline—a\nsolution which contains pemetrexed and chloride anions\nalongside sodium and tromethamine cations—is not “ad-\nministration of pemetrexed disodium.” Hospira also notes\nthat such a solution will, in any case, contain far more than\ntwo sodium cations per pemetrexed anion. Finally, Hos-\npira appears to make a perfunctory argument that, in the\nalternative, we should reverse the district court’s construc-\ntion and hold that the term encompasses any route of ad-\nministering pemetrexed disodium, not just liquid, as the\ndistrict court’s construction requires.\n Lilly counters that Hospira’s view improperly imposes\na “source limitation,” requiring that the pemetrexed diso-\ndium salt exist in solid form before administration, even\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 13\n\n\n\nthough Hospira’s proposed product label, like that of Ali-\nmta®, calls for administration of a solution containing\npemetrexed anions and sodium cations. Lilly also contends\nthat Hospira’s claim construction arguments are irrelevant\nbecause Hospira’s proposed product will be administered\nintravenously anyway.\n We agree with Hospira. It was clearly erroneous for\nthe district court to hold that the “administration of\npemetrexed disodium” step was met because Hospira’s\npemetrexed ditromethamine product will be dissolved in\nsaline before administration. A solution of pemetrexed and\nchloride anions and tromethamine and sodium cations can-\nnot be deemed pemetrexed disodium simply because some\nassortment of the ions in the solution consists of\npemetrexed and two sodium cations. As Lilly acknowl-\nedges throughout its brief, pemetrexed disodium is a salt.\nSee, e.g., Lilly Br. I 12 (pemetrexed toxicity is caused “by\npemetrexed itself once dissociated in solution,” not\npemetrexed disodium); see also Hospira J.A. 1596 (October\n2017 Alimta® Label referring to the drug substance as the\n“disodium salt” of pemetrexed). Once diluted, the salt’s\ncrystalline structure dissolves, and the individual ions dis-\nsociate. See Hospira J.A. 2820 (declaration of Lilly’s ex-\npert). In other words, pemetrexed disodium no longer\nexists once dissolved in solution, and, as a corollary, a dif-\nferent salt of pemetrexed dissolved in saline is not\npemetrexed disodium.\n We conclude that to literally practice the “administra-\ntion of pemetrexed disodium” step under the district court’s\nclaim construction, the pemetrexed disodium salt must be\nitself administered. See DRL Summary Judgment Deci-\nsion, 2017 WL 6387316, at *4 (“‘[A]dministration of\npemetrexed disodium’ . . . refer[s] to a liquid administration\nof pemetrexed disodium. . . ., accomplished by dissolving\nthe solid compound pemetrexed disodium into solu-\ntion . . . .”); see also Tex. Instruments Inc. v. Cypress Semi-\nconductor Corp., 90 F.3d 1558, 1563 (Fed. Cir. 1996) (“To\n\f14 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\nliterally infringe, the accused . . . process must contain\nevery limitation of the asserted claim.” (citing Laitram\nCorp. v. Rexnord, Inc., 939 F.2d 1533, 1535 (Fed. Cir.\n1991))). There is no dispute that Hospira has only sought\napproval to market pemetrexed ditromethamine, Lilly Br.\nI 4, and that neither its proposed product nor methods of\nadministering it will constitute administering the\npemetrexed disodium salt. Accordingly, Hospira will not\npractice the step of “administration of pemetrexed diso-\ndium,” and the district court’s finding of literal infringe-\nment must be reversed.\n B. Doctrine of Equivalents\n Few propositions of patent law have been so consist-\nently sustained by the Supreme Court as the doctrine of\nequivalents. See Festo Corp. v. Shoketsu Kinzoku Kogyo\nKabushki Co., 535 U.S. 722, 733 (2002) (“Festo VIII”)\n(“[E]quivalents remain a firmly entrenched part of the set-\ntled rights protected by the patent.”); Warner-Jenkinson\nCo. v. Hilton Davis Chem. Co., 520 U.S. 17, 40 (1997) (“[W]e\nadhere to the doctrine of equivalents.”); Graver Tank &\nMfg. Co. v. Linde Air Prods. Co., 339 U.S. 605, 608 (1950)\n(“Originating almost a century ago in the case of Winans v.\nDenmead, [56 U.S. 330 (1853)] . . . [the doctrine of equiva-\nlents] has been consistently applied by this Court and the\nlower federal courts, and continues today ready and avail-\nable for utilization when the proper circumstances for its\napplication arise.”). It is settled that a patentee is entitled\n“in all cases to invoke to some extent the doctrine of equiv-\nalents,” Seymour v. Osborne, 78 U.S. 516, 555 (1870), with-\nout a “judicial exploration of the equities of a case”\nbeforehand. See Warner-Jenkinson, 520 U.S. at 34.\n Yet the Supreme Court has also acknowledged that the\ndoctrine of equivalents, “when applied broadly, conflicts\nwith the definitional and public-notice functions of the stat-\nutory claiming requirement,” Warner-Jenkinson, 520 U.S.\nat 29, and that, without the proper balance between these\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 15\n\n\n\ntwo imperatives, the doctrine may “take[] on a life of its\nown, unbounded by the patent claims.” See id. at 28–29.\nWe have emphasized, moreover, that the doctrine of equiv-\nalents is “the exception, however, not the rule,” and not\nmerely “the second prong of every infringement charge,\nregularly available to extend protection beyond the scope\nof the claims.” London v. Carson Pirie Scott & Co., 946 F.2d\n1534, 1538 (Fed. Cir. 1991). Patent infringement is princi-\npally determined by examining whether the accused sub-\nject matter falls within the scope of the claims.\n To that end, courts have placed important limitations\non a patentee’s ability to assert infringement under the\ndoctrine of equivalents. See, e.g., Festo VIII, 535 U.S. at\n737–41 (prosecution history estoppel); Warner-Jenkinson,\n520 U.S. at 39 n.8 (“[A] theory of equivalence [cannot] en-\ntirely vitiate a particular claim element . . . .”); Graver\nTank, 339 U.S. at 608 (accused equivalent cannot differ\nsubstantially from the claimed invention); Johnson &\nJohnston Assocs. Inc. v. R.E. Serv. Co., 285 F.3d 1046, 1054\n(Fed. Cir. 2002) (en banc) (subject matter disclosed but not\nclaimed is dedicated to the public) (citing Maxwell v. J.\nBaker, Inc., 86 F.3d 1098 (Fed. Cir. 1996)); Wilson Sporting\nGoods Co. v. David Geoffrey & Assocs., 904 F.2d 677, 683\n(Fed. Cir. 1990) (“[T]he asserted scope of equivalency [can-\nnot] encompass the prior art . . . .” (Rich, J.) (citations omit-\nted)). These appeals implicate several of these limitations.\n 1. Prosecution History Estoppel\n The main dispute in these appeals is whether Lilly has\nrebutted the presumption of prosecution history estoppel\nthat attached to its amendment in the ’821 application.\nProsecution history estoppel arises when a patent appli-\ncant narrows the scope of his claims during prosecution for\na reason “substantial[ly] relating to patentability.” See\ngenerally Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki\nCo., 344 F.3d 1359, 1366–67 (Fed. Cir. 2003) (en banc)\n(“Festo X”). Such a narrowing amendment is presumed to\n\f16 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\nbe a surrender of all equivalents within “the territory be-\ntween the original claim and the amended claim,” but the\npresumption is overcome if the patentee can show the ap-\nplicability of one of the few exceptions identified by the Su-\npreme Court. Festo VIII, 535 U.S. at 740–41 (citing Exhibit\nSupply Co. v. Ace Patents Corp., 315 U.S. 126, 136–37\n(1942)). Whether prosecution history estoppel applies to\nbar a doctrine of equivalents claim is a question of law, re-\nviewed de novo. See Regents of Univ. of Cal. v. Dakocyto-\nmation Cal., Inc., 517 F.3d 1364, 1371 (Fed. Cir. 2008)\n(citing Pharmacia & Upjohn Co. v. Mylan Pharm., Inc., 170\nF.3d 1373, 1376 (Fed. Cir. 1999)).\n Lilly does not dispute that the amendment in question\nwas both narrowing and made for a substantial reason re-\nlating to patentability. Lilly Br. II 21. Furthermore, Lilly\nrelies on only one exception to giving effect to the presump-\ntion as to the scope of surrender: that the rationale of its\namendment “[bore] no more than a tangential relation to\nthe equivalent in question.” Festo VIII, 535 U.S. at 740. As\na result, the parties’ dispute about whether prosecution\nhistory estoppel applies is confined to whether Lilly’s\namendment narrowing “an antifolate” to “pemetrexed diso-\ndium” was only tangential to pemetrexed ditromethamine,\nwhich is the accused compound. Whether the tangential\nexception applies is a question of law, Integrated Tech.\nCorp. v. Rudolph Techs., Inc., 734 F.3d 1352, 1356 (Fed.\nCir. 2013), and a patentee seeking to use the exception\n“must base his arguments solely upon the public record of\nthe patent’s prosecution.” Festo X, 344 F.3d at 1369–70 (ci-\ntation omitted).\n The Appellants argue that Lilly failed to explain why\nit did not pursue a narrower amendment literally encom-\npassing pemetrexed ditromethamine, and they emphasize\nour statement that the tangential exception is “very nar-\nrow.” Integrated, 734 F.3d at 1358 (quoting Cross Med.\nProds., Inc. v. Medtronic Sofamor Danek, Inc., 480 F.3d\n1335, 1342 (Fed. Cir. 2007)). The Appellants further point\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 17\n\n\n\nout that Lilly cannot be said to have “lacked the words to\ndescribe” pemetrexed ditromethamine, see Festo VIII, 535\nU.S. at 734, because Lilly’s previous patents, as well as the\nEuropean companion to the ’209 patent, claimed\npemetrexed salts generally and pemetrexed disodium in a\ndependent claim. They also assert that the district court\nerred by focusing on whether Lilly actually needed to relin-\nquish pemetrexed ditromethamine to overcome the Arsen-\nyan anticipation rejection because “the tangential\nexception is not a patentee’s-buyer’s-remorse exception.”\nDRL Br. 39.\n In response, Lilly argues that the district court\nproperly held that the reason for its amendment was to dis-\ntinguish pemetrexed from antifolates generally and that\nthe different salt type is a merely tangential change with\nno consequence for pemetrexed’s administration or mecha-\nnism of action within the body. Lilly also contends that it\nis not barred from asserting the tangential exception\nsimply because pemetrexed ditromethamine is within “the\nterritory between the original claim and the amended\nclaim.” Festo VIII, 535 U.S. at 740. Finally, Lilly argues\nthat Appellants’ view that courts must “consider hypothet-\nical alternative amendments” that would literally encom-\npass the alleged equivalent “would eviscerate the\ntangentiality exception.” Lilly Br. II 44.\n We agree with Lilly. As a general matter, we find Ap-\npellants’ view of prosecution history estoppel, and the tan-\ngential exception in particular, too rigid. Tangential\nmeans “touching lightly or in the most tenuous way.” Web-\nster’s Third New International Dictionary (2002). The rea-\nson for Lilly’s amendment, as the district court concluded,\nwas to narrow original claim 2 to avoid Arsenyan, which\nonly discloses treatments using methotrexate, a different\nantifolate. See DRL J.A. 7879–80 (overcoming the Arsen-\nyan anticipation rejection by arguing that it “does not dis-\nclose pemetrexed disodium”). To overcome a clear\nanticipation, Lilly opted to narrow its original claim 2 and\n\f18 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\nits dependents to more accurately define what it actually\ninvented, an improved method of administering\npemetrexed. In other words, the particular type of salt to\nwhich pemetrexed is complexed relates only tenuously to\nthe reason for the narrowing amendment, which was to\navoid Arsenyan. We therefore hold that Lilly’s amendment\nwas merely tangential to pemetrexed ditromethamine be-\ncause the prosecution history, in view of the ’209 patent it-\nself, strongly indicates that the reason for the amendment\nwas not to cede other, functionally identical, pemetrexed\nsalts.\n The prosecution record confirms our understanding.\nOriginal claim 5, which, like all the current claims of the\n’209 patent, required supplementation with both vitamin\nB12 and folic acid, was never rejected as anticipated over\nArsenyan. Instead, the art cited against original claim 5\nand its dependent claims in the obviousness ground of re-\njection was replete with information about pemetrexed\ndisodium; John disclosed clinical trials using pemetrexed\ndisodium, reporting both its efficacy and its toxic side ef-\nfects, and in response, DRL J.A. 7869–70, Worzalla sug-\ngested folic acid supplementation to counteract these side\neffects, DRL J.A. 7870–71. The prosecution record implies\nthat Lilly’s amendment, inartful though it might have\nbeen, was prudential in nature and did not need or intend\nto cede other pemetrexed salts.\n Hospira argues that the amendment was made to over-\ncome the obviousness rejection over Ohmori and John and\nthat Lilly has provided no reason for the amendment rela-\ntive to that rejection. Like Lilly, we find this argument\nmakes little sense. John discloses the results of a clinical\ntrial of pemetrexed disodium and explicitly suggests the\ntoxicities caused by pemetrexed; as we concluded above,\nnarrowing “antifolate” to “pemetrexed disodium” could not\npossibly distinguish the art cited in the obviousness ground\nof rejection.\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 19\n\n\n\n DRL also insists that we have held that an applicant’s\nremorse at ceding more claim scope than necessary is not a\nreason for the tangential exception to apply. See, e.g., Lu-\ncent Techs., Inc. v. Gateway, Inc., 525 F.3d 1200, 1218 (Fed.\nCir. 2008); Schwarz Pharma, Inc. v. Paddock Labs., Inc.,\n504 F.3d 1371, 1377 (Fed. Cir. 2007). This is generally\ntrue, but DRL overreads the holdings of these cases. After\nall, the tangential exception only exists because applicants\nover-narrow their claims during prosecution. Amend-\nments are not construed to cede only that which is neces-\nsary to overcome the prior art, see Schwarz, 504 F.3d at\n1377, nor will the court “speculat[e]” whether an amend-\nment was necessary, see Kinzenbaw v. Deere & Co., 741\nF.2d 383, 389 (Fed. Cir. 1984). But the reason for an\namendment, where the tangential exception is invoked,\ncannot be determined without reference to the context in\nwhich it was made, including the prior art that might have\ngiven rise to the amendment in the first place. See Festo X,\n344 F.3d at 1370. Here, it is unlikely that a competitor\nwould have been “justified in assuming that if he [made an\nequivalent pemetrexed salt], he would not infringe [the\n’209 patent].” Kinzenbaw, 741 F.2d at 389; cf. Festo VIII,\n535 U.S. at 738 (“There is no reason why a narrowing\namendment should be deemed to relinquish equivalents\n. . . beyond a fair interpretation of what was surrendered.”).\n Furthermore, Appellants’ suggestion that Lilly must\nprove that it could not have drafted a claim that literally\nencompassed pemetrexed ditromethamine is unsupported\nby our precedent on prosecution history estoppel, not to\nmention excessive. We do not demand perfection from pa-\ntent prosecutors, and neither does the Supreme Court. See\nFesto VIII, 535 U.S. at 738 (“It does not follow . . . that [an]\namended claim becomes so perfect in its description that\nno one could devise an equivalent.”). Lilly’s burden was to\nshow that pemetrexed ditromethamine was “peripheral, or\nnot directly relevant,” to its amendment, Festo X, 344 F.3d\nat 1369. And as we concluded above, Lilly has done so.\n\f20 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\n In addition, the Appellants maintain that when a pa-\ntentee submits an amendment adding two claim limita-\ntions, it cannot later argue that the reason for the\namendment was tangential to an accused equivalent con-\ntaining only one of the added limitations simply because\nthe second limitation was unnecessary to overcome the\nprior art. They offer Felix v. American Honda Motor Co.,\n562 F.3d 1167 (Fed. Cir. 2009), as an illustration of this\nprinciple. 5 In that case, we held that prosecution history\nestoppel applied to a claim directed to a vehicle bed storage\nsystem—limited in response to a rejection to having a\nchannel with a flange and a gasket mounted on that\nflange—barring assertion of equivalence with respect to a\nproduct that met the channel aspect, but not the gasket as-\npect, of the limitation. Id. at 1184–85.\n But as Lilly points out, this holding was determined by\nthat patent’s prosecution history, Felix, 562 F.3d at 1184,\nand we have also held that prosecution history estoppel\ndoes not apply in similar circumstances, where the\n\n\n\n 5 The parties argue at length about which of our\ncases are properly analogous to the facts presented in these\nappeals. Here, in applying the Supreme Court’s frame-\nwork, we find the analogies to other cases less helpful than\na direct consideration of the specific record of this case and\nwhat it shows about the reason for amendment and the re-\nlation of that reason to the asserted equivalent. This case-\nspecific focus, within the governing framework, comports\nwith the equitable nature of prosecution history estoppel.\nSee Festo VIII, 535 U.S. at 738 (“[The Supreme Court has]\nconsistently applied the doctrine in a flexible way, not a\nrigid one.”); cf. Heckler v. Cmty. Health Servs. of Crawford\nCty., Inc., 467 U.S. 51, 59 (1984) (“Estoppel is an equitable\ndoctrine invoked to avoid injustice in particular cases. . . .\n[and] a hallmark of the doctrine is its flexible applica-\ntion . . . .”).\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 21\n\n\n\nprosecution record differed. See, e.g., Regents, 517 F.3d at\n1376–78 (amendment narrowing “disabling hybridization\ncapacity of [nucleic acid] sequences” to methods using a\n“blocking nucleic acid” was merely tangential to unclaimed\nrepetitive sequence nucleic acids); Insituform Techs., Inc.\nv. CAT Contracting, Inc., 385 F.3d 1360, 1368 (Fed. Cir.\n2004) (amendment narrowing method of inserting resin\ninto tube using a vacuum to one using “a cup” to do so was\nmerely tangential to a multiple cup embodiment because\nthe number of cups bore no relationship to the cited prior\nart or the rationale behind the narrowing amendment).\nThus, our cases demonstrate that prosecution history es-\ntoppel is resistant to the rigid legal formulae that Appel-\nlants seek to extract from them. See Intervet Inc. v. Merial\nLtd., 617 F.3d 1282, 1291 (Fed. Cir. 2010) (“[T]here is no\nhard-and-fast test for what is and what is not a tangential\nrelation . . . .”).\n Finally, DRL also contends that our precedent squarely\nforecloses Lilly’s tangentiality argument, and it invites us\nto read those cases to hold that “where the reason for the\namendment and the equivalent in question both relate to\nthe same claim element, the tangential exception does not\napply.” DRL Br. 47. We decline this invitation because\nsuch a bright-line rule is both contrary to the equitable na-\nture of prosecution history estoppel, as articulated in Festo\nVIII, 535 U.S. at 738, and inconsistent with the equitable\nspirit that animates the doctrine of equivalents, see Graver\nTank, 339 U.S. at 608–09 (the doctrine is one of “whole-\nsome realism”). Instead, we reaffirm that whether an\namendment was merely tangential to an equivalent must\nbe decided in the context of the invention disclosed in the\npatent and the prosecution history. Festo X, 344 F.3d at\n1370.\n DRL’s intuition—that an amendment that narrows an\nexisting claim element evinces an intention to relinquish\nthat claim scope—is often correct. Indeed, as we have\nfound in previous cases, it is a powerful indication that an\n\f22 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\namendment was not merely tangential. See, e.g., Honey-\nwell Int’l, Inc. v. Hamilton Sundstrand Corp., 523 F.3d\n1304, 1315–16 (Fed. Cir. 2008); Biagro W. Sales, Inc. v.\nGrow More, Inc., 423 F.3d 1296, 1306 (Fed. Cir. 2005). But\nhere, we conclude that this consideration is not dispositive\nbecause the rest of the prosecution history, and the ’209\npatent itself, show that it is implausible that the reason for\nLilly’s amendment was to surrender other pemetrexed\nsalts. Indeed, such a relinquishment would effectively ded-\nicate the entirety of Lilly’s invention to the public and\nthereby render the ’209 patent worthless, and it would\nhave been irrelevant for distinguishing the prior art.\nAgain, the prosecution history strongly indicates a less\nsweeping and more sensible reason for Lilly’s amendment:\nto surrender antifolates other than pemetrexed. Thus, we\nconclude on this prosecution record that Lilly’s amendment\nwas merely tangential to pemetrexed ditromethamine.\n 2. Disclosure-Dedication Rule\n DRL next argues that the disclosure-dedication rule\nbars Lilly from asserting infringement under the doctrine\nof equivalents. The ’209 patent sets forth its invention as\nan improved method of administering antifolates, ’209 pa-\ntent col. 2 ll. 47–58, and teaches that the derivatives de-\nscribed in the Akimoto patent are preferred examples of\nantifolates, id. col. 4 ll. 34–40. DRL contends that one of\nthese derivatives is pemetrexed ditromethamine and that\nit was dedicated to the public when Lilly declined to claim\nit. DRL asserts that the district court erred because it both\nrequired express incorporation of Akimoto by reference\ninto the ’209 patent and concluded that Akimoto does not\nspecifically disclose pemetrexed ditromethamine.\n Lilly counters that the disclosure-dedication rule re-\nquires express disclosure of the subject matter in question\nin the specification except in narrow circumstances, such\nas when that subject matter is disclosed in a priority appli-\ncation, see Abbott Labs. v. Sandoz, Inc., 566 F.3d 1282,\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 23\n\n\n\n1297 (Fed. Cir. 2009), or prior art expressly incorporated\nby reference, SanDisk, 695 F.3d at 1366. Lilly also argues\nthat the district court correctly determined that the rele-\nvant portion of Akimoto discloses only a generic formula\nfrom which a skilled artisan would not be able to recognize\npemetrexed ditromethamine.\n We agree with Lilly and hold that the disclosure-dedi-\ncation rule is inapplicable to this case because the ’209 pa-\ntent does not disclose methods of treatment using\npemetrexed ditromethamine, and, as a result, Lilly could\nnot have dedicated such a method to the public.\n Under the disclosure-dedication rule, subject matter\ndisclosed by a patentee, but not claimed, is considered ded-\nicated to the public. See Johnson & Johnston, 285 F.3d at\n1054. The reason for the doctrine is that members of the\npublic reading a disclosure of particular subject matter are\nentitled, absent a claim to it, to assume that it is not pa-\ntented and therefore dedicated to the public (unless, for ex-\nample, claimed in a continuation or other application based\non the disclosure). Cf. Maxwell, 86 F.3d at 1107 (failure to\nclaim inventive subject matter “is clearly contrary to 35\nU.S.C. § 112, which requires that a patent applicant ‘par-\nticularly point[] out and distinctly claim[] the subject mat-\nter which the applicant regards as his invention’”). Subject\nmatter is considered disclosed when a skilled artisan “can\nunderstand the unclaimed disclosed teaching upon reading\nthe written description,” but not “any generic reference . . .\nnecessarily dedicates all members of that particular ge-\nnus.” PSC Comput. Prod., Inc. v. Foxconn Int’l, Inc., 355\nF.3d 1353, 1360 (Fed. Cir. 2004).\n DRL further contends that the disclosure-dedication\nrule does not impose a § 112 requirement for sufficiency of\ndisclosure, see Toro Co. v. White Consol. Indus., Inc., 383\nF.3d 1326, 1334 (Fed. Cir. 2004), and that a skilled artisan\nreading the ’209 patent would both look for a disclosure of\npemetrexed in Akimoto, and also seek to use a well-known\n\f24 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\ncation like tromethamine, which it maintains is generically\ndisclosed in Akimoto in the form of “substituted ammo-\nnium” base salts.\n We are unpersuaded by DRL’s arguments. As the dis-\ntrict court noted, Akimoto’s formula, col. 1 l. 49–col. 2 l. 3,\nincludes seven functional group variables and encompasses\nthousands of compounds, and while Akimoto discloses\nabout fifty exemplary compounds, none of them is\npemetrexed. Moreover, Akimoto does not even disclose tro-\nmethamine expressly but only generically among dozens of\nother salts. At most, Akimoto discloses ammonium salts\ngenerally, which is far from a description of tromethamine.\nIn similar circumstances, we have held that “sufficient de-\nscription of a genus” requires that a skilled artisan be able\nto “‘visualize or recognize’ the members of the genus.” See\nAriad Pharm., Inc. v. Eli Lilly & Co., 598 F.3d 1336, 1350\n(Fed. Cir. 2010) (quoting Regents of the Univ. of Cal. v. Eli\nLilly & Co., 119 F.3d 1559, 1568–69 (Fed. Cir. 1997)).\nAkimoto does not so describe pemetrexed ditromethamine,\nand we see no reason why a skilled artisan would set out\non DRL’s winding path to cobble together pemetrexed dit-\nromethamine. While the ’209 patent teaches that\npemetrexed disodium is the “most preferred” antifolate,\nthat knowledge would not change the skilled artisan’s un-\nderstanding of what Akimoto discloses.\n Because Akimoto contains only a “generic reference” to\npemetrexed ditromethamine, PSC Comput., 355 F.3d at\n1360, we conclude that it was not dedicated to the public.\n 3. Merits\n A component in an accused product or process may be\nequivalent to a claim element if the two are insubstantially\ndifferent with respect to the “role played by [the] element\nin the context of the specific patent claim.” Warner-Jen-\nkinson, 520 U.S. at 39–40. Relevant differences can in-\nclude the function each serves, the way in which each\nworks, and the result each obtains, id. at 39, and, especially\n\fELI LILLY AND COMPANY v. HOSPIRA, INC. 25\n\n\n\nin biochemical cases, structural or pharmacological charac-\nteristics, Mylan Inst. LLC v. Aurobindo Pharm. Ltd., 857\nF.3d 858, 869 (Fed. Cir. 2017). “The determination of\nequivalency vel non is a question of fact,” Canton Bio Med.,\nInc. v. Integrated Liner Techs., Inc., 216 F.3d 1367, 1369\n(Fed. Cir. 2000) (citing Pall Corp. v. Micron Separations,\nInc., 66 F.3d 1211, 1218 (Fed. Cir. 1995)), which we review\nfor clear error in an appeal from a bench trial, Pfizer, Inc.\nv. Apotex, Inc., 480 F.3d 1348, 1359 (Fed. Cir. 2007).\n DRL argues that the district court erred in finding that\nits proposed pemetrexed ditromethamine product will be\nadministered in an insubstantially different way from the\nclaimed method. DRL maintains that the district court fo-\ncused on the fact that each product treats the same dis-\neases by delivering pemetrexed intravenously, when the\nrelevant context is the manner of administration. In DRL’s\nview, the chemical differences between sodium and tro-\nmethamine—e.g., pH, buffering capacity, or solubility—\nDRL Br. 20–21, render the methods in which each is ad-\nministered to a patient substantially different.\n Lilly responds that the relevant context is treatment of\na patient “in need of chemotherapeutic treatment.” ’209\npatent claim 12. Lilly agrees with the district court that\nthe chemical differences between sodium and trometham-\nine are clinically irrelevant because each undisputedly\nlacks therapeutic activity.\n We see no clear error in the district court’s findings. As\nthe district court found, DRL’s product will accomplish an\nidentical aim, furnishing the same amount of pemetrexed\nto active sites in the body; in exactly the same way, by di-\nluting a pemetrexed salt in an aqueous solution for intra-\nvenous administration. Indeed, after dilution and\nimmediately before administration, DRL’s product is func-\ntionally identical to Lilly’s in that it contains the same\namount of diluted pemetrexed anion. DRL J.A. 8557. And\nDRL declines to identify the relevance of any of the\n\f26 ELI LILLY AND COMPANY v. HOSPIRA, INC.\n\n\n\n\nchemical differences it identifies. See UCB, Inc. v. Watson\nLabs. Inc., 927 F.3d 1272, 1284–86 (Fed. Cir. 2019) (chem-\nical differences may not be relevant if the equivalent has\nknown interchangeability in the context of the claimed\ncomposition). We find DRL’s arguments unconvincing and\ntherefore affirm the district court’s findings.\n In summary, these cases are eminently suitable for ap-\nplication of the doctrine of equivalents, and we conclude\nthat neither prosecution history estoppel nor the disclo-\nsure-dedication rule bars Lilly from asserting infringement\nthrough equivalence.\n CONCLUSION\n We have fully considered each party’s further argu-\nments but find them unpersuasive. For the foregoing rea-\nsons, we reverse the district court’s finding of literal\ninfringement in the Hospira Decision but affirm its judg-\nment of infringement under the doctrine of equivalents.\nThe judgment of infringement under the doctrine of equiv-\nalents in the DRL Decision is likewise affirmed.\nAFFIRMED-IN-PART AND REVERSED-IN-PART IN\n APPEAL NOS. 2018-2126, 2018-2127\n AFFIRMED IN APPEAL NO. 2018-2128\n COSTS\n Each party shall bear its own costs.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: ELI LILLY AND COMPANY, Plaintiff-Appellee v. HOSPIRA, INC., Defendant-Appellant Eli Lilly and Company, Plaintiff-Appellee v. Dr. Reddy's Laboratories, Ltd., Dr. Reddy's Laboratories, Inc., Defendants-Appellants\\nCourt: U.S. Court of Appeals for the Federal Circuit\\nDate decided: 2019-08-09\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n ELI LILLY AND COMPANY,\\n Plaintiff-Appellee\\n\\n v.\\n\\n HOSPIRA, INC.,\\n Defendant-Appellant\\n ______________________\\n\\n 2018-2126, 2018-2127\\n ______________________\\n\\n Appeals from the United States District Court for the\\nSouthern District of Indiana in No. 1:16-cv-03460-TWP-\\nMPB, Judge Tanya Walton Pratt.\\n --------------------------------------------------\\n ELI LILLY AND COMPANY,\\n Plaintiff-Appellee\\n\\n v.\\n\\n DR. REDDY'S LABORATORIES, LTD., DR.\\n REDDY'S LABORATORIES, INC.,\\n Defendants-Appellants\\n ______________________\\n\\n 2018-2128\\n ______________________\\n\\f2 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\n Appeal from the United States District Court for the\\nSouthern District of Indiana in No. 1:16-cv-00308-TWP-\\nMPB, Judge Tanya Walton Pratt.\\n ______________________\\n\\n Decided: August 9, 2019\\n ______________________\\n\\n ADAM LAWRENCE PERLMAN, Williams & Connolly LLP,\\nWashington, DC, argued for plaintiff-appellee in 2018-2126\\nand 2018-2128. Also represented by GALINA I. FOMENKOVA,\\nDOV PHILIP GROSSMAN, DAVID M. KRINSKY, ANDREW P.\\nLEMENS, CHARLES MCCLOUD; JAMES PATRICK LEEDS, Eli\\nLilly and Company, Indianapolis, IN.\\n\\n BRADFORD PETER LYERLA, Jenner & Block LLP, Chi-\\ncago, IL, argued for defendant-appellant in 2018-2126.\\nAlso represented by YUSUF ESAT, SARA TONNIES HORTON;\\nADAM G. UNIKOWSKY, Washington, DC.\\n\\n JOHN C. O'QUINN, Kirkland & Ellis LLP, Washington,\\nDC, argued for defendants-appellants in 2018-2128. Also\\nrepresented by WILLIAM H. BURGESS, CALVIN ALEXANDER\\nSHANK; JEFFERY B. ARNOLD, Holland & Knight LLP, At-\\nlanta, GA; MERRI C. MOKEN, CHARLES A. WEISS, ERIC H.\\nYECIES, New York, NY.\\n\\n BRIAN TIMOTHY BURGESS, Goodwin Procter LLP, Wash-\\nington, DC, for amicus curiae Actavis LLC in 2018-2128.\\nAlso represented by EDWINA CLARKE, EMILY L. RAPALINO,\\nDARYL L. WIESEN, Boston, MA; LINNEA P. CIPRIANO, New\\nYork, NY.\\n ______________________\\n\\n\\n Before LOURIE, MOORE, and TARANTO, Circuit Judges.\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 3\\n\\n\\n\\nLOURIE, Circuit Judge.\\n Hospira Inc. (“Hospira”), Dr. Reddy’s Laboratories\\nLtd., and Dr. Reddy’s Laboratories Inc. (collectively,\\n“DRL”) appeal from two judgments of the United States\\nDistrict Court for the Southern District of Indiana in two\\ninfringement suits brought by Eli Lilly & Company\\n(“Lilly”) under the Hatch-Waxman Act, 21 U.S.C. § 355.\\nThe district court held in each case that the defendant’s\\nsubmission of a New Drug Application pursuant to 21\\nU.S.C. § 355(b)(2) infringed U.S. Patent 7,772,209 (the\\n“’209 patent”) under 35 U.S.C. § 271(e)(2). See Eli Lilly &\\nCo. v. Hospira, Inc., No. 1:16-cv-03460-TWP-MPB, 2018\\nWL 3008570 (S.D. Ind. June 15, 2018) (“Hospira Decision”);\\nEli Lilly & Co. v. Dr. Reddy’s Labs., Ltd., 323 F. Supp. 3d\\n1042 (S.D. Ind. 2018) (“DRL Decision”); see also Eli Lilly &\\nCo. v. Dr. Reddy’s Labs., Ltd., No. 1:16-cv-00308-TWP-\\nMPB, 2017 WL 6387316 (S.D. Ind. Dec. 14, 2017) (“DRL\\nSummary Judgment Decision”). Accordingly, the district\\ncourt entered orders under 35 U.S.C. § 271(e)(4)(A) prohib-\\niting FDA approval of the products at issue until the expi-\\nration of the ’209 patent. Eli Lilly & Co. v. Hospira, Inc.,\\nNo. 1:16-cv-03460-TWP-MPB (S.D. Ind. June 27, 2018),\\nECF No. 94; Eli Lilly & Co. v. Dr. Reddy’s Labs., Ltd., No.\\n1:16-cv-00308-TWP-MPB, 2018 WL 3616715 (S.D. Ind.\\nJuly 27, 2018). We decide these appeals together in this\\ncombined opinion. 1\\n We reverse the district court’s finding of literal in-\\nfringement in the Hospira Decision as clearly erroneous in\\nlight of the court’s claim construction of “administration of\\npemetrexed disodium.” Because the district court did not\\nerr in its application of the doctrine of equivalents in either\\n\\n\\n 1 We refer to the joint appendices in these appeals by\\nreference to each appellant. Lilly’s brief in the Hospira ap-\\npeal is referred to as “Lilly Br. I” and its brief in the DRL\\nappeal as “Lilly Br. II.”\\n\\f4 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\ndecision, we affirm both judgments of infringement. Thus,\\nthe Hospira Decision is affirmed-in-part and reversed-in-\\npart, and the DRL Decision is affirmed.\\n BACKGROUND\\n Lilly markets the compound pemetrexed in the form of\\na disodium salt as Alimta®, which is indicated, both alone\\nand in combination with other active agents, for treating\\ncertain types of non-small cell lung cancer and mesotheli-\\noma. Pemetrexed is an antifolate, a class of molecules\\nwhich, at the time of the invention in 2001, was “one of the\\nmost thoroughly studied classes of antineoplastic agents.”\\n’209 patent col. 1 ll. 19–20. Antifolates are structurally\\nsimilar to folic acid and work by competitively binding to\\ncertain enzymes that use folic acid metabolites as cofactors\\nin several steps of de novo nucleotide synthesis. Id. col. 1\\nll. 40–41. Unlike folic acid, antifolates do not enable these\\nsynthetic steps, but instead inhibit them. Pemetrexed in-\\nhibits several of these enzymes, including thymidylate syn-\\nthase, which methylates deoxyuridine in the final step of\\ndeoxythymidine synthesis. Id. col. 1 ll. 59–61. By inhibit-\\ning the creation of these nucleotides, antifolates slow down\\nDNA and RNA synthesis, and with it, cell growth and divi-\\nsion. Cancer cells tend to grow rapidly, so antifolate ther-\\napy affects them disproportionately, but healthy cells can\\nalso be damaged.\\n Pemetrexed had been known for at least a decade in\\n2001. Lilly’s U.S. Patent 5,344,932 (“Taylor”) disclosed\\nthat certain glutamic acid derivatives with pyrrolo[2,3-\\nd]pyrimidine heterocyclic ring structures, exemplified by\\npemetrexed, are “particularly active . . . inhibitors of thy-\\nmidylate synth[ase],” Taylor col. 1 ll. 59–60; see also id. col.\\n19 l. 37–col. 20 l. 25 (disclosing data indicating that\\npemetrexed inhibits thymidylate synthase activity in vitro\\nin human cell lines and in vivo in mice). The Taylor patent\\nalso disclosed that its compounds could be employed as\\n“pharmaceutically acceptable salt[s],” id. col. 2 l. 35, and\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 5\\n\\n\\n\\nthat the disodium salt form was particularly advantageous,\\nid. col. 2 ll. 47–48. U.S. Patent 4,997,838 (“Akimoto”), to\\nwhich Lilly took a license, disclosed a large genus of com-\\npounds containing pyrrolo[2,3-d]pyrimidine heterocyclic\\nring structures and a glutamic acid functional group, and\\nthat encompassed pemetrexed. The Akimoto patent dis-\\ncloses nearly fifty exemplary compounds, col. 14 l. 61–col.\\n16 l. 48, none of which is pemetrexed. Akimoto further dis-\\ncloses that its compounds may be prepared as salts of\\n“pharmaceutically acceptable bases,” such as “alkali met-\\nals, alkali earth metals, non-toxic metals, ammonium, and\\nsubstituted ammonium.” Id. col. 14 ll. 44–47.\\n By 2001, Lilly had also published the results of several\\nclinical trials investigating the use of pemetrexed disodium\\nas a treatment for different types of cancer. See, e.g., W.\\nJohn et al., “Activity of Multitargeted Antifolate\\n(Pemetrexed Disodium, LY231514) in Patients with Ad-\\nvanced Colorectal Carcinoma: Results from a Phase II\\nStudy,” Cancer, 88(8):1807–13 (2000). In the course of con-\\nducting these studies, Lilly discovered that pemetrexed\\ndisodium caused severe hematologic and immunologic side\\neffects, resulting in infections, nausea, rashes, and even\\nsome deaths. See id.; see also Neptune Generics, LLC v. Eli\\nLilly & Co., 921 F.3d 1372, 1377–78 (Fed. Cir. 2019) (dis-\\ncussing Lilly’s response to adverse clinical data), and Nep-\\ntune Generics, LLC v. Eli Lilly & Co., No. IPR2016-00240,\\n2017 WL 4466557, at *28–30 (P.T.A.B. Oct. 5, 2017) (same).\\nAs the ’209 patent teaches, such side effects are not uncom-\\nmon among antifolates. See ’209 patent col. 1 ll. 11–14.\\nSome researchers hypothesized that folic acid deficiency\\ncaused these side effects and suggested supplementing\\npemetrexed disodium treatment with folic acid. DRL J.A.\\n7870 (citing J.F. Worzalla et al., “Role of Folic Acid in Mod-\\nulating the Toxicity and Efficacy of the Multitargeted An-\\ntifolate, LY231514,” Anticancer Research, 18:3235–40\\n(1998)).\\n\\f6 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\n The invention of the ’209 patent is an improved method\\nof treatment with antifolates, particularly pemetrexed\\ndisodium, through supplementation with a methylmalonic\\nacid lowering agent and folic acid. Doing so, according to\\nthe patent, lessens antifolate toxicity without sacrificing ef-\\nficacy. See ’209 patent col. 10 ll. 17–53 (reporting that pre-\\nsupplementation regimen of vitamin B12 and folic acid in\\nclinical studies substantially reduced pemetrexed-induced\\ntoxicity and deaths while delivering a superior chemother-\\napeutic response rate). The ’209 patent lists preferred an-\\ntifolates, including some then-existing antifolate therapies,\\nas well as “derivatives described in” several patents includ-\\ning the Akimoto patent, and “most preferred, Pemetrexed\\nDisodium.” Id. col. 4 ll. 28–43. Each of the claims of the\\n’209 patent requires administration of pemetrexed diso-\\ndium following administration of folic acid and a\\nmethylmalonic acid lowering agent, specified in some\\nclaims, as well as the Alimta® label, as vitamin B12. Claim\\n12 is representative2:\\n 12. An improved method for administering\\n pemetrexed disodium to a patient in need of chemo-\\n therapeutic treatment, wherein the improvement\\n comprises:\\n a) administration of between about 350 µg\\n and about 1000 µg of folic acid prior to the\\n first administration of pemetrexed diso-\\n dium;\\n\\n\\n\\n\\n 2 The district court treated claim 12 as representa-\\ntive, DRL Summary Judgment Decision, 2017 WL\\n6387316, at *1–2; Hospira Decision, 2018 WL 3008570, at\\n*2, and no party has disputed that determination on ap-\\npeal. See, e.g., DRL Opening Br. 8–9; Hospira Opening Br.\\n23.\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 7\\n\\n\\n\\n b) administration of about 500 µg to about\\n 1500 µg of vitamin B12, prior to the first\\n administration of pemetrexed disodium;\\n and\\n c) administration of pemetrexed disodium.\\n In a parent application, Application 10/297,821 (the\\n“’821 application”), Lilly originally sought broad claims to\\nmethods of administering an antifolate in conjunction with\\na methylmalonic acid lowering agent, with or without folic\\nacid. The original independent claims 2 and 5 read:\\n 2. (Original) A method of reducing the toxicity as-\\n sociated with the administration of an antifolate to\\n a mammal comprising\\n administering to said mammal an effective\\n amount of said antifolate in combination\\n with a methylmalonic acid lowering agent.\\n 5. (Original) A method of reducing the toxicity as-\\n sociated with the administration of an antifolate to\\n a mammal comprising\\n administering to said mammal an effective\\n amount of said antifolate in combination\\n with a methylmalonic acid lowering agent\\n and FBP binding agent.\\nDRL J.A. 7860. A dependent claim further limited the an-\\ntifolate to pemetrexed disodium. Id. at 7861.\\n Claim 2 was rejected as anticipated by F.G. Arsenyan\\net al., “Influence of Methylcobalamin on the Antineoplastic\\nActivity of Methotrexate,” Onkol. Nauchn., 12(10):1299-\\n1303 (1978), which disclosed experiments treating mice\\nwith various tumors with a combination of methotrexate,\\nan antifolate, and methylcobalamin, a vitamin B12 deriva-\\ntive. The rest of the pending claims, including Claim 5,\\nwere rejected as obvious over a collection of references: U.S.\\nPatent 5,431,925 (“Ohmori”)—which taught treatment of\\n\\f8 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\nchemotherapeutically-induced immunosuppression with a\\ncombination of vitamins that could include folic acid and\\nvitamin B12—Worzalla, John, and Arsenyan. ’821 appli-\\ncation, Sept. 27, 2004, Office Action; DRL J.A. 7868–72.\\n In response, Lilly amended both claims to narrow “an-\\ntifolate” to “pemetrexed disodium” and cancelled its de-\\npendent claim limited to pemetrexed disodium. ’821\\napplication, Jan. 25, 2005, Response to Office Action; DRL\\nJ.A. 7877–84. In its remarks, Lilly asserted that the\\namendment to claim 2 overcame the anticipation rejection\\nbecause Arsenyan does not disclose pemetrexed disodium.\\nId. To overcome the obviousness rejection of claim 5 and\\nits dependents, Lilly generally argued that, while John dis-\\ncloses hematologic and immunologic toxicities from admin-\\nistration of pemetrexed disodium, it never suggests\\nvitamin supplementation, and none of the other references\\n“teach the use of [vitamin B12] to reduce toxicities associ-\\nated with an antifolate.” Id. The examiner then withdrew\\nthe anticipation rejection and later withdrew the obvious-\\nness rejection. The ’821 application issued as U.S. Patent\\n7,053,065, and the ’209 patent later issued from a continu-\\nation application.\\n These appeals were taken from cases which are among\\nthe latest in a series of patent disputes about Alimta® that\\nreaches back more than a decade. 3 In this most recent\\nchapter, DRL, Hospira, and Actavis 4 submitted New Drug\\n\\n\\n 3 This is the fourth appeal we have decided concern-\\ning Alimta® and the third specifically concerning the ’209\\npatent. See Neptune Generics, 921 F.3d 1372; Eli Lilly &\\nCo. v. Teva Parenteral Meds., Inc., 845 F.3d 1357 (Fed. Cir.\\n2017); Eli Lilly & Co. v. Teva Parenteral Meds., Inc., 689\\nF.3d 1368 (Fed. Cir. 2012).\\n 4 Lilly also sued Actavis LLC (“Actavis”) for infringe-\\nment of the ’209 patent, Eli Lilly & Co. v. Actavis LLC, No.\\n1:17-cv-00982-TWP-MPB (S.D. Ind. Mar. 30, 2017), ECF\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 9\\n\\n\\n\\nApplications under § 505(b)(2) of the Federal Food, Drug,\\nand Cosmetic Act, 21 U.S.C. § 355(b)(2), relying on Lilly’s\\nclinical data for pemetrexed disodium. But each applicant\\nseeks to market different pemetrexed salts—in DRL’s and\\nHospira’s applications, pemetrexed ditromethamine. Both\\nDRL and Hospira represented to the FDA that their choice\\nof the tromethamine cation was immaterial because\\npemetrexed dissociates from its counterion in solution,\\nDRL J.A. 8555–57; Hospira J.A. 124, and tromethamine\\nwas known to be safe for pharmaceutical use, DRL J.A.\\n8555, 8557.\\n Lilly then asserted the ’209 patent against each of\\nthese NDA applicants in the United States District Court\\nfor the Southern District of Indiana. In the DRL case, the\\ndistrict court construed the phrase “administration of\\npemetrexed disodium” to mean “liquid administration of\\npemetrexed disodium,” which “is accomplished by dissolv-\\ning the solid compound pemetrexed disodium into solu-\\ntion.” DRL Summary Judgment Decision, 2017 WL\\n6387316, at *4. The district court denied DRL’s motion for\\nsummary judgment of noninfringement, holding that pros-\\necution history estoppel does not bar Lilly from asserting\\nthat DRL’s proposed pemetrexed ditromethamine product\\nwould infringe through the doctrine of equivalents because\\nthe reason for Lilly’s amendment was to distinguish other\\nantifolates and was therefore only tangential to\\npemetrexed ditromethamine. Id. at *6–7. The district\\ncourt also rejected DRL’s argument that Lilly dedicated\\n\\n\\n\\nNo. 1, but the parties stipulated to be bound by the district\\ncourt’s decision in the DRL case that neither prosecution\\nhistory estoppel nor the disclosure-dedication rule bars\\nLilly’s assertion of infringement through the doctrine of\\nequivalents. Actavis Br. 2. Actavis filed a brief in the DRL\\nappeal as amicus curiae requesting reversal of that portion\\nof the district court’s decision.\\n\\f10 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\npemetrexed ditromethamine to the public under the disclo-\\nsure-dedication rule through its reference to Akimoto’s an-\\ntifolate compounds because Akimoto is not incorporated by\\nreference into the ’209 patent and in any event discloses\\npemetrexed ditromethamine only within a genus of thou-\\nsands of compounds, which the district court held does not\\nconstitute the requisite disclosure of an identifiable alter-\\nnative under this court’s precedent. Id. at *7–8; see, e.g.,\\nSanDisk Corp. v. Kingston Tech. Co., 695 F.3d 1348, 1363\\n(Fed. Cir. 2012).\\n Following a bench trial, the district court’s opinion\\nlargely followed its rationale in the DRL Summary Judg-\\nment Decision with respect to the applicability of prosecu-\\ntion history estoppel and the disclosure-dedication rule.\\nDRL Decision, 323 F. Supp. 3d at 1046–48. In addition, the\\ncourt found that DRL’s proposed product would be admin-\\nistered in a manner that would meet the “administration\\nof pemetrexed disodium” step of the asserted claims under\\nthe doctrine of equivalents, id. at 1049, regardless of the\\n“differences in chemical properties between pemetrexed\\ndisodium and pemetrexed ditromethamine,” id. at 1050.\\n In the Hospira case, the parties similarly disputed the\\ndoctrine of equivalents, but Lilly also asserted literal in-\\nfringement because Hospira’s proposed product label al-\\nlows reconstitution of its pemetrexed ditromethamine salt\\nin saline. Hospira Decision, 2018 WL 3008570, at *2–3;\\nHospira J.A. 229. After the district court issued the DRL\\nSummary Judgment Decision, Hospira conceded, contin-\\ngent upon its right to appeal, that its product would in-\\nfringe under the claim construction of “administration of\\npemetrexed disodium” set forth in that opinion and that its\\ndoctrine of equivalents arguments were likewise fore-\\nclosed. Hospira Br. 18. The district court, “rel[ying] heav-\\nily” on the DRL Summary Judgment Decision, granted\\nLilly’s motion for summary judgment of infringement, both\\nliterally and under the doctrine of equivalents. Hospira\\nDecision, 2018 WL 3008570, at *1 n.2, *6.\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 11\\n\\n\\n\\n These appeals followed. We have jurisdiction under 28\\nU.S.C. § 1295(a)(1).\\n DISCUSSION\\n We review a district court’s grant of summary judg-\\nment according to the law of the regional circuit. Kaneka\\nCorp. v. Xiamen Kingdomway Grp. Co., 790 F.3d 1298,\\n1303 (Fed. Cir. 2015) (citing Halo Elecs., Inc. v. Pulse El-\\necs., Inc., 769 F.3d 1371, 1377 (Fed. Cir. 2014)). In the Sev-\\nenth Circuit, summary judgment is reviewed de novo, con-\\nstruing all facts and drawing all inferences in favor of the\\nnon-movant. Wis. Alumni Research Found. v. Apple Inc.,\\n905 F.3d 1341, 1352 (Fed. Cir. 2018) (citing Austin v.\\nWalgreen Co., 885 F.3d 1085, 1087 (7th Cir. 2018)). On ap-\\npeal from a bench trial, we review a district court’s conclu-\\nsions of law de novo and its findings of fact for clear error.\\nBraintree Labs., Inc. v. Novel Labs., Inc., 749 F.3d 1349,\\n1358 (Fed. Cir. 2014) (citing Brown & Williamson Tobacco\\nCorp. v. Philip Morris Inc., 229 F.3d 1120, 1123 (Fed. Cir.\\n2000)). A factual finding is clearly erroneous if, despite\\nsome supporting evidence, we are left with the definite and\\nfirm conviction that a mistake has been made. United\\nStates v. U.S. Gypsum Co., 333 U.S. 364, 395 (1948).\\n Claim construction is ultimately an issue of law, which\\nwe review de novo. Shire Dev., LLC v. Watson Pharm., Inc.,\\n787 F.3d 1359, 1364 (Fed. Cir. 2015). We review de novo\\nthe district court’s findings of fact on evidence “intrinsic to\\nthe patent (the patent claims and specification[], along\\nwith the patent’s prosecution history),” and review for clear\\nerror extrinsic findings of fact. Teva Pharm. USA, Inc. v.\\nSandoz, Inc., 135 S. Ct. 831, 841 (2015). While infringe-\\nment is a question of fact, Lucent Techs., Inc. v. Gateway,\\nInc., 580 F.3d 1301, 1309 (Fed. Cir. 2009), we review de\\nnovo the district court’s grant of summary judgment of non-\\ninfringement, Unwired Planet, LLC v. Apple Inc., 829 F.3d\\n1353, 1356 (Fed. Cir. 2016). To prove infringement, a pa-\\ntentee “must supply sufficient evidence to prove that the\\n\\f12 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\naccused product or process contains, either literally or un-\\nder the doctrine of equivalents, every limitation of the\\nproperly construed claim.” Seal-Flex, Inc. v. Athletic Track\\n& Court Const., 172 F.3d 836, 842 (Fed. Cir. 1999). The\\npatentee has the burden of proving infringement by a pre-\\nponderance of the evidence. SmithKline Diagnostics, Inc.\\nv. Helena Labs. Corp., 859 F.2d 878, 889 (Fed. Cir. 1988).\\n Hospira requests reversal of the district court’s finding\\nthat its submission of a § 505(b)(2) NDA for its pemetrexed\\nproduct literally infringed the claims of the ’209 patent.\\nDRL and Hospira both argue, as does the amicus curiae\\nActavis, that the district court erred as a matter of law by\\nrefusing to apply prosecution history estoppel to bar Lilly’s\\ndoctrine of equivalents claim, and DRL further contends\\nthat the disclosure-dedication rule precludes Lilly’s equiv-\\nalents claim. Finally, DRL disputes the district court’s\\nfinding that administration of pemetrexed ditromethamine\\nis equivalent to the claim element “administration of\\npemetrexed disodium.” We address each argument in turn.\\n A. Literal Infringement\\n Hospira argues that it cannot literally infringe the\\nclaims of the ’209 patent because intravenous administra-\\ntion of pemetrexed ditromethamine dissolved in saline—a\\nsolution which contains pemetrexed and chloride anions\\nalongside sodium and tromethamine cations—is not “ad-\\nministration of pemetrexed disodium.” Hospira also notes\\nthat such a solution will, in any case, contain far more than\\ntwo sodium cations per pemetrexed anion. Finally, Hos-\\npira appears to make a perfunctory argument that, in the\\nalternative, we should reverse the district court’s construc-\\ntion and hold that the term encompasses any route of ad-\\nministering pemetrexed disodium, not just liquid, as the\\ndistrict court’s construction requires.\\n Lilly counters that Hospira’s view improperly imposes\\na “source limitation,” requiring that the pemetrexed diso-\\ndium salt exist in solid form before administration, even\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 13\\n\\n\\n\\nthough Hospira’s proposed product label, like that of Ali-\\nmta®, calls for administration of a solution containing\\npemetrexed anions and sodium cations. Lilly also contends\\nthat Hospira’s claim construction arguments are irrelevant\\nbecause Hospira’s proposed product will be administered\\nintravenously anyway.\\n We agree with Hospira. It was clearly erroneous for\\nthe district court to hold that the “administration of\\npemetrexed disodium” step was met because Hospira’s\\npemetrexed ditromethamine product will be dissolved in\\nsaline before administration. A solution of pemetrexed and\\nchloride anions and tromethamine and sodium cations can-\\nnot be deemed pemetrexed disodium simply because some\\nassortment of the ions in the solution consists of\\npemetrexed and two sodium cations. As Lilly acknowl-\\nedges throughout its brief, pemetrexed disodium is a salt.\\nSee, e.g., Lilly Br. I 12 (pemetrexed toxicity is caused “by\\npemetrexed itself once dissociated in solution,” not\\npemetrexed disodium); see also Hospira J.A. 1596 (October\\n2017 Alimta® Label referring to the drug substance as the\\n“disodium salt” of pemetrexed). Once diluted, the salt’s\\ncrystalline structure dissolves, and the individual ions dis-\\nsociate. See Hospira J.A. 2820 (declaration of Lilly’s ex-\\npert). In other words, pemetrexed disodium no longer\\nexists once dissolved in solution, and, as a corollary, a dif-\\nferent salt of pemetrexed dissolved in saline is not\\npemetrexed disodium.\\n We conclude that to literally practice the “administra-\\ntion of pemetrexed disodium” step under the district court’s\\nclaim construction, the pemetrexed disodium salt must be\\nitself administered. See DRL Summary Judgment Deci-\\nsion, 2017 WL 6387316, at *4 (“‘[A]dministration of\\npemetrexed disodium’ . . . refer[s] to a liquid administration\\nof pemetrexed disodium. . . ., accomplished by dissolving\\nthe solid compound pemetrexed disodium into solu-\\ntion . . . .”); see also Tex. Instruments Inc. v. Cypress Semi-\\nconductor Corp., 90 F.3d 1558, 1563 (Fed. Cir. 1996) (“To\\n\\f14 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\nliterally infringe, the accused . . . process must contain\\nevery limitation of the asserted claim.” (citing Laitram\\nCorp. v. Rexnord, Inc., 939 F.2d 1533, 1535 (Fed. Cir.\\n1991))). There is no dispute that Hospira has only sought\\napproval to market pemetrexed ditromethamine, Lilly Br.\\nI 4, and that neither its proposed product nor methods of\\nadministering it will constitute administering the\\npemetrexed disodium salt. Accordingly, Hospira will not\\npractice the step of “administration of pemetrexed diso-\\ndium,” and the district court’s finding of literal infringe-\\nment must be reversed.\\n B. Doctrine of Equivalents\\n Few propositions of patent law have been so consist-\\nently sustained by the Supreme Court as the doctrine of\\nequivalents. See Festo Corp. v. Shoketsu Kinzoku Kogyo\\nKabushki Co., 535 U.S. 722, 733 (2002) (“Festo VIII”)\\n(“[E]quivalents remain a firmly entrenched part of the set-\\ntled rights protected by the patent.”); Warner-Jenkinson\\nCo. v. Hilton Davis Chem. Co., 520 U.S. 17, 40 (1997) (“[W]e\\nadhere to the doctrine of equivalents.”); Graver Tank &\\nMfg. Co. v. Linde Air Prods. Co., 339 U.S. 605, 608 (1950)\\n(“Originating almost a century ago in the case of Winans v.\\nDenmead, [56 U.S. 330 (1853)] . . . [the doctrine of equiva-\\nlents] has been consistently applied by this Court and the\\nlower federal courts, and continues today ready and avail-\\nable for utilization when the proper circumstances for its\\napplication arise.”). It is settled that a patentee is entitled\\n“in all cases to invoke to some extent the doctrine of equiv-\\nalents,” Seymour v. Osborne, 78 U.S. 516, 555 (1870), with-\\nout a “judicial exploration of the equities of a case”\\nbeforehand. See Warner-Jenkinson, 520 U.S. at 34.\\n Yet the Supreme Court has also acknowledged that the\\ndoctrine of equivalents, “when applied broadly, conflicts\\nwith the definitional and public-notice functions of the stat-\\nutory claiming requirement,” Warner-Jenkinson, 520 U.S.\\nat 29, and that, without the proper balance between these\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 15\\n\\n\\n\\ntwo imperatives, the doctrine may “take[] on a life of its\\nown, unbounded by the patent claims.” See id. at 28–29.\\nWe have emphasized, moreover, that the doctrine of equiv-\\nalents is “the exception, however, not the rule,” and not\\nmerely “the second prong of every infringement charge,\\nregularly available to extend protection beyond the scope\\nof the claims.” London v. Carson Pirie Scott & Co., 946 F.2d\\n1534, 1538 (Fed. Cir. 1991). Patent infringement is princi-\\npally determined by examining whether the accused sub-\\nject matter falls within the scope of the claims.\\n To that end, courts have placed important limitations\\non a patentee’s ability to assert infringement under the\\ndoctrine of equivalents. See, e.g., Festo VIII, 535 U.S. at\\n737–41 (prosecution history estoppel); Warner-Jenkinson,\\n520 U.S. at 39 n.8 (“[A] theory of equivalence [cannot] en-\\ntirely vitiate a particular claim element . . . .”); Graver\\nTank, 339 U.S. at 608 (accused equivalent cannot differ\\nsubstantially from the claimed invention); Johnson &\\nJohnston Assocs. Inc. v. R.E. Serv. Co., 285 F.3d 1046, 1054\\n(Fed. Cir. 2002) (en banc) (subject matter disclosed but not\\nclaimed is dedicated to the public) (citing Maxwell v. J.\\nBaker, Inc., 86 F.3d 1098 (Fed. Cir. 1996)); Wilson Sporting\\nGoods Co. v. David Geoffrey & Assocs., 904 F.2d 677, 683\\n(Fed. Cir. 1990) (“[T]he asserted scope of equivalency [can-\\nnot] encompass the prior art . . . .” (Rich, J.) (citations omit-\\nted)). These appeals implicate several of these limitations.\\n 1. Prosecution History Estoppel\\n The main dispute in these appeals is whether Lilly has\\nrebutted the presumption of prosecution history estoppel\\nthat attached to its amendment in the ’821 application.\\nProsecution history estoppel arises when a patent appli-\\ncant narrows the scope of his claims during prosecution for\\na reason “substantial[ly] relating to patentability.” See\\ngenerally Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki\\nCo., 344 F.3d 1359, 1366–67 (Fed. Cir. 2003) (en banc)\\n(“Festo X”). Such a narrowing amendment is presumed to\\n\\f16 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\nbe a surrender of all equivalents within “the territory be-\\ntween the original claim and the amended claim,” but the\\npresumption is overcome if the patentee can show the ap-\\nplicability of one of the few exceptions identified by the Su-\\npreme Court. Festo VIII, 535 U.S. at 740–41 (citing Exhibit\\nSupply Co. v. Ace Patents Corp., 315 U.S. 126, 136–37\\n(1942)). Whether prosecution history estoppel applies to\\nbar a doctrine of equivalents claim is a question of law, re-\\nviewed de novo. See Regents of Univ. of Cal. v. Dakocyto-\\nmation Cal., Inc., 517 F.3d 1364, 1371 (Fed. Cir. 2008)\\n(citing Pharmacia & Upjohn Co. v. Mylan Pharm., Inc., 170\\nF.3d 1373, 1376 (Fed. Cir. 1999)).\\n Lilly does not dispute that the amendment in question\\nwas both narrowing and made for a substantial reason re-\\nlating to patentability. Lilly Br. II 21. Furthermore, Lilly\\nrelies on only one exception to giving effect to the presump-\\ntion as to the scope of surrender: that the rationale of its\\namendment “[bore] no more than a tangential relation to\\nthe equivalent in question.” Festo VIII, 535 U.S. at 740. As\\na result, the parties’ dispute about whether prosecution\\nhistory estoppel applies is confined to whether Lilly’s\\namendment narrowing “an antifolate” to “pemetrexed diso-\\ndium” was only tangential to pemetrexed ditromethamine,\\nwhich is the accused compound. Whether the tangential\\nexception applies is a question of law, Integrated Tech.\\nCorp. v. Rudolph Techs., Inc., 734 F.3d 1352, 1356 (Fed.\\nCir. 2013), and a patentee seeking to use the exception\\n“must base his arguments solely upon the public record of\\nthe patent’s prosecution.” Festo X, 344 F.3d at 1369–70 (ci-\\ntation omitted).\\n The Appellants argue that Lilly failed to explain why\\nit did not pursue a narrower amendment literally encom-\\npassing pemetrexed ditromethamine, and they emphasize\\nour statement that the tangential exception is “very nar-\\nrow.” Integrated, 734 F.3d at 1358 (quoting Cross Med.\\nProds., Inc. v. Medtronic Sofamor Danek, Inc., 480 F.3d\\n1335, 1342 (Fed. Cir. 2007)). The Appellants further point\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 17\\n\\n\\n\\nout that Lilly cannot be said to have “lacked the words to\\ndescribe” pemetrexed ditromethamine, see Festo VIII, 535\\nU.S. at 734, because Lilly’s previous patents, as well as the\\nEuropean companion to the ’209 patent, claimed\\npemetrexed salts generally and pemetrexed disodium in a\\ndependent claim. They also assert that the district court\\nerred by focusing on whether Lilly actually needed to relin-\\nquish pemetrexed ditromethamine to overcome the Arsen-\\nyan anticipation rejection because “the tangential\\nexception is not a patentee’s-buyer’s-remorse exception.”\\nDRL Br. 39.\\n In response, Lilly argues that the district court\\nproperly held that the reason for its amendment was to dis-\\ntinguish pemetrexed from antifolates generally and that\\nthe different salt type is a merely tangential change with\\nno consequence for pemetrexed’s administration or mecha-\\nnism of action within the body. Lilly also contends that it\\nis not barred from asserting the tangential exception\\nsimply because pemetrexed ditromethamine is within “the\\nterritory between the original claim and the amended\\nclaim.” Festo VIII, 535 U.S. at 740. Finally, Lilly argues\\nthat Appellants’ view that courts must “consider hypothet-\\nical alternative amendments” that would literally encom-\\npass the alleged equivalent “would eviscerate the\\ntangentiality exception.” Lilly Br. II 44.\\n We agree with Lilly. As a general matter, we find Ap-\\npellants’ view of prosecution history estoppel, and the tan-\\ngential exception in particular, too rigid. Tangential\\nmeans “touching lightly or in the most tenuous way.” Web-\\nster’s Third New International Dictionary (2002). The rea-\\nson for Lilly’s amendment, as the district court concluded,\\nwas to narrow original claim 2 to avoid Arsenyan, which\\nonly discloses treatments using methotrexate, a different\\nantifolate. See DRL J.A. 7879–80 (overcoming the Arsen-\\nyan anticipation rejection by arguing that it “does not dis-\\nclose pemetrexed disodium”). To overcome a clear\\nanticipation, Lilly opted to narrow its original claim 2 and\\n\\f18 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\nits dependents to more accurately define what it actually\\ninvented, an improved method of administering\\npemetrexed. In other words, the particular type of salt to\\nwhich pemetrexed is complexed relates only tenuously to\\nthe reason for the narrowing amendment, which was to\\navoid Arsenyan. We therefore hold that Lilly’s amendment\\nwas merely tangential to pemetrexed ditromethamine be-\\ncause the prosecution history, in view of the ’209 patent it-\\nself, strongly indicates that the reason for the amendment\\nwas not to cede other, functionally identical, pemetrexed\\nsalts.\\n The prosecution record confirms our understanding.\\nOriginal claim 5, which, like all the current claims of the\\n’209 patent, required supplementation with both vitamin\\nB12 and folic acid, was never rejected as anticipated over\\nArsenyan. Instead, the art cited against original claim 5\\nand its dependent claims in the obviousness ground of re-\\njection was replete with information about pemetrexed\\ndisodium; John disclosed clinical trials using pemetrexed\\ndisodium, reporting both its efficacy and its toxic side ef-\\nfects, and in response, DRL J.A. 7869–70, Worzalla sug-\\ngested folic acid supplementation to counteract these side\\neffects, DRL J.A. 7870–71. The prosecution record implies\\nthat Lilly’s amendment, inartful though it might have\\nbeen, was prudential in nature and did not need or intend\\nto cede other pemetrexed salts.\\n Hospira argues that the amendment was made to over-\\ncome the obviousness rejection over Ohmori and John and\\nthat Lilly has provided no reason for the amendment rela-\\ntive to that rejection. Like Lilly, we find this argument\\nmakes little sense. John discloses the results of a clinical\\ntrial of pemetrexed disodium and explicitly suggests the\\ntoxicities caused by pemetrexed; as we concluded above,\\nnarrowing “antifolate” to “pemetrexed disodium” could not\\npossibly distinguish the art cited in the obviousness ground\\nof rejection.\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 19\\n\\n\\n\\n DRL also insists that we have held that an applicant’s\\nremorse at ceding more claim scope than necessary is not a\\nreason for the tangential exception to apply. See, e.g., Lu-\\ncent Techs., Inc. v. Gateway, Inc., 525 F.3d 1200, 1218 (Fed.\\nCir. 2008); Schwarz Pharma, Inc. v. Paddock Labs., Inc.,\\n504 F.3d 1371, 1377 (Fed. Cir. 2007). This is generally\\ntrue, but DRL overreads the holdings of these cases. After\\nall, the tangential exception only exists because applicants\\nover-narrow their claims during prosecution. Amend-\\nments are not construed to cede only that which is neces-\\nsary to overcome the prior art, see Schwarz, 504 F.3d at\\n1377, nor will the court “speculat[e]” whether an amend-\\nment was necessary, see Kinzenbaw v. Deere & Co., 741\\nF.2d 383, 389 (Fed. Cir. 1984). But the reason for an\\namendment, where the tangential exception is invoked,\\ncannot be determined without reference to the context in\\nwhich it was made, including the prior art that might have\\ngiven rise to the amendment in the first place. See Festo X,\\n344 F.3d at 1370. Here, it is unlikely that a competitor\\nwould have been “justified in assuming that if he [made an\\nequivalent pemetrexed salt], he would not infringe [the\\n’209 patent].” Kinzenbaw, 741 F.2d at 389; cf. Festo VIII,\\n535 U.S. at 738 (“There is no reason why a narrowing\\namendment should be deemed to relinquish equivalents\\n. . . beyond a fair interpretation of what was surrendered.”).\\n Furthermore, Appellants’ suggestion that Lilly must\\nprove that it could not have drafted a claim that literally\\nencompassed pemetrexed ditromethamine is unsupported\\nby our precedent on prosecution history estoppel, not to\\nmention excessive. We do not demand perfection from pa-\\ntent prosecutors, and neither does the Supreme Court. See\\nFesto VIII, 535 U.S. at 738 (“It does not follow . . . that [an]\\namended claim becomes so perfect in its description that\\nno one could devise an equivalent.”). Lilly’s burden was to\\nshow that pemetrexed ditromethamine was “peripheral, or\\nnot directly relevant,” to its amendment, Festo X, 344 F.3d\\nat 1369. And as we concluded above, Lilly has done so.\\n\\f20 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\n In addition, the Appellants maintain that when a pa-\\ntentee submits an amendment adding two claim limita-\\ntions, it cannot later argue that the reason for the\\namendment was tangential to an accused equivalent con-\\ntaining only one of the added limitations simply because\\nthe second limitation was unnecessary to overcome the\\nprior art. They offer Felix v. American Honda Motor Co.,\\n562 F.3d 1167 (Fed. Cir. 2009), as an illustration of this\\nprinciple. 5 In that case, we held that prosecution history\\nestoppel applied to a claim directed to a vehicle bed storage\\nsystem—limited in response to a rejection to having a\\nchannel with a flange and a gasket mounted on that\\nflange—barring assertion of equivalence with respect to a\\nproduct that met the channel aspect, but not the gasket as-\\npect, of the limitation. Id. at 1184–85.\\n But as Lilly points out, this holding was determined by\\nthat patent’s prosecution history, Felix, 562 F.3d at 1184,\\nand we have also held that prosecution history estoppel\\ndoes not apply in similar circumstances, where the\\n\\n\\n\\n 5 The parties argue at length about which of our\\ncases are properly analogous to the facts presented in these\\nappeals. Here, in applying the Supreme Court’s frame-\\nwork, we find the analogies to other cases less helpful than\\na direct consideration of the specific record of this case and\\nwhat it shows about the reason for amendment and the re-\\nlation of that reason to the asserted equivalent. This case-\\nspecific focus, within the governing framework, comports\\nwith the equitable nature of prosecution history estoppel.\\nSee Festo VIII, 535 U.S. at 738 (“[The Supreme Court has]\\nconsistently applied the doctrine in a flexible way, not a\\nrigid one.”); cf. Heckler v. Cmty. Health Servs. of Crawford\\nCty., Inc., 467 U.S. 51, 59 (1984) (“Estoppel is an equitable\\ndoctrine invoked to avoid injustice in particular cases. . . .\\n[and] a hallmark of the doctrine is its flexible applica-\\ntion . . . .”).\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 21\\n\\n\\n\\nprosecution record differed. See, e.g., Regents, 517 F.3d at\\n1376–78 (amendment narrowing “disabling hybridization\\ncapacity of [nucleic acid] sequences” to methods using a\\n“blocking nucleic acid” was merely tangential to unclaimed\\nrepetitive sequence nucleic acids); Insituform Techs., Inc.\\nv. CAT Contracting, Inc., 385 F.3d 1360, 1368 (Fed. Cir.\\n2004) (amendment narrowing method of inserting resin\\ninto tube using a vacuum to one using “a cup” to do so was\\nmerely tangential to a multiple cup embodiment because\\nthe number of cups bore no relationship to the cited prior\\nart or the rationale behind the narrowing amendment).\\nThus, our cases demonstrate that prosecution history es-\\ntoppel is resistant to the rigid legal formulae that Appel-\\nlants seek to extract from them. See Intervet Inc. v. Merial\\nLtd., 617 F.3d 1282, 1291 (Fed. Cir. 2010) (“[T]here is no\\nhard-and-fast test for what is and what is not a tangential\\nrelation . . . .”).\\n Finally, DRL also contends that our precedent squarely\\nforecloses Lilly’s tangentiality argument, and it invites us\\nto read those cases to hold that “where the reason for the\\namendment and the equivalent in question both relate to\\nthe same claim element, the tangential exception does not\\napply.” DRL Br. 47. We decline this invitation because\\nsuch a bright-line rule is both contrary to the equitable na-\\nture of prosecution history estoppel, as articulated in Festo\\nVIII, 535 U.S. at 738, and inconsistent with the equitable\\nspirit that animates the doctrine of equivalents, see Graver\\nTank, 339 U.S. at 608–09 (the doctrine is one of “whole-\\nsome realism”). Instead, we reaffirm that whether an\\namendment was merely tangential to an equivalent must\\nbe decided in the context of the invention disclosed in the\\npatent and the prosecution history. Festo X, 344 F.3d at\\n1370.\\n DRL’s intuition—that an amendment that narrows an\\nexisting claim element evinces an intention to relinquish\\nthat claim scope—is often correct. Indeed, as we have\\nfound in previous cases, it is a powerful indication that an\\n\\f22 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\namendment was not merely tangential. See, e.g., Honey-\\nwell Int’l, Inc. v. Hamilton Sundstrand Corp., 523 F.3d\\n1304, 1315–16 (Fed. Cir. 2008); Biagro W. Sales, Inc. v.\\nGrow More, Inc., 423 F.3d 1296, 1306 (Fed. Cir. 2005). But\\nhere, we conclude that this consideration is not dispositive\\nbecause the rest of the prosecution history, and the ’209\\npatent itself, show that it is implausible that the reason for\\nLilly’s amendment was to surrender other pemetrexed\\nsalts. Indeed, such a relinquishment would effectively ded-\\nicate the entirety of Lilly’s invention to the public and\\nthereby render the ’209 patent worthless, and it would\\nhave been irrelevant for distinguishing the prior art.\\nAgain, the prosecution history strongly indicates a less\\nsweeping and more sensible reason for Lilly’s amendment:\\nto surrender antifolates other than pemetrexed. Thus, we\\nconclude on this prosecution record that Lilly’s amendment\\nwas merely tangential to pemetrexed ditromethamine.\\n 2. Disclosure-Dedication Rule\\n DRL next argues that the disclosure-dedication rule\\nbars Lilly from asserting infringement under the doctrine\\nof equivalents. The ’209 patent sets forth its invention as\\nan improved method of administering antifolates, ’209 pa-\\ntent col. 2 ll. 47–58, and teaches that the derivatives de-\\nscribed in the Akimoto patent are preferred examples of\\nantifolates, id. col. 4 ll. 34–40. DRL contends that one of\\nthese derivatives is pemetrexed ditromethamine and that\\nit was dedicated to the public when Lilly declined to claim\\nit. DRL asserts that the district court erred because it both\\nrequired express incorporation of Akimoto by reference\\ninto the ’209 patent and concluded that Akimoto does not\\nspecifically disclose pemetrexed ditromethamine.\\n Lilly counters that the disclosure-dedication rule re-\\nquires express disclosure of the subject matter in question\\nin the specification except in narrow circumstances, such\\nas when that subject matter is disclosed in a priority appli-\\ncation, see Abbott Labs. v. Sandoz, Inc., 566 F.3d 1282,\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 23\\n\\n\\n\\n1297 (Fed. Cir. 2009), or prior art expressly incorporated\\nby reference, SanDisk, 695 F.3d at 1366. Lilly also argues\\nthat the district court correctly determined that the rele-\\nvant portion of Akimoto discloses only a generic formula\\nfrom which a skilled artisan would not be able to recognize\\npemetrexed ditromethamine.\\n We agree with Lilly and hold that the disclosure-dedi-\\ncation rule is inapplicable to this case because the ’209 pa-\\ntent does not disclose methods of treatment using\\npemetrexed ditromethamine, and, as a result, Lilly could\\nnot have dedicated such a method to the public.\\n Under the disclosure-dedication rule, subject matter\\ndisclosed by a patentee, but not claimed, is considered ded-\\nicated to the public. See Johnson & Johnston, 285 F.3d at\\n1054. The reason for the doctrine is that members of the\\npublic reading a disclosure of particular subject matter are\\nentitled, absent a claim to it, to assume that it is not pa-\\ntented and therefore dedicated to the public (unless, for ex-\\nample, claimed in a continuation or other application based\\non the disclosure). Cf. Maxwell, 86 F.3d at 1107 (failure to\\nclaim inventive subject matter “is clearly contrary to 35\\nU.S.C. § 112, which requires that a patent applicant ‘par-\\nticularly point[] out and distinctly claim[] the subject mat-\\nter which the applicant regards as his invention’”). Subject\\nmatter is considered disclosed when a skilled artisan “can\\nunderstand the unclaimed disclosed teaching upon reading\\nthe written description,” but not “any generic reference . . .\\nnecessarily dedicates all members of that particular ge-\\nnus.” PSC Comput. Prod., Inc. v. Foxconn Int’l, Inc., 355\\nF.3d 1353, 1360 (Fed. Cir. 2004).\\n DRL further contends that the disclosure-dedication\\nrule does not impose a § 112 requirement for sufficiency of\\ndisclosure, see Toro Co. v. White Consol. Indus., Inc., 383\\nF.3d 1326, 1334 (Fed. Cir. 2004), and that a skilled artisan\\nreading the ’209 patent would both look for a disclosure of\\npemetrexed in Akimoto, and also seek to use a well-known\\n\\f24 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\ncation like tromethamine, which it maintains is generically\\ndisclosed in Akimoto in the form of “substituted ammo-\\nnium” base salts.\\n We are unpersuaded by DRL’s arguments. As the dis-\\ntrict court noted, Akimoto’s formula, col. 1 l. 49–col. 2 l. 3,\\nincludes seven functional group variables and encompasses\\nthousands of compounds, and while Akimoto discloses\\nabout fifty exemplary compounds, none of them is\\npemetrexed. Moreover, Akimoto does not even disclose tro-\\nmethamine expressly but only generically among dozens of\\nother salts. At most, Akimoto discloses ammonium salts\\ngenerally, which is far from a description of tromethamine.\\nIn similar circumstances, we have held that “sufficient de-\\nscription of a genus” requires that a skilled artisan be able\\nto “‘visualize or recognize’ the members of the genus.” See\\nAriad Pharm., Inc. v. Eli Lilly & Co., 598 F.3d 1336, 1350\\n(Fed. Cir. 2010) (quoting Regents of the Univ. of Cal. v. Eli\\nLilly & Co., 119 F.3d 1559, 1568–69 (Fed. Cir. 1997)).\\nAkimoto does not so describe pemetrexed ditromethamine,\\nand we see no reason why a skilled artisan would set out\\non DRL’s winding path to cobble together pemetrexed dit-\\nromethamine. While the ’209 patent teaches that\\npemetrexed disodium is the “most preferred” antifolate,\\nthat knowledge would not change the skilled artisan’s un-\\nderstanding of what Akimoto discloses.\\n Because Akimoto contains only a “generic reference” to\\npemetrexed ditromethamine, PSC Comput., 355 F.3d at\\n1360, we conclude that it was not dedicated to the public.\\n 3. Merits\\n A component in an accused product or process may be\\nequivalent to a claim element if the two are insubstantially\\ndifferent with respect to the “role played by [the] element\\nin the context of the specific patent claim.” Warner-Jen-\\nkinson, 520 U.S. at 39–40. Relevant differences can in-\\nclude the function each serves, the way in which each\\nworks, and the result each obtains, id. at 39, and, especially\\n\\fELI LILLY AND COMPANY v. HOSPIRA, INC. 25\\n\\n\\n\\nin biochemical cases, structural or pharmacological charac-\\nteristics, Mylan Inst. LLC v. Aurobindo Pharm. Ltd., 857\\nF.3d 858, 869 (Fed. Cir. 2017). “The determination of\\nequivalency vel non is a question of fact,” Canton Bio Med.,\\nInc. v. Integrated Liner Techs., Inc., 216 F.3d 1367, 1369\\n(Fed. Cir. 2000) (citing Pall Corp. v. Micron Separations,\\nInc., 66 F.3d 1211, 1218 (Fed. Cir. 1995)), which we review\\nfor clear error in an appeal from a bench trial, Pfizer, Inc.\\nv. Apotex, Inc., 480 F.3d 1348, 1359 (Fed. Cir. 2007).\\n DRL argues that the district court erred in finding that\\nits proposed pemetrexed ditromethamine product will be\\nadministered in an insubstantially different way from the\\nclaimed method. DRL maintains that the district court fo-\\ncused on the fact that each product treats the same dis-\\neases by delivering pemetrexed intravenously, when the\\nrelevant context is the manner of administration. In DRL’s\\nview, the chemical differences between sodium and tro-\\nmethamine—e.g., pH, buffering capacity, or solubility—\\nDRL Br. 20–21, render the methods in which each is ad-\\nministered to a patient substantially different.\\n Lilly responds that the relevant context is treatment of\\na patient “in need of chemotherapeutic treatment.” ’209\\npatent claim 12. Lilly agrees with the district court that\\nthe chemical differences between sodium and trometham-\\nine are clinically irrelevant because each undisputedly\\nlacks therapeutic activity.\\n We see no clear error in the district court’s findings. As\\nthe district court found, DRL’s product will accomplish an\\nidentical aim, furnishing the same amount of pemetrexed\\nto active sites in the body; in exactly the same way, by di-\\nluting a pemetrexed salt in an aqueous solution for intra-\\nvenous administration. Indeed, after dilution and\\nimmediately before administration, DRL’s product is func-\\ntionally identical to Lilly’s in that it contains the same\\namount of diluted pemetrexed anion. DRL J.A. 8557. And\\nDRL declines to identify the relevance of any of the\\n\\f26 ELI LILLY AND COMPANY v. HOSPIRA, INC.\\n\\n\\n\\n\\nchemical differences it identifies. See UCB, Inc. v. Watson\\nLabs. Inc., 927 F.3d 1272, 1284–86 (Fed. Cir. 2019) (chem-\\nical differences may not be relevant if the equivalent has\\nknown interchangeability in the context of the claimed\\ncomposition). We find DRL’s arguments unconvincing and\\ntherefore affirm the district court’s findings.\\n In summary, these cases are eminently suitable for ap-\\nplication of the doctrine of equivalents, and we conclude\\nthat neither prosecution history estoppel nor the disclo-\\nsure-dedication rule bars Lilly from asserting infringement\\nthrough equivalence.\\n CONCLUSION\\n We have fully considered each party’s further argu-\\nments but find them unpersuasive. For the foregoing rea-\\nsons, we reverse the district court’s finding of literal\\ninfringement in the Hospira Decision but affirm its judg-\\nment of infringement under the doctrine of equivalents.\\nThe judgment of infringement under the doctrine of equiv-\\nalents in the DRL Decision is likewise affirmed.\\nAFFIRMED-IN-PART AND REVERSED-IN-PART IN\\n APPEAL NOS. 2018-2126, 2018-2127\\n AFFIRMED IN APPEAL NO. 2018-2128\\n COSTS\\n Each party shall bear its own costs.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4198,"cells":{"id":{"kind":"number","value":4647451,"string":"4,647,451"},"name":{"kind":"string","value":"MTD PRODUCTS INC., Appellant v. Andrei IANCU, Under Secretary of Commerce for Intellectual Property and Director of the United States Patent and Trademark Office, Intervenor"},"name_abbreviation":{"kind":"string","value":"Mtd Products Inc. v. Iancu"},"decision_date":{"kind":"string","value":"2019-08-12"},"docket_number":{"kind":"string","value":"2017-2292"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Federal Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Reyna, Taranto, Stoll\", \"parties\": \"\", \"opinions\": [{\"author\": \"STOLL, Circuit Judge.\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n MTD PRODUCTS INC.,\\n Appellant\\n\\n v.\\n\\n ANDREI IANCU, UNDER SECRETARY OF\\n COMMERCE FOR INTELLECTUAL PROPERTY\\n AND DIRECTOR OF THE UNITED STATES\\n PATENT AND TRADEMARK OFFICE,\\n Intervenor\\n ______________________\\n\\n 2017-2292\\n ______________________\\n\\n Appeal from the United States Patent and Trademark\\nOffice, Patent Trial and Appeal Board in No. IPR2016-\\n00194.\\n ______________________\\n\\n Decided: August 12, 2019\\n ______________________\\n\\n JOHN SALVATORE CIPOLLA, Calfee, Halter & Griswold\\nLLP, Cleveland, OH, argued for appellant. Also repre-\\nsented by ANDREW ALEXANDER, TRACY SCOTT JOHNSON,\\nMARK MCDOUGALL.\\n\\n PETER JOHN SAWERT, Office of the Solicitor, United\\nStates Patent and Trademark Office, Alexandria, VA, ar-\\ngued for intervenor. Also represented by THOMAS W.\\nKRAUSE, PHILIP J. WARRICK.\\n\\f2 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\n ______________________\\n\\n Before REYNA, TARANTO, and STOLL, Circuit Judges.\\nSTOLL, Circuit Judge.\\n The Toro Company sought inter partes review of\\nclaims 1–16 of U.S. Patent No. 8,011,458 before the U.S.\\nPatent and Trademark Office’s Patent Trial and Appeal\\nBoard. The Board instituted review and, in its final writ-\\nten decision, held the challenged claims obvious under\\n35 U.S.C. § 103. Critical to its decision, the Board deter-\\nmined that the claim term “mechanical control assem-\\nbly . . . configured to” perform certain functions is not a\\nmeans-plus-function term subject to 35 U.S.C. § 112, ¶ 6.\\nMTD Products Inc., owner of the ’458 patent, appeals the\\nBoard’s decision.\\n We conclude that the Board erred by conflating corre-\\nsponding structure in the specification with a structural\\ndefinition for the term, and by misinterpreting certain\\nstatements in the prosecution history. Under the appropri-\\nate legal framework, we conclude that the term “mechani-\\ncal control assembly” is a means-plus-function term\\ngoverned by § 112, ¶ 6. We therefore vacate the Board’s\\nobviousness conclusion, which was predicated on its incor-\\nrect claim construction, and remand for further proceed-\\nings consistent with this opinion. Because we are\\npersuaded by MTD’s primary argument, we do not reach\\nits alternative arguments.\\n BACKGROUND\\n I\\n The ’458 patent discloses a steering and driving system\\nfor zero turn radius (“ZTR”) vehicles, with specific refer-\\nence to ZTR lawn mowers. ’458 patent col. 1 ll. 17–21. The\\npatented system is designed to provide a more intuitive\\nsteering mechanism to operators of ZTR vehicles. Id. at\\ncol. 1 ll. 20–38. In contrast to prior art systems that\\n\\fMTD PRODS. INC. v. IANCU 3\\n\\n\\n\\nreverse in the opposite direction of a forward motion turn,\\nthe claimed invention permits ZTR vehicles to turn in the\\nsame direction both forward and backwards. Id. at col. 1\\nll. 20–47. The claimed steering mechanism thus mimics\\nthe forward and backward movements of an automobile.\\n The term “mechanical control assembly” appears in\\nboth claims 1 and 9, the only independent claims of the\\n’458 patent. Claim 1 recites:\\n 1. A vehicle capable of making a small radius turn,\\n comprising:\\n a frame;\\n a left drive wheel and a right drive wheel,\\n both coupled to the frame;\\n two independent left and right drive units,\\n the left drive unit coupled to the left drive\\n wheel via an axle and the right drive unit\\n coupled to the right drive wheel via another\\n axle;\\n a steering device coupled to the frame;\\n a speed control member coupled to the\\n frame; and\\n a mechanical control assembly coupled to\\n the left and right drive units that is config-\\n ured to actuate the left and right drive\\n units based on a steering input received\\n from the steering device and a speed input\\n received from the speed control member;\\n the mechanical control assembly being con-\\n figured such that if the speed control mem-\\n ber is shifted from (a) a forward position in\\n which the left drive wheel is rotating in a\\n forward direction at a first forward speed\\n and the right drive wheel is rotating in a\\n\\f4 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\n forward direction at a second forward\\n speed that is less than the first forward\\n speed as a result of the steering device be-\\n ing in a first right turn position to (b) a re-\\n verse position while the first right turn\\n position of the steering device is main-\\n tained, then the left drive wheel will rotate\\n in a reverse direction at a first reverse\\n speed and the right drive wheel will rotate\\n in a reverse direction at a second reverse\\n speed that is less than the first reverse\\n speed.\\nId. at col. 7 l. 63–col. 8 l. 24 (emphasis added to highlight\\nportion of disputed claim term). Claim 9 is identical to\\nclaim 1 in substantial part, adding only the further limita-\\ntion of:\\n the mechanical control assembly also being config-\\n ured to cause the vehicle to execute a zero-radius\\n turn when the speed control member is in a maxi-\\n mum forward position and the steering device is in\\n a maximum turn position.\\nId. at col. 9 ll. 13–16 (emphasis added).\\n While the patent specification does not expressly refer\\nto a “mechanical control assembly,” it discloses a preferred\\nembodiment that includes a “ZTR control assembly.” Id. at\\ncol. 3 ll. 41–42. The specification describes components of\\nthe ZTR control assembly and its inputs, outputs, and link-\\nages. Id. at col. 3 l. 41–col. 4 l. 57.\\n II\\n Toro petitioned for inter partes review of the ’458 pa-\\ntent in November 2015, arguing that the challenged claims\\nwere invalid as anticipated or obvious. MTD responded\\nthat the term “mechanical control assembly” is a means-\\nplus-function term, and that the asserted prior art did not\\ndisclose the claim term’s corresponding structure. In\\n\\fMTD PRODS. INC. v. IANCU 5\\n\\n\\n\\nsupport of its argument, MTD introduced expert testimony\\nindicating that “mechanical control assembly” has no rea-\\nsonably well-understood meaning in the art. Specifically,\\nMTD’s expert testified that “mechanical control assembly”\\nis a nonce term that is not used in common parlance and\\ndoes not bring to mind any specific structure to a person of\\nordinary skill in the art. J.A. 1366. He explained that the\\nterm is used as a black box recitation for structure and, at\\nmost, amounts to a collection of various parts. J.A. 1248,\\n1366. He further demonstrated that the term is used in\\nvarious prior patents and publications to describe a wide\\nvariety of structures with varying functions. J.A. 1367–69\\n(noting that “mechanical control assembly” is used generi-\\ncally to describe mechanisms for infusion pumps, digital\\nfiring systems, flush tanks, endoscopes, transmissions, and\\nengine outputs).\\n Toro did not expressly contradict MTD’s evidence that\\n“mechanical control assembly” did not have a well-under-\\nstood structural meaning. Instead, Toro responded that a\\nperson of ordinary skill in the art would understand the\\nterm to denote a specific structure in the context of the\\n’458 patent specification. Specifically, Toro argued that the\\n“ZTR control assembly” disclosed in the specification pro-\\nvides an express structural definition for the claimed “me-\\nchanical control assembly.” J.A. 2201–03. Toro also\\nargued that MTD admitted that the term “mechanical con-\\ntrol assembly” conveys particular structure when it distin-\\nguished the patent claims from a prior art reference during\\nprosecution. J.A. 2203.\\n The Board initially agreed with MTD, stating that\\nwhen viewed “in isolation, the genericness of this term\\nbears similarities to other words or phrases that have been\\nheld to be subject to § 112, ¶ 6 . . . such as ‘mechanism,’ ‘el-\\nement,’ ‘device,’ ‘link member,’ and ‘control mechanism.’”\\nToro Co. v. MTD Prods. Inc., No. IPR2016-00194,\\n2017 WL 1969747, at *9 (P.T.A.B. May 10, 2017) (first cit-\\ning Williamson v. Citrix Online, LLC, 792 F.3d 1339, 1350\\n\\f6 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\n(Fed. Cir. 2015) (en banc); then citing Mas-Hamilton Grp.\\nv. LaGard, Inc., 156 F.3d 1206, 1215 (Fed. Cir. 1998); and\\nthen citing Toro Co. v. Deere & Co., 355 F.3d 1313, 1325\\n(Fed. Cir. 2004)). The Board also determined that the “lan-\\nguage reciting what the mechanical control assembly is\\n‘configured to’ do . . . fits the mold of functional language\\nbecause it describes the mechanical control assembly by\\nwhat it does.” Id. The Board thus concluded that “the\\nclaim language of the disputed phrase is primarily, but not\\nentirely, functional, which tends to favor [MTD]’s position\\nthat § 112, ¶ 6 applies.” Id.\\n The Board agreed with Toro, however, that the ’458 pa-\\ntent specification weighed against application of § 112, ¶ 6.\\nId. at *10. Citing the parties’ briefs, the Board stated that\\n“[t]he parties agree that the claimed ‘mechanical control\\nassembly’ is referred to in the specification as a ‘ZTR con-\\ntrol assembly.’” Id. According to the Board, a person of\\nordinary skill in the art would understand “mechanical\\ncontrol assembly” to denote structure because “the specifi-\\ncation illustrates and describes the specific structure that\\nmakes up the ZTR control assembly, and how it connects to\\nand operates with other components.” Id.\\n The most persuasive piece of evidence to the Board,\\nhowever, was the prosecution history. Id. at *9 (“The factor\\nthat weighs most heavily in [the] determination is the pros-\\necution history.”). According to the Board, MTD admitted\\nthat the term “mechanical control assembly” connotes spe-\\ncific structure by asserting that the claims recite “a me-\\nchanical control assembly that is structurally different\\nfrom what [the asserted prior art] discloses.” Id. The\\nBoard emphasized MTD’s statements that “the claim lan-\\nguage at issue concerns the configuration of the claimed\\nmechanical control assembly” and “the claimed configura-\\ntion is indeed structural.” Id. (emphases in original). The\\nBoard concluded that these statements “present[] strong\\nevidence that the disputed phrase should be understood as\\na structural limitation rather than a means-plus-function\\n\\fMTD PRODS. INC. v. IANCU 7\\n\\n\\n\\nlimitation under § 112 ¶ 6.” Id. at *11. Relying on the spec-\\nification and prosecution history, the Board ultimately de-\\ntermined that “mechanical control assembly” is not\\ngoverned by § 112, ¶ 6. Id.\\n DISCUSSION\\n I\\n This appeal requires us to address whether a particu-\\nlar claim limitation is drafted in means-plus-function for-\\nmat. Whether claim language invokes 35 U.S.C. § 112, ¶ 6 1\\nis a legal question of claim construction that we review de\\nnovo. Williamson, 792 F.3d at 1346. We review the\\nBoard’s factual findings underlying this inquiry for sub-\\nstantial evidence. EnOcean, GmbH v. Face Int’l Corp.,\\n742 F.3d 955, 959 (Fed. Cir. 2014).\\n Under this court’s guidance in Williamson, we begin by\\nasking whether the claim limitation employs the word\\n“means.” Williamson, 792 F.3d at 1348. If it does not, we\\napply a rebuttable presumption that the term conveys suf-\\nficiently definite structure and is not subject to § 112, ¶ 6.\\nId. A challenger can rebut the presumption by demonstrat-\\ning “that the claim term fails to ‘recite sufficiently definite\\nstructure’ or else recites ‘function without reciting suffi-\\ncient structure for performing that function.’” Id. (quoting\\nWatts v. XL Sys., Inc., 232 F.3d 877, 880 (Fed. Cir. 2000)).\\nThe “essential inquiry is not merely the presence or ab-\\nsence of the word ‘means’ but whether the words of the\\n\\n\\n 1 Because the issue date of the ’458 patent is Septem-\\nber 6, 2011, and neither the ’458 patent nor the application\\nfrom which it issued ever contained a claim with an effec-\\ntive filing date on or after September 16, 2012, the version\\nof 35 U.S.C. § 112 that applies here is the one preceding\\nthe changes made by the America Invents Act. See Leahy-\\nSmith America Invents Act, Pub. L. No. 112-29, 125 Stat.\\n284, 296, § 4(c) (2011).\\n\\f8 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\nclaim are understood by persons of ordinary skill in the art\\nto have a sufficiently definite meaning as the name for\\nstructure.” Id.\\n One way to demonstrate that a claim limitation fails to\\nrecite sufficiently definite structure is to show that, alt-\\nhough not employing the word “means,” the claim limita-\\ntion uses a similar “nonce word that can operate as a\\nsubstitute for ‘means’ in the context of § 112, para. 6.” Id.\\nat 1350. Generic terms like “module,” “mechanism,” “ele-\\nment,” and “device” are commonly used as verbal con-\\nstructs that operate, like “means,” to claim a particular\\nfunction rather than describe a “sufficiently definite struc-\\nture.” Id. Our case law is replete with guidance on\\nwhether or not a particular claim term is a “nonce” term.\\nSee, e.g., Zeroclick, LLC v. Apple Inc., 891 F.3d 1003, 1008\\n(Fed. Cir. 2018) (holding that “program” and “user inter-\\nface code” are not nonce words because they are “used not\\nas generic terms or black box recitations of structure or ab-\\nstractions, but rather as specific references to conventional\\ngraphical user interface programs or code”); Williamson,\\n792 F.3d at 1350 (noting that “module” is a nonce term be-\\ncause it “sets forth the same black box recitation of struc-\\nture for providing the same specified function as if the term\\n‘means’ had been used”). In each case, a critical question\\nis whether “the claim term is used in common parlance or\\nby persons of skill in the pertinent art to designate struc-\\nture,” including either a particular structure or a class of\\nstructures. Skky, Inc. v. MindGeek, s.a.r.l., 859 F.3d 1014,\\n1019 (Fed. Cir. 2017) (citing TecSec, Inc. v. Int’l Bus.\\nMachs. Corp., 731 F.3d 1336, 1347 (Fed. Cir. 2013)) (hold-\\ning that the claim term “wireless device means” does not\\ninvoke § 112, ¶ 6 because it denotes a class of structures).\\n In addition, even if the claims recite a nonce term fol-\\nlowed by functional language, other language in the claim\\n“might inform the structural character of the limitation-in-\\nquestion or otherwise impart structure” to the claim term.\\nWilliamson, 792 F.3d at 1351. In assessing whether the\\n\\fMTD PRODS. INC. v. IANCU 9\\n\\n\\n\\nclaim limitation is in means-plus-function format, we do\\nnot merely consider the introductory phrase (e.g., “mechan-\\nical control assembly”) in isolation, but look to the entire\\npassage including functions performed by the introductory\\nphrase. See Apex Inc. v. Raritan Computer, Inc., 325 F.3d\\n1364 (Fed. Cir. 2003). The ultimate question is whether\\n“the claim language, read in light of the specification, re-\\ncites sufficiently definite structure to avoid § 112, ¶ 6.” Me-\\ndia Rights Techs. Inc. v. Capital One Fin. Corp., 800 F.3d\\n1366, 1372 (Fed. Cir. 2015) (citing Robert Bosch, LLC v.\\nSnap-On Inc., 769 F.3d 1094, 1099 (Fed. Cir. 2014)).\\n In Apex, for example, the court concluded that the term\\n“circuit” recited sufficient structure in the context of the\\nclaims at issue. 325 F.3d at 1372–73 (considering the use\\nof the term “circuit” in the claim limitation “a first interface\\ncircuit for receiving keyboard and cursor control device sig-\\nnals from the workstation”). Based on a dictionary defini-\\ntion of the word “circuit,” the court reasoned that “the term\\n‘circuit’ by itself connotes some structure,” and that “the\\nterm ‘circuit with an appropriate identifier such as ‘inter-\\nface,’ ‘programming,’ and ‘logic,’ certainly identifies some\\nstructural meaning to one of ordinary skill in the art.” Id.\\nat 1373. The court noted that the extrinsic evidence did\\nnot show that the term “circuit” was not understood to have\\nstructure, but rather “only that the term ‘circuit’ is under-\\nstood . . . as a very broad term.” Id. at 1374. As neither the\\nspecification nor the prosecution history used the term “in\\na manner clearly inconsistent with the ordinary meaning,”\\nthe court held that the defendant failed to rebut the pre-\\nsumption that § 112, ¶ 6 did not apply. Id. at 1373–74.\\n In contrast, in Diebold Nixdorf, Inc. v. International\\nTrade Commission, the court held that the term “cheque\\nstandby unit” for performing certain specified functions\\nwas governed by § 112, ¶ 6. 899 F.3d 1291, 1300 (Fed. Cir.\\n2018). The court noted that “there is no evidence—in the\\nform of dictionary definitions or otherwise—that ‘cheque\\nstandby unit’ was reasonably well understood by persons\\n\\f10 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\nof ordinary skill in the art to refer to a structure or class of\\nstructures.” Id. at 1302. Instead, the extrinsic evidence\\ndemonstrated only that “a skilled artisan would under-\\nstand the functional term ‘cheque standby unit’ to be any\\nstructure capable of performing the claimed function.” Id.\\nat 1301. Further, neither the words of the claim nor the\\nspecification suggested a “structural limitation that might\\nserve to cabin the scope of the functional term,” thus sup-\\nporting the conclusion that the claim limitation was writ-\\nten in means-plus-function format. Id.\\n Finally, we note that “[c]laims are interpreted in light\\nof the written description supporting them, and that is true\\nwhether or not the claim construction involves interpreting\\na ‘means’ clause.” Inventio AG v. ThyssenKrupp Elevator\\nAms. Corp., 649 F.3d 1350, 1356 (Fed. Cir. 2011), overruled\\non other grounds by Williamson, 792 F.3d at 1339. For ex-\\nample, a patentee may avoid application of § 112, ¶ 6 by\\nacting as a lexicographer and providing its own structural\\ndefinition of a nonce term in the specification by “‘clearly\\nset[ting] forth a definition of the disputed claim term’ other\\nthan its plain and ordinary meaning.” Thorner v. Sony\\nComput. Entm’t Am. LLC, 669 F.3d 1362, 1365 (Fed. Cir.\\n2012) (citing CCS Fitness, Inc. v. Brunswick Corp.,\\n288 F.3d 1359, 1366 (Fed. Cir. 2002)). Just as it is im-\\nproper to “import[] limitations from the specification into\\nthe claims,” Phillips v. AWH Corp., 415 F.3d 1303, 1323\\n(Fed. Cir. 2005), however, a preferred embodiment dis-\\nclosed in the specification cannot impart structure to a\\nterm that otherwise has none. As with all lexicography,\\n“[i]t is not enough for a patentee to simply disclose a single\\nembodiment.” Thorner, 669 F.3d at 1365. Rather, “the pa-\\ntentee must ‘clearly express an intent’ to redefine the\\nterm.” Id. (quoting Helmsderfer v. Bobrick Washroom\\nEquip., Inc., 527 F.3d 1379, 1381 (Fed. Cir. 2008)).\\n With these background principles in mind, we turn to\\nthe claim language at issue in this case.\\n\\fMTD PRODS. INC. v. IANCU 11\\n\\n\\n\\n II\\n The disputed claim limitation is lengthy and recites:\\n a mechanical control assembly coupled to the left\\n and right drive units that is configured to actuate\\n the left and right drive units based on a steering\\n input received from the steering device and a speed\\n input received from the speed control member;\\n the mechanical control assembly being configured\\n such that if the speed control member is shifted\\n from (a) a forward position in which the left drive\\n wheel is rotating in a forward direction at a first\\n forward speed and the right drive wheel is rotating\\n in a forward direction at a second forward speed\\n that is less than the first forward speed as a result\\n of the steering device being in a first right turn po-\\n sition to (b) a reverse position while the first right\\n turn position of the steering device is maintained,\\n then the left drive wheel will rotate in a reverse di-\\n rection at a first reverse speed and the right drive\\n wheel will rotate in a reverse direction at a second\\n reverse speed that is less than the first reverse\\n speed.\\n’458 patent col. 8 ll. 7–24.\\n At the outset, we agree with the Board that the term\\n“mechanical control assembly” is similar to other generic,\\nblack-box words that this court has held to be nonce terms\\nsimilar to “means” and subject to § 112, ¶ 6 because the\\nterm does not connote sufficiently definite structure to one\\nof ordinary skill in the art. Toro, 2017 WL 1969747, at *9.\\nWe also agree with the Board that the rest of the “claim\\nlanguage of the disputed phrase is primarily, but not en-\\ntirely, functional.” Id. While the claim language reciting\\nthat the mechanical control assembly is “coupled to the left\\nand right drive units” connotes structure, the claim lan-\\nguage reciting what the mechanical control assembly is\\n\\f12 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\n“configured to” do is functional. In this respect, as the\\nBoard correctly recognized, the claim format tends to favor\\nMTD’s position that § 112, ¶ 6 applies. 2\\n We also discern no error in the Board’s acceptance of\\nMTD’s extrinsic evidence as showing that the term “me-\\nchanical control assembly” does not have an established\\nmeaning in the art and instead merely operates as a ge-\\nneric label for a collection of parts. Toro did not dispute\\nMTD’s expert testimony that, in common parlance, “me-\\nchanical control assembly” does not bring to mind any spe-\\ncific structure to a person of ordinary skill in the art. J.A.\\n1366. Toro likewise did not dispute MTD’s reliance on var-\\nious prior patents and publications that used “mechanical\\ncontrol assembly” to describe a wide variety of structures\\nwith varying functions. J.A. 1367–69 (noting that\\n\\n\\n\\n 2 The Board also stated that construing the disputed\\nphrase as a means-plus-function limitation “would seem to\\nrender the second part [of the claim’s recitation of a ‘me-\\nchanical control assembly’] . . . superfluous” as it “would no\\nlonger serve to define functionally the structural features\\nof the mechanical control assembly, as those features\\nwould be part and parcel of what [MTD] alleges is the cor-\\nresponding structure.” Toro, 2017 WL 1969747, at *9. We\\ndisagree. Both the first and second limitations following\\nthe two occurrences of the term “mechanical control assem-\\nbly” recite functions associated with the “mechanical con-\\ntrol assembly”; i.e., (1) actuate the left and right drive units\\nand (2) rotate the wheel in a particular direction based on\\nthe position of the speed control member. As we have held,\\n“[w]here there are multiple claimed functions, as we have\\nhere, the patentee must disclose adequate corresponding\\nstructure to perform all of the claim functions.” William-\\nson, 792 F.3d at 1351–52. Thus, the corresponding struc-\\nture for “mechanical control assembly” in the specification\\nmust perform both of these functions.\\n\\fMTD PRODS. INC. v. IANCU 13\\n\\n\\n\\n“mechanical control assembly” is used generically to de-\\nscribe mechanisms for infusion pumps, digital firing sys-\\ntems, flush tanks, endoscopes, transmissions, and engine\\noutputs).\\n We conclude that the Board erred, however, when it re-\\nlied on the specification’s description of a “ZTR control as-\\nsembly” to conclude that the claim term “mechanical\\ncontrol assembly” has an established structural meaning.\\nWhile the parties agreed that the ZTR control assembly in\\nthe specification is the structure “corresponding to” the\\nclaimed mechanical control assembly, MTD did not agree\\nthat the specification expressly defines the claim term “me-\\nchanical control assembly.” That the specification discloses\\na structure corresponding to an asserted means-plus-func-\\ntion claim term does not necessarily mean that the claim\\nterm is understood by persons of ordinary skill in the art to\\nconnote a specific structure or a class of structures.\\n Interpretation of an asserted means-plus-function lim-\\nitation involves two steps. First, we determine if the claim\\nlimitation is drafted in means-plus-function format. As\\npart of this step, we consider whether the claim limitation\\nconnotes “sufficiently definite structure” to a person of or-\\ndinary skill in the art. If we conclude that the limitation is\\nin means-plus-function format, the second step requires us\\nto review the specification to identify the structure that\\nperforms the claimed function(s) and thus “corresponds to”\\nthe claimed means. While related, these two inquiries are\\ndistinct. In this case, however, the Board conflated these\\ndistinct inquiries, holding that the specification’s disclo-\\nsure of corresponding structure demonstrates that the al-\\nleged means-plus-function term is sufficiently definite so\\nas to not invoke § 112, ¶ 6. The Board’s analysis implies\\nthat so long as a claim term has corresponding structure in\\nthe specification, it is not a means-plus-function limitation.\\nThis is not consistent with our prior decisions. Indeed, this\\nview would seem to leave § 112, ¶ 6 without any applica-\\ntion: any means-plus-function limitation that met the\\n\\f14 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\nstatutory requirements, i.e., which includes having corre-\\nsponding structure in the specification, would end up not\\nbeing a means-plus-function limitation at all.\\n While we agree with the Board that the specification\\nplays a role in assessing whether particular claim language\\ninvokes § 112, ¶ 6, we do not agree that the patent specifi-\\ncation at issue here renders the nonce term “mechanical\\ncontrol assembly” sufficiently structural to a person of or-\\ndinary skill in the art. The specification does not demon-\\nstrate that the patentee intended to act as its own\\nlexicographer and define the nonce term “mechanical con-\\ntrol assembly” as the “ZTR control assembly” of the pre-\\nferred embodiment. Indeed, the specification does not even\\nrefer to a “mechanical control assembly.” Furthermore, the\\nfunctional language in the claim limitation suggests a\\nbroader meaning of the generic term “mechanical control\\nassembly,” as it specifically adds to the “mechanical control\\nassembly” limitation the ability to execute a zero radius\\nturn. ’458 patent col. 9 ll. 13–16. Interpreting the “me-\\nchanical control assembly” as the “ZTR”—or zero-turn-ra-\\ndius—control assembly would render this functional\\nlanguage superfluous.\\n We are also not persuaded by the Board’s interpreta-\\ntion of the prosecution history. While it would have\\navoided uncertainty and argument had MTD shared its\\ncurrent view that the claim limitation is written in means-\\nplus-function format during the original prosecution,\\nMTD’s statements did not clearly disclaim such an inter-\\npretation. Rather, MTD’s statements indicated that the\\nphrase “mechanical control assembly configured to” per-\\nform certain functions must be given weight because it con-\\nnotes structure and thus is not merely an intended use.\\nThese statements were not made within the context of\\n§ 112, ¶ 6. Moreover, stating that the limitation connotes\\nstructure and has weight is not inconsistent with claiming\\nin means-plus-function format since means-plus-function\\nlimitations connote structure (i.e., corresponding structure\\n\\fMTD PRODS. INC. v. IANCU 15\\n\\n\\n\\nand their equivalents) and have weight. Furthermore, as\\nMTD explained, its interpretation of the claims as being in\\nmeans-plus-function format during inter partes review was\\nbased on this court’s intervening law in Williamson.\\nJ.A. 1204. Given the lack of any clear and undisputed\\nstatement foreclosing application of § 112, ¶ 6, we conclude\\nthat the Board erred in giving dispositive weight to the\\nequivocal statements it cited in the prosecution history.\\n CONCLUSION\\n We conclude that the Board erred by using the exist-\\nence of corresponding structure in the specification to con-\\nclude that “mechanical control assembly” has a sufficiently\\ndefinite structure to evade § 112, ¶ 6. The Board also erred\\nby giving improper weight to out-of-context statements in\\nthe prosecution history. We hold that the remaining evi-\\ndence and the Board’s factual findings demonstrate that\\nthe term “mechanical control assembly . . . configured to”\\nperform certain functions in independent claims 1 and 9 of\\nthe ’458 patent is governed by § 112, ¶ 6. We therefore va-\\ncate the Board’s decision and remand for further proceed-\\nings consistent with this opinion.\\n VACATED AND REMANDED\\n COSTS\\n Costs to Appellant.\", \"resource_uri\": \"https://www.courtlistener.com/api/rest/v4/opinions/4424704/\", \"author_raw\": \"STOLL, Circuit Judge.\"}]}"},"judge1":{"kind":"string","value":"REYNA"},"judge2":{"kind":"string","value":"TARANTO"},"judge3":{"kind":"string","value":"STOLL"},"exactly_three_caps":{"kind":"number","value":1,"string":"1"},"extra_judges":{"kind":"string","value":""},"judge_designations":{"kind":"string","value":"{}"},"opinions_count":{"kind":"number","value":1,"string":"1"},"majority_count":{"kind":"number","value":0,"string":"0"},"dissent_count":{"kind":"number","value":0,"string":"0"},"concurrence_count":{"kind":"number","value":0,"string":"0"},"other_count":{"kind":"number","value":1,"string":"1"},"primary_citation":{"kind":"string","value":""},"reporter":{"kind":"string","value":""},"volume":{"kind":"null"},"resource_url":{"kind":"string","value":"https://www.courtlistener.com/api/rest/v4/clusters/4647451/"},"precedential_status":{"kind":"string","value":"Published"},"010combined_count":{"kind":"number","value":1,"string":"1"},"020lead_count":{"kind":"number","value":0,"string":"0"},"030concurrence_count":{"kind":"number","value":0,"string":"0"},"035concurrenceinpart_count":{"kind":"number","value":0,"string":"0"},"040dissent_count":{"kind":"number","value":0,"string":"0"},"year":{"kind":"number","value":2019,"string":"2,019"},"month":{"kind":"number","value":8,"string":"8"},"messages":{"kind":"list like","value":[{"content":"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\n\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\nobject that encodes the following variables:\n\n- Petitioner (name, description, and general category)\n- Respondent (name, description, and general category)\n- Disposition of Case\n- Winning Party\n- Issue Area\n- Legal Provisions\n- Decision Direction\n- Lower Court Direction\n- Declaration of Unconstitutionality\n- Method (panel, en banc, or rehearing)\n- Who Appealed (original plaintiff vs. defendant)\n- Dissent Basis (procedural vs. merits — only if dissent exists)\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\n\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\nrules as closely as possible, treating the **Court of Appeals as \"the Court\"** and assuming\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\n\nTo encode these variables, start by determining the petitioner and respondent, following the\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\n\n==================================================\n1. INPUT FORMAT\n==================================================\n\nYou will receive input in two sections:\n\n[CASE_METADATA]\nCase name: <case name as in the caption>\nCourt: <name of the Court of Appeals, e.g. \"United States Court of Appeals for the Ninth Circuit\">\nDate decided: <YYYY-MM-DD or best approximation>\nOpinion type: <e.g. \"majority\", \"per curiam\", \"on-the-merits\", \"rehearing\", \"en banc\", etc.>\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\nAdditional notes: <any other metadata or summary, if provided>\n\n[CASE_TEXT]\n<full or excerpted opinion text here>\n[END_CASE]\n\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\nsubstantive discussion in the opinion.\n\n==================================================\n2. OUTPUT FORMAT\n==================================================\n\nYou must output ONLY a single JSON object with exactly these keys:\n\n- \"petitioner\"\n- \"respondent\"\n- \"case_disposition\"\n- \"winning_party\"\n- \"issue_area\"\n- \"legal_provisions\"\n- \"decision_direction\"\n- \"lower_court_direction\"\n- \"declaration_of_unconstitutionality\"\n- \"method\"\n- \"who_appealed\"\n- \"dissent_basis\" (only if the case has a dissenting opinion)\n- \"concurrence_type\" (only if the case has a concurring opinion)\n- \"concurrence_basis\" (only if the case has a concurring opinion)\n- \"justification\"\n\nSchema:\n\n- \"petitioner\": {\n \"name\": string, // short party name from caption; usually the appellant\n \"description\": string, // role and identity, e.g. \"federal criminal defendant appealing conviction\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"respondent\": {\n \"name\": string,\n \"description\": string, // e.g. \"United States as appellee\", \"private employer\", \"federal agency\"\n \"category\": integer, // 1–7 general litigant type (see Section 3.1)\n \"category_label\": string // label for the category code\n }\n\n- \"case_disposition\": {\n \"code\": integer, // must be one of the codes below\n \"label\": string\n }\n\n- \"winning_party\": {\n \"code\": integer, // 0, 1, or 2 as defined below\n \"label\": string\n }\n\n- \"issue_area\": {\n \"code\": integer, // 1–14 from the list below\n \"label\": string // short description of the main issue area\n }\n\n- \"issue_area_alt\": {\n \"code\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\n \"label\": string\n }\n\n- \"legal_provisions\": [\n {\n \"area_code\": integer, // 1–9 (law area codes below)\n \"area_label\": string,\n \"provision_text\": string // e.g. \"Fourth Amendment, search and seizure\", \"18 U.S.C. § 924(c)\"\n },\n ...\n ]\n\n- \"decision_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"lower_court_direction\": {\n \"code\": integer, // 1, 2, or 3 as defined below\n \"label\": string\n }\n\n- \"declaration_of_unconstitutionality\": {\n \"code\": integer, // 0 or 1 (see Section 3.8)\n \"label\": string // \"no\" or \"yes\"\n }\n\n- \"method\": {\n \"code\": integer, // 1–4 (see Section 3.9)\n \"label\": string\n }\n\n- \"who_appealed\": {\n \"code\": integer, // 1–4 (see Section 3.10)\n \"label\": string\n }\n\n- \"dissent_basis\": { // ONLY include if \"Has dissenting opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"concurrence_type\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"code\": integer, // 1–3 (see Section 3.12)\n \"label\": string\n }\n\n- \"concurrence_basis\": { // ONLY include if \"Has concurring opinion: Yes\" in metadata\n \"procedural\": integer, // 0–3 (see Section 3.11)\n \"merits\": integer // 0–3 (see Section 3.11)\n }\n\n- \"justification\": string // 2–4 sentences summarizing why you chose these codes.\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\n // (c) how that combination implied the decision_direction.code you chose, and\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\n\nYour entire response must be exactly one JSON object in this schema and nothing else.\n\n==================================================\n3. VARIABLE INTERPRETATION\n==================================================\n\n3.1 Petitioner and Respondent\n\n- Use the case caption, opinion type, and text to determine who is appealing.\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\n- Respondent = the party opposing the appeal (typically the appellee).\n- The \"name\" should be a reasonably short label from the caption.\n- The \"description\" should describe their role and identity (e.g., \"federal prisoner challenging sentence\",\n \"state official in civil-rights suit\", \"private corporation defendant\", \"federal agency\").\n\nFor each party, also assign a \"category\" code using this 7-category typology:\n\n1 = private business or corporation (including partnerships, sole proprietorships)\n2 = nonprofit organization, interest group, or union (as an institutional party)\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\n4 = sub-state government (city, county, municipality, school district, local official)\n5 = state government (state, governor, state agency, state official in official capacity)\n6 = tribal government or Indian nation\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\n\nChoose the category based on who the party IS, not what kind of case it is. When a government\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\n\n\n3.2 Disposition of Case\n\nCodes (Songer-style numbering):\n\n1 = affirmed (includes modified and affirmed)\n2 = reversed\n3 = reversed and remanded\n4 = vacated and remanded\n5 = affirmed in part, reversed (or vacated) in part\n6 = affirmed in part, reversed (or vacated) in part, and remanded\n7 = vacated\n8 = petition denied or appeal dismissed\n9 = stay, petition, or motion granted\n10 = certification to or from a lower court\n11 = affirmed, vacated, and remanded (without reversal)\n0 = no disposition\n\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\n or original authority (the word \"remand\" must appear in the opinion text).\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\n or vacatur (7) of the prior judgment.\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\"no disposition\").\n\n\n3.3 Winning Party\n\nCodes:\n\n0 = no favorable disposition for petitioning party apparent\n1 = petitioning party received a favorable disposition\n2 = favorable disposition for petitioning party unclear\n\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\n- A \"favorable disposition\" is one that materially improves the petitioner's position relative to the lower court judgment\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\n regardless of the case disposition.\n\n\n3.4 Issue Area\n\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\n\n1 = Criminal Procedure\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\n\n2 = Civil Rights\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\n\n3 = First Amendment\n (e.g., freedom of speech, press, association, religion, and related expression claims)\n\n4 = Due Process\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\n\n5 = Privacy\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\n\n6 = Attorneys\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\n\n7 = Unions\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\n\n8 = Economic Activity\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\n\n9 = Judicial Power\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\n\n10 = Federalism\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\n\n11 = Interstate Relations\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\n\n12 = Federal Taxation\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\n\n13 = Miscellaneous\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\n\n14 = Private Action\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\n\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\nIf multiple areas appear, choose the one most emphasized in the reasoning.\n\nIMPORTANT — AVOID MISCELLANEOUS (13):\n\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\napproximately. Before choosing 13, reconsider the following common mappings:\n\n - Medical malpractice, products liability, personal injury, breach of contract,\n admiralty, property disputes → 14 (Private Action)\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\n - Attorneys' fees, bar discipline → 6 (Attorneys)\n - Federal preemption of state law → 10 (Federalism)\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\n\nIf you DO choose issue_area.code = 13, you MUST also fill in \"issue_area_alt\" with your\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\nset \"issue_area_alt\" to the same code and label as \"issue_area\".\n\nISSUE AREA TIEBREAKING RULES:\n\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\n\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\n a civil-rights statute was violated.\n\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\n termination of benefits, property seizure in a civil case).\n\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\n primarily about whether it has the power to hear the case — i.e., the court resolves a\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\n\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\n statute preempts state regulation, use 10. If the holding is about the substance of a\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\n use 8.\n\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\n procedural fairness or property/liberty deprivation without an underlying discrimination\n allegation.\n\n\n3.5 Legal Provisions\n\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \"legal_provisions\".\n\nLaw Area codes (SCDB varLawArea):\n\n1 = Constitution (original text)\n An Article or Section of the original Constitution (not an Amendment).\n Examples: Article I (congressional power), Article II (executive power),\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\n and Credit Clause, Contracts Clause.\n\n2 = Constitutional Amendment\n Any Amendment to the Constitution, including the Bill of Rights.\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\n punishment), Fourteenth Amendment (equal protection, due process).\n\n3 = Federal Statute\n A statute enacted by Congress, cited by title and section of the U.S. Code or\n by its popular name. This is the most common category.\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\n\n4 = Court Rules\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\n Procedure, as well as local circuit rules.\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\n (prejudicial evidence).\n\n5 = Other\n Common law doctrines, equitable principles, or other legal authorities not fitting\n the categories above.\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\n of laches, implied contracts.\n\n6 = Infrequently litigated statutes\n Federal statutes that appear rarely in appellate litigation and are not well-known\n by popular name. Use this when the statute is obscure or highly specialized.\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\n obscure provisions of the Internal Revenue Code.\n\n8 = State or local law or regulation\n A state statute, state constitutional provision, local ordinance, or state/local\n regulation. Use this when the court applies or interprets state or local law rather\n than federal law.\n Examples: state tort law, state contract law, state criminal statute, local zoning\n ordinance, state environmental regulation.\n\n9 = No Legal Provision\n The opinion does not meaningfully rely on any identifiable legal provision. Very\n rare — use only when the court's analysis is purely factual or procedural without\n citing specific legal authority.\n\nFor each entry:\n- `area_code`: choose the single best area for the provision.\n- `area_label`: the corresponding label (e.g., \"Constitution\", \"Federal Statute\").\n- `provision_text`: identify the specific provision (e.g., \"Fourth Amendment, search and seizure\",\n \"42 U.S.C. § 1983\", \"Federal Rule of Civil Procedure 12(b)(6)\", \"state tort law\").\n\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\n\nDistinguishing area codes:\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\n- Use 5 for common law doctrines that are not codified in any statute or rule.\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\n\n\n3.6 Decision Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\nCRITICAL: The codes \"conservative\" and \"liberal\" are MECHANICAL LABELS defined by the SCDB\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\nTaxation case, regardless of the policy context. A decision upholding a government regulation\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\nbusiness. Do NOT let your general understanding of politics override these rules.\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\n\nissue_area.code | benefiting side | direction\n----------------|------------------------------------------------------------|----------\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\n | regulation, or small entities against large business |\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\n10 (Federalism) | federal authority over state authority | 2 (liberal)\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\n | with one of the patterns above |\n\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\nissue_area, and LAST look up the decision_direction from the table above by matching the\nissue_area.code with the side that benefited. Do NOT reason about direction from first\nprinciples — use the table.\n\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\n\n\n3.6.1 Worked Examples\n\nThe following three examples demonstrate correct mechanical application of the direction\nlookup table, including cases where the SCDB coding may seem counterintuitive:\n\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\n\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\n\n Coding:\n issue_area.code = 12 (Federal Taxation)\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\n says \"taxpayer → 1 (conservative)\"\n\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\n codes this as conservative (1) because the taxpayer benefited.\n\n\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\n\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\n\n Coding:\n issue_area.code = 1 (Criminal Procedure)\n winning_party.code = 0 (no favorable disposition for petitioner)\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\n row for code 1 says \"prosecution → 1 (conservative)\"\n\n\nEXAMPLE C — Economic Activity (issue_area.code = 8)\n\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\n affirms, upholding the regulation as within the agency's statutory authority.\n\n Coding:\n issue_area.code = 8 (Economic Activity)\n winning_party.code = 0 (no favorable disposition for petitioner/company)\n decision_direction = 2 (liberal) ← government regulation upheld against business;\n lookup table row for code 8 says \"govt regulation → 2 (liberal)\"\n\n NOTE: Even if one might view upholding a regulation as a \"pro-business\" outcome (e.g., it\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\n over a business challenge as liberal (2).\n\n\n3.6.2 Internal Consistency Check\n\nBefore you output the JSON, you MUST run the following consistency check:\n\n1. Identify issue_area.code.\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\n4. Verify that your decision_direction.code matches the table entry.\n5. If it does not match, CORRECT decision_direction.code to match the table.\n\n\n3.7 Lower Court Direction\n\nCodes:\n\n1 = conservative\n2 = liberal\n3 = unspecifiable\n\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\ncourt's judgment and the appellate court's disposition.\n\nUse this rule of thumb:\n\n1. First, identify the appellate disposition (case_disposition.code).\n2. Next, identify the decision_direction.code for the Court of Appeals.\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\n\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \"affirmed\"\n or 8 \"petition denied or appeal dismissed\"), then the lower court's direction is the SAME as decision_direction.code.\n\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \"reversed\",\n 3 \"reversed and remanded\", 4 \"vacated and remanded\", or 7 \"vacated\"), then the lower court's direction is the\n OPPOSITE of decision_direction.code.\n\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\n the lower court's direction should usually be the SAME as decision_direction.code.\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\n\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\n the lower court if possible; if not, use 3 (unspecifiable).\n\nWhen you set lower_court_direction, remember:\n- \"SAME\" means lower_court_direction.code = decision_direction.code.\n- \"OPPOSITE\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\n\n3.8 Declaration of Unconstitutionality\n\nCodes:\n\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\n1 = the court DID declare a statute, regulation, or government action unconstitutional\n\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\nregulations, state statutes, state constitutional provisions, and local ordinances.\n\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\nhold that something is unconstitutional. If the court upholds the constitutionality of the\nchallenged provision, code 0.\n\n\n3.9 Method\n\nCodes:\n\n1 = standard panel decision (typically 3 judges, first hearing)\n2 = panel rehearing (same panel reconsidering its prior decision)\n3 = en banc (full court or en banc panel rehearing a case)\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\n\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\nstatements like \"before the court en banc,\" \"on rehearing,\" \"upon reconsideration,\" or\n\"per curiam.\" If the metadata indicates more than 3 judges participated, this strongly\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\n\n\n3.10 Who Appealed\n\nCodes:\n\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\n3 = federal government on behalf of the original plaintiff\n4 = federal government on behalf of the original defendant\n\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\nhistory. The \"original plaintiff\" is the party who filed suit in the lower court; the \"original\ndefendant\" is the party who was sued. In criminal cases, the original plaintiff is typically the\ngovernment (prosecution) and the original defendant is the accused.\n\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\nordering (appellant is typically listed first).\n\n\n3.11 Dissent/Concurrence Basis (Conditional)\n\nIf the case metadata states \"Has dissenting opinion: Yes\" or \"Has concurring opinion: Yes\",\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\n\nFor each dimension (procedural / merits), use one of these codes:\n\n0 = the dissent/concurrence exists but does NOT address this type of issue\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\n3 = addresses an issue NOT discussed by the majority opinion\n\n\"dissent_basis\" (include ONLY if \"Has dissenting opinion: Yes\"):\n - \"procedural\": Does the dissent address procedural questions (jurisdiction, standing,\n standard of review, procedural default, timeliness, etc.)?\n - \"merits\": Does the dissent address the substantive legal merits (statutory interpretation,\n constitutional analysis, factual sufficiency, etc.)?\n\n\"concurrence_basis\" (include ONLY if \"Has concurring opinion: Yes\"):\n - \"procedural\": Does the concurrence address procedural questions?\n - \"merits\": Does the concurrence address the substantive legal merits?\n\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\n(the longest or most substantive).\n\nIf the metadata says \"Has dissenting opinion: No\", OMIT \"dissent_basis\" entirely from your output.\nIf the metadata says \"Has concurring opinion: No\", OMIT \"concurrence_type\" and \"concurrence_basis\" entirely from your output.\n\n\n3.12 Concurrence Type (Conditional)\n\nIf the case metadata states \"Has concurring opinion: Yes\", classify the concurrence as one of\nthe following types:\n\nCodes:\n\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\n\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\n like \"concurring in the result,\" \"concurring in the judgment,\" \"concurring in the result\n only,\" or \"I concur in the judgment but write separately because I disagree with the\n majority's analysis.\" The key indicator: the judge explicitly declines to join the majority\n opinion or distances themselves from the majority's reasoning.\n\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\n the majority opinion or only the result.\n\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\nwrites a special concurrence is qualitatively different from one where all judges agree on\nboth outcome and reasoning.\n\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\nmost substantive). If one is a regular concurrence and another is a special concurrence,\ncode based on the special concurrence (code 2), as it is more analytically informative.\n\n\n==================================================\n4. GENERAL RULES\n==================================================\n\n- Use metadata (case name, court, date, opinion type) to:\n - distinguish main merits opinions from rehearing or procedural opinions,\n - interpret how the appellate disposition relates to the prior judgment,\n - infer whether the case is criminal, civil rights, tax, etc.\n- If some information is missing, use the appropriate \"unspecifiable\" or \"no disposition\" code\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\n- Carefully follow the steps laid out in the introduction in order. In particular,\n decide who won the case (winning_party). Then, decide what the issue_area is.\n Only choose decision_direction once you have completed those steps, and follow the exact\n directions in Section 3.6.\n- In the justification, explicitly state who won, why (according to the disposition), and\n how that winner and the issue_area code together implied the direction code you chose.\n Reference the DIRECTION LOOKUP TABLE row you applied.\n- ALWAYS choose exactly one code for each coded variable where required.\n- NEVER output any keys other than the ones defined in the schema above.\n- Include \"dissent_basis\", \"concurrence_type\", and \"concurrence_basis\" ONLY when the metadata\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\n- NEVER output any numeric codes outside the permitted lists above.\n\n\n==================================================\n5. FINAL REQUIREMENT\n==================================================\n\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\n","role":"system"},{"content":"[CASE_METADATA]\nCase name: MTD PRODUCTS INC., Appellant v. Andrei IANCU, Under Secretary of Commerce for Intellectual Property and Director of the United States Patent and Trademark Office, Intervenor\nCourt: U.S. Court of Appeals for the Federal Circuit\nDate decided: 2019-08-12\nOpinion type: majority\nHas dissenting opinion: No\nHas concurring opinion: No\nAdditional notes: \n\n[CASE_TEXT]\nUnited States Court of Appeals\n for the Federal Circuit\n ______________________\n\n MTD PRODUCTS INC.,\n Appellant\n\n v.\n\n ANDREI IANCU, UNDER SECRETARY OF\n COMMERCE FOR INTELLECTUAL PROPERTY\n AND DIRECTOR OF THE UNITED STATES\n PATENT AND TRADEMARK OFFICE,\n Intervenor\n ______________________\n\n 2017-2292\n ______________________\n\n Appeal from the United States Patent and Trademark\nOffice, Patent Trial and Appeal Board in No. IPR2016-\n00194.\n ______________________\n\n Decided: August 12, 2019\n ______________________\n\n JOHN SALVATORE CIPOLLA, Calfee, Halter & Griswold\nLLP, Cleveland, OH, argued for appellant. Also repre-\nsented by ANDREW ALEXANDER, TRACY SCOTT JOHNSON,\nMARK MCDOUGALL.\n\n PETER JOHN SAWERT, Office of the Solicitor, United\nStates Patent and Trademark Office, Alexandria, VA, ar-\ngued for intervenor. Also represented by THOMAS W.\nKRAUSE, PHILIP J. WARRICK.\n\f2 MTD PRODS. INC. v. IANCU\n\n\n\n\n ______________________\n\n Before REYNA, TARANTO, and STOLL, Circuit Judges.\nSTOLL, Circuit Judge.\n The Toro Company sought inter partes review of\nclaims 1–16 of U.S. Patent No. 8,011,458 before the U.S.\nPatent and Trademark Office’s Patent Trial and Appeal\nBoard. The Board instituted review and, in its final writ-\nten decision, held the challenged claims obvious under\n35 U.S.C. § 103. Critical to its decision, the Board deter-\nmined that the claim term “mechanical control assem-\nbly . . . configured to” perform certain functions is not a\nmeans-plus-function term subject to 35 U.S.C. § 112, ¶ 6.\nMTD Products Inc., owner of the ’458 patent, appeals the\nBoard’s decision.\n We conclude that the Board erred by conflating corre-\nsponding structure in the specification with a structural\ndefinition for the term, and by misinterpreting certain\nstatements in the prosecution history. Under the appropri-\nate legal framework, we conclude that the term “mechani-\ncal control assembly” is a means-plus-function term\ngoverned by § 112, ¶ 6. We therefore vacate the Board’s\nobviousness conclusion, which was predicated on its incor-\nrect claim construction, and remand for further proceed-\nings consistent with this opinion. Because we are\npersuaded by MTD’s primary argument, we do not reach\nits alternative arguments.\n BACKGROUND\n I\n The ’458 patent discloses a steering and driving system\nfor zero turn radius (“ZTR”) vehicles, with specific refer-\nence to ZTR lawn mowers. ’458 patent col. 1 ll. 17–21. The\npatented system is designed to provide a more intuitive\nsteering mechanism to operators of ZTR vehicles. Id. at\ncol. 1 ll. 20–38. In contrast to prior art systems that\n\fMTD PRODS. INC. v. IANCU 3\n\n\n\nreverse in the opposite direction of a forward motion turn,\nthe claimed invention permits ZTR vehicles to turn in the\nsame direction both forward and backwards. Id. at col. 1\nll. 20–47. The claimed steering mechanism thus mimics\nthe forward and backward movements of an automobile.\n The term “mechanical control assembly” appears in\nboth claims 1 and 9, the only independent claims of the\n’458 patent. Claim 1 recites:\n 1. A vehicle capable of making a small radius turn,\n comprising:\n a frame;\n a left drive wheel and a right drive wheel,\n both coupled to the frame;\n two independent left and right drive units,\n the left drive unit coupled to the left drive\n wheel via an axle and the right drive unit\n coupled to the right drive wheel via another\n axle;\n a steering device coupled to the frame;\n a speed control member coupled to the\n frame; and\n a mechanical control assembly coupled to\n the left and right drive units that is config-\n ured to actuate the left and right drive\n units based on a steering input received\n from the steering device and a speed input\n received from the speed control member;\n the mechanical control assembly being con-\n figured such that if the speed control mem-\n ber is shifted from (a) a forward position in\n which the left drive wheel is rotating in a\n forward direction at a first forward speed\n and the right drive wheel is rotating in a\n\f4 MTD PRODS. INC. v. IANCU\n\n\n\n\n forward direction at a second forward\n speed that is less than the first forward\n speed as a result of the steering device be-\n ing in a first right turn position to (b) a re-\n verse position while the first right turn\n position of the steering device is main-\n tained, then the left drive wheel will rotate\n in a reverse direction at a first reverse\n speed and the right drive wheel will rotate\n in a reverse direction at a second reverse\n speed that is less than the first reverse\n speed.\nId. at col. 7 l. 63–col. 8 l. 24 (emphasis added to highlight\nportion of disputed claim term). Claim 9 is identical to\nclaim 1 in substantial part, adding only the further limita-\ntion of:\n the mechanical control assembly also being config-\n ured to cause the vehicle to execute a zero-radius\n turn when the speed control member is in a maxi-\n mum forward position and the steering device is in\n a maximum turn position.\nId. at col. 9 ll. 13–16 (emphasis added).\n While the patent specification does not expressly refer\nto a “mechanical control assembly,” it discloses a preferred\nembodiment that includes a “ZTR control assembly.” Id. at\ncol. 3 ll. 41–42. The specification describes components of\nthe ZTR control assembly and its inputs, outputs, and link-\nages. Id. at col. 3 l. 41–col. 4 l. 57.\n II\n Toro petitioned for inter partes review of the ’458 pa-\ntent in November 2015, arguing that the challenged claims\nwere invalid as anticipated or obvious. MTD responded\nthat the term “mechanical control assembly” is a means-\nplus-function term, and that the asserted prior art did not\ndisclose the claim term’s corresponding structure. In\n\fMTD PRODS. INC. v. IANCU 5\n\n\n\nsupport of its argument, MTD introduced expert testimony\nindicating that “mechanical control assembly” has no rea-\nsonably well-understood meaning in the art. Specifically,\nMTD’s expert testified that “mechanical control assembly”\nis a nonce term that is not used in common parlance and\ndoes not bring to mind any specific structure to a person of\nordinary skill in the art. J.A. 1366. He explained that the\nterm is used as a black box recitation for structure and, at\nmost, amounts to a collection of various parts. J.A. 1248,\n1366. He further demonstrated that the term is used in\nvarious prior patents and publications to describe a wide\nvariety of structures with varying functions. J.A. 1367–69\n(noting that “mechanical control assembly” is used generi-\ncally to describe mechanisms for infusion pumps, digital\nfiring systems, flush tanks, endoscopes, transmissions, and\nengine outputs).\n Toro did not expressly contradict MTD’s evidence that\n“mechanical control assembly” did not have a well-under-\nstood structural meaning. Instead, Toro responded that a\nperson of ordinary skill in the art would understand the\nterm to denote a specific structure in the context of the\n’458 patent specification. Specifically, Toro argued that the\n“ZTR control assembly” disclosed in the specification pro-\nvides an express structural definition for the claimed “me-\nchanical control assembly.” J.A. 2201–03. Toro also\nargued that MTD admitted that the term “mechanical con-\ntrol assembly” conveys particular structure when it distin-\nguished the patent claims from a prior art reference during\nprosecution. J.A. 2203.\n The Board initially agreed with MTD, stating that\nwhen viewed “in isolation, the genericness of this term\nbears similarities to other words or phrases that have been\nheld to be subject to § 112, ¶ 6 . . . such as ‘mechanism,’ ‘el-\nement,’ ‘device,’ ‘link member,’ and ‘control mechanism.’”\nToro Co. v. MTD Prods. Inc., No. IPR2016-00194,\n2017 WL 1969747, at *9 (P.T.A.B. May 10, 2017) (first cit-\ning Williamson v. Citrix Online, LLC, 792 F.3d 1339, 1350\n\f6 MTD PRODS. INC. v. IANCU\n\n\n\n\n(Fed. Cir. 2015) (en banc); then citing Mas-Hamilton Grp.\nv. LaGard, Inc., 156 F.3d 1206, 1215 (Fed. Cir. 1998); and\nthen citing Toro Co. v. Deere & Co., 355 F.3d 1313, 1325\n(Fed. Cir. 2004)). The Board also determined that the “lan-\nguage reciting what the mechanical control assembly is\n‘configured to’ do . . . fits the mold of functional language\nbecause it describes the mechanical control assembly by\nwhat it does.” Id. The Board thus concluded that “the\nclaim language of the disputed phrase is primarily, but not\nentirely, functional, which tends to favor [MTD]’s position\nthat § 112, ¶ 6 applies.” Id.\n The Board agreed with Toro, however, that the ’458 pa-\ntent specification weighed against application of § 112, ¶ 6.\nId. at *10. Citing the parties’ briefs, the Board stated that\n“[t]he parties agree that the claimed ‘mechanical control\nassembly’ is referred to in the specification as a ‘ZTR con-\ntrol assembly.’” Id. According to the Board, a person of\nordinary skill in the art would understand “mechanical\ncontrol assembly” to denote structure because “the specifi-\ncation illustrates and describes the specific structure that\nmakes up the ZTR control assembly, and how it connects to\nand operates with other components.” Id.\n The most persuasive piece of evidence to the Board,\nhowever, was the prosecution history. Id. at *9 (“The factor\nthat weighs most heavily in [the] determination is the pros-\necution history.”). According to the Board, MTD admitted\nthat the term “mechanical control assembly” connotes spe-\ncific structure by asserting that the claims recite “a me-\nchanical control assembly that is structurally different\nfrom what [the asserted prior art] discloses.” Id. The\nBoard emphasized MTD’s statements that “the claim lan-\nguage at issue concerns the configuration of the claimed\nmechanical control assembly” and “the claimed configura-\ntion is indeed structural.” Id. (emphases in original). The\nBoard concluded that these statements “present[] strong\nevidence that the disputed phrase should be understood as\na structural limitation rather than a means-plus-function\n\fMTD PRODS. INC. v. IANCU 7\n\n\n\nlimitation under § 112 ¶ 6.” Id. at *11. Relying on the spec-\nification and prosecution history, the Board ultimately de-\ntermined that “mechanical control assembly” is not\ngoverned by § 112, ¶ 6. Id.\n DISCUSSION\n I\n This appeal requires us to address whether a particu-\nlar claim limitation is drafted in means-plus-function for-\nmat. Whether claim language invokes 35 U.S.C. § 112, ¶ 6 1\nis a legal question of claim construction that we review de\nnovo. Williamson, 792 F.3d at 1346. We review the\nBoard’s factual findings underlying this inquiry for sub-\nstantial evidence. EnOcean, GmbH v. Face Int’l Corp.,\n742 F.3d 955, 959 (Fed. Cir. 2014).\n Under this court’s guidance in Williamson, we begin by\nasking whether the claim limitation employs the word\n“means.” Williamson, 792 F.3d at 1348. If it does not, we\napply a rebuttable presumption that the term conveys suf-\nficiently definite structure and is not subject to § 112, ¶ 6.\nId. A challenger can rebut the presumption by demonstrat-\ning “that the claim term fails to ‘recite sufficiently definite\nstructure’ or else recites ‘function without reciting suffi-\ncient structure for performing that function.’” Id. (quoting\nWatts v. XL Sys., Inc., 232 F.3d 877, 880 (Fed. Cir. 2000)).\nThe “essential inquiry is not merely the presence or ab-\nsence of the word ‘means’ but whether the words of the\n\n\n 1 Because the issue date of the ’458 patent is Septem-\nber 6, 2011, and neither the ’458 patent nor the application\nfrom which it issued ever contained a claim with an effec-\ntive filing date on or after September 16, 2012, the version\nof 35 U.S.C. § 112 that applies here is the one preceding\nthe changes made by the America Invents Act. See Leahy-\nSmith America Invents Act, Pub. L. No. 112-29, 125 Stat.\n284, 296, § 4(c) (2011).\n\f8 MTD PRODS. INC. v. IANCU\n\n\n\n\nclaim are understood by persons of ordinary skill in the art\nto have a sufficiently definite meaning as the name for\nstructure.” Id.\n One way to demonstrate that a claim limitation fails to\nrecite sufficiently definite structure is to show that, alt-\nhough not employing the word “means,” the claim limita-\ntion uses a similar “nonce word that can operate as a\nsubstitute for ‘means’ in the context of § 112, para. 6.” Id.\nat 1350. Generic terms like “module,” “mechanism,” “ele-\nment,” and “device” are commonly used as verbal con-\nstructs that operate, like “means,” to claim a particular\nfunction rather than describe a “sufficiently definite struc-\nture.” Id. Our case law is replete with guidance on\nwhether or not a particular claim term is a “nonce” term.\nSee, e.g., Zeroclick, LLC v. Apple Inc., 891 F.3d 1003, 1008\n(Fed. Cir. 2018) (holding that “program” and “user inter-\nface code” are not nonce words because they are “used not\nas generic terms or black box recitations of structure or ab-\nstractions, but rather as specific references to conventional\ngraphical user interface programs or code”); Williamson,\n792 F.3d at 1350 (noting that “module” is a nonce term be-\ncause it “sets forth the same black box recitation of struc-\nture for providing the same specified function as if the term\n‘means’ had been used”). In each case, a critical question\nis whether “the claim term is used in common parlance or\nby persons of skill in the pertinent art to designate struc-\nture,” including either a particular structure or a class of\nstructures. Skky, Inc. v. MindGeek, s.a.r.l., 859 F.3d 1014,\n1019 (Fed. Cir. 2017) (citing TecSec, Inc. v. Int’l Bus.\nMachs. Corp., 731 F.3d 1336, 1347 (Fed. Cir. 2013)) (hold-\ning that the claim term “wireless device means” does not\ninvoke § 112, ¶ 6 because it denotes a class of structures).\n In addition, even if the claims recite a nonce term fol-\nlowed by functional language, other language in the claim\n“might inform the structural character of the limitation-in-\nquestion or otherwise impart structure” to the claim term.\nWilliamson, 792 F.3d at 1351. In assessing whether the\n\fMTD PRODS. INC. v. IANCU 9\n\n\n\nclaim limitation is in means-plus-function format, we do\nnot merely consider the introductory phrase (e.g., “mechan-\nical control assembly”) in isolation, but look to the entire\npassage including functions performed by the introductory\nphrase. See Apex Inc. v. Raritan Computer, Inc., 325 F.3d\n1364 (Fed. Cir. 2003). The ultimate question is whether\n“the claim language, read in light of the specification, re-\ncites sufficiently definite structure to avoid § 112, ¶ 6.” Me-\ndia Rights Techs. Inc. v. Capital One Fin. Corp., 800 F.3d\n1366, 1372 (Fed. Cir. 2015) (citing Robert Bosch, LLC v.\nSnap-On Inc., 769 F.3d 1094, 1099 (Fed. Cir. 2014)).\n In Apex, for example, the court concluded that the term\n“circuit” recited sufficient structure in the context of the\nclaims at issue. 325 F.3d at 1372–73 (considering the use\nof the term “circuit” in the claim limitation “a first interface\ncircuit for receiving keyboard and cursor control device sig-\nnals from the workstation”). Based on a dictionary defini-\ntion of the word “circuit,” the court reasoned that “the term\n‘circuit’ by itself connotes some structure,” and that “the\nterm ‘circuit with an appropriate identifier such as ‘inter-\nface,’ ‘programming,’ and ‘logic,’ certainly identifies some\nstructural meaning to one of ordinary skill in the art.” Id.\nat 1373. The court noted that the extrinsic evidence did\nnot show that the term “circuit” was not understood to have\nstructure, but rather “only that the term ‘circuit’ is under-\nstood . . . as a very broad term.” Id. at 1374. As neither the\nspecification nor the prosecution history used the term “in\na manner clearly inconsistent with the ordinary meaning,”\nthe court held that the defendant failed to rebut the pre-\nsumption that § 112, ¶ 6 did not apply. Id. at 1373–74.\n In contrast, in Diebold Nixdorf, Inc. v. International\nTrade Commission, the court held that the term “cheque\nstandby unit” for performing certain specified functions\nwas governed by § 112, ¶ 6. 899 F.3d 1291, 1300 (Fed. Cir.\n2018). The court noted that “there is no evidence—in the\nform of dictionary definitions or otherwise—that ‘cheque\nstandby unit’ was reasonably well understood by persons\n\f10 MTD PRODS. INC. v. IANCU\n\n\n\n\nof ordinary skill in the art to refer to a structure or class of\nstructures.” Id. at 1302. Instead, the extrinsic evidence\ndemonstrated only that “a skilled artisan would under-\nstand the functional term ‘cheque standby unit’ to be any\nstructure capable of performing the claimed function.” Id.\nat 1301. Further, neither the words of the claim nor the\nspecification suggested a “structural limitation that might\nserve to cabin the scope of the functional term,” thus sup-\nporting the conclusion that the claim limitation was writ-\nten in means-plus-function format. Id.\n Finally, we note that “[c]laims are interpreted in light\nof the written description supporting them, and that is true\nwhether or not the claim construction involves interpreting\na ‘means’ clause.” Inventio AG v. ThyssenKrupp Elevator\nAms. Corp., 649 F.3d 1350, 1356 (Fed. Cir. 2011), overruled\non other grounds by Williamson, 792 F.3d at 1339. For ex-\nample, a patentee may avoid application of § 112, ¶ 6 by\nacting as a lexicographer and providing its own structural\ndefinition of a nonce term in the specification by “‘clearly\nset[ting] forth a definition of the disputed claim term’ other\nthan its plain and ordinary meaning.” Thorner v. Sony\nComput. Entm’t Am. LLC, 669 F.3d 1362, 1365 (Fed. Cir.\n2012) (citing CCS Fitness, Inc. v. Brunswick Corp.,\n288 F.3d 1359, 1366 (Fed. Cir. 2002)). Just as it is im-\nproper to “import[] limitations from the specification into\nthe claims,” Phillips v. AWH Corp., 415 F.3d 1303, 1323\n(Fed. Cir. 2005), however, a preferred embodiment dis-\nclosed in the specification cannot impart structure to a\nterm that otherwise has none. As with all lexicography,\n“[i]t is not enough for a patentee to simply disclose a single\nembodiment.” Thorner, 669 F.3d at 1365. Rather, “the pa-\ntentee must ‘clearly express an intent’ to redefine the\nterm.” Id. (quoting Helmsderfer v. Bobrick Washroom\nEquip., Inc., 527 F.3d 1379, 1381 (Fed. Cir. 2008)).\n With these background principles in mind, we turn to\nthe claim language at issue in this case.\n\fMTD PRODS. INC. v. IANCU 11\n\n\n\n II\n The disputed claim limitation is lengthy and recites:\n a mechanical control assembly coupled to the left\n and right drive units that is configured to actuate\n the left and right drive units based on a steering\n input received from the steering device and a speed\n input received from the speed control member;\n the mechanical control assembly being configured\n such that if the speed control member is shifted\n from (a) a forward position in which the left drive\n wheel is rotating in a forward direction at a first\n forward speed and the right drive wheel is rotating\n in a forward direction at a second forward speed\n that is less than the first forward speed as a result\n of the steering device being in a first right turn po-\n sition to (b) a reverse position while the first right\n turn position of the steering device is maintained,\n then the left drive wheel will rotate in a reverse di-\n rection at a first reverse speed and the right drive\n wheel will rotate in a reverse direction at a second\n reverse speed that is less than the first reverse\n speed.\n’458 patent col. 8 ll. 7–24.\n At the outset, we agree with the Board that the term\n“mechanical control assembly” is similar to other generic,\nblack-box words that this court has held to be nonce terms\nsimilar to “means” and subject to § 112, ¶ 6 because the\nterm does not connote sufficiently definite structure to one\nof ordinary skill in the art. Toro, 2017 WL 1969747, at *9.\nWe also agree with the Board that the rest of the “claim\nlanguage of the disputed phrase is primarily, but not en-\ntirely, functional.” Id. While the claim language reciting\nthat the mechanical control assembly is “coupled to the left\nand right drive units” connotes structure, the claim lan-\nguage reciting what the mechanical control assembly is\n\f12 MTD PRODS. INC. v. IANCU\n\n\n\n\n“configured to” do is functional. In this respect, as the\nBoard correctly recognized, the claim format tends to favor\nMTD’s position that § 112, ¶ 6 applies. 2\n We also discern no error in the Board’s acceptance of\nMTD’s extrinsic evidence as showing that the term “me-\nchanical control assembly” does not have an established\nmeaning in the art and instead merely operates as a ge-\nneric label for a collection of parts. Toro did not dispute\nMTD’s expert testimony that, in common parlance, “me-\nchanical control assembly” does not bring to mind any spe-\ncific structure to a person of ordinary skill in the art. J.A.\n1366. Toro likewise did not dispute MTD’s reliance on var-\nious prior patents and publications that used “mechanical\ncontrol assembly” to describe a wide variety of structures\nwith varying functions. J.A. 1367–69 (noting that\n\n\n\n 2 The Board also stated that construing the disputed\nphrase as a means-plus-function limitation “would seem to\nrender the second part [of the claim’s recitation of a ‘me-\nchanical control assembly’] . . . superfluous” as it “would no\nlonger serve to define functionally the structural features\nof the mechanical control assembly, as those features\nwould be part and parcel of what [MTD] alleges is the cor-\nresponding structure.” Toro, 2017 WL 1969747, at *9. We\ndisagree. Both the first and second limitations following\nthe two occurrences of the term “mechanical control assem-\nbly” recite functions associated with the “mechanical con-\ntrol assembly”; i.e., (1) actuate the left and right drive units\nand (2) rotate the wheel in a particular direction based on\nthe position of the speed control member. As we have held,\n“[w]here there are multiple claimed functions, as we have\nhere, the patentee must disclose adequate corresponding\nstructure to perform all of the claim functions.” William-\nson, 792 F.3d at 1351–52. Thus, the corresponding struc-\nture for “mechanical control assembly” in the specification\nmust perform both of these functions.\n\fMTD PRODS. INC. v. IANCU 13\n\n\n\n“mechanical control assembly” is used generically to de-\nscribe mechanisms for infusion pumps, digital firing sys-\ntems, flush tanks, endoscopes, transmissions, and engine\noutputs).\n We conclude that the Board erred, however, when it re-\nlied on the specification’s description of a “ZTR control as-\nsembly” to conclude that the claim term “mechanical\ncontrol assembly” has an established structural meaning.\nWhile the parties agreed that the ZTR control assembly in\nthe specification is the structure “corresponding to” the\nclaimed mechanical control assembly, MTD did not agree\nthat the specification expressly defines the claim term “me-\nchanical control assembly.” That the specification discloses\na structure corresponding to an asserted means-plus-func-\ntion claim term does not necessarily mean that the claim\nterm is understood by persons of ordinary skill in the art to\nconnote a specific structure or a class of structures.\n Interpretation of an asserted means-plus-function lim-\nitation involves two steps. First, we determine if the claim\nlimitation is drafted in means-plus-function format. As\npart of this step, we consider whether the claim limitation\nconnotes “sufficiently definite structure” to a person of or-\ndinary skill in the art. If we conclude that the limitation is\nin means-plus-function format, the second step requires us\nto review the specification to identify the structure that\nperforms the claimed function(s) and thus “corresponds to”\nthe claimed means. While related, these two inquiries are\ndistinct. In this case, however, the Board conflated these\ndistinct inquiries, holding that the specification’s disclo-\nsure of corresponding structure demonstrates that the al-\nleged means-plus-function term is sufficiently definite so\nas to not invoke § 112, ¶ 6. The Board’s analysis implies\nthat so long as a claim term has corresponding structure in\nthe specification, it is not a means-plus-function limitation.\nThis is not consistent with our prior decisions. Indeed, this\nview would seem to leave § 112, ¶ 6 without any applica-\ntion: any means-plus-function limitation that met the\n\f14 MTD PRODS. INC. v. IANCU\n\n\n\n\nstatutory requirements, i.e., which includes having corre-\nsponding structure in the specification, would end up not\nbeing a means-plus-function limitation at all.\n While we agree with the Board that the specification\nplays a role in assessing whether particular claim language\ninvokes § 112, ¶ 6, we do not agree that the patent specifi-\ncation at issue here renders the nonce term “mechanical\ncontrol assembly” sufficiently structural to a person of or-\ndinary skill in the art. The specification does not demon-\nstrate that the patentee intended to act as its own\nlexicographer and define the nonce term “mechanical con-\ntrol assembly” as the “ZTR control assembly” of the pre-\nferred embodiment. Indeed, the specification does not even\nrefer to a “mechanical control assembly.” Furthermore, the\nfunctional language in the claim limitation suggests a\nbroader meaning of the generic term “mechanical control\nassembly,” as it specifically adds to the “mechanical control\nassembly” limitation the ability to execute a zero radius\nturn. ’458 patent col. 9 ll. 13–16. Interpreting the “me-\nchanical control assembly” as the “ZTR”—or zero-turn-ra-\ndius—control assembly would render this functional\nlanguage superfluous.\n We are also not persuaded by the Board’s interpreta-\ntion of the prosecution history. While it would have\navoided uncertainty and argument had MTD shared its\ncurrent view that the claim limitation is written in means-\nplus-function format during the original prosecution,\nMTD’s statements did not clearly disclaim such an inter-\npretation. Rather, MTD’s statements indicated that the\nphrase “mechanical control assembly configured to” per-\nform certain functions must be given weight because it con-\nnotes structure and thus is not merely an intended use.\nThese statements were not made within the context of\n§ 112, ¶ 6. Moreover, stating that the limitation connotes\nstructure and has weight is not inconsistent with claiming\nin means-plus-function format since means-plus-function\nlimitations connote structure (i.e., corresponding structure\n\fMTD PRODS. INC. v. IANCU 15\n\n\n\nand their equivalents) and have weight. Furthermore, as\nMTD explained, its interpretation of the claims as being in\nmeans-plus-function format during inter partes review was\nbased on this court’s intervening law in Williamson.\nJ.A. 1204. Given the lack of any clear and undisputed\nstatement foreclosing application of § 112, ¶ 6, we conclude\nthat the Board erred in giving dispositive weight to the\nequivocal statements it cited in the prosecution history.\n CONCLUSION\n We conclude that the Board erred by using the exist-\nence of corresponding structure in the specification to con-\nclude that “mechanical control assembly” has a sufficiently\ndefinite structure to evade § 112, ¶ 6. The Board also erred\nby giving improper weight to out-of-context statements in\nthe prosecution history. We hold that the remaining evi-\ndence and the Board’s factual findings demonstrate that\nthe term “mechanical control assembly . . . configured to”\nperform certain functions in independent claims 1 and 9 of\nthe ’458 patent is governed by § 112, ¶ 6. We therefore va-\ncate the Board’s decision and remand for further proceed-\nings consistent with this opinion.\n VACATED AND REMANDED\n COSTS\n Costs to Appellant.\n[END_CASE]","role":"user"}],"string":"[\n {\n \"content\": \"You are an expert legal coding assistant trained to classify U.S. federal Courts of Appeals\\ncases using an adaptation of the Supreme Court Database (SCDB_2023_01) codebook. You follow the coding procedure\\nin the codebook step by step and use the precise definitions of terms presented in the codebook below.\\n\\nYour task: Given metadata and text for a single federal appellate case, output a single JSON\\nobject that encodes the following variables:\\n\\n- Petitioner (name, description, and general category)\\n- Respondent (name, description, and general category)\\n- Disposition of Case\\n- Winning Party\\n- Issue Area\\n- Legal Provisions\\n- Decision Direction\\n- Lower Court Direction\\n- Declaration of Unconstitutionality\\n- Method (panel, en banc, or rehearing)\\n- Who Appealed (original plaintiff vs. defendant)\\n- Dissent Basis (procedural vs. merits — only if dissent exists)\\n- Concurrence Type (regular vs. special/result-only — only if concurrence exists)\\n- Concurrence Basis (procedural vs. merits — only if concurrence exists)\\n\\nYou are working with U.S. Courts of Appeals, not the Supreme Court. Apply the SCDB coding\\nrules as closely as possible, treating the **Court of Appeals as \\\"the Court\\\"** and assuming\\nthat the **lower court is, e.g., a federal district court** whose decision is being reviewed.\\n\\nTo encode these variables, start by determining the petitioner and respondent, following the\\ninstructions in Section 3.1. Then, find the disposition of the case and code it according\\nto Section 3.2. Knowing the disposition will guide your next step: determining if the petitioner\\nwas the winning party, following the directions in Section 3.3. Next, use the overall opinion content\\nto code the issue area as described in Section 3.4. This step will also inform the legal provisions\\ncoding as laid out in Section 3.5. With the winning party, disposition, and issue area fixed,\\ncarefully and precisely follow the directions in Section 3.6 to assign a decision direction for the opinion.\\nBased on that decision direction and the disposition, you can assign a lower court direction as well,\\nfollowing Section 3.7. Finally, justify your decisions in the manner described below.\\n\\n==================================================\\n1. INPUT FORMAT\\n==================================================\\n\\nYou will receive input in two sections:\\n\\n[CASE_METADATA]\\nCase name: <case name as in the caption>\\nCourt: <name of the Court of Appeals, e.g. \\\"United States Court of Appeals for the Ninth Circuit\\\">\\nDate decided: <YYYY-MM-DD or best approximation>\\nOpinion type: <e.g. \\\"majority\\\", \\\"per curiam\\\", \\\"on-the-merits\\\", \\\"rehearing\\\", \\\"en banc\\\", etc.>\\nHas dissenting opinion: <Yes or No, based on whether the input data contains a separate dissent>\\nHas concurring opinion: <Yes or No, based on whether the input data contains a separate concurrence>\\nAdditional notes: <any other metadata or summary, if provided>\\n\\n[CASE_TEXT]\\n<full or excerpted opinion text here>\\n[END_CASE]\\n\\nUse both metadata and text to code the variables. If metadata and text conflict, prefer the\\nsubstantive discussion in the opinion.\\n\\n==================================================\\n2. OUTPUT FORMAT\\n==================================================\\n\\nYou must output ONLY a single JSON object with exactly these keys:\\n\\n- \\\"petitioner\\\"\\n- \\\"respondent\\\"\\n- \\\"case_disposition\\\"\\n- \\\"winning_party\\\"\\n- \\\"issue_area\\\"\\n- \\\"legal_provisions\\\"\\n- \\\"decision_direction\\\"\\n- \\\"lower_court_direction\\\"\\n- \\\"declaration_of_unconstitutionality\\\"\\n- \\\"method\\\"\\n- \\\"who_appealed\\\"\\n- \\\"dissent_basis\\\" (only if the case has a dissenting opinion)\\n- \\\"concurrence_type\\\" (only if the case has a concurring opinion)\\n- \\\"concurrence_basis\\\" (only if the case has a concurring opinion)\\n- \\\"justification\\\"\\n\\nSchema:\\n\\n- \\\"petitioner\\\": {\\n \\\"name\\\": string, // short party name from caption; usually the appellant\\n \\\"description\\\": string, // role and identity, e.g. \\\"federal criminal defendant appealing conviction\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"respondent\\\": {\\n \\\"name\\\": string,\\n \\\"description\\\": string, // e.g. \\\"United States as appellee\\\", \\\"private employer\\\", \\\"federal agency\\\"\\n \\\"category\\\": integer, // 1–7 general litigant type (see Section 3.1)\\n \\\"category_label\\\": string // label for the category code\\n }\\n\\n- \\\"case_disposition\\\": {\\n \\\"code\\\": integer, // must be one of the codes below\\n \\\"label\\\": string\\n }\\n\\n- \\\"winning_party\\\": {\\n \\\"code\\\": integer, // 0, 1, or 2 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"issue_area\\\": {\\n \\\"code\\\": integer, // 1–14 from the list below\\n \\\"label\\\": string // short description of the main issue area\\n }\\n\\n- \\\"issue_area_alt\\\": {\\n \\\"code\\\": integer, // 1–14; your best non-Misc alternative if issue_area.code == 13, else same as issue_area.code\\n \\\"label\\\": string\\n }\\n\\n- \\\"legal_provisions\\\": [\\n {\\n \\\"area_code\\\": integer, // 1–9 (law area codes below)\\n \\\"area_label\\\": string,\\n \\\"provision_text\\\": string // e.g. \\\"Fourth Amendment, search and seizure\\\", \\\"18 U.S.C. § 924(c)\\\"\\n },\\n ...\\n ]\\n\\n- \\\"decision_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"lower_court_direction\\\": {\\n \\\"code\\\": integer, // 1, 2, or 3 as defined below\\n \\\"label\\\": string\\n }\\n\\n- \\\"declaration_of_unconstitutionality\\\": {\\n \\\"code\\\": integer, // 0 or 1 (see Section 3.8)\\n \\\"label\\\": string // \\\"no\\\" or \\\"yes\\\"\\n }\\n\\n- \\\"method\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.9)\\n \\\"label\\\": string\\n }\\n\\n- \\\"who_appealed\\\": {\\n \\\"code\\\": integer, // 1–4 (see Section 3.10)\\n \\\"label\\\": string\\n }\\n\\n- \\\"dissent_basis\\\": { // ONLY include if \\\"Has dissenting opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"concurrence_type\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"code\\\": integer, // 1–3 (see Section 3.12)\\n \\\"label\\\": string\\n }\\n\\n- \\\"concurrence_basis\\\": { // ONLY include if \\\"Has concurring opinion: Yes\\\" in metadata\\n \\\"procedural\\\": integer, // 0–3 (see Section 3.11)\\n \\\"merits\\\": integer // 0–3 (see Section 3.11)\\n }\\n\\n- \\\"justification\\\": string // 2–4 sentences summarizing why you chose these codes.\\n // You MUST explicitly say: (a) who won and why, (b) what issue_area.code is,\\n // (c) how that combination implied the decision_direction.code you chose, and\\n // (d) why the lower_court_direction is the same as or opposite to decision_direction.\\n\\nYour entire response must be exactly one JSON object in this schema and nothing else.\\n\\n==================================================\\n3. VARIABLE INTERPRETATION\\n==================================================\\n\\n3.1 Petitioner and Respondent\\n\\n- Use the case caption, opinion type, and text to determine who is appealing.\\n- Petitioner = the party seeking relief in the Court of Appeals (typically the appellant).\\n- Respondent = the party opposing the appeal (typically the appellee).\\n- The \\\"name\\\" should be a reasonably short label from the caption.\\n- The \\\"description\\\" should describe their role and identity (e.g., \\\"federal prisoner challenging sentence\\\",\\n \\\"state official in civil-rights suit\\\", \\\"private corporation defendant\\\", \\\"federal agency\\\").\\n\\nFor each party, also assign a \\\"category\\\" code using this 7-category typology:\\n\\n1 = private business or corporation (including partnerships, sole proprietorships)\\n2 = nonprofit organization, interest group, or union (as an institutional party)\\n3 = federal government (including U.S. as party, federal agencies like DOJ, IRS, EPA, SEC, NLRB, etc.)\\n4 = sub-state government (city, county, municipality, school district, local official)\\n5 = state government (state, governor, state agency, state official in official capacity)\\n6 = tribal government or Indian nation\\n7 = natural person (individual, including pro se litigants and individual officials sued personally)\\n\\nChoose the category based on who the party IS, not what kind of case it is. When a government\\nofficial is sued in their official capacity, code them as the relevant government level (3, 4, or 5).\\nWhen sued in their personal/individual capacity, code them as 7 (natural person).\\n\\n\\n3.2 Disposition of Case\\n\\nCodes (Songer-style numbering):\\n\\n1 = affirmed (includes modified and affirmed)\\n2 = reversed\\n3 = reversed and remanded\\n4 = vacated and remanded\\n5 = affirmed in part, reversed (or vacated) in part\\n6 = affirmed in part, reversed (or vacated) in part, and remanded\\n7 = vacated\\n8 = petition denied or appeal dismissed\\n9 = stay, petition, or motion granted\\n10 = certification to or from a lower court\\n11 = affirmed, vacated, and remanded (without reversal)\\n0 = no disposition\\n\\n- Use the text and opinion type (majority, rehearing, etc.) to identify what the Court of Appeals actually did.\\n- Only choose options 3, 4, or 6 if the opinion contains explicit instructions to remand the matter to the lower court\\n or original authority (the word \\\"remand\\\" must appear in the opinion text).\\n- For rehearing petitions, be explicit about whether the petition is granted (9), denied (8), or leads to modification\\n or vacatur (7) of the prior judgment.\\n- If the Court takes no substantive action on the merits (e.g., purely procedural opinion), use 0 (\\\"no disposition\\\").\\n\\n\\n3.3 Winning Party\\n\\nCodes:\\n\\n0 = no favorable disposition for petitioning party apparent\\n1 = petitioning party received a favorable disposition\\n2 = favorable disposition for petitioning party unclear\\n\\n- Treat the **petitioner** as the party seeking relief in the Court of Appeals.\\n- A \\\"favorable disposition\\\" is one that materially improves the petitioner's position relative to the lower court judgment\\n (e.g., reversal, vacatur, remand requiring relief, sentence reduction).\\n- For mixed outcomes (e.g., affirmed in part and reversed in part), assess whether, on balance, the petitioner is better off.\\n If clearly better off, use 1; if clearly not better off, use 0; if ambiguous, use 2.\\n- Note that an outcome favoring the petitioning party may be either an affirmation (if the lower court had ruled in favor of the petitioner)\\n OR a reversal (if the lower court had ruled against the petitioner). You should focus on whether the petitioner is better off,\\n regardless of the case disposition.\\n\\n\\n3.4 Issue Area\\n\\nChoose exactly one code and label from this list, based on the main legal/policy issue in the appeal:\\n\\n1 = Criminal Procedure\\n (e.g., search and seizure, arrest, interrogation, right to counsel, trial and sentencing rules, habeas)\\n\\n2 = Civil Rights\\n (e.g., equal protection, discrimination based on race or other protected traits, voting rights, school desegregation)\\n\\n3 = First Amendment\\n (e.g., freedom of speech, press, association, religion, and related expression claims)\\n\\n4 = Due Process\\n (e.g., non-criminal procedural fairness, property or liberty deprivations, economic regulation challenges framed as due process)\\n\\n5 = Privacy\\n (e.g., contraception, abortion, family relationships, intimate or sexual privacy, personal autonomy)\\n\\n6 = Attorneys\\n (e.g., attorney's fees, sanctions, bar admission and discipline, representation issues)\\n\\n7 = Unions\\n (e.g., labor relations, union membership and internal affairs, collective bargaining, union security)\\n\\n8 = Economic Activity\\n (e.g., antitrust, securities, regulation of business, commercial law, bankruptcy, financial regulation)\\n\\n9 = Judicial Power\\n (e.g., jurisdiction, standing, mootness, justiciability, sovereign immunity, standards of review, appellate procedure)\\n\\n10 = Federalism\\n (e.g., conflicts between federal and state authority, preemption, Tenth Amendment issues, state sovereign immunity vs federal power)\\n\\n11 = Interstate Relations\\n (e.g., disputes between states, boundary and water disputes, interstate compacts and obligations)\\n\\n12 = Federal Taxation\\n (e.g., federal income, corporate, estate, excise, and other federal tax issues)\\n\\n13 = Miscellaneous\\n (cases that genuinely do not fit any other issue area — see IMPORTANT note below)\\n\\n14 = Private Action\\n (e.g., private law disputes in property, contracts, torts, wills, estates, and similar matters)\\n\\nPick the **single best** issue area based on what the Court of Appeals treats as the central question.\\nIf multiple areas appear, choose the one most emphasized in the reasoning.\\n\\nIMPORTANT — AVOID MISCELLANEOUS (13):\\n\\nCode 13 should be used ONLY as a genuine last resort when NO other category applies even\\napproximately. Before choosing 13, reconsider the following common mappings:\\n\\n - Medical malpractice, products liability, personal injury, breach of contract,\\n admiralty, property disputes → 14 (Private Action)\\n - Workers' compensation (LHWCA, FELA, Longshore Act) → 7 (Unions) or 14 (Private Action)\\n - Bankruptcy, securities, antitrust, business regulation → 8 (Economic Activity)\\n - Social Security, disability, welfare benefits → 2 (Civil Rights)\\n - Attorneys' fees, bar discipline → 6 (Attorneys)\\n - Federal preemption of state law → 10 (Federalism)\\n - Procedural rehearing or jurisdictional questions → 9 (Judicial Power)\\n - Separation of powers, legislative veto → 13 (Miscellaneous) — this is a genuine use\\n\\nIf you DO choose issue_area.code = 13, you MUST also fill in \\\"issue_area_alt\\\" with your\\nbest-guess alternative code (the closest non-Misc category). If issue_area.code is not 13,\\nset \\\"issue_area_alt\\\" to the same code and label as \\\"issue_area\\\".\\n\\nISSUE AREA TIEBREAKING RULES:\\n\\nWhen a case implicates more than one issue area, apply these rules to select the primary one:\\n\\n- Criminal Procedure (1) vs. Civil Rights (2): If the court's holding concerns criminal\\n conviction, sentencing, or procedure itself (e.g., suppression of evidence, habeas review of\\n a conviction), use 1 even if discrimination or civil-rights statutes are mentioned. Use 2\\n only when the holding primarily addresses whether discriminatory treatment occurred or whether\\n a civil-rights statute was violated.\\n\\n- Criminal Procedure (1) vs. Due Process (4): If the party is a criminal defendant raising\\n due-process claims about trial procedure, sentencing, or conditions of confinement, use 1.\\n Use 4 only when the due-process claim arises in a non-criminal context (e.g., government\\n termination of benefits, property seizure in a civil case).\\n\\n- Judicial Power (9) vs. any other area: Many cases discuss jurisdiction, standing, or\\n justiciability before reaching the merits. Use 9 ONLY when the court's holding is\\n primarily about whether it has the power to hear the case — i.e., the court resolves a\\n jurisdictional or justiciability question and that resolution IS the outcome. If the court\\n establishes jurisdiction and then rules on the merits, use the substantive merits area instead.\\n\\n- Economic Activity (8) vs. Federalism (10): If the holding is about whether a federal\\n statute preempts state regulation, use 10. If the holding is about the substance of a\\n federal regulatory scheme (e.g., upholding or striking down an EPA rule on the merits),\\n use 8.\\n\\n- Civil Rights (2) vs. Due Process (4): If the claim involves alleged discrimination under\\n equal protection or a civil-rights statute, use 2. Use 4 only when the claim is about\\n procedural fairness or property/liberty deprivation without an underlying discrimination\\n allegation.\\n\\n\\n3.5 Legal Provisions\\n\\nFor each major legal provision the Court of Appeals applies or interprets, add an entry to \\\"legal_provisions\\\".\\n\\nLaw Area codes (SCDB varLawArea):\\n\\n1 = Constitution (original text)\\n An Article or Section of the original Constitution (not an Amendment).\\n Examples: Article I (congressional power), Article II (executive power),\\n Article III (judicial power), Supremacy Clause, Commerce Clause, Full Faith\\n and Credit Clause, Contracts Clause.\\n\\n2 = Constitutional Amendment\\n Any Amendment to the Constitution, including the Bill of Rights.\\n Examples: First Amendment (speech, religion), Fourth Amendment (search and\\n seizure), Fifth Amendment (due process, takings, self-incrimination), Sixth\\n Amendment (right to counsel, jury trial), Eighth Amendment (cruel and unusual\\n punishment), Fourteenth Amendment (equal protection, due process).\\n\\n3 = Federal Statute\\n A statute enacted by Congress, cited by title and section of the U.S. Code or\\n by its popular name. This is the most common category.\\n Examples: 42 U.S.C. § 1983 (civil rights), 28 U.S.C. § 1331 (federal question\\n jurisdiction), Title VII of the Civil Rights Act, Sherman Act § 1 (antitrust),\\n 18 U.S.C. § 924(c) (firearms), ERISA, CERCLA, ADA, FOIA, Sentencing Guidelines.\\n\\n4 = Court Rules\\n Federal Rules of Civil Procedure, Criminal Procedure, Evidence, or Appellate\\n Procedure, as well as local circuit rules.\\n Examples: Fed. R. Civ. P. 12(b)(6) (failure to state a claim), Fed. R. Civ. P.\\n 56 (summary judgment), Fed. R. Crim. P. 11 (plea agreements), Fed. R. Evid. 403\\n (prejudicial evidence).\\n\\n5 = Other\\n Common law doctrines, equitable principles, or other legal authorities not fitting\\n the categories above.\\n Examples: common law fraud, maritime law (general), equitable estoppel, doctrine\\n of laches, implied contracts.\\n\\n6 = Infrequently litigated statutes\\n Federal statutes that appear rarely in appellate litigation and are not well-known\\n by popular name. Use this when the statute is obscure or highly specialized.\\n Examples: Railway Labor Act, Packers and Stockyards Act, Suits in Admiralty Act,\\n obscure provisions of the Internal Revenue Code.\\n\\n8 = State or local law or regulation\\n A state statute, state constitutional provision, local ordinance, or state/local\\n regulation. Use this when the court applies or interprets state or local law rather\\n than federal law.\\n Examples: state tort law, state contract law, state criminal statute, local zoning\\n ordinance, state environmental regulation.\\n\\n9 = No Legal Provision\\n The opinion does not meaningfully rely on any identifiable legal provision. Very\\n rare — use only when the court's analysis is purely factual or procedural without\\n citing specific legal authority.\\n\\nFor each entry:\\n- `area_code`: choose the single best area for the provision.\\n- `area_label`: the corresponding label (e.g., \\\"Constitution\\\", \\\"Federal Statute\\\").\\n- `provision_text`: identify the specific provision (e.g., \\\"Fourth Amendment, search and seizure\\\",\\n \\\"42 U.S.C. § 1983\\\", \\\"Federal Rule of Civil Procedure 12(b)(6)\\\", \\\"state tort law\\\").\\n\\nInclude the one to three provisions central to the court's analysis. If the opinion does not meaningfully\\nrely on any legal provision (very rare), use a single entry with area_code 9 and a brief explanation.\\n\\nDistinguishing area codes:\\n- If the court interprets the CONSTITUTION ITSELF (e.g., Commerce Clause), use 1.\\n- If the court interprets a CONSTITUTIONAL AMENDMENT (e.g., Fourth Amendment), use 2.\\n- If the court interprets a FEDERAL STATUTE (the most common scenario), use 3.\\n- If the dispute is about a PROCEDURAL RULE (e.g., Rule 12(b)(6) motion), use 4.\\n- Use 8 for STATE LAW even if the case is in federal court on diversity jurisdiction.\\n- Use 5 for common law doctrines that are not codified in any statute or rule.\\n- Use 6 only for genuinely obscure federal statutes that would not be recognized by name.\\n\\n\\n3.6 Decision Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\n>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\\nCRITICAL: The codes \\\"conservative\\\" and \\\"liberal\\\" are MECHANICAL LABELS defined by the SCDB\\ncodebook. They are NOT your own assessment of political ideology. You MUST apply them exactly\\nas the lookup table below specifies, even when the result seems counterintuitive. For example,\\na decision favoring a taxpayer over the IRS is ALWAYS coded conservative (1) in a Federal\\nTaxation case, regardless of the policy context. A decision upholding a government regulation\\nis ALWAYS coded liberal (2) in an Economic Activity case, even if the regulation benefits\\nbusiness. Do NOT let your general understanding of politics override these rules.\\n<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\\n\\nDIRECTION LOOKUP TABLE — apply mechanically based on issue_area.code and who benefits:\\n\\nissue_area.code | benefiting side | direction\\n----------------|------------------------------------------------------------|----------\\n1 (Crim. Pro.) | accused / convicted person / criminal defendant | 2 (liberal)\\n1 (Crim. Pro.) | prosecution / law enforcement / government | 1 (conservative)\\n2 (Civ. Rts.) | civil rights or civil liberties claimant | 2 (liberal)\\n2 (Civ. Rts.) | government defendant / party opposing civil rights claim | 1 (conservative)\\n3 (1st Amend.) | person asserting First Amendment protections | 2 (liberal)\\n3 (1st Amend.) | government restricting speech, press, religion, etc. | 1 (conservative)\\n4 (Due Proc.) | individual challenging deprivation of liberty or property | 2 (liberal)\\n4 (Due Proc.) | government upholding deprivation | 1 (conservative)\\n5 (Privacy) | person asserting privacy / autonomy / reproductive choice | 2 (liberal)\\n5 (Privacy) | government restricting privacy or autonomy | 1 (conservative)\\n7 (Unions) | union / workers / collective bargaining interests | 2 (liberal)\\n7 (Unions) | employer / management / anti-union interests | 1 (conservative)\\n8 (Econ. Act.) | workers, consumers, injured parties, environment, govt | 2 (liberal)\\n | regulation, or small entities against large business |\\n8 (Econ. Act.) | employers, large business, property owners, deregulation | 1 (conservative)\\n9 (Jud. Power) | broader court access / standing / review | 2 (liberal)\\n9 (Jud. Power) | narrower court access / standing / review | 1 (conservative)\\n10 (Federalism) | federal authority over state authority | 2 (liberal)\\n10 (Federalism) | state sovereignty over federal power | 1 (conservative)\\n12 (Fed. Tax.) | United States / IRS / federal taxing authority | 2 (liberal)\\n12 (Fed. Tax.) | taxpayer | 1 (conservative)\\n11,14 | use 3 (unspecifiable) unless decision clearly aligns | 3 (unspecifiable)\\n | with one of the patterns above |\\n\\nPROCEDURE: You MUST FIRST determine who won the case (winning_party), THEN identify the\\nissue_area, and LAST look up the decision_direction from the table above by matching the\\nissue_area.code with the side that benefited. Do NOT reason about direction from first\\nprinciples — use the table.\\n\\nIf, after applying the table, you cannot determine which side benefited, use 3 (unspecifiable).\\n\\n\\n3.6.1 Worked Examples\\n\\nThe following three examples demonstrate correct mechanical application of the direction\\nlookup table, including cases where the SCDB coding may seem counterintuitive:\\n\\nEXAMPLE A — Federal Taxation (issue_area.code = 12)\\n\\n Facts: A taxpayer appeals an IRS assessment to the Court of Appeals. The court reverses the\\n Tax Court, holding that the taxpayer's deductions were proper and the assessment was incorrect.\\n\\n Coding:\\n issue_area.code = 12 (Federal Taxation)\\n winning_party.code = 1 (petitioner/taxpayer received favorable disposition)\\n decision_direction = 1 (conservative) ← taxpayer won; lookup table row for code 12\\n says \\\"taxpayer → 1 (conservative)\\\"\\n\\n NOTE: Even though the taxpayer prevailed against the government, the SCDB mechanically\\n codes this as conservative (1) because the taxpayer benefited.\\n\\n\\nEXAMPLE B — Criminal Procedure (issue_area.code = 1)\\n\\n Facts: A criminal defendant appeals his conviction, arguing that evidence was obtained\\n through an unlawful search. The Court of Appeals affirms, holding the search was lawful.\\n\\n Coding:\\n issue_area.code = 1 (Criminal Procedure)\\n winning_party.code = 0 (no favorable disposition for petitioner)\\n decision_direction = 1 (conservative) ← prosecution/government benefited; lookup table\\n row for code 1 says \\\"prosecution → 1 (conservative)\\\"\\n\\n\\nEXAMPLE C — Economic Activity (issue_area.code = 8)\\n\\n Facts: A chemical company challenges an EPA water-quality regulation. The Court of Appeals\\n affirms, upholding the regulation as within the agency's statutory authority.\\n\\n Coding:\\n issue_area.code = 8 (Economic Activity)\\n winning_party.code = 0 (no favorable disposition for petitioner/company)\\n decision_direction = 2 (liberal) ← government regulation upheld against business;\\n lookup table row for code 8 says \\\"govt regulation → 2 (liberal)\\\"\\n\\n NOTE: Even if one might view upholding a regulation as a \\\"pro-business\\\" outcome (e.g., it\\n creates regulatory certainty), the SCDB mechanically codes government regulation prevailing\\n over a business challenge as liberal (2).\\n\\n\\n3.6.2 Internal Consistency Check\\n\\nBefore you output the JSON, you MUST run the following consistency check:\\n\\n1. Identify issue_area.code.\\n2. Identify which side won in the case (petitioner vs respondent) and what kind of party they are.\\n3. Look up the corresponding row in the DIRECTION LOOKUP TABLE above.\\n4. Verify that your decision_direction.code matches the table entry.\\n5. If it does not match, CORRECT decision_direction.code to match the table.\\n\\n\\n3.7 Lower Court Direction\\n\\nCodes:\\n\\n1 = conservative\\n2 = liberal\\n3 = unspecifiable\\n\\nUse the same DIRECTION LOOKUP TABLE as for Decision Direction (see Section 3.6), but applied to\\nthe **lower court's outcome**. You MUST determine the lower court's direction by comparing the lower\\ncourt's judgment and the appellate court's disposition.\\n\\nUse this rule of thumb:\\n\\n1. First, identify the appellate disposition (case_disposition.code).\\n2. Next, identify the decision_direction.code for the Court of Appeals.\\n3. Last, infer who benefited in the lower court and assign lower_court_direction as follows:\\n\\n- If the Court of Appeals clearly **leaves the lower court judgment in place** (e.g., case_disposition.code = 1 \\\"affirmed\\\"\\n or 8 \\\"petition denied or appeal dismissed\\\"), then the lower court's direction is the SAME as decision_direction.code.\\n\\n- If the Court of Appeals clearly **undoes the lower court judgment** (e.g., case_disposition.code = 2 \\\"reversed\\\",\\n 3 \\\"reversed and remanded\\\", 4 \\\"vacated and remanded\\\", or 7 \\\"vacated\\\"), then the lower court's direction is the\\n OPPOSITE of decision_direction.code.\\n\\n- For mixed outcomes (case_disposition.code = 5 or 6), compare who won more in the lower court vs on appeal:\\n * If, on balance, the petitioner is better off after the appeal than under the lower court judgment, then\\n the lower court's direction should usually be the OPPOSITE of decision_direction.code.\\n * If, on balance, the petitioner is worse off or no better off than under the lower court judgment, then\\n the lower court's direction should usually be the SAME as decision_direction.code.\\n * If you cannot determine who benefited more from the lower court judgment even approximately, use 3 (unspecifiable).\\n\\n- For stays, motions, certifications, or purely procedural dispositions (codes 9, 10, 0), infer who benefited in\\n the lower court if possible; if not, use 3 (unspecifiable).\\n\\nWhen you set lower_court_direction, remember:\\n- \\\"SAME\\\" means lower_court_direction.code = decision_direction.code.\\n- \\\"OPPOSITE\\\" means: if decision_direction.code = 1 (conservative), then lower_court_direction.code = 2 (liberal);\\n if decision_direction.code = 2 (liberal), then lower_court_direction.code = 1 (conservative).\\n\\n3.8 Declaration of Unconstitutionality\\n\\nCodes:\\n\\n0 = the court did NOT strike down any statute, regulation, or government action as unconstitutional\\n1 = the court DID declare a statute, regulation, or government action unconstitutional\\n\\nCode 1 if the court holds that a law, regulation, executive action, or government practice violates\\nthe Constitution (whether facial or as-applied). This includes striking down acts of Congress, federal\\nregulations, state statutes, state constitutional provisions, and local ordinances.\\n\\nDo NOT code 1 merely because the court discusses constitutional issues. The court must actually\\nhold that something is unconstitutional. If the court upholds the constitutionality of the\\nchallenged provision, code 0.\\n\\n\\n3.9 Method\\n\\nCodes:\\n\\n1 = standard panel decision (typically 3 judges, first hearing)\\n2 = panel rehearing (same panel reconsidering its prior decision)\\n3 = en banc (full court or en banc panel rehearing a case)\\n4 = per curiam (unsigned opinion — may overlap with codes 1 or 3)\\n\\nUse the opinion text, metadata, and case caption to determine the method. Look for explicit\\nstatements like \\\"before the court en banc,\\\" \\\"on rehearing,\\\" \\\"upon reconsideration,\\\" or\\n\\\"per curiam.\\\" If the metadata indicates more than 3 judges participated, this strongly\\nsuggests en banc (code 3). If none of these indicators are present, use 1 (standard panel).\\n\\n\\n3.10 Who Appealed\\n\\nCodes:\\n\\n1 = original plaintiff (or the equivalent, e.g. the Title VII complainant, habeas petitioner)\\n2 = original defendant (or the equivalent, e.g. the employer in a Title VII suit)\\n3 = federal government on behalf of the original plaintiff\\n4 = federal government on behalf of the original defendant\\n\\nDetermine who initiated the appeal by examining the case caption, opinion text, and procedural\\nhistory. The \\\"original plaintiff\\\" is the party who filed suit in the lower court; the \\\"original\\ndefendant\\\" is the party who was sued. In criminal cases, the original plaintiff is typically the\\ngovernment (prosecution) and the original defendant is the accused.\\n\\nIf the opinion describes a cross-appeal, code based on who filed the PRIMARY appeal (the one\\nthe court addresses as the main appeal). If you cannot determine who appealed, use 1 as default\\nfor criminal defendants (who are most commonly appellants) or your best judgment from the caption\\nordering (appellant is typically listed first).\\n\\n\\n3.11 Dissent/Concurrence Basis (Conditional)\\n\\nIf the case metadata states \\\"Has dissenting opinion: Yes\\\" or \\\"Has concurring opinion: Yes\\\",\\ncode whether that separate opinion addresses procedural issues, merits issues, or both.\\n\\nFor each dimension (procedural / merits), use one of these codes:\\n\\n0 = the dissent/concurrence exists but does NOT address this type of issue\\n1 = addresses the SAME issue as the majority and DISAGREES with the majority's resolution\\n2 = addresses the SAME issue as the majority and AGREES with the majority's resolution\\n3 = addresses an issue NOT discussed by the majority opinion\\n\\n\\\"dissent_basis\\\" (include ONLY if \\\"Has dissenting opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the dissent address procedural questions (jurisdiction, standing,\\n standard of review, procedural default, timeliness, etc.)?\\n - \\\"merits\\\": Does the dissent address the substantive legal merits (statutory interpretation,\\n constitutional analysis, factual sufficiency, etc.)?\\n\\n\\\"concurrence_basis\\\" (include ONLY if \\\"Has concurring opinion: Yes\\\"):\\n - \\\"procedural\\\": Does the concurrence address procedural questions?\\n - \\\"merits\\\": Does the concurrence address the substantive legal merits?\\n\\nIf multiple dissenting or concurring opinions exist, code based on the PRIMARY one\\n(the longest or most substantive).\\n\\nIf the metadata says \\\"Has dissenting opinion: No\\\", OMIT \\\"dissent_basis\\\" entirely from your output.\\nIf the metadata says \\\"Has concurring opinion: No\\\", OMIT \\\"concurrence_type\\\" and \\\"concurrence_basis\\\" entirely from your output.\\n\\n\\n3.12 Concurrence Type (Conditional)\\n\\nIf the case metadata states \\\"Has concurring opinion: Yes\\\", classify the concurrence as one of\\nthe following types:\\n\\nCodes:\\n\\n1 = regular concurrence — the concurring judge JOINED the majority opinion AND wrote or joined\\n a separate concurrence to add additional reasoning, emphasize a point, or note a limitation.\\n The key indicator: the judge agrees with BOTH the outcome AND the reasoning of the majority.\\n\\n2 = special concurrence / concurrence in result only — the concurring judge agrees with the\\n OUTCOME (disposition) but does NOT join the majority opinion's reasoning. Look for language\\n like \\\"concurring in the result,\\\" \\\"concurring in the judgment,\\\" \\\"concurring in the result\\n only,\\\" or \\\"I concur in the judgment but write separately because I disagree with the\\n majority's analysis.\\\" The key indicator: the judge explicitly declines to join the majority\\n opinion or distances themselves from the majority's reasoning.\\n\\n3 = cannot determine — the opinion text does not make clear whether the concurring judge joined\\n the majority opinion or only the result.\\n\\nThis distinction matters because special concurrences (code 2) signal reasoning disagreement\\nwithin a nominally unanimous disposition. A case where all judges agree on the outcome but one\\nwrites a special concurrence is qualitatively different from one where all judges agree on\\nboth outcome and reasoning.\\n\\nIf multiple concurring opinions exist, code based on the PRIMARY concurrence (the longest or\\nmost substantive). If one is a regular concurrence and another is a special concurrence,\\ncode based on the special concurrence (code 2), as it is more analytically informative.\\n\\n\\n==================================================\\n4. GENERAL RULES\\n==================================================\\n\\n- Use metadata (case name, court, date, opinion type) to:\\n - distinguish main merits opinions from rehearing or procedural opinions,\\n - interpret how the appellate disposition relates to the prior judgment,\\n - infer whether the case is criminal, civil rights, tax, etc.\\n- If some information is missing, use the appropriate \\\"unspecifiable\\\" or \\\"no disposition\\\" code\\n (e.g., 3 for direction, 11 for case_disposition); do not invent facts.\\n- Carefully follow the steps laid out in the introduction in order. In particular,\\n decide who won the case (winning_party). Then, decide what the issue_area is.\\n Only choose decision_direction once you have completed those steps, and follow the exact\\n directions in Section 3.6.\\n- In the justification, explicitly state who won, why (according to the disposition), and\\n how that winner and the issue_area code together implied the direction code you chose.\\n Reference the DIRECTION LOOKUP TABLE row you applied.\\n- ALWAYS choose exactly one code for each coded variable where required.\\n- NEVER output any keys other than the ones defined in the schema above.\\n- Include \\\"dissent_basis\\\", \\\"concurrence_type\\\", and \\\"concurrence_basis\\\" ONLY when the metadata\\n indicates the corresponding opinion type exists. Omit them entirely otherwise.\\n- NEVER output any numeric codes outside the permitted lists above.\\n\\n\\n==================================================\\n5. FINAL REQUIREMENT\\n==================================================\\n\\nAfter reading [CASE_METADATA] and [CASE_TEXT], respond ONLY with the JSON object in the schema\\nabove. Do not include any extra commentary, markdown, or explanation outside the JSON.\\n\",\n \"role\": \"system\"\n },\n {\n \"content\": \"[CASE_METADATA]\\nCase name: MTD PRODUCTS INC., Appellant v. Andrei IANCU, Under Secretary of Commerce for Intellectual Property and Director of the United States Patent and Trademark Office, Intervenor\\nCourt: U.S. Court of Appeals for the Federal Circuit\\nDate decided: 2019-08-12\\nOpinion type: majority\\nHas dissenting opinion: No\\nHas concurring opinion: No\\nAdditional notes: \\n\\n[CASE_TEXT]\\nUnited States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n MTD PRODUCTS INC.,\\n Appellant\\n\\n v.\\n\\n ANDREI IANCU, UNDER SECRETARY OF\\n COMMERCE FOR INTELLECTUAL PROPERTY\\n AND DIRECTOR OF THE UNITED STATES\\n PATENT AND TRADEMARK OFFICE,\\n Intervenor\\n ______________________\\n\\n 2017-2292\\n ______________________\\n\\n Appeal from the United States Patent and Trademark\\nOffice, Patent Trial and Appeal Board in No. IPR2016-\\n00194.\\n ______________________\\n\\n Decided: August 12, 2019\\n ______________________\\n\\n JOHN SALVATORE CIPOLLA, Calfee, Halter & Griswold\\nLLP, Cleveland, OH, argued for appellant. Also repre-\\nsented by ANDREW ALEXANDER, TRACY SCOTT JOHNSON,\\nMARK MCDOUGALL.\\n\\n PETER JOHN SAWERT, Office of the Solicitor, United\\nStates Patent and Trademark Office, Alexandria, VA, ar-\\ngued for intervenor. Also represented by THOMAS W.\\nKRAUSE, PHILIP J. WARRICK.\\n\\f2 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\n ______________________\\n\\n Before REYNA, TARANTO, and STOLL, Circuit Judges.\\nSTOLL, Circuit Judge.\\n The Toro Company sought inter partes review of\\nclaims 1–16 of U.S. Patent No. 8,011,458 before the U.S.\\nPatent and Trademark Office’s Patent Trial and Appeal\\nBoard. The Board instituted review and, in its final writ-\\nten decision, held the challenged claims obvious under\\n35 U.S.C. § 103. Critical to its decision, the Board deter-\\nmined that the claim term “mechanical control assem-\\nbly . . . configured to” perform certain functions is not a\\nmeans-plus-function term subject to 35 U.S.C. § 112, ¶ 6.\\nMTD Products Inc., owner of the ’458 patent, appeals the\\nBoard’s decision.\\n We conclude that the Board erred by conflating corre-\\nsponding structure in the specification with a structural\\ndefinition for the term, and by misinterpreting certain\\nstatements in the prosecution history. Under the appropri-\\nate legal framework, we conclude that the term “mechani-\\ncal control assembly” is a means-plus-function term\\ngoverned by § 112, ¶ 6. We therefore vacate the Board’s\\nobviousness conclusion, which was predicated on its incor-\\nrect claim construction, and remand for further proceed-\\nings consistent with this opinion. Because we are\\npersuaded by MTD’s primary argument, we do not reach\\nits alternative arguments.\\n BACKGROUND\\n I\\n The ’458 patent discloses a steering and driving system\\nfor zero turn radius (“ZTR”) vehicles, with specific refer-\\nence to ZTR lawn mowers. ’458 patent col. 1 ll. 17–21. The\\npatented system is designed to provide a more intuitive\\nsteering mechanism to operators of ZTR vehicles. Id. at\\ncol. 1 ll. 20–38. In contrast to prior art systems that\\n\\fMTD PRODS. INC. v. IANCU 3\\n\\n\\n\\nreverse in the opposite direction of a forward motion turn,\\nthe claimed invention permits ZTR vehicles to turn in the\\nsame direction both forward and backwards. Id. at col. 1\\nll. 20–47. The claimed steering mechanism thus mimics\\nthe forward and backward movements of an automobile.\\n The term “mechanical control assembly” appears in\\nboth claims 1 and 9, the only independent claims of the\\n’458 patent. Claim 1 recites:\\n 1. A vehicle capable of making a small radius turn,\\n comprising:\\n a frame;\\n a left drive wheel and a right drive wheel,\\n both coupled to the frame;\\n two independent left and right drive units,\\n the left drive unit coupled to the left drive\\n wheel via an axle and the right drive unit\\n coupled to the right drive wheel via another\\n axle;\\n a steering device coupled to the frame;\\n a speed control member coupled to the\\n frame; and\\n a mechanical control assembly coupled to\\n the left and right drive units that is config-\\n ured to actuate the left and right drive\\n units based on a steering input received\\n from the steering device and a speed input\\n received from the speed control member;\\n the mechanical control assembly being con-\\n figured such that if the speed control mem-\\n ber is shifted from (a) a forward position in\\n which the left drive wheel is rotating in a\\n forward direction at a first forward speed\\n and the right drive wheel is rotating in a\\n\\f4 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\n forward direction at a second forward\\n speed that is less than the first forward\\n speed as a result of the steering device be-\\n ing in a first right turn position to (b) a re-\\n verse position while the first right turn\\n position of the steering device is main-\\n tained, then the left drive wheel will rotate\\n in a reverse direction at a first reverse\\n speed and the right drive wheel will rotate\\n in a reverse direction at a second reverse\\n speed that is less than the first reverse\\n speed.\\nId. at col. 7 l. 63–col. 8 l. 24 (emphasis added to highlight\\nportion of disputed claim term). Claim 9 is identical to\\nclaim 1 in substantial part, adding only the further limita-\\ntion of:\\n the mechanical control assembly also being config-\\n ured to cause the vehicle to execute a zero-radius\\n turn when the speed control member is in a maxi-\\n mum forward position and the steering device is in\\n a maximum turn position.\\nId. at col. 9 ll. 13–16 (emphasis added).\\n While the patent specification does not expressly refer\\nto a “mechanical control assembly,” it discloses a preferred\\nembodiment that includes a “ZTR control assembly.” Id. at\\ncol. 3 ll. 41–42. The specification describes components of\\nthe ZTR control assembly and its inputs, outputs, and link-\\nages. Id. at col. 3 l. 41–col. 4 l. 57.\\n II\\n Toro petitioned for inter partes review of the ’458 pa-\\ntent in November 2015, arguing that the challenged claims\\nwere invalid as anticipated or obvious. MTD responded\\nthat the term “mechanical control assembly” is a means-\\nplus-function term, and that the asserted prior art did not\\ndisclose the claim term’s corresponding structure. In\\n\\fMTD PRODS. INC. v. IANCU 5\\n\\n\\n\\nsupport of its argument, MTD introduced expert testimony\\nindicating that “mechanical control assembly” has no rea-\\nsonably well-understood meaning in the art. Specifically,\\nMTD’s expert testified that “mechanical control assembly”\\nis a nonce term that is not used in common parlance and\\ndoes not bring to mind any specific structure to a person of\\nordinary skill in the art. J.A. 1366. He explained that the\\nterm is used as a black box recitation for structure and, at\\nmost, amounts to a collection of various parts. J.A. 1248,\\n1366. He further demonstrated that the term is used in\\nvarious prior patents and publications to describe a wide\\nvariety of structures with varying functions. J.A. 1367–69\\n(noting that “mechanical control assembly” is used generi-\\ncally to describe mechanisms for infusion pumps, digital\\nfiring systems, flush tanks, endoscopes, transmissions, and\\nengine outputs).\\n Toro did not expressly contradict MTD’s evidence that\\n“mechanical control assembly” did not have a well-under-\\nstood structural meaning. Instead, Toro responded that a\\nperson of ordinary skill in the art would understand the\\nterm to denote a specific structure in the context of the\\n’458 patent specification. Specifically, Toro argued that the\\n“ZTR control assembly” disclosed in the specification pro-\\nvides an express structural definition for the claimed “me-\\nchanical control assembly.” J.A. 2201–03. Toro also\\nargued that MTD admitted that the term “mechanical con-\\ntrol assembly” conveys particular structure when it distin-\\nguished the patent claims from a prior art reference during\\nprosecution. J.A. 2203.\\n The Board initially agreed with MTD, stating that\\nwhen viewed “in isolation, the genericness of this term\\nbears similarities to other words or phrases that have been\\nheld to be subject to § 112, ¶ 6 . . . such as ‘mechanism,’ ‘el-\\nement,’ ‘device,’ ‘link member,’ and ‘control mechanism.’”\\nToro Co. v. MTD Prods. Inc., No. IPR2016-00194,\\n2017 WL 1969747, at *9 (P.T.A.B. May 10, 2017) (first cit-\\ning Williamson v. Citrix Online, LLC, 792 F.3d 1339, 1350\\n\\f6 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\n(Fed. Cir. 2015) (en banc); then citing Mas-Hamilton Grp.\\nv. LaGard, Inc., 156 F.3d 1206, 1215 (Fed. Cir. 1998); and\\nthen citing Toro Co. v. Deere & Co., 355 F.3d 1313, 1325\\n(Fed. Cir. 2004)). The Board also determined that the “lan-\\nguage reciting what the mechanical control assembly is\\n‘configured to’ do . . . fits the mold of functional language\\nbecause it describes the mechanical control assembly by\\nwhat it does.” Id. The Board thus concluded that “the\\nclaim language of the disputed phrase is primarily, but not\\nentirely, functional, which tends to favor [MTD]’s position\\nthat § 112, ¶ 6 applies.” Id.\\n The Board agreed with Toro, however, that the ’458 pa-\\ntent specification weighed against application of § 112, ¶ 6.\\nId. at *10. Citing the parties’ briefs, the Board stated that\\n“[t]he parties agree that the claimed ‘mechanical control\\nassembly’ is referred to in the specification as a ‘ZTR con-\\ntrol assembly.’” Id. According to the Board, a person of\\nordinary skill in the art would understand “mechanical\\ncontrol assembly” to denote structure because “the specifi-\\ncation illustrates and describes the specific structure that\\nmakes up the ZTR control assembly, and how it connects to\\nand operates with other components.” Id.\\n The most persuasive piece of evidence to the Board,\\nhowever, was the prosecution history. Id. at *9 (“The factor\\nthat weighs most heavily in [the] determination is the pros-\\necution history.”). According to the Board, MTD admitted\\nthat the term “mechanical control assembly” connotes spe-\\ncific structure by asserting that the claims recite “a me-\\nchanical control assembly that is structurally different\\nfrom what [the asserted prior art] discloses.” Id. The\\nBoard emphasized MTD’s statements that “the claim lan-\\nguage at issue concerns the configuration of the claimed\\nmechanical control assembly” and “the claimed configura-\\ntion is indeed structural.” Id. (emphases in original). The\\nBoard concluded that these statements “present[] strong\\nevidence that the disputed phrase should be understood as\\na structural limitation rather than a means-plus-function\\n\\fMTD PRODS. INC. v. IANCU 7\\n\\n\\n\\nlimitation under § 112 ¶ 6.” Id. at *11. Relying on the spec-\\nification and prosecution history, the Board ultimately de-\\ntermined that “mechanical control assembly” is not\\ngoverned by § 112, ¶ 6. Id.\\n DISCUSSION\\n I\\n This appeal requires us to address whether a particu-\\nlar claim limitation is drafted in means-plus-function for-\\nmat. Whether claim language invokes 35 U.S.C. § 112, ¶ 6 1\\nis a legal question of claim construction that we review de\\nnovo. Williamson, 792 F.3d at 1346. We review the\\nBoard’s factual findings underlying this inquiry for sub-\\nstantial evidence. EnOcean, GmbH v. Face Int’l Corp.,\\n742 F.3d 955, 959 (Fed. Cir. 2014).\\n Under this court’s guidance in Williamson, we begin by\\nasking whether the claim limitation employs the word\\n“means.” Williamson, 792 F.3d at 1348. If it does not, we\\napply a rebuttable presumption that the term conveys suf-\\nficiently definite structure and is not subject to § 112, ¶ 6.\\nId. A challenger can rebut the presumption by demonstrat-\\ning “that the claim term fails to ‘recite sufficiently definite\\nstructure’ or else recites ‘function without reciting suffi-\\ncient structure for performing that function.’” Id. (quoting\\nWatts v. XL Sys., Inc., 232 F.3d 877, 880 (Fed. Cir. 2000)).\\nThe “essential inquiry is not merely the presence or ab-\\nsence of the word ‘means’ but whether the words of the\\n\\n\\n 1 Because the issue date of the ’458 patent is Septem-\\nber 6, 2011, and neither the ’458 patent nor the application\\nfrom which it issued ever contained a claim with an effec-\\ntive filing date on or after September 16, 2012, the version\\nof 35 U.S.C. § 112 that applies here is the one preceding\\nthe changes made by the America Invents Act. See Leahy-\\nSmith America Invents Act, Pub. L. No. 112-29, 125 Stat.\\n284, 296, § 4(c) (2011).\\n\\f8 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\nclaim are understood by persons of ordinary skill in the art\\nto have a sufficiently definite meaning as the name for\\nstructure.” Id.\\n One way to demonstrate that a claim limitation fails to\\nrecite sufficiently definite structure is to show that, alt-\\nhough not employing the word “means,” the claim limita-\\ntion uses a similar “nonce word that can operate as a\\nsubstitute for ‘means’ in the context of § 112, para. 6.” Id.\\nat 1350. Generic terms like “module,” “mechanism,” “ele-\\nment,” and “device” are commonly used as verbal con-\\nstructs that operate, like “means,” to claim a particular\\nfunction rather than describe a “sufficiently definite struc-\\nture.” Id. Our case law is replete with guidance on\\nwhether or not a particular claim term is a “nonce” term.\\nSee, e.g., Zeroclick, LLC v. Apple Inc., 891 F.3d 1003, 1008\\n(Fed. Cir. 2018) (holding that “program” and “user inter-\\nface code” are not nonce words because they are “used not\\nas generic terms or black box recitations of structure or ab-\\nstractions, but rather as specific references to conventional\\ngraphical user interface programs or code”); Williamson,\\n792 F.3d at 1350 (noting that “module” is a nonce term be-\\ncause it “sets forth the same black box recitation of struc-\\nture for providing the same specified function as if the term\\n‘means’ had been used”). In each case, a critical question\\nis whether “the claim term is used in common parlance or\\nby persons of skill in the pertinent art to designate struc-\\nture,” including either a particular structure or a class of\\nstructures. Skky, Inc. v. MindGeek, s.a.r.l., 859 F.3d 1014,\\n1019 (Fed. Cir. 2017) (citing TecSec, Inc. v. Int’l Bus.\\nMachs. Corp., 731 F.3d 1336, 1347 (Fed. Cir. 2013)) (hold-\\ning that the claim term “wireless device means” does not\\ninvoke § 112, ¶ 6 because it denotes a class of structures).\\n In addition, even if the claims recite a nonce term fol-\\nlowed by functional language, other language in the claim\\n“might inform the structural character of the limitation-in-\\nquestion or otherwise impart structure” to the claim term.\\nWilliamson, 792 F.3d at 1351. In assessing whether the\\n\\fMTD PRODS. INC. v. IANCU 9\\n\\n\\n\\nclaim limitation is in means-plus-function format, we do\\nnot merely consider the introductory phrase (e.g., “mechan-\\nical control assembly”) in isolation, but look to the entire\\npassage including functions performed by the introductory\\nphrase. See Apex Inc. v. Raritan Computer, Inc., 325 F.3d\\n1364 (Fed. Cir. 2003). The ultimate question is whether\\n“the claim language, read in light of the specification, re-\\ncites sufficiently definite structure to avoid § 112, ¶ 6.” Me-\\ndia Rights Techs. Inc. v. Capital One Fin. Corp., 800 F.3d\\n1366, 1372 (Fed. Cir. 2015) (citing Robert Bosch, LLC v.\\nSnap-On Inc., 769 F.3d 1094, 1099 (Fed. Cir. 2014)).\\n In Apex, for example, the court concluded that the term\\n“circuit” recited sufficient structure in the context of the\\nclaims at issue. 325 F.3d at 1372–73 (considering the use\\nof the term “circuit” in the claim limitation “a first interface\\ncircuit for receiving keyboard and cursor control device sig-\\nnals from the workstation”). Based on a dictionary defini-\\ntion of the word “circuit,” the court reasoned that “the term\\n‘circuit’ by itself connotes some structure,” and that “the\\nterm ‘circuit with an appropriate identifier such as ‘inter-\\nface,’ ‘programming,’ and ‘logic,’ certainly identifies some\\nstructural meaning to one of ordinary skill in the art.” Id.\\nat 1373. The court noted that the extrinsic evidence did\\nnot show that the term “circuit” was not understood to have\\nstructure, but rather “only that the term ‘circuit’ is under-\\nstood . . . as a very broad term.” Id. at 1374. As neither the\\nspecification nor the prosecution history used the term “in\\na manner clearly inconsistent with the ordinary meaning,”\\nthe court held that the defendant failed to rebut the pre-\\nsumption that § 112, ¶ 6 did not apply. Id. at 1373–74.\\n In contrast, in Diebold Nixdorf, Inc. v. International\\nTrade Commission, the court held that the term “cheque\\nstandby unit” for performing certain specified functions\\nwas governed by § 112, ¶ 6. 899 F.3d 1291, 1300 (Fed. Cir.\\n2018). The court noted that “there is no evidence—in the\\nform of dictionary definitions or otherwise—that ‘cheque\\nstandby unit’ was reasonably well understood by persons\\n\\f10 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\nof ordinary skill in the art to refer to a structure or class of\\nstructures.” Id. at 1302. Instead, the extrinsic evidence\\ndemonstrated only that “a skilled artisan would under-\\nstand the functional term ‘cheque standby unit’ to be any\\nstructure capable of performing the claimed function.” Id.\\nat 1301. Further, neither the words of the claim nor the\\nspecification suggested a “structural limitation that might\\nserve to cabin the scope of the functional term,” thus sup-\\nporting the conclusion that the claim limitation was writ-\\nten in means-plus-function format. Id.\\n Finally, we note that “[c]laims are interpreted in light\\nof the written description supporting them, and that is true\\nwhether or not the claim construction involves interpreting\\na ‘means’ clause.” Inventio AG v. ThyssenKrupp Elevator\\nAms. Corp., 649 F.3d 1350, 1356 (Fed. Cir. 2011), overruled\\non other grounds by Williamson, 792 F.3d at 1339. For ex-\\nample, a patentee may avoid application of § 112, ¶ 6 by\\nacting as a lexicographer and providing its own structural\\ndefinition of a nonce term in the specification by “‘clearly\\nset[ting] forth a definition of the disputed claim term’ other\\nthan its plain and ordinary meaning.” Thorner v. Sony\\nComput. Entm’t Am. LLC, 669 F.3d 1362, 1365 (Fed. Cir.\\n2012) (citing CCS Fitness, Inc. v. Brunswick Corp.,\\n288 F.3d 1359, 1366 (Fed. Cir. 2002)). Just as it is im-\\nproper to “import[] limitations from the specification into\\nthe claims,” Phillips v. AWH Corp., 415 F.3d 1303, 1323\\n(Fed. Cir. 2005), however, a preferred embodiment dis-\\nclosed in the specification cannot impart structure to a\\nterm that otherwise has none. As with all lexicography,\\n“[i]t is not enough for a patentee to simply disclose a single\\nembodiment.” Thorner, 669 F.3d at 1365. Rather, “the pa-\\ntentee must ‘clearly express an intent’ to redefine the\\nterm.” Id. (quoting Helmsderfer v. Bobrick Washroom\\nEquip., Inc., 527 F.3d 1379, 1381 (Fed. Cir. 2008)).\\n With these background principles in mind, we turn to\\nthe claim language at issue in this case.\\n\\fMTD PRODS. INC. v. IANCU 11\\n\\n\\n\\n II\\n The disputed claim limitation is lengthy and recites:\\n a mechanical control assembly coupled to the left\\n and right drive units that is configured to actuate\\n the left and right drive units based on a steering\\n input received from the steering device and a speed\\n input received from the speed control member;\\n the mechanical control assembly being configured\\n such that if the speed control member is shifted\\n from (a) a forward position in which the left drive\\n wheel is rotating in a forward direction at a first\\n forward speed and the right drive wheel is rotating\\n in a forward direction at a second forward speed\\n that is less than the first forward speed as a result\\n of the steering device being in a first right turn po-\\n sition to (b) a reverse position while the first right\\n turn position of the steering device is maintained,\\n then the left drive wheel will rotate in a reverse di-\\n rection at a first reverse speed and the right drive\\n wheel will rotate in a reverse direction at a second\\n reverse speed that is less than the first reverse\\n speed.\\n’458 patent col. 8 ll. 7–24.\\n At the outset, we agree with the Board that the term\\n“mechanical control assembly” is similar to other generic,\\nblack-box words that this court has held to be nonce terms\\nsimilar to “means” and subject to § 112, ¶ 6 because the\\nterm does not connote sufficiently definite structure to one\\nof ordinary skill in the art. Toro, 2017 WL 1969747, at *9.\\nWe also agree with the Board that the rest of the “claim\\nlanguage of the disputed phrase is primarily, but not en-\\ntirely, functional.” Id. While the claim language reciting\\nthat the mechanical control assembly is “coupled to the left\\nand right drive units” connotes structure, the claim lan-\\nguage reciting what the mechanical control assembly is\\n\\f12 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\n“configured to” do is functional. In this respect, as the\\nBoard correctly recognized, the claim format tends to favor\\nMTD’s position that § 112, ¶ 6 applies. 2\\n We also discern no error in the Board’s acceptance of\\nMTD’s extrinsic evidence as showing that the term “me-\\nchanical control assembly” does not have an established\\nmeaning in the art and instead merely operates as a ge-\\nneric label for a collection of parts. Toro did not dispute\\nMTD’s expert testimony that, in common parlance, “me-\\nchanical control assembly” does not bring to mind any spe-\\ncific structure to a person of ordinary skill in the art. J.A.\\n1366. Toro likewise did not dispute MTD’s reliance on var-\\nious prior patents and publications that used “mechanical\\ncontrol assembly” to describe a wide variety of structures\\nwith varying functions. J.A. 1367–69 (noting that\\n\\n\\n\\n 2 The Board also stated that construing the disputed\\nphrase as a means-plus-function limitation “would seem to\\nrender the second part [of the claim’s recitation of a ‘me-\\nchanical control assembly’] . . . superfluous” as it “would no\\nlonger serve to define functionally the structural features\\nof the mechanical control assembly, as those features\\nwould be part and parcel of what [MTD] alleges is the cor-\\nresponding structure.” Toro, 2017 WL 1969747, at *9. We\\ndisagree. Both the first and second limitations following\\nthe two occurrences of the term “mechanical control assem-\\nbly” recite functions associated with the “mechanical con-\\ntrol assembly”; i.e., (1) actuate the left and right drive units\\nand (2) rotate the wheel in a particular direction based on\\nthe position of the speed control member. As we have held,\\n“[w]here there are multiple claimed functions, as we have\\nhere, the patentee must disclose adequate corresponding\\nstructure to perform all of the claim functions.” William-\\nson, 792 F.3d at 1351–52. Thus, the corresponding struc-\\nture for “mechanical control assembly” in the specification\\nmust perform both of these functions.\\n\\fMTD PRODS. INC. v. IANCU 13\\n\\n\\n\\n“mechanical control assembly” is used generically to de-\\nscribe mechanisms for infusion pumps, digital firing sys-\\ntems, flush tanks, endoscopes, transmissions, and engine\\noutputs).\\n We conclude that the Board erred, however, when it re-\\nlied on the specification’s description of a “ZTR control as-\\nsembly” to conclude that the claim term “mechanical\\ncontrol assembly” has an established structural meaning.\\nWhile the parties agreed that the ZTR control assembly in\\nthe specification is the structure “corresponding to” the\\nclaimed mechanical control assembly, MTD did not agree\\nthat the specification expressly defines the claim term “me-\\nchanical control assembly.” That the specification discloses\\na structure corresponding to an asserted means-plus-func-\\ntion claim term does not necessarily mean that the claim\\nterm is understood by persons of ordinary skill in the art to\\nconnote a specific structure or a class of structures.\\n Interpretation of an asserted means-plus-function lim-\\nitation involves two steps. First, we determine if the claim\\nlimitation is drafted in means-plus-function format. As\\npart of this step, we consider whether the claim limitation\\nconnotes “sufficiently definite structure” to a person of or-\\ndinary skill in the art. If we conclude that the limitation is\\nin means-plus-function format, the second step requires us\\nto review the specification to identify the structure that\\nperforms the claimed function(s) and thus “corresponds to”\\nthe claimed means. While related, these two inquiries are\\ndistinct. In this case, however, the Board conflated these\\ndistinct inquiries, holding that the specification’s disclo-\\nsure of corresponding structure demonstrates that the al-\\nleged means-plus-function term is sufficiently definite so\\nas to not invoke § 112, ¶ 6. The Board’s analysis implies\\nthat so long as a claim term has corresponding structure in\\nthe specification, it is not a means-plus-function limitation.\\nThis is not consistent with our prior decisions. Indeed, this\\nview would seem to leave § 112, ¶ 6 without any applica-\\ntion: any means-plus-function limitation that met the\\n\\f14 MTD PRODS. INC. v. IANCU\\n\\n\\n\\n\\nstatutory requirements, i.e., which includes having corre-\\nsponding structure in the specification, would end up not\\nbeing a means-plus-function limitation at all.\\n While we agree with the Board that the specification\\nplays a role in assessing whether particular claim language\\ninvokes § 112, ¶ 6, we do not agree that the patent specifi-\\ncation at issue here renders the nonce term “mechanical\\ncontrol assembly” sufficiently structural to a person of or-\\ndinary skill in the art. The specification does not demon-\\nstrate that the patentee intended to act as its own\\nlexicographer and define the nonce term “mechanical con-\\ntrol assembly” as the “ZTR control assembly” of the pre-\\nferred embodiment. Indeed, the specification does not even\\nrefer to a “mechanical control assembly.” Furthermore, the\\nfunctional language in the claim limitation suggests a\\nbroader meaning of the generic term “mechanical control\\nassembly,” as it specifically adds to the “mechanical control\\nassembly” limitation the ability to execute a zero radius\\nturn. ’458 patent col. 9 ll. 13–16. Interpreting the “me-\\nchanical control assembly” as the “ZTR”—or zero-turn-ra-\\ndius—control assembly would render this functional\\nlanguage superfluous.\\n We are also not persuaded by the Board’s interpreta-\\ntion of the prosecution history. While it would have\\navoided uncertainty and argument had MTD shared its\\ncurrent view that the claim limitation is written in means-\\nplus-function format during the original prosecution,\\nMTD’s statements did not clearly disclaim such an inter-\\npretation. Rather, MTD’s statements indicated that the\\nphrase “mechanical control assembly configured to” per-\\nform certain functions must be given weight because it con-\\nnotes structure and thus is not merely an intended use.\\nThese statements were not made within the context of\\n§ 112, ¶ 6. Moreover, stating that the limitation connotes\\nstructure and has weight is not inconsistent with claiming\\nin means-plus-function format since means-plus-function\\nlimitations connote structure (i.e., corresponding structure\\n\\fMTD PRODS. INC. v. IANCU 15\\n\\n\\n\\nand their equivalents) and have weight. Furthermore, as\\nMTD explained, its interpretation of the claims as being in\\nmeans-plus-function format during inter partes review was\\nbased on this court’s intervening law in Williamson.\\nJ.A. 1204. Given the lack of any clear and undisputed\\nstatement foreclosing application of § 112, ¶ 6, we conclude\\nthat the Board erred in giving dispositive weight to the\\nequivocal statements it cited in the prosecution history.\\n CONCLUSION\\n We conclude that the Board erred by using the exist-\\nence of corresponding structure in the specification to con-\\nclude that “mechanical control assembly” has a sufficiently\\ndefinite structure to evade § 112, ¶ 6. The Board also erred\\nby giving improper weight to out-of-context statements in\\nthe prosecution history. We hold that the remaining evi-\\ndence and the Board’s factual findings demonstrate that\\nthe term “mechanical control assembly . . . configured to”\\nperform certain functions in independent claims 1 and 9 of\\nthe ’458 patent is governed by § 112, ¶ 6. We therefore va-\\ncate the Board’s decision and remand for further proceed-\\nings consistent with this opinion.\\n VACATED AND REMANDED\\n COSTS\\n Costs to Appellant.\\n[END_CASE]\",\n \"role\": \"user\"\n }\n]","isConversational":true}}},{"rowIdx":4199,"cells":{"id":{"kind":"number","value":4647452,"string":"4,647,452"},"name":{"kind":"string","value":"IRIDESCENT NETWORKS, INC., Plaintiff-Appellant v. AT&T MOBILITY, LLC, Ericsson Inc., Defendants-Appellees"},"name_abbreviation":{"kind":"string","value":"Iridescent Networks, Inc. v. At&t Mobility, LLC"},"decision_date":{"kind":"string","value":"2019-08-12"},"docket_number":{"kind":"string","value":"2018-1449"},"court":{"kind":"string","value":"U.S. Court of Appeals for the Federal Circuit"},"casebody":{"kind":"string","value":"{\"judges\": \"Prost, Reyna, Taranto\", \"parties\": \"\", \"opinions\": [{\"author\": \"REYNA, Circuit Judge.\", \"type\": \"010combined\", \"text\": \"United States Court of Appeals\\n for the Federal Circuit\\n ______________________\\n\\n IRIDESCENT NETWORKS, INC.,\\n Plaintiff-Appellant\\n\\n v.\\n\\n AT&T MOBILITY, LLC, ERICSSON INC.,\\n Defendants-Appellees\\n ______________________\\n\\n 2018-1449\\n ______________________\\n\\n Appeal from the United States District Court for the\\nEastern District of Texas in No. 6:16-cv-01003-RWS-JDL,\\nJudge Robert Schroeder, III.\\n ______________________\\n\\n Decided: August 12, 2019\\n ______________________\\n\\n SHAWN DANIEL BLACKBURN, Susman Godfrey LLP,\\nHouston, TX, argued for plaintiff-appellant. Also repre-\\nsented by PARKER C. FOLSE, III, IAN B. CROSBY, Seattle,\\nWA; ERIC J. ENGER, ALDEN HARRIS, LESLIE PAYNE, Heim,\\nPayne & Chorush, LLP, Houston, TX.\\n\\n MICHAEL HAWES, Baker Botts, LLP, Houston, TX, ar-\\ngued for defendants-appellees. Also represented by\\nDOUGLAS M. KUBEHL, BETHANY ROSE FORD, JEFFERY SCOTT\\nBECKER, Dallas, TX. Defendant-appellee AT&T Mobility,\\nLLC also represented by BRYANT C. BOREN, JR., Palo Alto,\\nCA.\\n\\f2 IRIDESCENT NETWORKS, INC. v. AT&T MOBILITY, LLC\\n\\n\\n\\n\\n ______________________\\n\\n Before PROST, Chief Judge, REYNA and TARANTO, Circuit\\n Judges.\\nREYNA, Circuit Judge.\\n Iridescent Networks, Inc. sued AT&T Mobility, LLC\\nand Ericsson Inc. in the U.S. District Court for the Eastern\\nDistrict of Texas for infringement of U.S. Patent\\nNo. 8,036,119. Following claim construction, the parties\\njointly stipulated to noninfringement, and the district\\ncourt entered judgment in favor of AT&